Simon Conway Morris-Life's Solution-Cambridge University Press (2003) PDF

Life’s Solution
Inevitable Humans in a Lonely Universe
The assassin’s bullet misses, the Archduke’s carriage moves forward,

and a catastrophic war is avoided. So too with the history of life. Rerun
the tape of life, as Stephen J. Gould claimed, and the outcome must be
entirely different: an alien world, without humans and maybe not even
intelligence. The history of life is littered with accidents: any twist or
turn may lead to a completely different world. Now this view is being
challenged. Simon Conway Morris explores the evidence demonstrating
life’s almost eerie ability to navigate to the correct solution, repeatedly.
Eyes, brains, tools, even culture: all are very much on the cards. So if
these are all evolutionary inevitabilities, where are our counterparts
across the Galaxy? The tape of life can run only on a suitable planet, and
it seems that such Earth-like planets may be much rarer than is hoped.
Inevitable humans, yes, but in a lonely Universe.
s i m o n c o n w a y m o r r i s is Professor of Evolutionary Palaeobiology at

the University of Cambridge. He was elected a fellow of the Royal
Society in 1990, and presented the Royal Institution Christmas lectures
in 1996. His work on Cambrian soft-bodied faunas has taken him to
China, Mongolia, Greenland, and Australia, and inspired his previous
book The Crucible of Creation (1998).
Pre-publication praise for Life’s Solution:
‘Having spent four centuries taking the world to bits and trying to find
out what makes it tick, in the twenty-first century scientists are now
trying to fit the pieces together and understand why the whole is greater
than the sum of its parts. Simon Conway Morris provides the best
overview, from a biological viewpoint, of how complexity on the large
scale arises from simple laws on the small scale, and why creatures like
us may not be the accidents that many suppose. This is the most
important book about evolution since The Selfish Gene; essential
reading for everyone who has wondered about why we are here in a
universe that seems tailor-made for life.’
John Gribbin, author of Science: A History
‘Are human beings the insignificant products of countless quirky

biological accidents, or the expected result of evolutionary patterns
deeply embedded in the structure of natural selection? Drawing upon
diverse biological evidence, Conway Morris convincingly argues that the
general features of our bodies and minds are indeed written into the laws
of the Universe. This is a truly inspiring book, and a welcome antidote
to the bleak nihilism of the ultra-Darwinists.’
Paul Davies, author of How to Build a Time Machine
‘Is intelligent life in the Universe common or incredibly rare? Are even
planets like the Earth rare? We won’t really know until our searches are
further advanced, but until then these debates pivot on the tension
between contingency and convergence. Advocates of the first point to
the unlikelihood of particular historical paths, while those favoring the
second emphasize multiple paths to similar functional outcomes. In
Life’s Solution Conway Morris argues that the evidence from life on
Earth supports a variety of paths leading toward intelligence. Our
searches for life elsewhere are informed by such insights into life here.’
Christopher Chyba, Stanford University and the SETI Institute
Life’s Solution
Inevitable Humans in a
Lonely Universe
simon conway morris

University of Cambridge
  
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo
Cambridge University Press

The Edinburgh Building, Cambridge  , United Kingdom
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521827041
© Simon Conway Morris 2003
This book is in copyright. Subject to statutory exception and to the provision of

relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2003
- ---- eBook (NetLibrary)

- --- eBook (NetLibrary)
- ---- hardback

- --- hardback
Cambridge University Press has no responsibility for the persistence or accuracy of

s for external or third-party internet websites referred to in this book, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
For Zoë, with love
Contents
Preface. The Cambridge sandwich page xi

Acknowledgements xvii
List of abbreviations xix
1. Looking for Easter Island 1

Inherency: where is the ground plan in evolution? 5
The navigation of protein hyperspace 8
The game of life 10
Eerie perfection 13
Finding Easter Island 19
2. Can we break the great code? 22

The ground floor 23
DNA: the strangest of all molecules? 27
3. Universal goo: life as a cosmic principle? 32

A Martini the size of the Pacific 33
Goo from the sky 34
Back to deep space 38
A life-saving rain? 42
4. The origin of life: straining the soup or our

credulity? 44
Finding its path 47
Problems with experiments 49
On the flat 53
Back to the test tube 58
A sceptic’s charter 63
5. Uniquely lucky? The strangeness of Earth 69

The shattered orb 69
Battering the Earth 71
The Mars express 75
viii contents
Making the Solar System 77

Rare Moon 87
Just the right size 92
Jupiter and the comets 93
Just the right place 99
A cosmic fluke? 105
6. Converging on the extreme 106

Universal chlorophyll? 106
The wheels of life? 111
Fortean bladders 112
A silken convergence 115
Matrices and skeletons 117
Play it again! 120
Attacking convergence 126
Convergence: on the ground, above the ground, under
the ground 134
7. Seeing convergence 147

A balancing act 148
Returning the gaze 151
Eyes of an alien? 158
Clarity and colour vision 166
Universal rhodopsin? 170
Smelling convergence 173
The echo of convergence 181
Shocking convergence 182
Hearing convergence 190
Thinking convergence? 194
8. Alien convergences? 197

Down in the farm 198
Military convergence 200
Convergent complexities 205
Hearts and minds 214
Honorary mammals 218
Giving birth to convergence 220
Warming to convergence, singing of convergence,
chewing convergence 223
contents ix
9. The non-prevalence of humanoids? 229

Interstellar nervous systems? 230
The conceptualizing pancake 231
The bricks and mortar of life 234
Genes and networks 237
Jack, the railway baboon 242
Giant brains 243
Grasping convergence 261
Converging on the humanoid 264
Converging on the ultimate 271
10. Evolution bound: the ubiquity of convergence 283

Ubiquitous convergence 284
Respiratory convergence 287
Freezing convergence, photosynthetic convergence 290
The molecules converge 295
Convergence and evolution 298
Converging trends 304
A possible research programme 308
11. Towards a theology of evolution 311

An evolutionary embedment 312
Darwin’s priesthood 314
Heresy! Heresy!! 318
Genetic fundamentalism 323
A path to recovery? 326
Converging on convergence 328
12. Last word 331
Notes 333
Index 446
Preface. The Cambridge
sandwich
Writing in the New York Review of Books,1 John Maynard Smith,

one of Britain’s greatest biologists, remarked ‘If one was able to
re-play the whole evolution of animals, starting at the bottom of the
Cambrian (and, to satisfy Laplace, moving one of the individual
animals two feet to its left), there is no guarantee – indeed no
likelihood – that the result would be the same. There might be no
conquest of the land, no emergence of mammals, and certainly no
human beings’.2 This review, written with characteristic flair and
economy, was addressing three books on evolution, two by S. J.
Gould and the third by E. Mayr. Maynard Smith was raising this
issue because both the authors under review have been forthright in
claiming that the emergence of human intelligence during the
course of evolution has a vanishingly small probability. The logic of
the argument, that because we are unique on this planet then
nothing like us can occur elsewhere, is gently checked by Maynard
Smith: ‘This argument seems to me so manifestly false that I fear I
must have misunderstood it’.3 However, he, Mayr and Gould, and I
imagine almost anyone else, would agree that the likelihood of
‘exactly the same cognitive creatures – with five fingers on each
hand, a vermiform appendix, thirty-two teeth, and so on’4 evolving
again if, somehow, the Cambrian explosion could be rerun is remote
in the extreme.
What, however, of the emergence of more general biological
properties? In considering some earlier views of R. C. Lewontin, who
was uncertain as to whether ‘general principles of biological
organization’5 existed, Maynard Smith was more upbeat: ‘In seeking
a theory of biological form, I would probably place greater emphasis
than Lewontin on the principles of engineering design. I suspect that
there are only a limited number of ways in which eyes can possibly
work, and, maybe only a limited number of ways in which brains
can work. But I agree that it would be good to know whether such
principles exist, and, if so, what they are’.6 Even though neither
Lewontin nor Maynard Smith thought ‘A description of all the
xii preface. the cambridge sandwich
organisms that have ever been’7 could settle this issue,8 Life’s
Solution sets out to demonstrate that what we already know gives
some strong indicators of what must be: even in this book pigs
don’t fly.
The central theme of this book depends on the realities of
evolutionary convergence: the recurrent tendency of biological
organization to arrive at the same ‘solution’ to a particular ‘need’.
Perhaps the best-known example is the similarity between the
camera-like eye of the octopus and the human eye (or that of any
other vertebrate). As we shall see in this particular instance, where
the camera-like eye has evolved independently at least six times,
Maynard Smith’s premise that ‘only a limited number of ways in
which eyes can possibly work’ is amply confirmed. If this book
happens to serve no other purpose than act as a compilation of
evolutionary convergences, be it head-banging in mole rats and
termites or matriarchal social structure in sperm whales and
elephants, then that will be sufficient. But, of course, the net is in
pursuit of a much bigger prey. Its main, but not ultimate, aim is to
argue that, contrary to received wisdom, the emergence of human
intelligence is a near-inevitability. My purpose is not to demonstrate
the inevitability of a five-fingered organism, although in this context
it is amusing to note that the famous panda’s ‘thumb’ is, in one
sense, convergent.9 Nor is it my aim to find repeated examples of
species with 32 teeth, even though we might note that there are a
number of fascinating examples of dental convergence. And it is this
that matters, not five of this or 32 of that, but the recurrent
emergence of various biological properties.
This book has its anecdotes, from baboons operating railway
signals to a harbour seal that spoke like an inebriated Bostonian, but
there is a serious argument that takes us from the apparently arcane,
such as the natural (and convergent) gyroscopes of insects, through
to the convergences of the sensory modalities (vision, of course, but
also olfaction, hearing and echolocation, electroreception, and so on)
to agriculture, brain size, and culture. And there are four
conclusions. First, what we regard as complex is usually inherent in
simpler systems: the real and in part unanswered question in
evolution is not novelty per se, but how it is that things are put
together. Second, the number of evolutionary end-points is limited:
by no means everything is possible. Third, what is possible has
preface. the cambridge sandwich xiii
usually been arrived at multiple times, meaning that the emergence

of the various biological properties is effectively inevitable. Finally,
all this takes time. What was impossible billions of years ago
becomes increasingly inevitable: evolution has trajectories (trends, if
you prefer) and progress is not some noxious by-product of the
terminally optimistic, but simply part of our reality.
There is, however, a paradox. If we, in a sense, are
evolutionarily inevitable, as too are animals with compound eyes or
tiny organelles that make hydrogen, then where are our equivalents,
out there, across the galaxy? After all, the Milky Way has been
available for colonization for at least a billion years, so in Enrico
Fermi’s famous words concerning putative extraterrestrials: ‘Where
are they?’ To paraphrase much of this book, life may be a universal
principle, but we can still be alone. In other words, once you are on
the path it is pretty straightforward, but finding a suitable planet and
maybe getting the right recipe for life’s origination could be
exceedingly difficult: inevitable humans in a lonely Universe. Now,
if this happens to be the case, that in turn might be telling us
something very interesting indeed. Either we are a cosmic accident,
without either meaning or purpose, or alternatively . . .
Enough of backgrounds; what specifically is this book about?
Here is a brief outline. Overall it is a sandwich. The central meat on
convergences is in Chapters 6 to 10, flanked by thinner expositions
in the form of the first five chapters and two end chapters, the last
very short indeed. So, the first two chapters are introductory. They
look at two extraordinarily effective biological systems. The first
concerns the genetic code, how the building blocks of protein, the
amino acids, are read off the DNA. This code is eerily effective,
indeed it has been argued to be ‘one in a million’. This raises the
question of how life navigates to such precise end-points, an analogy
being how the Polynesians in the great diaspora across the Pacific
ever managed to find that remote speck of land that we call Easter
Island. This is followed in the second chapter by a consideration of
DNA, a molecule of iconic if not totemic significance. But for all its
familiarity, DNA also turns out to be one of the strangest molecules
in the Universe. A rather useful invention.
The next two chapters (3 and 4) consider how easy it is to
make the molecules necessary for life, but paradoxically how
difficult it is to make life itself. To some the universality of organic
xiv preface. the cambridge sandwich
material, from immense interstellar gas clouds rich in carbon

compounds to questing bipeds plodding around out-of-the-way
planets, almost suggests the cosmos ‘breathes’ life; a Universe
seeded with vital possibilities. Maybe so, but the trillion upon
trillion tonnes of interstellar organics may still be a universal ‘goo’.
To be sure, they could be the essential ingredient for getting life
started in terms of basic supplies, but the question of just how
inanimate became animate has proved stubbornly recalcitrant. It
should all be rather simple, especially if you worship at the crowded
shrine of self-organization. Yet, somewhere, somehow the right
question has not yet been asked, and not for want of trying.
So confident, however, is the majority that the emergence of
life is a pre-ordained inevitability that the question of whether
beyond the Earth there are any planetary homes available has only
recently emerged. Thus Chapter 5 looks at what we know of the
many peculiarities of our Solar System. Planets there will be aplenty,
but suitable abodes for organic evolution might require very special
sets of circumstances. This is an area that has been reviewed by such
workers as Peter Ward and Donald Brownlee10 and Stuart Ross
Taylor,11 but here I take the argument further as the ferment of
discussion continues.
Chapters 6 to 10 are, as already mentioned, the heart of the
book. They effectively track the story of evolutionary convergence,
starting with the classic cases familiar to biologists as well as some
very intriguing experiments, using bacteria, which allow
evolutionary history to be rerun. That provides a framework of a sort,
but the goal is to argue for the inevitable emergence of sentience.
This is achieved by first (Chapter 7) looking, in some detail, at the
sensory modalities. Eyes provide a superb story, but so too in their
different ways do such features as balance, hearing, olfaction,
echolocation, and electrogeneration: all are rampantly convergent.
These complex systems can arise from very different starting
positions, but again and again converge on the same evolutionary
solution. Chapters 8 and 9 develop the story by seeing how certain
features that we believe are peculiarly human, such as agriculture,
human brains, and even advanced culture are each convergent.
This is not, emphatically, to say that humans are the only
evolutionary outcome worth considering: clearly they are not. And
preface. the cambridge sandwich xv
this leads to the last two chapters (10 and 11), and a brief coda
(Chapter 12). Too often evolutionary convergence is regarded as
simply anecdotal, good for a bedtime story. Its importance is surely
underestimated, and for two reasons. The first is scientific. Ideas on
evolution about such features as adaptation and trends have been
under fierce attack, especially by those who believe that if
contingent happenstance dogs every step of evolution then assuredly
the emergence of humans is a cosmic accident, leaving us free to
make the world as we will, with such happy results as are plain to
see. Yet convergence tells us two things: that evolutionary trends are
real, and that adaptation is not some occasional cog in the organic
machine, but is central to the explanation of how we came to be
here. In principle, such ideas are in themselves so unremarkable as
to require no comment, were it not for the ferocious attacks by such
writers as S. J. Gould. What, one wonders, did he get so excited
about, and how, one may ask, has our understanding of evolution
really changed despite more than forty years of polemic?
Yet, convergence also opens another door. If the emergence of
our sentience was effectively inevitable, then perhaps we should
take rather more seriously the sentiences of other species? So too
perhaps we should stand back and consider what a very odd set-up it
is we inhabit, from the eerily efficient genetic code, to the deeply
peculiar molecule DNA, to a set of biological organizations that
repeatedly throw up complex structures, not least the brain. The late
Fred Hoyle, no friend of most biologists, carried some strange ideas
about the origins of biological complexity to his grave, yet his
remark that the Universe was a set-up job rings strangely true.
Having said that, if you happen to be a ‘creation scientist’ (or
something of that kind) and have read this far, may I politely suggest
that you put this book back on the shelf. It will do you no good.
Evolution is true, it happens, it is the way the world is, and we too
are one of its products. This does not mean that evolution does not
have metaphysical implications; I remain convinced that this is the
case. To deny, however, the reality of evolution and more seriously
to distort deliberately the scientific evidence in support of
fundamentalist tenets is inadmissible. Contrary to popular belief,
the science of evolution does not belittle us. As I argue, something
like ourselves is an evolutionary inevitability, and our existence also
xvi preface. the cambridge sandwich
reaffirms our one-ness with the rest of Creation. Nevertheless, the

free will we are given allows us to make a choice. Of course, it might
all be a glorious accident; but alternatively perhaps now is the time
to take some of the implications of evolution and the world in which
we find ourselves a little more seriously. If you haven’t put Life’s
Solution back on the shelf, please read on.
Acknowledgements
‘To copy one paper is plagiarism, to copy many is scholarship’: few

academics are unfamiliar with one or other version of this gentle
jibe. Moreover, given that my one area of vague scientific knowledge
concerns fossil worms from the Cambrian it will be self-evident that
to have been able to write this book I am heavily dependant on the
expertise, knowledge, and enthusiasm of hundreds of workers. For
this reason I have drawn upon a number of their quotations, which
are of course fully acknowledged. This is not to say that the
researchers I have cited would necessarily agree with the overall
theme of Life’s Solution, but I trust that in each case the context is
clear and fair. Thus I hope that a book that flits from extraterrestrial
amino acids to dolphin brains, from the eyes of spiders to the
discovery of a Roman terracotta head in pre-Columbia Mexico, or
Francis Galton calculating by smells, is understood as an exploration
along a common theme rather than simply a jumble of half-digested
facts. So first I must acknowledge the many authors whose work I
have drawn on liberally. So too I thank the following friends for
reading one or other section, and in a few cases the entire draft at
one stage or another. Thus I record my gratitude to the following
friends: Ken Catania, Stephen Clark, Rob Foley, Stephen Freeland,
Jack Lissauer, Ken McKinney, Lori Marino, Ulrich Mueller, and Nick
Strausfeld for their detailed critiques. In addition, many other
colleagues provided illustrative material (also acknowledged in
specific figures), particular insights, and information. Again, I am
most grateful, and specifically I thank Rachelle Adams, Tim
Bayliss-Smith, Curtis Bell, Yfke van Bergen, Quentin Bone, Graham
Budd, Hynek Burda, John Chambers, Jenny Clack, Rod Conway
Morris, James Crampton, Cameron Currie, Nick Davies, Eric
Denton, Laurence Doyle, Doug Erwin, Albert Eschenmoser, Richard
Felger, Russ Fernald, Larry Field, Siegfried Franck, Adrian Friday,
Linda Gamlin, Liz Harper, Carl Hopkins, Ken Joysey, Harvey Karten,
Jeyaraney Kathirithamby, Richard Keynes, Kuno Kirschfeld, David
Kistner, Mike Land, Charley Lineweaver, Ken McNamara, Charles
xviii acknowledgements
Melville, Eviator Nevo, Dan Nilsson, Euan Nisbet, David Norman,

Ray Norris, Beth Okamura, Art Popper, Christopher Pynes, Simon
Reader, Neill Reid, Michael Ruse, John Taylor, Nigel Veitch, Tom
Waller, Michael Wilson, and Rachel Wood.
It is oxymoronic to say that the mistakes that remain are
mine: of course they are. Nor can I promise that everything is up to
date; it can’t be, nor are my references intended to be exhaustive;
they aren’t. I hope, however, they are sufficient for the interested
reader to begin to explore the literature.
The source of this book was the invitation by Trinity College,
Cambridge, to deliver the Tarner Lectures for 1999, and I thank the
Master and Fellows, especially Boyd Hilton, for their encouragement
and support. I owe an enormous debt to several other people. First, I
wish to thank wholeheartedly Sandra Last for her patience and
stamina as smoothly and flawlessly draft after draft emerged. Next, I
owe a debt of gratitude to the University of Cambridge and
especially the Department of Earth Sciences, for allowing me time
for such an enterprise. I also specifically wish to thank Sharon
Capon and Dudley Simons for assistance with drafting and
photography, and also to acknowledge the superb libraries in many
departments, the University Library, and the unfailingly helpful
librarians. So, too, I give thanks to Cambridge University Press,
especially Sally Thomas, Alison Litherland, and Robert Whitelock,
and to Bruce Wilcock and his skills in disentanglement.
In one way or another, support has been provided by my
college in Cambridge, St John’s, the Leverhulme Foundation, the
SETI Institute, NASA–Ames, and the Royal Society. Finally, I want
to thank my wife Zoë for her interest as I droned on about star-nosed
moles, dolphin brains, or electric fish. To maintain such an attention
span tells me something rather important, and to her I dedicate this
book.
The publisher has used its best endeavours to ensure that the URLs
for external websites referred to in this book are correct and active at
the time of going to press. However, the publisher has no
responsibility for the websites and can make no guarantee that a site
will remain live or that the content is or will remain appropriate.
Abbreviations
general
ATP adenosine 5 -triphosphate, the triphosphate of the
nucleotide adenosine, which plays a key role in the
energetics of the cell. See also p. 25.
BP before the present; by convention taken as before 1950.
CHZ the Circumstellar Habitable Zone, the zone surrounding a
star in which the evolution of life is both possible, and can
be maintained for protracted intervals of time. See also pp.
83, 99–100.
DNA deoxyribonucleic acid, the nucleic acid that forms the basis
of genetic inheritance in nearly all organisms. See also
pp. 4, 23–24.
EOD electric organ discharge; the discharge of electricity from
specialized tissues in fish. See p. 184.
IDO the enzyme indoleamine 2,3-dioxygenase.
JAR jamming avoidance response, exhibited by fish that use
electrogeneration. See p. 186.
K/T the boundary between the end of the Cretaceous (K) period
and the beginning of the Tertiary (T) era at about 65 Ma
ago. The K/T event that occurred at this time resulted in a
mass extinction. See pp. 94–95.
LPTM the late Paleocene thermal maximum, a warm interval that
occurred during the Paleocene period at c. 55 Ma.
OZMA (Project) the first radio-telescope project to search for
extraterrestrial signals, so named by Frank Drake in
reference to organisms as strange as the Wizard of Oz. See
p. 231.
PAHs polycyclic aromatic hydrocarbons; organic compounds with
a carbon “network” that are abundant in the Universe. See
p. 43.
xx list of abbreviations
RNA ribonucleic acid, a polynucleotide that conveys genetic

information to the proteins in the cell. There are three
forms: messenger RNA (mRNA), ribosomal RNA (rRNA),
and transfer RNA (tRNA). See also pp. 4, 13, 44.
SETI the Search for Extraterrestrial Intelligence. See pp. 231–232.
TNA threo-nucleic acid. See pp. 52–53.
UV ultraviolet; electromagnetic radiation in the range between
visible (violet) light and X-rays, i.e. with wavelengths from
about 400 nm to 4 nm. Ultraviolet radiation is invisible to
the human eye, but not to many animals.
some abbreviations for units

Length
cm centimetre, one-hundredth of a metre (0.3937 inch)
m metre (39.37 inches)
km kilometre, 1000 metres (approx. 0.621 mile)
nm nanometre, 10−9 metre, i.e. one millionth of a millimetre
(0.03937 millionths of an inch).
AU astronomical unit, equal to the mean distance of the Earth
from the Sun; 1.496 × 108 km or approx. 93 million miles.
Mass
g gram (0.03527 ounce)
kg kilogram, 1000 grams (2.2046 pounds)
Time
s second
Ga billion years (109 years)
Ma million years (109 years)
Frequency
Hz herz, frequency per second
kHz kiloherz, 1000 Hz
MHz megaherz, 106 herz.
list of abbreviations xxi
Temperature
◦
C degree Celsius (0 ◦ C is the freezing point of water, 100 ◦ C is
the boiling point of water).
K temperature on the Kelvin (thermodynamic) absolute scale
(with 0 K as absolute zero); 1 degree K = 1 degree C; 0 ◦ C is
about 273 K and 100 ◦ C is about 373 K.
Pressure
Pa Pascal, SI unit of pressure, equivalent to the pressure
produced by a force of one newton applied (uniformly) over
an area of one square metre; 105 Pa (100 kPa) is equivalent to
1 bar, or roughly 1 atmosphere.
1 Looking for Easter Island
I am a bipedal hominid, of average cranial capacity, write my

manuscripts with a fountain pen, and loathe jogging. Thanks to years
of work by innumerable biologists I, or anyone else, can tell you to
a fair degree of accuracy when the ability to walk upright began, the
rate at which our brain increased to its present and seemingly aston-
ishing size, and the origin of the five-fingered forelimb whose present
versatility allows me to hold a pen, not to mention the fishy origin of
those lungs that make such a noise as the joggers pass me early in the
morning on Cambridge’s Midsummer Common.
It is obvious that the entire fabric of evolution is imprinted on
and through our bodies, from the architecture of our bony skeleton,
to the proteins carrying the oxygen surging through our arteries, and
our eyes that even unaided can see at least two million years into the
past – the amount of time it has taken for the light to travel from the
Andromeda Galaxy. In every case – whether for hand or brain – we can
trace an ancestry that extends backwards for hundreds of millions, if
not billions, of years. Yet, for all that, both the processes and the impli-
cations of organic evolution remain controversial. Now at first sight
this is rather odd, because it is not immediately clear what is being
called into question. Certainly not the fact of evolution, at least as a
historical narrative: very crudely, first bacteria, then dinosaurs, now
humans. More specifically in terms of process, Darwin’s formulation
of the mechanisms of evolution is not only straightforward, but seem-
ingly irrefutable. Organisms live in a real world, and evolve to fit their
environment by a process of continuous adaptation. This is achieved
by a constant winnowing through the operation of natural selection
that scrutinizes the available variation to confer reproductive success
on those that, by one yardstick or another, are fitter in the struggle for
survival.
So is that all there is to say? The recipe for evolution just given
is a decidedly bald summary. One intuitively senses that it is an in-
herently feeble response to an extraordinarily rich history that has
brought forth an immense coruscation of form and diversity. Among
2 looking for easter island
living forms this ranges across many scales of complexity, from bac-
teria that build colonies like miniature trees1 to immense societies
of ants whose populations run into the millions and, independent of
us, have stumbled across the advantages of agriculture (Chapter 8).
And it is a history that is by no means confined to the complex-
ity of colonies or the limpidity of a geometric shell. It is as much
in the range, scope, and acuity of living organisms. They may be
mere machines, but consider those owls whose hearing can pinpoint
within a two-degree arc the rustling made by a mouse,2 the naviga-
tional abilities of albatrosses across the seemingly trackless South-
ern Ocean3 (Fig. 1.1), or even Nellie the cat that smelled Madagascar
across more than two hundred miles of ocean.4 But despite our admi-
ration, wonder, and – if we are candid – even awe, surely we can still
offer the following paraphrase: evolution happens, this bone evolved
from that one, this molecule from that one. To be sure, not every
transformation and transition will be elucidated, but we are confi-
dent this is because of a lack of information rather than a failure of the
method.
Yet despite the reality that, as it happens, we humans evolved
from apes rather than, say, lizards, let alone tulips, the interpretations
surrounding the brute fact of evolution remain contentious, contro-
versial, fractious, and acrimonious. Why should this be so? The heart
of the problem, I believe, is to explain how it might be that we, a prod-
uct of evolution, possess an overwhelming sense of purpose and moral
identity yet arose by processes that were seemingly without meaning.
If, however, we can begin to demonstrate that organic evolution con-
tains deeper structures and potentialities, if not inevitabilities, then
perhaps we can begin to move away from the dreary materialism of
much current thinking with its agenda of a world now open to lim-
itless manipulation. Nor need this counter-attack be anti-scientific:
far from it. First, evolution may simply be a fact, yet it is in need
of continuous interpretation. The study of evolution surely retains its
fascination, not because it offers a universal explanation, even though
this may appeal to fundamentalists (of all persuasions), but because
evolution is both riven with ambiguities and, paradoxically, is also
rich in implications. In my opinion the sure sign of the right road is
a limitless prospect of deeper knowledge: what was once baffling is
now clear, what seemed absurdly important is now simply childish,
yet still the journey is unfinished.
looking for easter island 3
figure 1.1 Two trackways, obtained by satellite monitoring, of the

Wandering Albatross across the Southern Ocean. Dots indicate data
intercepts, and arrows direction of travel. The upper panel is a departure
from South Georgia, on its 13-day trip it passed the Falkland Islands and
subsequently Tierra del Fuego. Apart from the distance covered, note
the near-straight-line intercept for home. Lower panel is an excursion
from Crozet Islands; note how close are the outward and return
pathways. (Redrawn from fig. 4b of P. A. Prince et al. (1992), Satellite
tracking of wandering albatrosses (Diomedea exulans) in the South
Atlantic, Antarctic Science, vol. 4, pp. 31–6 (upper panel) and fig. 8A of
H. Weimerskirch et al. (1993), Foraging strategy of Wandering
Albatrosses through the breeding season: A study using satellite
telemetry, The Auk, vol. 110, pp. 325–42 (lower panel), with permission
of the authors, Cambridge University Press, and The Auk.)
One such ambiguity is how life itself may have originated.

As we shall see (in Chapter 4) there is no reason to doubt that it oc-
curred by natural means, but despite the necessary simplicity of the
process, the details remain strangely elusive. Life itself is underpinned
by a rather simple array of building blocks. Most notable are the four
(or more accurately five) nucleotides (that is molecules, such as ade-
nine, consisting of a ring of carbon atoms with an attached nitrogen,
a phosphate, and a sugar) that comprise the DNA (and RNA). The
other key building blocks are twenty-odd amino acids that when ar-
ranged in chains form the polypeptides and ultimately the proteins.
Yet, from this, by various elaborations, has arisen the immense diver-
sity of life. At first sight this would seem to encapsulate the entire
process of evolution, yet it soon becomes clear that we hardly under-
stand in any detail the links between the molecular substrate and the
nature of the organism. To be sure, there is some crude correlation be-
tween the total number of genes and the complexity of the organism,
but when we learn that the ‘worm’ of molecular biologists (the ne-
matode Caenorhabditis elegans), which has a relatively simple body
plan with a fixed number of cells, has more genes than the ‘fly’ (the
fruit-fly Drosophila) with its complex form and behaviour, then there
should be pause for thought (see Chapter 9).
One response is to reconsider what we mean by ‘the gene’. In
particular, it is time to move away from a crippling atomistic portrayal
and rethink our views. As has been pointed out by numerous workers,
the concept of the gene is without meaning unless it is put into the
context of what it is coding for, not least an extremely sophisticated
biochemistry. Nor are these the only complications. It is well known
that significant quantities of DNA, at least in the eukaryotic cell (that
is a cell with a defined nucleus and organelles such as mitochondria),
are never employed in the process of coding. Pejoratively labelled as
‘junk DNA’ or ‘parasitic DNA’, it may be just that, silent and surplus
DNA churned out by repeated rounds of duplication of genetic ma-
terial, like an assembly line commandeered by lunatic robots.5 Such
a view fits well with the notion that evolution is a process of blind
stupidity, a meaningless trek from primordial pond to glassy oceans
dying beneath a swollen Sun.
So, beyond the brute fact that evolution happens, the mecha-
nisms and the consequences remain the subject of the liveliest debate
and not infrequently acrimony. But, contrary to the desires and beliefs
inherency: where is the ground plan in evolution? 5
of creation ‘science’, the reality of evolution as a historical process is

not in dispute. And whatever the divergences of opinion, which as
often as not have a tacit ideological agenda concerning the origins of
human uniqueness, there is a uniform consensus that vitalism was
safely buried many years ago, and the slight shaking of the earth above
the grave marking the resting place of teleology is certainly an optical
illusion. But is it an illusion? Perhaps as the roots and the branches of
the Tree of Life are more fully explored our perspectives will begin to
shift. Evolution is manifestly true, but that does not necessarily mean
we should take it for granted: the end results, be it the immense com-
plexity of a biochemical system or the fluid grace of a living organism,
are genuinely awe-inspiring. Could it be that attempts to reinstall or
reinject notions of awe and wonder are not simply delusions of some
deracinated super-ape, but rather reopen the portals to our finding
a metaphysic for evolution? And this in turn might at last allow a
conversation with religious sensibilities rather than the more charac-
teristic response of either howling abuse or lofty condescension.
inherency: where is the ground plan

in evolution?
Although much of this book will be concerned with retelling the
minutiae of biological detail in support of the general thesis of the
ubiquity of evolutionary convergence and, what is more important,
its implications, here is a brief overview of what strike me as the basic
tensions in evolution. The first is what, for want of a better name, I
might term ‘inherency’. A hard-boiled reductionist will dismiss this
as a non-problem, but I am not so sure. Perhaps the first obvious clue
was the result, surprising at the time, of the minimal genetic differ-
ence between ourselves and the chimps. In terms of structural genes
the much-quoted difference amounts, it is said, to about 0.4%. If there
were any residual doubt of the closeness between Homo and Pan, then
other indicators of similarity, such as the fact that the string of amino
acids that make up the protein haemoglobin is identical in number and
sequence, are surely a sufficient indicator of our evolutionary prox-
imity. This, of course, confirms the obvious: we and the chimps share
an ancestor, probably between about 6 and 12 million years ago, and
indeed there is much we have in common. But in other respects we
are poles apart. I’m told that chimps driving cars (or at least go-karts)
have the time of their lives, but we are neither likely to see a chimp
designing a car, nor for that matter mixing the driest of Martinis, let
alone being haunted by existentialist doubts.
This problem of inherency, however, is far more prevalent and
pervasive than the local quirk that chimps and humans are genom-
ically almost identical, but otherwise separated by an immense gulf
of differences. Let us look, for example, at a much deeper stage in our
evolution, effectively at the time of the ancestors of the fish. Enter
the moderately undistinguished animal known as the lancelet worm
or amphioxus (Branchiostoma and its relatives, Fig. 1.2). By general
agreement this beast is the nearest living approximation to the stage
in evolution that preceded the fish, which in turn clambered on to
land, moved to using the egg, grew fur, and in one lineage developed
into socially alert arborealists. All these changes and shifts must have
been accompanied by genetic changes, but if we look back to am-
phioxus we see a genetic architecture in place that seemingly has no
obvious counterpart in its anatomy. To give just one example: the cen-
tral nervous system of amphioxus is really rather simple. It consists
of an elongate nerve cord stretching back along the body, above the
precursor of the vertebral column (our backbone, consisting of a row
of vertebrae) and a so-called brain. The brain can only be described
as a disappointment. It is little more than an anterior swelling (it is
called the cerebral vesicle) and has no obvious sign in terms of its
morphology of even the beginnings of the characteristic threefold di-
vision seen in the vertebrate brain of hind-, mid- and fore-sections. Yet
the molecular evidence,6 which is also backed up by some exquisitely
fine studies of microanatomy,7 suggests that, cryptically, the brain of
amphioxus has regions equivalent to the tripartite division seen in
the vertebrates.
The clear implication of this is that folded within the seem-
ingly simple brain of amphioxus is what can almost be described as
a template for the equivalent organ of the vertebrates: in some sense
amphioxus carries the inherent potential for intelligence. Quite how
the more complex brain emerges is yet to be established. The evi-
dence that a key development in the molecular architecture of the
vertebrates was episodes of gene duplication,8 that is, doubling up of
a gene, could well give one clue. This is because the ‘surplus’ gene
is then potentially available for some new function. It could alterna-
tively be claimed that amphioxus is secondarily simplified (the con-
dition sometimes referred to as regressive), but it retained genes for
inherency: where is the ground plan in evolution? 7
figure 1.2 The amphioxus animal. Upper, entire animal. The anterior
end is to the right, with the ‘brain’ located towards upper side.
Prominent white units are gonads. Lower, detail of anterior with
prominent feeding (buccal) tentacles and more posteriorly gill bars. The
notochord is the longitudinal structure slightly above the mid-line, with
closely spaced vertical lines. The nerve cord lies above the notochord,
with minimal enlargement at the anterior. (Courtesy of Dale Stokes,
Scripps Institution of Oceanography (upper) and Thurston Lacalli,
University of Victoria, British Columbia (lower).)
vital functions, although ones no longer specifically connected to the

coding for a complex brain. Unfortunately the rather limited infor-
mation on the earliest amphioxus-like animals, from the Cambrian
period (c. 545–500 Ma (million years) ago)9 does not extend to seeing
their brains. In general, however, the genomic evidence suggests that
the living amphioxus is not in some sense degenerate but is genuinely
primitive.
Revealing the foundations of the molecular architecture that
underpins our brains and sentience gives us not only a feeling of emer-
gence, but underlines how little we really know about why and how
organic complexity arises. Nor is this example of the amphioxus brain
and its molecular inherency in any way unusual. Equally instructive

examples can be culled from the most primitive animals, such as the
sponges and Hydra (the latter is a relative of the sea-anemones and
corals), in which genes (or proteins) that are essential for complex ac-
tivities in more advanced animals are already present. Doubtless they
have their functions, but what these are and how they have been re-
deployed, co-opted, or realigned in more advanced animals is for the
most part still unknown. The unravelling of these evolutionary stories
is going to be one of the most fascinating episodes in recent biological
history, but what will almost certainly be more extraordinary is how
much of organic complexity will be seen to be latent in more prim-
itive organisms. Or perhaps not that extraordinary: it is sometimes
forgotten that the main principle of evolution, beyond selection and
adaptation, is not the drawing of new plans but relying on the tried
and trusted building blocks of organic architecture.
the navigation of protein hyperspace

Life, then, is full of inherencies. We might legitimately enquire
whether there is anything in the human condition that could prefigure
some future level of complexity that with the virtue of hindsight will,
no doubt, seem to have been inevitable, but to us remains unimag-
inable. Yet whatever privileges exist for untrammelled speculation,
there is a story to be told which will occupy the rest of this book.
My critics will, I imagine, complain at its eclectic, if not unorthodox,
nature; and given that the topics covered will include such matters as
extraterrestrial amino acids and ants pursuing warfare it is advisable
to try to explain the underlying thread of the argument. Here we can
do no better than to look at a stimulating and thoughtful essay written
by Temple Smith and Harold Morowitz,10 which is an exploration of
the tension between the predictabilities of physical systems and the
seemingly contingent muddle that we call history. In brief, and their
paper contains many other insights, they remind us of the simplicity
of the basic building blocks of life, by which is meant such molecules
as amino acids (which go to build the proteins, perhaps familiar as
collagen or haemoglobin), or sugars (which when joined together can
form carbohydrates). In the case of amino acids, however, even with
the rather modest total of the 20 available variants and taking a rela-
tively simple protein – consisting, say, of 100 such building blocks – it
is immediately apparent that the potential number of combinations in
the navigation of protein hyperspace 9
which this protein could be assembled is absolutely gigantic. Specifi-

cally it is 20100 , which is equivalent to 10130 .
This is an uncomfortably large number,11 and, as Smith and
Morowitz see it, this immensity of possibilities confers an inherent
unpredictability on the process of evolution. Taking the figures given
above, that is the 20 available amino acids and their random inclu-
sion in a protein composed of a total of 100 amino acids, Smith and
Morowitz then apply some apparently stringent criteria to the selec-
tion of those proteins that will actually work. The specific function
they have in mind for proteins is as natural catalysts, that is, to func-
tion as the organic molecules known as enzymes,12 which serve to
accelerate metabolic processes. The alternative, of course, is that a
hypothetical protein will be non-functional, failing in one way or an-
other. We know that in principle this is perfectly feasible, because
there are many examples known where only a handful of changes, and
sometimes even the substitution of a single amino acid for another
one, will render the entire protein inoperative and thereby biologically
useless. Let us then suppose that only one in a million proteins will be
soluble, a necessary prerequisite for the watery milieu of a cell. Let us
further suppose, and again the figure seems reasonable, that of these
again only one in a million has a configuration suitable for it to be
chemically active. How many potentially enzymatically active solu-
ble proteins with an amino acid length totalling 100 could we expect
to be available to life? A few thousand, perhaps even a few million?
In fact, the total far exceeds the number of stars in the universe.
As Smith and Morowitz dryly note, ‘It is quite clear from such
numerology that the domain of possible organisms is enormously
large if not infinite’,13 especially when we recall that many proteins
are substantially longer than 100 amino acids. The only way we can be-
gin to envisage such a protein domain is in the abstract terms of a kind
of hyperspace. Mathematically this will encompass all the measure-
ments that together serve to define the totality of this ‘protein space’.
As Smith and Morowitz point out, with such an immense number
of potential possibilities the number of proteins known to exist on
Earth can only be an infinitesmally small fraction of this vast total.
As they say, notwithstanding ‘the immensity of the dimensionality
of the descriptive hyperspace’, the world we know and the evolution-
ary processes that define it have ‘produced a very sparsely sampled
hyperspace in the actual living world’.14
One inference that might be immediately drawn from this is

that in principle the likelihood of any other world employing an area
of ‘protein space’ that is even remotely close to that found on Earth
should be vanishingly small. At this early stage of the argument we
can leave aside, for the moment, the distinct likelihood that Earth-like
planets are going to be in exceedingly short supply (Chapter 5), and
simply remind ourselves that even as our net of exploration spreads
first across the Milky Way and then from galaxy to galaxy, so each time
a protein chemist steps on to the surface of a new planet only another
tiny fraction of this immense ‘hyperspace’ will be documented. The
combinatorial possibilities are so much more immense than all the
planets with all their biospheres that most proteins will for ever be
only hypothetical constructs. That, at least, is the expectation and it
would seem difficult to refute. All other worlds might be expected to
be truly alien, at least in so far as the occupation of protein ‘space’ is
concerned. That is, at least, the assumption.
the game of life

Nevertheless, despite Smith and Morowitz’s calculus of immensity,
matters are probably much less alien than might at first be imagined.
This is because at one level the strings of amino acids and their ex-
act sequence are irrelevant, so long as the protein works effectively.
To be sure, specific regions of a protein may be exceedingly sensitive
to which amino acid is present, but we also know that various pro-
teins have evolved independently of each other to perform a similar,
if not identical, function. Such examples of molecular convergences
are examined in more detail later (Chapter 10). What matters here is
that these convergences emphatically do not depend on arriving at
a closely similar sequence of amino acids, which given the size of
protein ‘hyperspace’ would be almost a miracle. Navigation through
this ‘hyperspace’ depends rather on two principal properties that, as
it happens, underpin all life. The first concerns the remarkable speci-
ficities of particular sites within the protein that confer the necessary
function, for example in those microbial pathogens whose existence
depends on precise molecular mimicry to outwit a host’s defences.
The secondary property is that the complex functions that character-
ize proteins depend not only, in many cases, on highly specific sites,
but also on particular architectural forms that are highly recurrent.15
As we shall see in at least some protein designs, such as those that ren-
der tissue transparent (as in an eye lens), transport or store oxygen (for
the game of life 11
respiration), or are sensitive to light (rhodopsin and cryptochromes),

the same solution to the biological need has been arrived at indepen-
dently several times.
The implications are far-reaching, because the ‘nodes’ of bi-
ological possibility may, because of physical constraints, be much
more limited than is usually supposed. As Michael Denton and Craig
Marshall remark,
If forms as complex as the protein folds are intrinsic features of

nature, might some of the higher architecture of life also be
determined by physical law? The robustness of certain
cytoplasmic forms . . . suggests that [they] may also represent
uniquely stable and energetically favoured structures . . . If it does
turn out that a substantial amount of higher biological form is
natural, then the implications will be radical and far-reaching. It
will mean that physical laws must have had a far greater role in
the evolution of biological form than is generally assumed. And it
will mean . . . that underlying all the diversity of life is a finite set
of natural forms that will recur over and over again anywhere in
the cosmos where there is carbon-based life.16
I agree. Not all is possible, options are limited, and different starting
points converge repeatedly on the same destinations. Any such evo-
lutionary journey, including navigation through protein ‘hyperspace’
must presuppose intermediary stages. And here there may be further
constraints because seemingly ‘sensible’ paths may turn out to be
non-functional.17 The ‘landscape’ of biological form, be it at the level
of proteins, organisms, or social systems, may in principle be almost
infinitely rich, but in reality the number of ‘roads’ through it may be
much, much more restricted.
This is not to say that there are no alternatives: patently there
are, and the world is a diverse place. Smith and Morowitz remind us
that despite these potential immensities the actual ‘Game of Life’,
as they call it, is still going to be played the same way everywhere.
Here are the four basic rules, which incidentally presuppose variation
(which is offered by the different alleles (determining characters) of a
gene) and subsequent process of selection.
(1) Hindsight and foresight are strictly forbidden. Of course we are

fully entitled to hug ourselves with delight as we trace, for ex-
ample, the multiple evolutionary origins of the electric organs
in certain fish (a topic I return to in Chapter 7), but we can only

retrodict and not predict.
(2) Minor changes are easier than major changes. That’s something
all biologists recognize,18 and why, for example, there is a deep-
seated distrust of macroevolutionary ‘jumps’ that allow a fully
fledged body plan to emerge from some strikingly dissimilar
ancestor.
(3) Resources are not unlimited: the world is finite, and ultimately
energy and space are in restricted supply.
(4) Life has no option but to carry on; it must always play the best
hand it can no matter how poor and disastrous the hand might
be, and no matter who or what offers the challenge.
In this way Smith and Morowitz neatly encapsulate what evo-

lution is all about. They suggest that given these four basic rules for
the Game of Life we should not be surprised to see the emergence
both of evolutionary trends and of emerging complexity; Smith and
Morowitz also remind us that symbioses19 and sex are two good ways
of speeding up the game. And that is all there is to it? Not quite. Try-
ing to keep the surprise out of their joint voices, Smith and Morowitz
then continue:
There is at least one major evolutionary trend not immediately

explained by our strategy rules [i.e. their Game of Life]. That being
the numerous examples of morphological convergence. Why, in
the sparsely sampled genetic space, have there been so many cases
of apparent convergence or parallelism? It is surprising in the light
of the high probability for novelty to find, even in similar niches,
high morphological similarity in distinctly different genetic
lines . . . there may be additional rules operating at coarser levels of
the genetic space which are less statistical than those discussed.20
This prescient statement prefigures the main purpose of this book,

that evolution is indeed constrained, if not bound. Despite the im-
mensity of biological hyperspace I shall argue that nearly all of it
must remain for ever empty, not because our chance drunken walk
failed to wander into one domain rather than another but because the
door could never open, the road was never there, the possibilities were
from the beginning for ever unavailable. This implies that we may not
only be on the verge of glimpsing a deeper structure to life, but that
eerie perfection 13
it matters little what our starting point may have been: the different
routes will not prevent a convergence to similar ends.21
eerie perfection
The understanding of the genetic code was, after the elucidation of the
structure of DNA with its four bases and famous double helix, the next
triumph in the field of molecular biology. As already noted, proteins
are built from the twenty available amino acids,22 although it has long
been known that particular examples, such as the protein collagen
that goes to form such structures as tendons (Achilles’ heel) or the
silk proteins that form the spider’s web, are enriched in particular
amino acids which reflect, in ways that even now are not completely
understood, the functional and structural properties of these and other
proteins. Thus collagens are enriched in such amino acids as proline,
while spider-silks possess notable quantities of alanine and glycine.
Each of the amino acids is coded for by a set of three nucleotide
base pairs, accordingly known as a triplet. The original code is, of
course, stored in the DNA of the chromosomes, but the actual synthe-
sis of the amino acids occurs through the agency of the RNA in minute
structures within the cell known as the ribosomes. Thus, in RNA the
four bases are adenine (A), cytosine (C), guanine (G), and uracil (U),
the last of which substitutes for thymine (T), which is found in DNA
only. With a triplet code and four base pairs there are of course 64 pos-
sible combinations. This implies that with only 20 amino acids there
is a considerable degree of redundancy, even with the assignment of
certain codons to signal ‘Start’ and ‘Stop’. In fact we see that only two
amino acids (methionine (abbreviated M) and tryptophan (W)) rely on a
single codon each (respectively coded for by AUG and UGG), whereas
the remaining 18 amino acids are able to call upon from two to six
codons. (For example, histidine (H) uses either CAC or CAU; arginine
(R) employs CGU, CGC, CGA, CGG, AGA, and AGG.) It has long
been known that this redundancy means that mistakes in coding may
not be detrimental; if a substitution within the codon fails to result in
the identical amino acid, it stands a good chance of producing another
amino acid with similar properties. Amino acids with similar proper-
ties, of which their affinity to or repulsion from water (the property of
polarity) is particularly important, also tend to have similar pathways
of biosynthesis. Here, too, if errors occur then the mistake need not
be lethal. For these and other reasons, therefore, it is clear that the
genetic code is excellently adapted to the needs of reliably providing

the amino acids that underpin protein construction.
But how good is good? The rule of thumb in evolution is ‘good
enough to do the job in most circumstances’, but not to waste time
building a Rolls-Royce of an organism, or, to put it more flippantly, no
supersonic albatrosses. Even so, measuring this ‘goodness’ for purpose
is not so easy: organisms themselves are rubbery, slippery, and pliable
and non-invasive techniques of investigation are time-consuming and
often difficult. One way to address this problem is to look at the design
tolerance of an organism, that is, to see the margins of safety built
into such a structure as a bone. A powerful analogy, as Jared Diamond
reminds us,23 is to think of a lift in a prestigious building dedicated
to the serious accumulation and worship of money. ‘Room for one
more’, says the lift attendant, before the cage shuts, shoots skywards
towards the 59th floor, which it never reaches because at the 48th
floor the cable snaps . . . Such instances are, in the absence of malice,
mercifully rare because the safety factor of such a lift cable, measured
as the ratio between its ultimate capacity and maximum load imposed
in normal use, is almost 12 times. The equivalent ratio for a cable in
a dumb waiter ascending with its cargo of brown Windsor soup and
claret is about five; for a bridge, engineers are content to allow a safety
factor of only two. In this last case, however, Henry Petroski reminds
us that the safety factors for some modern bridges may in reality be
perilously small.24
It is perhaps not surprising that by and large the safety factors
adopted by organisms25 are closer to those of the dumb waiter and the
bridge. Thus the silk dragline of a spider has a modest safety factor
of only 1.5, whereas the factor for the leg bone of a kangaroo hop-
ping through the Australian outback is 3. There is an additional and
quite important point that many safety factors may in themselves
be sub-optimal – spider silk does snap and kangaroos can break their
legs – but the margins of safety are necessarily a compromise between
strength and many other vital functions in the organism. Even so,
over-design does provide an important safety margin, especially when
an organism encounters an unpredictable and rare circumstance. In
assessing this and other reasons for such safety margins Carl Gans
also makes the point that such tolerances may facilitate the occu-
pation of a hitherto untested adaptive zone.26 One of the examples
he gives is the New Zealand parrot known as the kea. This is a
eerie perfection 15
fascinating bird with highly adaptable feeding habits. The kea also
has a penchant for trashing cars, and its behavioural characteristics in-
clude delinquent gangs of young birds.27 In passing I should also men-
tion that notwithstanding the overwhelming evidence for adaptation
and functional demands faced by organisms there remain some ex-
amples of structures whose significance still baffles biologists. John
Currey gives a nice example in the form of the rostral bone in the
snout of Blainville’s beaked whale (Mesoplodon densirostris).28 As the
species name suggests, this bone is incredibly dense, but why? One
can speculate that it might be employed in fighting, but this rostral
bone is very brittle, a consequence of its very low organic content.
Alternatively, it might act as ballast,29 but Currey is candid when he
writes, ‘At the moment, its function, in this rarely found whale, is a
mystery’.30
By this stage you will be wondering what possible connection
could exist between the safety factors of a kangaroo, let alone the
rostral bone of a rare whale, and the efficiency of the genetic code.
The point, simply, is that given the realities of the physical world and
adaptation, organisms and their components should be designed to do
the job adequately, but no more. Humans shudder at the prospect of
hurtling to their doom down a lift shaft, and so incorporate a safety
margin that seems to be found very seldom in organisms. And at first
sight this is what we should see in the genetic code: it certainly isn’t
random; in fact it is really rather good. But in recent years a group
of molecular biologists, notably Steve Freeland and Laurence Hurst,
have been trying to arrive at a more precise answer.31
Their approach is computer-based, and the basic aim is to ran-
domize the genetic code and then compare the efficiency of a certain
fraction of the vast number of alternative codes the computer can
generate with the real one, here on Earth. There is, of course, the im-
plicit assumption that a genetic alphabet composed of two base pairs
(that is AT/CG),32 as well as the system of triplet codons and the
20 amino acids33 available for protein construction found in all ter-
restrial life represents some sort of norm. Alternatives to codon usage
and the number and type of amino acids can, of course, be envisaged,
but Arthur Weber and Stanley Miller have gone so far as to suggest
that ‘If life were to arise on another planet, we would expect that . . .
about 75% of the amino acids would be the same as on the earth.’34
Naturally we need to be cautious in assuming that even if proteins
are universal they necessarily depend on the terrestrial mechanism of

codons35 and the same battery of amino acids. Yet there still may be
constraints. Codons built as doublets, i.e. only two base pairs (e.g. AA
or AU) to code for an amino acid, would probably be rather vulner-
able, while quartet or quintet (e.g. AAAA or AUAUA) codons might
be getting cumbersome. There are, of course, many more amino acids
known than are actually employed in the proteins and, as we shall see
(Chapter 3), some of these are best known from meteorites and have no
biological equivalents. Even so, given that the simplest amino acids
(such as glycine, serine, and alanine) are probably the most readily
synthesized anywhere in the Universe, it is possible that they predis-
pose the biosynthetic pathways that lead to the more complex amino
acids.36 So, perhaps both the genetic code and protein construction
‘out there’ are not so very different.
There is, however, a second difficulty in deciding just how ef-
fective the terrestrial code might be. This is because randomizing the
existing genetic code leads to an astronomical number of alternative
possibilities: Freeland and his co-workers suggest a figure of about
1018 , which, as they helpfully remind us, is ten times as many sec-
onds as have elapsed since the formation of the Earth. It is another big
number (see note 11), and echoes the point I raised in discussing the
essay by Smith and Morowitz (see note 10), that with the immensity
of a protein, or in this case, genetic ‘hyperspace’, it would not only be
a priori exceedingly unlikely that any two biospheres – separated also
by a gulf of many light years – would arrive at the same evolutionary
solution, but it would be even more fantastically improbable that the
solution achieved was not only good (the process of natural selection
should see to that) but in fact the very best. Yet, this appears to be the
implication in the work by Freeland and his colleagues.
Their work, as is customary, has proceeded in several stages.
Well aware of the preceding work already indicating the general effi-
ciency of the genetic code, they examined a million alternative codes
(Figure 1.3). To the first approximation the distribution can be com-
pared to the familiar bell-shaped curve that, it is said, describes the dis-
tribution of human intelligence (IQ): a few stupid people and equally
few geniuses, with most of us somewhere in the middle. So, too, with
the distributions of alternative genetic codes: there is a wide range of
efficiencies; some alternatives are extremely inefficient (‘disastrous’)
and, perhaps not surprisingly, the majority are quite efficient but not
eerie perfection 17
2500
2000
Number of codes
15 000 Natural code
10 000
5000
0
2.00 5.00 8.00 11.00 13.40
Relative efficiency of code
figure 1.3 Eerie perfection. The relative efficiency of randomized

genetic codes, ranging from disastrous on the right to increasingly
competent to the left. Note the approximately bell-shaped curve: most
codes are pretty good, a few terrible, and a few very good. Also note
where this planet’s genetic code falls: far, far to the left. (Reproduced
with permission from Journal of Molecular Evolution, from the article
The genetic code is one in a million, by S.J. Freeland and L.D. Hurst,
vol. 47, pp. 238–248, fig.7; 1998, copyright Springer-Verlag, and also with
the permission of the authors.)
remarkably effective. Very few of the alternatives are really impres-

sive, but note where in Figure 1.3 the real or natural code falls. Freeland
and Hurst have difficulty in keeping the surprise out of their report,
even given the proviso that their approach necessitates a number of
assumptions. They write: ‘the natural genetic code shows startling
[my emphasis] evidence of optimization, two orders of magnitude
higher than has been suggested previously. Though the precise quan-
tification used here may be questioned, the overall result seems fairly
clear: under our model, of 1 million random variant codes produced,
only 1 was better . . . than the natural code – our genetic code is quite
literally “1 in a million”.’ 37
This result, however, needs to be put into a wider context, be-
cause the million (106 ) alternatives that Freeland and Hurst looked
at is only a small fraction of the total number of possibilities, which,
as already noted, they estimate to be about 1018 . On this basis there

could still be an astronomically large number of alternative genetic
codes, each of which in its ‘local’ context could also prove to be very
good indeed when compared to a randomly chosen set of a million
other codes. In their analysis of the million alternatives Freeland and
Hurst specifically noted that the one code that in principle might be
better than the natural one had, as one might expect, little similarity
to the one used by life on Earth. It seems, however, that the potential
figure of 1018 alternatives is, in reality, inflated. This is because not all
the biosynthetic pathways used to construct the 20 different amino
acids are in themselves viable. In a subsequent analysis Freeland and
his co-workers suggest that the number of alternative codes that over-
all are realistically functional is relatively small. They estimate that
this number might be about 270 million; and taking into account the
similarities between certain amino acids they conclude, again in my
opinion startlingly, ‘that nature’s choice [on Earth] might indeed be
the best possible code’.38
In one way we should hardly be surprised at the efficiency of
the genetic code.39 It is difficult to believe that the genetic code is not
a product of selection, but to arrive at the best of all possible codes
selection has to be more than powerful, it has to be overwhelmingly
effective. The reason for saying this is that with some minor, and ev-
idently secondary, exceptions,40 the genetic code is universal to life:
you, the primrose on the table, and the bacteria in your gut all employ
the same code. The earliest evidence for life is about 3.8 billion years
ago and these forms are presumably directly ancestral to all groups
still alive today. If so, this indicates that whatever changes occurred
as the genetic code evolved towards its stable state must have been
achieved still earlier; the genetic code would not otherwise be uni-
versal. Yet, as we shall see (Chapter 4), life itself may not be older
than about 4 billion years. Two hundred million years (and possibly
much less) to navigate to the best of all possible codes, or at least from
the 270 million alternatives? Part of the explanation, as is so often the
case in evolution, may be to look for a step-like arrangement: once one
stage is achieved, other things then become so much more likely.41
Yet, there is also a sense that given a world of DNA and amino acids,
then perhaps the genetic code we know is more or less an inevitable
outcome. And if this is true, then what else might be inevitable, both
here on Earth and elsewhere?
finding easter island 19
This is not the only way to look at inevitabilities in evolution.

The argument from the genetic code looks to a potentially gigantic
‘hyperspace’ of alternative possibilities, yet the evidence suggests that
rapidly and with extraordinary effectiveness a very good, perhaps even
the best, code is arrived at. It is as if the Blind Watchmaker takes off
her sunglasses and decides to visit her brother Chronos. Off she sets,
crossing streets roaring with traffic driven by psychotics, through the
entrails of the subway system of a megalopolis, and, after catching a
series of intercontinental express trains with connection times of two
minutes each, she arrives at Chronos’ front door at 4 p.m. prompt, just
in time for a relaxing cup of tea.
finding easter island

Hence to an explanation for the title of this first chapter. Easter Island
is the remotest speck of land on Earth, surrounded by the vastness of
the Pacific Ocean. At first sight it seems quite extraordinary that it
could have been encountered by the seafaring Polynesians, however
audacious. Surely, one would suppose, it was a chance discovery, per-
haps by mariners who had been blown far off course, which led to the
prows of the first canoes accidentally grating onto a beach of Easter
Island perhaps some 1500 years ago. Another quirk of history? Very
probably not. Easter Island may have marked one of the furthest points
in this great human diaspora, but its discovery was inevitable given
the sophisticated search strategy of the Polynesians. As Geoffrey Irwin
has shown,42 not only were these people superb navigators, but they
developed a method of quartering the ocean that aimed to find new
lands. Century by century their net of exploration widened. When a
particular season failed in the objective they had a sure way of finding
their way home to safety. Their vessels were designed for protracted
journeys, but the key to their success was to head against the prevail-
ing winds on the outward journey. At the limit of their range on any
one journey, the sternward winds rapidly returned them towards their
home and safety. And how was home, another speck in the ocean, ar-
rived at? In the sky above the boats the net of stars provided the clues
to celestial navigation, and as the constellations fitted into place so an
increasingly familiar starlit sky provided the beacons for a successful
homecoming.
So, too, in evolution. Isolated ‘islands’ provide havens of biolog-
ical possibility in an ocean of maladaptedness (Fig. 1.4). No wonder
figure 1.4 A metaphorical view of protein ‘hyperspace’, in which

functional proteins project above an immense ‘ocean’ that submerges
non-functional alternatives. (Reprinted from Journal of Molecular
Biology, vol. 301, D.D. Axe, Extreme functional sensitivity to
conservative amino acid changes on enzyme exteriors, pp. 585–596,
fig. 5. Copyright 2000, with permission from Elsevier Science, and also
with the permission of the author.)
the arguments for design and intelligent planning have such a peren-
nial appeal. Whether it be by navigation across the hyperdimensional
vastness of protein space, the journey to a genetic code of almost eerie
efficiency, or the more familiar examples of superb adaptation, life has
an extraordinary propensity for its metaphorical hand to fit the glove.
Life depends both on a suitable chemistry, whose origins are literally
cosmic, and on the realities of evolutionary adaptation. The chemistry
is acknowledged but largely ignored; the adaptation is often derided as
a wishful fantasy. As with the audacious and intelligent Polynesians,
so life shows a kind of homing instinct. Its central paradox revolves
around the fact that despite its fecundity and baroque richness life
is also strongly constrained. The net result is a genuine creation, al-
most unimaginably rich and beautiful, but one also with an underlying
structure in which, given enough time, the inevitable must happen.
finding easter island 21
To conclude this chapter – and to anticipate the main theme of

this book – let us accept that the genetic code must be spectacularly
efficient, driven to a one-in-a-hundred-million alternatives by the re-
morseless action of selection. All life shares this one code, but this
commonality has not stifled the creative potentials of life, as both the
fossil record and the exuberance of the living world so clearly demon-
strate. Yet for all this exuberance and flair there are constraints: con-
vergence is inevitable, yet paradoxically the net result is not one of
sterile returns to worn-out themes; rather there is also a patent trend
of increased complexity. Some cosmologists like to speculate that the
Universe is designed to be the home of life, to which some biologists
might add ‘Yes, and not only that but we have a pretty shrewd idea of
what was on the cards.’ But to see how the hand is played, we need
first to see how life itself might have originated – and a very odd story
is now emerging.
2 Can we break the great code?
One of the paradoxes of science is that its very greatness as an intel-

lectual adventure is perversely mirrored by a crippling diminution of
what it is to be human. Emerging from the slime, our animal instincts
barely controlled, we are informed in gloating terms of our complete
and utter insignificance.1 Trapped in an out-of-the-way solar system,
the galaxy around us may well pulsate with sentient activity set in
a dazzling array of civilizations, but if we are ever visited it will be
either to view one more zoo (or lunatic asylum) or to stock up on a de-
pleted larder. This is hardly an encouraging view, but, as many others
have pointed out, our genocidal and destructive tendencies may make
a plea for a lenient sentence sound hollow.2 And, in principle at least,
this possibility should be taken seriously. After all, the scientific view
is that the emergence of life is surely inevitable, and there is no short-
age of planets on which the long climb from pond-scum to shopping
must surely be taking place – or, happy thought, perhaps twinkling
just out of sight is a planet now covered with shopping malls, parking
lots, and internet cafes.
The question of whether such a climb need necessarily lead to
a sentient species is the topic of a later chapter (9), although conceiv-
ably your reading of this section may be interrupted by the roar of
descending spacecraft or, perhaps more plausibly, the dramatic news
of the success of the SETI enterprise as the long-awaited extraterres-
trial signals finally confirm that indeed we are not alone. Alterna-
tively, it may be that sentience is rare indeed, and pond-scum is the
rule.3 No matter: orthodoxy states that life is ubiquitous. Indeed, it
may be elevated to a cosmic principle.4 Christian de Duve, for ex-
ample, writes, ‘the Universe was pregnant with life . . . we belong
to a Universe of which life is a necessary component, not a freak
manifestation.’5
In the absence of direct evidence for extraterrestrial life, the
question of the ubiquity of life in the cosmos really hinges on dis-
covering how easy it might be to synthesize life and on reassuring
ourselves that there are mansions a-plenty to allow a near-infinitude
the ground floor 23
of evolutionary experiments. Yet, as I shall try to demonstrate in

Chapter 4, the scientific annexe dedicated to the problem of the origin
of life has not been marked by a series of sweeping and spectacular
advances – the norm for science – but has lurched indecisively across
a landscape dotted with stumbling blocks and crevasses. Nor does the
discovery that the processes of organic chemistry permeate the galaxy
offer any more than a crumb of hope. The next chapter (3) is entitled
‘Universal goo’, and with good reason. And is there really a plurality
of worlds, a trillion planets endowed with an unimaginable range of
biological diversity, while on a favoured handful – let’s be conserva-
tive and propose a mere ten thousand – millions of alien ‘children’
wait for the ‘bedtime’ story as the spectacular sunset of Threga IX6
floods the landscape? Distant solar systems are now being discovered
at a remarkable rate (see Chapter 5), but, as we shall see, the results so
far are distinctly discouraging: worlds without number, but strange,
hostile, and most probably uninhabitable.
the ground floor

What then might be ground rules for life, here or anywhere? At first
sight this second chapter may seem something of a diversion. This is
especially so because it looks to what preceded even the genetic code,
discussed in the previous chapter: that is, to consider the structure of
DNA. But there is a purpose to this approach because it can be argued
that DNA shows some very peculiar features. This, in turn, is the
first hint that in terms of its fundamental building blocks there may
be few, if any, alternatives available to life, here on Earth or at the far
end of the Universe. Yet the basic building blocks of DNA (composed
in essence of the nucleotides and a sugar), and for that matter those
of proteins (the amino acids), are really rather simple. There is a para-
dox here, which Maitland Edey and Don Johanson rightly regard as
‘breathtaking’.7 Their words have a peculiar resonance in this book.
Speaking of DNA and life itself they write:
on the one hand [there is] the endless complexity of the process,
on the other the simplicity of the principle. To make everything
that can be called alive, to monitor the development of every fern
and feather on earth, to direct their growth, to enable them to
function, to replace worn-out parts, to turn things on, to turn them
off – for all those activities to be orchestrated by just four kinds of
24 can we break the great code?
small molecules [that is, the nucleotide bases] is awe-inspiring. It

is that magisterial power of DNA, the power to direct, that
commands our attention. Its molecules are absurdly simple. They
are not alive. In many ways they resemble crystals. But they can
do things no crystal ever dreamed of . . . They wave their
instructional wands . . . proteins grow, enzymes snip and patch.
Lo, there emerges a bacterium, a flower, a fish, a Frenchman.8
From remarkable simplicity arises immense complexity, yet a basic

theme still emerges which confers on evolution a broad predictability.
If indeed we can delineate the architecture of life, then two tantalizing
prospects arise. Perhaps we can really begin to explore the reality of
alternatives, of evolutionary counterfactuals. And possibly, to antici-
pate the last chapter (12), we shall discover in the end that there are
none. And, despite the almost crass simplicity of life’s building blocks,
perhaps we can discern inherent within this framework the inevitable
and pre-ordained trajectories of evolution? This second topic, of life’s
directions, will occupy the second half of this book, but now let us
look briefly at some of the underpinnings of life.
At a fundamental level matters seem to be rather simple. The
principal elements required for life, at least as we know it, are car-
bon, hydrogen, oxygen, nitrogen, and phosphorus. These are all read-
ily available, and carbon certainly has an almost uncanny knack of
arranging itself in configurations that are both flexible and robust.
Not that we should fail to consider alternatives. Iris Fry, for example,
has an interesting discussion of both the definitions of life and the
likelihood of radical alternatives that regard the ‘carbaquist’ (i.e. car-
bon plus water) foundation as hopelessly parochial.9 An ever-popular
topic in this area is whether the element silicon, whose most familiar
natural manifestation is probably as quartz crystals, could ever act
as a substitute for carbon in providing the backbone of molecular
architecture, be it in some sort of equivalent to the carbohydrates,
lipids, or proteins of life. Silicon certainly possesses some advantages
over carbon, such as its much greater abundance. Yet these seem to
be more than offset by a number of drawbacks.10 For example the
silicon–silicon bond is unstable in the presence of water and oxygen,
which is something of a disadvantage for a water-based aerobic or-
ganism. In addition, silicon has a tendency to unite with oxygen to
form very stable silicate polymers. Fry reminds us that speculations
the ground floor 25
on alternatives need not be restricted to hypothetical silicon-based life

forms. Robert Shapiro and Gerald Feinberg go even further and conjure
up life in liquid ammonia, molten silicates, the interiors of stars, or
within interstellar clouds.11 Fry reminds us that however far-fetched
these ideas might be ‘the formal characteristics of hypothetical ex-
traterrestrial life [still derive] from the principles of organization of
life as we know it,’12 and she continues,
Every living system, as revealed particularly at the molecular

level, is organized in a much more complex way than any ordered
physical system known to us. The unique character of this
complexity lies in the ability of an organism to maintain and
reproduce its organization according to specific internal
instructions, or information, manifested in specific
macromolecules. This character is connected with the purposeful,
functional nature of biological organization, in which each part
serves the survival of the whole. The Shapiro and Feinberg
definition [effectively free flow of energy and emergence of order]
is silent on those prominent biological features.13
These comments are important in the context of what is, and is not,
likely in alien ‘biospheres’. It is important to stress that Fry does
not regard non-carbaquist life forms as impossible, but in some con-
texts, e.g. stellar interiors, their detection is problematic, and even
planetary-based systems that might involve liquid ammonia or silicon
‘backbones’ (note 10) run into difficulties. Thus, liquid ammonia
requires very low temperatures (less than −33 ◦ C) and its ice sinks,
unlike water ice, which is a remarkable solid in that it floats on its
own liquid. Fry concludes, together with most workers, that life else-
where, if there is any, is most probably carbon-based.
Somewhat similar arguments may also apply to a number
of elements vital for the operation of life, perhaps most notably
phosphorus.14 In the chemical form of phosphate it plays key, and
probably irreplaceable, roles. Of central importance is its employment
in the construction of the nucleic acids, the building blocks of DNA
and RNA. Phosphorus is also integral to the molecule ATP (adenosine
triphosphate), which is central to the storage and transfer of energy
within the cell. As with the case of the merits of silicon versus carbon,
it needs to be emphasized that it is not that alternatives simply can-
not be envisaged. Indeed, if humans had been detailed to get the whole
process going, the eager occupants of the First Laboratory might well
have passed by phosphates to consider such biochemical alternatives
as arsenic and silicic acid (note 14). As is repeatedly the case in looking
at the basis of life, however, these alternatives are substantially less
fit for the purpose. As we shall also see when discussing some current
problems in understanding the origin of life (Chapter 4), it is proba-
bly just as well that humans or their equivalents were not recruited
to the First Laboratory: my guess is that they would still be hard
at work.
There is another twist to the way life uses the elements of the
periodic table. Consider, for example, the rather unfamiliar element
known as molybdenum. Some years ago Francis Crick and Leslie Orgel
came up with a particularly intriguing idea that nicely reflected their
ingenuity and lateral thinking.15 In essence their argument was as
follows: molybdenum is an essential element for life. It plays, for
example, a key role in various enzymes. On Earth, however, the avail-
ability of this element seems to be drastically limited, occurring as
it does in only minute concentrations. Crick and Orgel’s imaginative
leap was to argue that this dependence on molybdenum was not an
unfortunate, if brute, fact, but actually it indicated that our ances-
try was far from Earth, derived from life forms where the precious
molybdenum was more freely available. The idea that we might rep-
resent marooned colonists – perhaps from a long-dead planet engulfed
in some stellar catastrophe – has a romantic appeal that taps a recur-
rent root in humans of displacement and longing. Not, of course, that
these hypothetical colonists would be anything more than bacteria
or some such equivalent. In any event, the history of life provides no
evidence (although perhaps it should) of any subsequent visitation, let
alone intervention, by extraterrestrials. Of course, getting even bacte-
ria across the interstellar wastes, those cubic parsecs of hard vacuum
drenched in radiation, is in itself so problematic that it may be rea-
sonable to suppose that if panspermia (that is, transport from one star
system to another16 ) occurs at all it can only be by a directed, that is,
an intelligent, activity. And this is what Crick and Orgel suggested:
from a simple observation, too little molybdenum, to a bold hypoth-
esis of a directed panspermia. The latter notion may still be credible,
but, alas, the original estimates of the amounts of molybdenum were
too pessimistic. It looks as though even on Earth there is actually
enough.17
dna: the strangest of all molecules? 27
It seems, therefore, that not only is life definitely terrestrial

(unless it is from Mars; see Chapter 5), but is firmly grounded in a
number of key elements of which carbon is pre-eminent. There are,
however, few elements in the periodic table that are not employed
somewhere or other in the panoply of life.18 With the exception of
carbon it would be stretching matters, at least from a biochemical
perspective, to describe the elements themselves as the real building
blocks of life. The standard view is that this epithet is better applied to
the nucleotide base pairs of DNA (and RNA), as well as the twenty-odd
amino acids that are employed to build the proteins, not to mention
the sugars, the basis for the carbohydrates and more complex polymers
such as cellulose and chitin, and the fatty acids (based on a hydrocar-
bon chain) that are the key component of the lipids employed, for
example, in the cell wall.
The DNA is, of course, the basis of the genetic information;
and through the further agency of RNA it provides the basic instruc-
tions that give rise to the scaffolding of life, most notably in the
proteins. There are two points to be made at this juncture. First it
seems reasonable to suppose that in unravelling at least one compo-
nent of the mystery of the origin of life (Chapter 4) we must ask not
only how the nucleotides and amino acids were synthesized in some
credible prebiotic milieu – be it as different as an interstellar cloud or
a tidal pool – but also how they came to be in a position of mutual
association. As we shall see, there remains a yawning gulf between
this unremarkable expectation and reality, at least in so far as experi-
mental data are concerned. The second point actually pertains to the
central theme of this book: how does evolutionary complexity emerge
and what viable alternatives are there to the world we know?
dna: the strangest of all molecules?

Consider again the nucleotides. As individual building blocks these
molecules are rather unremarkable, yet when they combine to form
DNA, they suddenly show some very strange properties indeed.
Consider, for example, these comments by Christopher Switzer
and colleagues:19 ‘The interaction between the two complementary
oligonucleotide strands [that entwined form the famous double helix]
remains one of the most remarkable examples of molecular recogni-
tion known to chemistry, especially because several features of nu-
cleic acids appear a priori to suit them poorly for this process,’20 that
is their quite superb performance in the process of replication, the

sine qua non of DNA. It is difficult to believe that any terrestrial
chemist, if handed those molecules and asked to make something
useful, would feel encouraged. As Switzer et al. point out the whole
problem is fraught with ambiguities. First, the oligonucleotides are a
type of polymer that is effectively unidimensional and also flexible.
Stringing them into the double helix involves unexpected distortions
to the polymer. In the technical language of the original article they
experience a ‘substantial loss of conformational entropy’.21 Put sim-
ply, polymers like these shouldn’t really behave like this. Moreover,
the double helix itself springs various surprises. The specificity of the
helix, with the pairing between the oligonucleotides (in DNA adenine
matches to thymine, and guanine to cytosine) is mediated by hydro-
gen bonds; yet, as Switzer et al. again point out, such an arrangement
is decidedly ‘problematic in an aqueous environment’.22 Here, too,
a hypothetical designer might be expected to construct a replicatory
system that would not face, potentially at least, disablement in the
watery interior of the cell. Yet despite these and other ambiguities23
the action of DNA is almost uncannily effective.
Nor does the versatility of the DNA molecule end here: it has
a whole series of other strange properties. Not only can the DNA
strand readily coil and twist, but it can also be assembled into so-
called Borromean rings. These are named after the motif of a noble
Italian family, and consist of several interlocked rings so arranged that
if any one ring is broken then the whole structure disintegrates.24
DNA might be most familiar as a double helix, but it can also be
arranged in two-dimensional sheets,25 and even configured into a sort
of octahedron.26 Not only that, but DNA can act as a template upon
which metals can be seeded, forming, for example, tiny gold crystals27
and silver wires.28
The very strangeness of DNA, of course, immediately begs the
question: all right, granted that it is certainly a decidedly odd mole-
cule, but surely we can envisage alternatives? The answer to this
question is that indeed we can inasmuch as considerable progress
has been made in constructing various sorts of ‘artificial’ DNA (and
RNA). The crucial question is whether, even if such a DNA shows su-
perior properties, however defined, such an alternative configuration
could have arisen by natural processes. The picture is still emerg-
ing, but what is not in doubt is that the five bases (adenine (A),
cytosine (C), guanine (G), thymine (T, in DNA)/uracil (U, in RNA))

found in nature can be replaced in the laboratory by other base pairs,
which for obvious reasons are referred to as ‘non-standard’. Quite a few
are available to the experimenter; typical examples are the molecules
isocytidine (iso-C) and isoguanosine (iso-G). When incorporated into
a short strand of DNA, these novel base pairs do not usually alter
the overall geometry of the double helix but the arrangement of the
hydrogen bonds does differ. Potentially we could envisage a new type
of DNA with not the four and possibly optimal (note 32, Chapter 1)
standard ‘letters’ (ACGT/U), but six (or perhaps even more?). Switzer
and his colleagues even wonder whether ‘In another, more whimsi-
cal form, we might ask whether the (iso-C).(iso-G) base pair might be
found as part of a coding system in extraterrestrial life.’29
This is not the first time we shall encounter apparent whimsy
as covering a potentially interesting evolutionary question. I would
argue that these remarks by Switzer et al. can be taken seriously:
it is surely important to explore at least some of the alternatives.
This is for at least two reasons. First, it may illuminate how DNA
(or more probably RNA if one follows the popular view of a prior
‘RNA world’) emerged in the prebiotic milieu. Second, the effort in
imagining alternatives may assist in the exercise of trying to con-
ceive what may approximate to an alien biochemistry. The fact that
these so-called ‘non-standard’ base pairs can be woven, by suitable
laboratory manipulation,30 into the double helix certainly excites
our curiosity. As Joseph Piccirilli and his co-workers31 have point-
edly remarked, however, despite their special interest to chemists
such novelties ‘apparently never have been pursued by nature.’32 At
present it seems that the reason why we do not find these alterna-
tive DNAs, at least on Earth, is not due to chance factors but for two
specific reasons. First, these ‘unnatural’ base pairs33 seem to be sig-
nificantly more difficult to prepare in plausible prebiotic conditions.
Thus Piccirilli et al. also note that such non-standard base pairs rep-
resent ‘Structures [that] are undoubtedly more difficult to prepare un-
der prebiotic conditions . . . perhaps explaining their absence from the
repertoire of natural nucleotides. But such molecules can be obtained
by synthetic chemistry which is not subject to prebiotic constraints.’34
Second, and perhaps more significantly, serious difficulties then
arise in replicating faithfully at least some of the non-standard
substitutions.35
Introducing peculiar base pairs is, of course, not the only way
of meddling with DNA (and RNA) to see what alternatives might ex-
ist, either in the laboratory or extraterrestrially. The question can, in
any event, be tackled from several directions. Here, too, the results
are neither comprehensive nor complete, but for the most part they
seem to suggest that finding viable alternatives to DNA in any sort
of natural context may be difficult, and perhaps even impossible. For
example, if a series of base pairs are linked to make a short strand, in
a process known as oligomerization, then if this is carried out in the
test tube, the results are interestingly different from those obtained
within the living cell. Although the molecules that are employed to
build DNA are relatively simple, the association of the various com-
ponents is precisely determined. For example, the phosphate unit and
the sugar (known as a ribose), both of which are essential compo-
nents of DNA, are each attached to the other principal unit, i.e. the
nucleotide (e.g. adenine) at two specific locations of sites. In ‘real’
DNA, which is made in organisms, the linkage is at sites 3 and 5
(the superscripts refer simply to positions on the molecule). In con-
trast, in the artificial situation in the laboratory, the linkage positions
are now at 2 and 5 .36 Perhaps the former arrangement (3 ,5 ) evolved
from the latter (2 ,5 ), abiotic leading to biotic. If it did, this may not
be surprising, because DNA showing the 2 ,5 linkage seems to be per-
ceptibly weaker and, at least under laboratory conditions viable only
when the salt content of the surrounding solution is elevated.37 In
the case of RNA matters are somewhat different. This is because here
an ‘abiotic’ 2 ,5 linkage does not seem to be prejudicial to its opera-
tion until, that is, DNA is involved in the system. Then everything
collapses.38
What else can we tinker with? Another key element is the sugar,
ribose in RNA and deoxyribose in DNA. As we shall see subsequently
(Chapter 4), one of the particular hurdles in understanding how life
itself might have originated is to find a believable pathway by which
ribose might be synthesized, at least in sufficient quantities to be of
any use. At this juncture all we need to ask, in the context of alter-
native worlds, is whether ribose (or deoxyribose) is the only molecule
in town? In fact, this topic has been explored in considerable detail
by Albert Eschenmoser and his team, in what is appropriately termed
an aetiology of nucleic acids,39 and what has been very appropriately
described as ‘a chemical tour de force.’40
The overall aim of Eschenmoser’s team is to find viable alter-

natives to the genetic code by the synthesis of what they call homo-
DNAs. Such work is important for at least two reasons. First, it might
conceivably give us some glimpse of an alien biochemistry or alterna-
tively confirm that, in a way analogous to the employment of opsins in
visual processes (see p. 170), DNA is the molecule of choice – on Earth,
across the galaxy, and at all points beyond. Second, given the complex-
ity of DNA and the improbability that it evolved de novo from the
prebiotic ‘soup’, alternatives that incorporate simpler building blocks
could give us some clue as to how DNA might have emerged. In brief,
many of the constructs devised by the Eschenmoser team seem to be
distinctly sub-optimal, if not catastrophically inept.41 For example, if
the sugar hexose is substituted for the normal ribose the nucleic acid
strands are effectively inoperable.42 On the other hand, employment
of a chemical derivative of a simpler sugar (a threose), with four car-
bon atoms, as against the five of ribose43 (and six of hexose), does pro-
duce a viable analogue of DNA.44 The synthesis of a key compound,
α-threofuranosyl, however, involves a complex laboratory procedure
that is far removed from any credible prebiotic milieu, a problem that
is returned to in Chapter 4.
Other types of homo-DNA show quite robust features that seem
to match and sometimes exceed, at least in certain respects, the de-
manding design specifications set by the natural product. It is still
too early to say whether genuine alternatives to DNA, either in the
nucleotide–sugar field or on some wilder shore of biochemistry, really
exist. Certainly some of the molecules used in the natural world are
strangely suitable. Also we need to remember, as Eschenmoser notes,
that the aim is optimization not maximization. Thus he writes how
the ‘moderate base-pairing strength . . . in RNA and resulting from
the high conformational flexibility of the ribofuranose backbone, was
essential for the evolution of a rich diversity of nucleic-acid-related
biological functions.’45 Even if any of these modified DNAs and RNAs
are found to show the full versatility of the original, it has yet to be
established whether such an alien replication system could emerge
from the primeval broth on some distant planet. But before we taste
any soup it is to deep space we first need to turn in our pursuit of life’s
origins, to wallow in goo.
3 Universal goo: life as a cosmic
principle?
For centuries, if not millennia, we have looked towards the night sky
and with the glory of the Milky Way, now enhanced beyond earlier
imagination by the extraordinary pictures captured by such great tele-
scopes as the Hubble and Keck, felt a deep resonance, a wider and
wilder reality, a glimpse of a cosmic architecture. Naturally enough
such an awestruck view has also enfolded the prospect of other habit-
able worlds. Yet it is only quite recently that a deeper, more pervasive
view has emerged that looks beyond the question of extraterrestrials
to invoke organic activity on a truly universal scale, life as a cosmic
imperative. In the previous chapter I mentioned Christian de Duve,1
and perhaps he best of all encapsulates this wider and more optimistic
view that the Universe is not a howling wilderness, but part of our
home. Thus he writes, the ‘Universe is a hotbed of organic synthe-
ses leading among others, to amino acids and other typical building
blocks of life. This ‘vital dust’ permeates the entire Universe and
most likely represents the chemical seeds from which life arose.’2
Nor is de Duve the only exponent of life as a cosmic principle. Cyril
Ponnamperuma, for example, has stated that ‘You look at the inter-
stellar molecules and you see cyanide and formaldehyde. These two
can provide the pathway for everything else. There is a simplicity
in the whole scheme – so much so that you practically feel that
the whole universe is trying to make life.’3 Such a cosmic dimen-
sion has a strong resonance with our sense of belonging somehow to
a wider order, but it still could be a serious misreading of the evi-
dence. Here I shall take a contrary view and shall argue on the basis
of several different approaches that ‘Life may be a universal princi-
ple, but we can still be alone.’4 In this regard, major stumbling blocks
might be the apparent difficulty in understanding how life actually
originated (Chapter 4) and the possible dearth of habitable planets
(Chapter 5), but first in pursuit of de Duve’s vision let us look into
outer space.
a martini the size of the pacific 33
a martini the size of the pacific

Among the many surprises of radio astronomy has been the detection
in the huge interstellar molecular clouds of various spectra that are
evidently compatible with the presence of a significant number of or-
ganic molecules.5 Outer space being what it is, the quantities available
are truly gigantic. The chemist Robert Shapiro6 remarks that there is
enough alcohol to make any number of dry martinis, each the size of
the Pacific Ocean. But what is the real significance of this interstellar
material? The first problem, of course, is that many organic com-
pounds may escape detection either because their spectra are masked
by the stronger vibrations of other molecules or because they occur
in very low concentrations and are difficult to recognize. It would not
be at all surprising, for example, if amino acids were to be detected
(and the presence of amino acids in carbonaceous meteorites is clear
enough evidence for their extraterrestrial formation), but so far not
even the simplest amino acid, known as glycine, has been definitely
recorded.7
However, not all the organic molecules in deep space are sim-
ple. Another surprise in this adventure has been the detection of
rather more complex forms. This includes the so-called PAHs, that
is, the polycyclic aromatic hydrocarbons.8 Being organic compounds
they have a carbon ‘backbone’, which typically shows a texture
like chicken wire. PAHs are probably the most abundant organic
molecules in the Universe, entirely dwarfing the total biomass of
the Earth, which weighs in at a mere 6 × 1027 grams. This molecu-
lar group has quite a wide diversity of configurations, a number of
which have been detected in deep space. One of these is a relatively
familiar compound, napthalene. Indeed, one might derive wry amuse-
ment from the fact that the abundance of napthalene means that the
entire Universe smells faintly of mothballs. As with the other inter-
stellar organic compounds there is, of course, no question of these
PAHs being made by some sort of organism, although on Earth they
may be a chemical end-product of once-living organisms. For exam-
ple, the napthalene in those old-fashioned mothballs is a derivative of
the chemical product coal tar, which in turn ultimately comes from
the plants and algae of the distant Carboniferous swamps.
There is also a famous case where the source of the PAHs is ex-
citingly ambiguous. This is connected to the controversy surrounding
the famous Allan Hills meteorite from Antarctica (ALH 84001). This
34 universal goo: life as a cosmic principle?
has been claimed to yield evidence consistent with Martian life, and
one of the lines of evidence is the recovery of undoubted PAHs.9 The
other lines of evidence include (a) tiny crystals of magnetite,10 said to
be similar to those made by certain terrestrial bacteria that act as tiny
intracellular magnets; (b) putative fossils; and (c) the geochemistry,11
especially of the carbonate,12 and the related question of the tempera-
ture at which it was formed. Unfortunately the enthusiasm of the orig-
inal team investigating this meteorite has not been matched by most
other scientists. Whatever the significance of these Martian PAHs,
which are also well known in carbonaceous meteorites,13 the general
consensus is that, as for the other lines of evidence, non-biological
processes are much more likely to be responsible. So, too, on a much
grander scale, the interstellar equivalents may tell us little of a Uni-
verse pregnant with life, but may rather point to the existence of a
universal ‘goo’. These PAHs are incredibly stable, going nowhere, an
end-product with no future. Is this too pessimistic a view? Not accord-
ing to received opinion, which not only embraces life as a cosmic prin-
ciple but also seizes upon the only strictly tangible evidence we have
for extraterrestrial organic matter, which is in the form of meteorites.
goo from the sky

So far as human activities are concerned, meteorites tend to be re-
garded principally as potential agents of destruction: nobody fancies
even a few kilograms of iron falling out of the sky on to an aeroplane
or garden party. Even so, it is sometimes claimed that there is no defi-
nite evidence of death by impact. There is, of course, the ill-fated dog
that is said to have been killed in 1911 by the Nakhla meteorite,14
which like the Allan Hills meteorite was subsequently shown have
come from Mars. Human casualties are usually thought, with some
reservation, to be limited to a handful of possible examples, including
a Franciscan monk in seventeenth-century Milan.
Yet, as Kevin Yau and his colleagues15 have shown, a study of
Chinese records, representing an exemplary bureaucracy, paints a far
more sombre picture of sudden destruction as tons of material rain
from the sky. One Ming document, for example, refers to events in
1490 when ‘Stones fell like rain in the Ch’ing-yang district of Shansi
province . . . They struck dead several tens of thousands of people.’16
Yau et al. question the size of the casualty list, and cannot rule out
entirely the alternative agency of hailstones, but there is certainly no
goo from the sky 35
reason to doubt that multiple deaths and injuries have been caused
by meteorites, although in many instances the casualties probably
resulted from the collapse of the building rather than from a direct hit.
Moreover, a more recent survey17 provides a list of buildings hit (one
through the observatory of an amateur astronomer), near misses, and
several deaths. There is little doubt that further mining of historical
records (see also note 15) will reveal more examples of meteoritic
casualties. So, too, one may expect, will archaeological evidence. Of
course, it is tempting to ascribe the collapse of great civilizations to
such cosmic interruptions, but to date there is little, if any, evidence
to support such ideas. There is, however, evidence for a major impact
(the Kaali event) on the Baltic island of Saaremaa (Estonia) during the
local Bronze age18 (c. 400–800 bc; a possible report by the famous
traveller Pytheas, who originated from the Greek colony of Marseille,
might narrow this event to c. 350 bc). The energy released was about
equivalent to the atomic bomb that destroyed Hiroshima, and as is
evident both from the crater and the pollen record the impact must
have led to widespread devastation.
Yet, from the prospect of the origins of life, certain meteorites,
and especially a rather peculiar class known as the carbonaceous chon-
drites, are now being taken as a key creative agency. As such, these
meteorites provide a terrestrial glimpse of the vital principle that, it is
suggested, permeates the entire Universe. Carbonaceous meteorites
are quite rare, but the most famous examples (such as the Allende,
Murchison, and Orgueil) have proved treasure troves of information
on extraterrestrial organic chemistry. There is, however, a potentially
major problem because not only do they quickly disintegrate, but there
is the ever-present danger of contamination from terrestrial sources.
Not surprisingly, new discoveries, for example, the Tagish Lake fall,19
excite considerable attention. The Tagish Lake find was especially
propitious because the meteorite came down at a low angle and its
fall was observed; and its rapid recovery from an ice-covered lake in
northern British Columbia ensured a remarkably pristine sample. It
has yielded many surprises. The fact that its chemistry is very dif-
ferent from that of other carbonaceous meteorites implies that our
overall understanding is far from complete.20 Most striking perhaps
is the low concentrations of amino acids, some of which are clearly
terrestrial contaminants.21 Even so, many of the organic molecules
discovered in this and other meteorites are indigenous and represent
the material necessary for life to emerge – in other words those build-
ing blocks that are basic to life on Earth and presumably anywhere
else. So, in contrast to the blocks of iron and stone wreaking destruc-
tion, here we have a potentially rich brew of chemicals that, at the
dawn of Earth history, might have helped to initiate the whole process
of organic evolution.
As we shall see, the importance of the carbonaceous meteorites,
as well as comets, in enabling the Earth to become habitable should
not be underestimated. A closer look, however, suggests that to regard
these meteorites as some sort of primitive crucible out of which life
could take its first steps may not merely be simplistic: it could be seri-
ously misleading. Consider, for example, the amino acids22 recovered
from various carbonaceous meteorites, such as the famous Murchison
fall that landed near the town of the same name, in Australia in 1969.
Certainly these compounds occur in some abundance, but it is very
significant that only the simplest types of amino acids are found and
that they represent two groups (the monoamino monocarboxylic and
monoamino dicarboxyl classes). There is, moreover, another equally
important observation that, of the various molecular configurations
that can exist in these two groups, effectively all are found in the
meteorites.23 The message is simple. These amino acids are by def-
inition organic, but the processes of formation are patently abiotic
and unrelated to the ways in which amino acids are made in a biol-
ogical cell. Exactly the same remarks apply to the more recent
discovery in carbonaceous meteorites of compounds related to the
sugars.24
The restriction to the simpler types of molecule and the explo-
ration of all possible variants is non-selective and in marked contrast
to the activities of life. These extraterrestrial amino acids are the prod-
uct of an unconstrained chemistry, with no hint of life’s strange speci-
ficities and subtle avenues of synthesis. Within the restricted range
of the simple amino acids, all available types have been produced in
these carbonaceous meteorites by the random combination of yet sim-
pler precursors, but then the chemical process comes to a full stop.
In such an environment it is scarcely surprising that we see an expo-
nential decline in the abundance of amino acids with a larger number
of carbon atoms in their framework. Again the message is abiologi-
cal: even slightly more complex molecules are much more difficult to
assemble.
goo from the sky 37
These extraterrestrial amino acids show other clear signatures

of an origin free from life’s influence or interference. For example,
a number of the amino acids, such as α-amino isobutyric acid and
isovaline, recovered from carbonaceous meteorites have effectively
no terrestrial counterparts.25 In addition, the isotopic signatures of
such elements as nitrogen, carbon, and hydrogen are also greatly at
variance with those found in living systems.26
The problems with latching on to the carbonaceous meteorites
as some sort of cosmic guide as to how the foundations, if not the first
floor, of life are put in place do not stop with the amino acids. To-
gether with these compounds there is an enormous range of other or-
ganic compounds, but most of them have no terrestrial counterparts.27
These include very complex mixtures of hydrocarbons, including both
the aliphatic and aromatic varieties (effectively defined as to whether
the carbons are in a chain or ring respectively). Given this welter of
forms it is hardly surprising that special enthusiasm is reserved for
molecules, however rare, that do play a key role in life. Take, for
example, the group known as the nitrogen heterocycles. They are cer-
tainly important because they include the compounds known as the
purines and pyrimidines, which are the main building blocks of DNA
(and also RNA). But once again any optimism that floating in space
is a sort of proto-genetic code needs, so to speak, to be brought firmly
down to Earth.28
It is, of course, understandable that the discovery of the key
components of proteins and DNA in carbonaceous meteorites rein-
forces the notion of life permeating the Universe. Yet this is at best
a selective reading. As already noted, most of the compounds have
no parallels in living systems, at least on Earth. More seriously, most
of the carbon occurs as a sort of featureless ‘gunk’: highly stable, in-
soluble, this complex entanglement of organic compounds is going
nowhere – it is another variant of the universal ‘goo’. The occurrence
of this extraterrestrial muck closely parallels, as it happens, many of
the origin of life experiments in terrestrial laboratories (Chapter 4)
where a plethora of unwanted by-products and tar-like goo give at
least the sceptic pause for thought.
But carbonaceous meteorites continue to exert their fascina-
tion. Surely they are telling us something? Many years ago there was
a flurry of excitement when seeming fossils were found, but their
biogenicity was never confirmed, except that is from some structures
that were clearly terrestrial contaminants. But are there other ways in
which more complex structures might emerge: perhaps some of these
extraterrestrial organic compounds exhibit self-organizing properties
that could give us at least a glimpse as to how life emerged from a
distinctly discouraging mess? Nearly all are agreed that at some point
a key step in making life had to be the invention of some sort of en-
velope, a proto-cell, in which the nascent biochemical machinery of
replication and molecular synthesis could be shielded from the chem-
ical chaos of the surrounding milieu. And there are some intriguing
hints as to how this might have happened. Among the hydrocarbons
found in the carbonaceous meteorites are some compounds quite like
the lipids. These are the fatty compounds that typically make up part
of the wall of a living cell. Experimental manipulation of such ex-
traterrestrial compounds typically leads to the formation of simple
droplets, as you might see if you splash olive oil into water. Under cer-
tain conditions, however, genuine vesicles, apparently with a double-
walled membrane (as in living cells), can be formed.29 These artificial
vesicles are very small, typically in the micron range and so compa-
rable in size to bacteria. The main snag, however, is that although
hydrocarbons of a suitable size (that is with chains ranging in length
from 12 to 20+ carbon atoms) occur in the carbonaceous meteorites,
as yet no plausible prebiotic milieu has been devised in the labora-
tory to synthesize hydrocarbons of a comparable length. The obvious
implication is that life might not have emerged unless these extrater-
restrial hydrocarbons had been delivered to the surface of the early
Earth.
back to deep space

Carbonaceous meteorites therefore represent the product of a com-
plex chemical factory that generated a cocktail of organic molecules,
of which only a handful might be relevant to the origin of life on
Earth. But there is a more profound problem: where exactly did these
compounds form? The general idea is to look at events early in the
history of the Solar System when a multitude of small planets formed
before being destroyed in larger collisions. Many of these fragments
were incorporated into the planets as we know them but others re-
mained free, and of these some have since been captured as meteorites
by the Earth. Early in the history of the Solar System, however, it is
likely that at least some of these micro-planets were warm and wet,
back to deep space 39
allowing a ferment of chemical activity to take place. Not only were

organic compounds formed in abundance, but perhaps life itself flick-
ered into existence before another planet-shattering event abruptly
terminated one more experiment. So we should not necessarily dis-
miss the possibility of one day finding extraterrestrial fossils in some
of these meteorites.
But before rekindling too much optimism we need to have an-
other look at these extraterrestrial organic compounds. Rather surpris-
ingly, it is now becoming increasingly clear that at least some of what
we regard as the building blocks of life were probably synthesized long
before the Solar System itself formed. They are products of an inter-
stellar environment far removed from anything we could envisage as
remotely habitable. Here the processes of organic synthesis take place
in conditions of extreme cold, a ‘hard’ vacuum that is bathed in radi-
ation. The evidence of this dramatic shift in location – from a warm
cradle in the early Solar System to interstellar wilderness – comes from
a surprising discovery concerning some of the extraterrestrial amino
acids. Like many organic molecules, most amino acids are asymmetri-
cal. This means that they can adopt one of two forms, each the mirror
image of the other. Our two hands, left and right, are a more familiar
example. The technical name for these mirror variants (also assigned
to a left and a right) is enantiomorphs, based on the Greek words for
opposite and shape. Despite their unfamiliar name, enantiomorphs
are far from being some arcane area of chemistry, with dome-headed,
white-coated scientists frowning at banks of equipment, and a silent,
breathless huddle of assistants waiting for the meter to sink below
0.82. In fact enantiomorphs are quite familiar, if not domestic. The
example that is probably most often quoted is the smell of oranges
and lemons, which is said to be the product of the same molecule, but
as respective mirror images. As it so happens, this is not quite cor-
rect. The molecule in question belongs to an interesting group known
as the terpenes, and is specifically known as limonene. In oranges it
is present as (+)-limonene but, as Nigel Veitch of Kew Gardens in-
forms me, its mirror counterpart (−)-limonene, is actually found in
pine needles and not lemons.
So what does all this have to do with extraterrestrial amino
acids, and where they were originally synthesized? In general the abi-
otic production of amino acids produces equal proportions of either
enantiomorph of a given molecule. This result applies as much to
the amino acid content of carbonaceous meteorites as it does to lab-

oratory experiments on Earth. This fifty–fifty arrangement, of equal
numbers of left- and right-handed molecules, is known as a racemic
mixture. One of the puzzles of life on Earth is that with very few ex-
ceptions the amino acids are consistently of one type, specifically the
left-handed variety.30 If by some sleight of hand we were to persuade a
cell to employ the identical amino acid, but in its right-handed enan-
tiomorph, the effects would be very deleterious: Alice passes through
the looking-glass at her peril. Once life embarked on a left-handed
history, in so far as the amino acids were concerned, matters were
presumably fixed. It is much less clear, however, whether the orig-
inal ‘decision’ was an accident of history or whether there is some
tendency to favour, even slightly, the left-handed enantiomorphs. At
first sight the latter possibility seems rather remote because, as al-
ready mentioned, racemic mixtures are the rule in abiotic syntheses.
The general implication is that when life ‘chose’ the left-hand option
it was some time after amino acids had begun to be synthesized. It
now seems possible that the ‘choice’ was not only made much earlier,
but before life itself evolved.
When the discovery of non-racemic ratios of extraterrestrial
amino acids in carbonaceous meteorites was first announced there
was understandably considerable scepticism.31 All prior expectations
of abiotic organic chemistry pointed towards racemic mixtures. In the
case, however, of at least two amino acids there is an excess of one
enantiomorph. Even more interestingly, it is the left-handed variety
that is in the majority.32 The crucial point to appreciate is that it
seems to be generally agreed that such a racemic imbalance cannot
have arisen by any known chemical process that could occur on some
proto-planet swirling around its newly formed sun. Rather, the shift
away from a racemic mixture must have been the product of inter-
stellar processes that occurred long before the Solar System formed.
It has to be admitted that exactly how these amino acids reached a
preponderance of the left-handed enantiomorph (values are c. 7–10%)
remains an unsolved question. One intriguing idea is that polarized
ultraviolet radiation from nearby stars is responsible.33 There is, there-
fore, a hint that interstellar processes might stamp a handedness on
all life, wherever it occurs. Many suspect that the DNA/protein ba-
sis of life is universal, and this too may apply to left-handed amino
acids.
back to deep space 41
Whether or not this is correct will depend in part on the suc-

cess in finding if there are other plausible mechanisms to break the
racemic ratio, a problem which continues to exercise a perennial fas-
cination. There has been some progress inasmuch as enantiomorphic
excesses have been produced by various mechanisms, which in the
laboratory include the operation of a magnetic field.34 The problem is
whether these experiments have much relevance to what may have
happened originally in deep space. This is not only because the com-
pound investigated (a chromium complex) is probably not relevant
to biological processes, but more importantly because it is not clear
whether natural magnetic fields could produce sufficient quantities
of one enantiomorph, given that the laboratory experiments produced
only a very small excess of either enantiomorph (depending on the
direction of the magnetic field). Other approaches to this problem
include specific chemical reactions,35 as well as sorting on crystal
surfaces.36 The latter approach, of course, excites the interest of those
who believe the selection of amino acid handedness might have oc-
curred in the surface of the early Earth. The use of a mineral surface,
in this case calcite, is important because, as we shall see in the next
chapter, there is considerable interest in the possibility that such sur-
faces play an important, if not a key, role in the origin of life. Even
so, experiments using calcite yield only a very small (of the order of
10−3 ) enantiomeric excess, in this case l- or d-aspartic acid.
The point of this discussion is not to pour cold water on the
search for an explanation of biological handedness; there is a strong
sense that the choice of left-handed amino acids is unlikely to be a
simple accident and a satisfactory explanation may well constrain
other problems in the origin of life. There are, however, two general
points relevant to both life in the Universe and the origin of life on
Earth (see Chapter 4). The first difficulty is the wildly unconstrained
theatre of operation: opinion varies between those who regard much
of the basis of life emerging as the Solar System formed, if not ear-
lier, while others regard the heavens as almost irrelevant as they peer
into one sort of muddy pool or another. The second difficulty is that
despite the attraction of using analogies, many of the experiments
designed to explain one or other step in the origin of life are either
of tenuous relevance to any believable prebiotic setting or involve an
experimental rig in which the hand of the researcher becomes for all
intents and purposes the hand of God.
a life-saving rain?
It seems that however and wherever extraterrestrial organic material
was, and still is being, formed it may be of limited relevance to un-
derstanding how the building blocks of life came to be assembled and
coordinated on the early Earth. Even so, there are still strong argu-
ments that without chunks of extraterrestrial material, rich not only
in organic compounds but also in water (as ice) and other volatiles,
in the form of both carbonaceous meteorites and comets, then life
on Earth might never have arisen. This in turn would have precluded
some billions of years of evolution during which at least one sentient
species could evolve to a position where it could begin to think about
the whole business.
So why should carbonaceous meteorites and comets play such
a key role? By general agreement the Solar System originated from a
huge cloud of interstellar dust that, perhaps nudged by the blast from
a nearby supernova, subsequently condensed into a rapidly spinning
disc. Out of this emerged the central star and a series of planets. In
our system only one body, our Sun, achieved sufficient mass to initi-
ate the thermonuclear reactions that provide the sunlight. The rapid
processes of accretion, and especially the gravitational maw of the
newly forming Sun, means that even as the planets continued to col-
lide and grow, the central star itself achieved a large enough size to
begin to shine. The prospect for those planets nearer to the Sun was
not good, at least for the future of life. This is because the effect of the
Sun’s radiation was to drive volatile compounds (necessary to form
an atmosphere and ocean) out to the cooler regions of the spinning
disc, to or beyond the so-called ‘snowline’ which approximates to the
present-day position of Jupiter. Stripped of their volatiles, planets like
the Earth would be reduced to balls of bone-dry rock, with the ele-
ments that are essential to life either driven far beyond the orbit of
Mars or locked up in crystal lattices within the interior of the planet.
There would certainly be no oceans, and presumably no chance for
life to emerge.
So what are we doing here? One possibility is that such a planet
as Earth did not accrete in its present position, but formed in an orbit
much further out where the volatiles were ready to provide it with the
mantle of atmosphere and ocean. Subsequently, under the influence
of the chaotic system of planetary movements, the Earth was flung
inwards, to occupy an orbit much closer to the Sun and – by strange,
a life-saving rain? 43
good fortune – comfortably within the so-called habitable zone (see

Chapter 5): not too close to the Sun to be reduced to a searing inferno
of a runaway greenhouse, as befell Venus, nor too remote to end up as
a sort of super-tundra, as was the fate of Mars.
An alternative, and perhaps more plausible, suggestion to the
Earth being catapulted across the Solar System is that it remained
close to the Sun, but was ‘rescued’ by a massive influx of volatiles
as carbonaceous meteorites, and more particularly comets,37 bom-
barded the early Earth. Without this cosmic infall which provided
the seas and an atmosphere it is difficult to imagine how life could
have arisen. Some workers take an almost ‘manna from heaven’ view
of this process, in which the Earth was drenched in a rain of fully
formed organic molecules: a mizzle of amino acids, hydrocarbons,
and other complex organic molecules such as the PAHs and deriva-
tives like the quinones38 wafting as a rain to the surface, accumulating
in warm seas and coating tidal flats. There is certainly some evidence
to suggest that some quite complex molecules might survive even
catastrophic impacts.39 The more general consensus, however, is that
comets brought in, not the building blocks of life, but the basic raw
materials, such as carbon, water, and nitrogen. Four billion years ago
the natural laboratory was here on Earth, and now owing to human
ingenuity we can replicate these processes. Never mind the universal
goo: in laboratories all over the world we are on the verge of seeing
how the spark of creation transmuted the inanimate to the animate.
Nothing could be further from the truth.
4 The origin of life: straining the
soup or our credulity?
This chapter is a story of abject scientific failure. It is also a story

of omission and simplification, but how else could one encapsulate
this enormous and unresolved area? Perhaps by way of introduction I
should mention two items. First, there are some excellent books de-
voted to the topic of life’s origins.1 There are also, of course, many
review papers in the literature,2 although almost invariably they re-
flect a particular school of thought so that while opposing schools re-
ceive raking fire the inadequacies of their own stance tend to receive
more muted criticism. Second, here I make effectively no mention
of the so-called ‘RNA world’.3 This ‘world’ is widely, but not uni-
versally, accepted as predating our familiar DNA-based world, and is
taken to represent a key staging post that attempts to solve a sort of
‘chicken-and-egg’ question as to what came first: a system of repli-
cation to code and transmit instructions or a chemically controlled
metabolism to channel energy using catalysts (a.k.a. enzymes). Pro-
ponents of the ‘RNA world’ solve this conundrum by answering this
question with the reply: ‘simultaneously’. This is on the basis of ear-
lier speculation, combined with subsequent experimental proof of two
facts: first, that RNA acts as an agent of replication (in nearly all or-
ganisms this is on the basis of instructions coded in the DNA and
then transferred to the RNA); second, that in some circumstances
RNA acts in an enzymatic capacity (ribozymes), a role that in practi-
cally all living cells depends on particular proteins (enzymes). So why
is there no discussion here of this putative ‘RNA world’? Simply, be-
cause by the time you have reached this stage in the early evolution of
life you are probably home and dry. Getting there, however, remains
much more problematic. The remarks concerning the ‘RNA world’,
made by David Bartel and Peter Unrau4 are particularly apposite when
they write, ‘its status as a hypothesis is easily forgotten. Problems re-
main, particularly the implausibility of prebiotic RNA synthesis and
stability’.5
And that gently questioning tone is the gist of this chapter, draw-
ing as it does on a sceptical tradition articulated most cogently by
the origin of life: straining the soup or our credulity? 45
Robert Shapiro (see note 16). Another individual for whom I have a
particular admiration in this context is Leslie Orgel, not only because
he is a key individual in the study of the origin of life, but as impor-
tantly because, in contrast with the sunny optimism displayed in so
much of the literature, his papers reveal a traveller who knows he
might be on the high road to success, but will not be totally surprised
if the path leads to a fetid bog. This stance is neatly encapsulated in
one his review articles,6 where he writes:
The problem of the origin of life on the earth has much in

common with a well-constructed detective story. There is no
shortage of clues pointing to the way in which the crime, the
contamination of the pristine environment of the early earth, was
committed. On the contrary, there are far too many clues and far
too many suspects. It would be hard to find two investigators who
agree on even the broad outline of events.7
Yet, the study of life’s origins is set in a paradox because the un-
derlying principles could not be more straightforward. This is because
the first chemical steps that led to the processes of self-replication (ul-
timately held in the thrall of DNA) and controlled cycles of chemistry
(what we call metabolism) must have been simple. Knowing these
foundations it surely cannot be that difficult to add storey to storey,
ultimately to remove the scaffolding, and so reveal the functioning
cell, a near-miracle of encapsulated design but arising by unremark-
able processes?
To be sure we must acknowledge preconditions, including the
fact that if the precursors of life had not rained from the heavens on
to the surface of the early Earth, supplying it again and again with
volatiles, then quite possibly there would have been no oceans or at-
mosphere. Without this cosmic drizzle it is quite possible that life
would never have had either the building blocks or the milieu neces-
sary for its origination. On the other hand, such an influx of water and
organic compounds is hardly relevant to the actual question, how did
life begin? This is because the simplest of the molecules required for
the beginning of organic chemistry, compounds such as formaldehyde
and hydrogen cyanide, are themselves readily synthesized in at least
some circumstances that might have occurred on the early Earth.8
In other words, it doesn’t really matter where such compounds
form: what does matter is what happens next. So, too, with the next
46 the origin of life: straining the soup or our credulity?
stage, that is, the assembly of such basic building blocks of life as
the amino acids, sugars, and hydrocarbons, as well as the nucleotides
that are fundamental to DNA. Such syntheses are widely regarded
as inevitable, and images of warm ponds, seething volcanic springs,
and massive thunderstorms rumbling across a deserted yet pregnant
landscape are used to feed the imagination. At this stage, life has
yet to exercise its peculiarly specific grip, its spinning of the ge-
netic code, its weaving of biochemical complexities; but no matter:
despite the vast pot-pourri of resultant chemicals, the nascent pro-
cesses of Darwinian selection are already winnowing and reaping,
the inappropriate is steadily devoured in chemical competition with
the winners. Metaphorically the molecules slug it out by tooth and
claw. Cycles develop, life emerges, and four billion years later one
species invokes the marvels of autocatalysis and emergent proper-
ties to cap the argument. These ideas are the bread and butter, so
to speak, of a substantial part of the origin-of-life industry. They
are well encapsulated in the proposals of Manfred Eigen.9 An im-
portant part of his argument is the emergence of primitive cycles
that combine, as so-called hypercycles, to permit replication. A cri-
tique of Eigen’s ideas is offered by John Maynard-Smith,10 who sug-
gests that the mechanism of hypercycles becomes plausible only if
it is housed within a cell, and the origin of the cell . . . ? Even so,
the drums and trumpets that regularly sound in support of self-
organization and autocatalysis as appropriate mechanisms that could
lead to stable and self-replicating systems form a regular musical ac-
companiment to those who believe that these precepts are a highly
desirable, if not necessary, goal,11 even though there are painfully few
experiments.12
And yet, something is clearly missing: life cannot be created in
the laboratory, nor is there any clear prospect of it happening. When
we return to the simpler stages of organic synthesis, the apparently
crisp outlines begin to blur. Even the arrangement of the experimental
apparatus, the literal disposition of the test tubes, can spell the differ-
ence between success – of a sort – and yet another failure. And yes,
certain compounds can be generated, but as often as not in minute
quantities, mixed up with all sorts of other goo and under conditions
which, to be frank, bear precious little resemblance to anything con-
ceivably like the early Earth.
finding its path 47
finding its path

It is certainly not my intention to suggest that the origin of life is a
scientifically intractable problem, but at this stage of the proceedings
simply to register mild surprise at the relative lack of experimental
success. There has, of course, been a succession of ingenious experi-
ments that aim to throw light on one or other facet of prebiotic chem-
istry. The real problem is getting past this first stage, from the early
organic ‘soup’ (however envisaged) to the metabolic and biochemical
highway, with a functioning cell as its destination. Self-evidently this
end product – the cell – is an immensely complex chemical factory.
The path leading to the living cell must have required simpler prede-
cessors, and most probably can be envisaged as analogous to building
with something like a modular construction kit. Parts that were for-
merly independent could then be ‘bolted on’ at various stages. The
difficulty remains: exactly which path, which kit? These questions
seem to be virtually intractable, at least at the moment.
Now to me this difficulty appears to be rather strange. Clearly
the conditions on the early Earth are difficult to judge with any accu-
racy. Just how much methane was in the atmosphere? How salty was
the ocean? Were there ice caps? And so on. In any of these cases, and
many others, the precise values or conditions are elusive. Yet the early
Earth must have had some sort of atmosphere, a sea bed with certain
clays, volcanic hot springs, and so forth. Strictly speaking the number
of variations is very large indeed, but many plausible combinations
can be assembled. Perhaps we need to get just the right balance: so
much carbon dioxide and no more; a particular mix of clay miner-
als. (‘Steady Trevor with that beaker of kaolinite!’); just a smidgeon
of a phosphatic mineral. Maybe, but this is almost getting us into the
position of arguing that life was effectively a fluke-like happening, a
one-in-a-trillion chance. Perhaps it was. Perhaps the emergence of life
was ‘almost a miracle’ (to anticipate a phrase; see note 89) and given
enough time, patience (and money), so too may we stumble on a sim-
ilar solution. Yet, the general consensus is that far from being a fluke
the emergence of life on any Earth-like planet is more or less a fore-
gone conclusion, an inevitability inherent in the self-organizational
properties of organic chemistry. So what is the evidence?
Certainly the area dealing with the origin of life has no short-
age of enthusiasts: zeal of purpose is combined with an unwavering
naturalism. Yet, as Klaus Dose13 reminds us, the research programme,

which was initiated by such giants of the field as the Soviet scien-
tist Alexander Oparin, the maverick genius J. B. S. Haldane, and sub-
sequently the chief rejuvenator of this field Stanley Miller, is still
fraught with problems. It should all be rather simple. Here are the
building blocks, like bricks: today we build something rather simple,
say a garden wall, and tomorrow we aim higher, perhaps a cathedral.
Despite the many disagreements about the most appropriate experi-
ments and about the environmental milieu, there is a deeply shared
belief in the power of the self-organization of prebiotic molecules.
In fact this belief in self-organization not only implies a sort of
‘biochemical predestination’ but, as Dose wryly notes, the research
programme has led to an ‘era of unconfessed vitalism’. Dose contin-
ues with some apt words of warning: ‘[It has become] abundantly clear
that the power of self-organization inherent in macromolecules syn-
thesised in cells is based on extremely subtle physical and chemical,
and particularly stereochemical, properties [which]’, he continues in
near-understatement, ‘have never been observed in this highly orga-
nized form in prebiotic molecules.’ He concludes bleakly, ‘It appears
that the field has now reached a stage of stalemate.’14
Others have reached similarly pessimistic conclusions. John
Horgan15 and more notably Robert Shapiro16 have discussed how dif-
ficulties crop up at every stage and each step. Not only that but, as
Horgan makes clear, the proponents of the various rival schemes of
investigation are happy to express scepticism about alternative ap-
proaches to the problem, yet for the most part retain quiet optimism
that somehow their chosen methodology – with a little bit of luck
and another injection of research money – will crack the problem.
For many years the cries have been repeatedly ‘Breakthrough! Nearly
there! A major step! Most significant!!’ Let us hope they are right,
even if the shouts of acclamation are beginning to sound somewhat
hollow. At least Stanley Miller is candid enough to tell Horgan that
‘The problem of the origin of life has turned out to be much more
difficult than I, and most other people, envisioned.’17
Yet, turning to the mainstream literature we see hardly any
inkling of these largely submerged doubts. The practitioners fully
realize that the problem cannot be solved by a single master-
demonstration, but will yield only to a patient, chip-by-chip assault on
the edifice. The research workers seem always to be frustrated in their
problems with experiments 49
quest for what they refer to as the ‘one-pot reaction’, in which chemi-
cals go in at one end and a replicating system emerges from the other.
A survey of this area, which occupies the remainder of this chapter,
reveals a picture that can only be described as distinctly discouraging.
Indeed, there is a quite general pattern of problems.
problems with experiments

One difficulty is that the great majority of experiments connected
to one or other aspect of the origin of life do not actually work. Not
surprisingly, such results are seldom reported. This is not any sort of
dishonesty; it is simply that in exploring prebiotic chemistry most
avenues are going to prove to be dead ends, tracks leading nowhere.
Well, that is quite typical in science, yet in the field of the origin of
life this problem indicates some more pervasive difficulties. An illu-
minating example is provided by Anthony Keefe and Stanley Miller,18
who consider the crucial role of the element phosphorus. As is well
known to biologists, phosphates play a key role in the biochemistry of
all organisms, perhaps most notably by the agency of the compound
ATP (adenosine triphosphate), which is usually referred to as the
‘currency’ for energy transfer. This is on the basis of its ability to
store substantial amounts of energy, specifically in the branch of the
molecule that contains the phosphoanhydride bonds. It seems almost
certain that before the ‘invention’ of ATP there must have been some
sort of prebiotic phosphate compound capable of a similar, if less
efficient, function. It had been long appreciated that the production
of such compounds was fraught with difficulties, at least if the exper-
iments were to be conducted in a believable prebiotic milieu. Keefe
and Miller therefore embarked on an extensive series of experiments
that took as their starting points a credible range of chemicals that
one might reasonably expect to exist on the early Earth. The net
result of this programme was effectively a catalogue of disasters, and
although there were local successes Keefe and Miller were forced to
conclude that none of the new processes they had investigated would
be ‘sufficiently robust to have been of importance in the prebiotic
ocean’.19
One of the problems Keefe and Miller encountered was that in
most cases the yields of the favoured compounds were very low. This
is one of the recurrent features of such experiments into the origin of
life: again and again a compound that is believed to hold a key role in
some significant path of biochemistry is indeed synthesized, but typ-

ically the reactions are highly inefficient and the quantities produced
are quite negligible. To be sure, one or more agencies of distillation or
concentration may be invoked, but unless life itself is asked to give
a helping hand – defeating, of course, the purpose of the exercise –
then the mechanisms employed are highly artificial, if not contrived.
There is also a flip side to the production of potentially important pre-
biotic molecules. Not only are the yields often disappointingly low,
even minuscule, but typically (and unsurprisingly) the experiments
produce a wide range of other chemicals that seemingly have no rel-
evance to the origin of life. In some instances a substantial quantity
of the organic material synthesized forms a tar-like ‘gunk’, reminis-
cent of the heterogeneous ‘goo’ found in the carbonaceous meteorites
(Chapter 3).
Many of the difficulties just referred to are exemplified by what
has come to be known as the ‘ribose problem’. Ribose is a sugar, one
of a large group of molecules related to the molecular family known
as aldopentoses. In its right-handed, or dextral, form it is one of the
key ingredients of life, most notably as a key component of the back-
bone of the genetic code, namely deoxyribonucleic acid (DNA) and
ribonucleic acid (RNA). Given this crucial role, it is hardly surprising
that the successful laboratory synthesis of ribose is a prime objective
for understanding the origin of life. The starting point is formaldehyde,
a simple molecule (H2 CO), which is readily synthesized abiotically.20
Starting with formaldehyde, the next step is to engineer a process that
will lead to a polymerization, that is, a stringing together of simple
molecules, whereby more complex sugars can be formed. In certain
circumstances one of these should be ribose.
So what is the ‘ribose problem’? In a paper reviewing this area
Robert Shapiro casts a withering eye on a number of earlier exper-
iments in which success was claimed.21 In discussing one proposed
pathway Shapiro bluntly remarks that the investigators are to be con-
gratulated for their ‘vivid use of imagination’.22 Among the various
procedures that can be adopted, particular attention has been paid to
what is called the formose reaction.23 This was first identified by a
chemist, A. M. Butlerov, and is effectively concerned with the way in
which complex compounds can be produced from the polymerization
of formaldehyde under certain conditions. The limits within which
this reaction operates are at least broadly consistent with what we
problems with experiments 51
Formose reaction
14 16 18 20 22 24 26 min
figure 4.1 The ‘ribose problem’ exemplified. The graph (a gas

chromatograph) shows the many products of the formose reaction, of
which ribose – an essential component of DNA – is marked by peak 8.
Such reactions, with a plethora of compounds, are the norm in prebiotic
reactions, raising the question as to how separation and sufficient
abundance of key compounds, such as ribose, was achieved. (Reprinted
from Journal of Chromatography, vol. 244, P. Decker, H. Schweer, and
R. Pohlmann, X. Identification of formose sugars, presumably prebiotic
metabolites, using capillary gas chromatography/gas chromatography–
mass spectrometry of n-butoxime trifluroacetates on O/V–225, pp. 281–
291, fig. 11.5. Copyright 1982, with permission from Elsevier Science.)
might expect on the early Earth, and indeed ribose is produced. But
there is a snag: the yields are very low and the formose reactions
produce a disturbingly large array of other compounds (Fig. 4.1). To
compound the difficulties the ribose is rather unstable24 and degra-
dation is particularly rapid if the solution becomes more alkaline.
It is, of course, possible to tinker with the reaction to engender a
greater degree of selectivity, but unfortunately this does not in itself
boost ribose production. So we seem to have been led into an impasse.
Nothing daunted, chemists have devised reaction pathways that can
produce reasonable quantities of ribose,25 but the sheer complexity of
the process and the careful manipulation of the many steps during the
reaction make one wonder about its applicability to the origin of life.
A key element in the hunt for chemical pathways that might be

relevant to prebiotic evolution is the search for natural catalysts that
could not only accelerate the reactions but ideally could also impose a
degree of specificity that might circumvent the problem of producing
a cocktail of molecules, if not more goo. Perhaps this is the way to
crack the ribose problem? G. Zubay26 tackled this question by looking
at a wide array of potential catalysts that might direct the formose
reaction in favour of particular sugars, including ribose. On his own
admission this work was in part spurred on by the disappointment
and frustration experienced in investigating whether widely available
compounds such as calcium or magnesium hydroxides could act as
catalysts. He found that they could not, but a search among other
potentially suitable catalytic agents was equally futile, with the single
exception of lead. This element certainly occurs in natural compounds
such as the sulphide mineral galena, but Zubay himself was cautious
as he added, ‘we have pondered the legitimacy of lead as a prebiotic
catalyst.’27
The most obvious way forward from such an apparent impasse
is to find a viable precursor that is chemically simpler and, ideally, eas-
ier to synthesize. This is an important and more general point. If such
a stage could be identified, it might even begin to show how nascent
life, or its molecular antecedents, began to exert its strange yet charac-
teristic specificity, its ability to select the one-in-a-million chemicals,
by metaphorically dipping into the organic ‘soup’ and selecting any
molecule that improves, be it ever so slightly, one or other biologi-
cal process. Perhaps we should stop worrying about ribose, and look
towards a more believable predecessor. Such might be the case with
a compound known as α-threofuranosyl. This can be derived from a
sugar (threose) with only four atoms (i.e. tetrose) and thus is simpler
than ribose and perhaps easier to synthesize. Of equal significance is
that α-threofuranosyl can be used in the building of a short molecular
strand that is directly analogous to DNA (and RNA). This molecule,
referred to as TNA (T for ‘threo’), can form a double strand (a duplex),
and, rather remarkably, can also pair with RNA and DNA. TNA could
thus, at least potentially, be capable of replication, and thereby act as
a genetic code.28 If so, this could obviously a very important step for-
ward, but it raises two general problems. It is all very well to by-pass
the difficulty of making ribose, but we need to ask whether the path-
ways of threose sugar synthesis (and those of the other component
on the flat 53
molecules of TNA) can themselves be reconstructed in a credible pre-

biotic milieu. The sequence of steps undertaken in this laboratory in-
vestigation is certainly very far removed from any ‘warm, little pond’.
The second problem is that while life must have originated in a series
of step-like processes, as one moves to ever simpler compounds, so
the real difficulty, that of bridging the gap between complex organic
chemistry and any sort of functioning system we choose to call living,
actually grows.
on the flat
Mineral surfaces have played a key role in many speculations concern-
ing the origin of life, and for rather simple reasons. One is the potential
for the binding and selective entrapment of key molecules, possibly
with the added bonus of orientating them in some sort of regular ar-
ray. Of particular importance would be a prebiotic process of stringing
together simpler building blocks, a process known as oligomerization.
What might be a demanding and problematic process in a free aque-
ous environment, with its continuous molecular jostling and recur-
rent danger of chemical interference by such processes as hydrolysis,
might be neatly short-circuited on the surface of some mineral such
as a clay or the phosphorus-rich apatite. Mineral surfaces potentially
have other advantages. These include an enormous surface area, espe-
cially if the mineral grains are small. Thus a cubic centimetre of clay
will contain particles whose total surface area, so Christopher Jeans
tells me, could, in principle, reach an astonishing 1800 m2 , roughly
equivalent to nine tennis courts. Moreover, if the mineral grains are
not too tightly packed then an interconnecting system of pores exists
so that fresh solutions can be swept in to supply a growing molecule.
And there is an added bonus if the mineral happens to have some sort
of catalytic property, capable of accelerating chemical reactions. It all
sounds very promising.
The experimental work is certainly interesting. Although most
efforts have been expended in persuading amino acids to link
together,29 the necessary step to form first a polypeptide and ul-
timately a protein, there have also been investigations into the
oligomerization of nucleotides,30 which would be analogous to build-
ing a strand of DNA. In these experiments typically only a very limited
variety of the potential building blocks is used, and quite often it is
the oligomerization of only one type of molecule that is attempted.
Moreover, as is so often the case in research on the origin of life, it

is difficult to avoid a strong artificiality in the experimental set-up.
Thus, on occasion, analogues of ‘real’ nucleotides are employed, while
the chosen substrate is of a chemical purity unlikely to be found in a
natural prebiotic environment.31 Given all this, it is perhaps not sur-
prising that the results of oligomerization experiments are variable.
Some substrates bind efficiently but others do not. The experimental
process typically entails successively feeding the solution across the
surface. In this way chains of up to 50 or so monomers can be linked
into an oligomer. There is, however, a snag inasmuch as the binding to
the mineral surface is typically an irreversible process: the oligomer
remains locked in position. It is tempting, of course, to see this as a
sort of staging post to life itself, with the mineral-bound strand then
reading off multiple copies that then jostle for selective advantage as
some sort of protolife. Perhaps so, but the investigators Rihe Liu and
Leslie Orgel note that concerning such oligomerizations ‘Many more
experiments will be needed.’32
It is also worth remarking that while the role of clays in the
origin of life has been seen largely in a positive light – be it as cat-
alytic surfaces or in terms of Graham Cairns-Smith’s ingenious ideas
of a clay-based information system preceding that of DNA33 – these
minerals may have unexpected drawbacks. Earlier in the history of the
Earth there were many radioactive isotopes that have now disappeared
because of their short half-lives. Of particular significance, perhaps,
is that the radioisotope potassium-40, which would have been much
more abundant at the time life was emerging. Given that potassium is
an integral component of many clays, then, as has been pointed out,34
if these minerals ‘acted as important nucleation sites, the protobio-
logical synthetic reactions had to proceed much more rapidly than
deleterious decomposition reactions to avoid the long-term effect of
radiolysis and hydrolysis within days.’35
But not all minerals are so radioactive, and the potential role of
their surfaces as the key to understanding the origins of life has an
almost mesmeric attraction that can be traced back to that uneasy ge-
nius J. D. Bernal.36 There has of late been considerable interest in the
mineral pyrite, popularly known as ‘fool’s gold’ a compound of iron
and sulphur (FeS2 ). This interest concerns both its chemistry and its
surface properties, as well as a possible setting in active hydrothermal
vents. For many people this is by far the most encouraging avenue in
on the flat 55
the pursuit of life’s origins, although it comes as little surprise that

those who have devoted their research energies in peering into other
crucibles adopt a duly critical stance.37 The chief exponent of the pos-
sible role of pyrite is Gunter Wächtershäuser, of whom the generally
(and correctly) sceptical Robert Shapiro38 wrote, ‘When I examined
his longest paper, I felt as if someone had handed me a biochemistry
text from the late twenty-first century. Most of the chemicals were
familiar, but they were organized into an evolutionary pattern I had
never seen before.’39
The gist of Wächtershäuser’s approach is to argue that when
pyrite is formed from the pre-existing ferrous sulphide (FeS) in combi-
nation with hydrogen sulphide energy is made available (by the release
of hydrogen). This energy could act to reduce carbon dioxide, so free-
ing the carbon for involvement in organic reactions. An important
corollary to this hypothesis is that the surface of the pyrite crystal
also plays a key role in controlling the ensuing chemical reactions
that are the first steps in a primitive metabolism.40 It is certainly
true that there have been some encouraging experimental results,41
although the experiments were performed under closely controlled
conditions, and other research workers have arrived at more discour-
aging conclusions.42 In addition, attempts to portray a longer series
of reactions in this hydrothermal setting look rather more precarious:
a number of key steps remain hypothetical.43 Still, these are early
days and there is a widespread sense that for all its uncertainties the
Wächtershäuser school is on to something important. One obvious at-
traction is the abundance of the main substrate, pyrite. As will be dis-
cussed below, the likelihood that the overall setting was in a volcanic
hydrothermal system is also attractive, especially as the heat budget
of the early Earth was substantially in excess of that of today. Another
reason for (cautious) enthusiasm is the possibility of identifying very
ancient biochemical traits and pathways: to watch a metabolism grow
out of a mineral surface. Even so, it is worth bringing to mind some
more general remarks of Andrew Ellington44 when he comments on
a disturbing trend in modern thinking about abiogenesis and the

evolution of metabolism . . . the results of these experiments [on
the origin of life] are often interpreted as being proof of a particular
scenario, rather than as general support for a more vague view of
our origins. While it would be extremely exciting to establish that
the RNA world evolved to a particular scenario, it is unfortunately

unlikely that sequence minutiae have remained unchanged and
unscathed following a several billion year journey through
multiple different types of organisms and biochemistries.45
The notion that pyrite (and other metal sulphides, especially

nickel) could play the key role in the initiation of life has attracted
attention for another potentially important reason. This is because
such sulphides are abundant in hydrothermal systems, most famously
in the form of the ‘black smokers’ found on oceanic spreading ridges.
As potential sites for the origin of life, these systems are attractive
for several reasons. These include their dynamic nature with strong
temperature gradients, the possibility of active mineral growth, and a
strong flux of both heated water and various chemicals. Such settings
are now the focus of investigations into various prebiotic syntheses
that might be alternative sources for some of the major building blocks
of life.46 Hydrothermal systems have another advantage in that, of all
the regions of the early Earth, they would have been the most immune
to the searingly powerful destructive forces released by a series of
violent impacts early in the history of our planet (see Chapter 5).
The possible association between very hot environments and
the earliest life has, in turn, led to a potentially fascinating possi-
bility that the most primitive types of bacterial life are those still
found today living in hot springs,47 such as those in both the mid-
ocean ridges adjacent to the ‘black smokers’ and other volcanic re-
gions such as the famous Yellowstone geyser and fumarole pools. This
does not necessarily mean that the very first life emerged in such tor-
rid circumstances, although proponents of pyrite-based metabolism
and hydrothermal systems will offer their own warm support. The
fact that nucleotides require temperatures nearer to freezing for long-
term stability (outside, that is, the functioning cell) has already been
mentioned (see note 37), and this is not the only difficulty with a
hot start to life.48 More interesting is the possibility that it was only
those bacteria capable of resisting extreme temperatures, imposed by
the series of colossal meteorite impacts (see Chapter 5), that could
survive these ‘thermal bottlenecks’ to repopulate the planet.49 At-
tractive and popular as this idea might be, there is also a vociferous
group50 who reject the notion that bacteria adapted to extreme heat,
the hyperthermophiles, represent the common ancestor of all extant
on the flat 57
life. This group argues that these adaptations to high temperatures

must be a secondary feature. Even so, despite this scepticism and
more general reservations,51 at the moment the idea that life’s ori-
gins are to be found in ancient environments nearer to the traditional
depictions of Hell – sulphur and boiling temperatures – is probably
the most influential idea in town.
Yet even a cursory knowledge of the investigations into the ori-
gin of life invites some caution. We have been here before: a new
locale, apparently realistic conditions, some intriguing results that
indeed confirm the possibility of replicating certain early stages – say
stringing amino acids into peptides – of the evolution of living sys-
tems. It is all very encouraging, but twin questions remain. First, can
we move from the carefully controlled environment of the laboratory
bench (‘all solutions were prepared from doubly distilled water’52 ), to
a more realistic setting? Second, research on the history of the origin
of life may repeat itself. Those first stages certainly require chemi-
cal skill and imagination, but the question remains as to whether the
further and crucial stages to reach a functioning metabolism are so
easily achievable. To put it bluntly, the very success of one system
in one context, say pyrite, makes it questionable whether a very dif-
ferent set of reactions will be catalysed with anything like the same
efficiency. Thus, while finding much to admire in Wächtershäuser’s
work, Leslie Orgel53 points out that there are difficulties in extrapo-
lation. He writes,
One must expect the results of mineral catalysis to be highly

idiosyncratic; most minerals will probably catalyze some reactions
and many reactions will no doubt be catalyzed by some minerals . . .
Nor would the situation be changed if the proposed participants
in a complex cyclic reaction scheme were synthesized in situ
on a mineral surface. If the products are mobile on the surface, the
situation is identical to that for adsorbed molecules. If they are not,
one must postulate a series of remarkable coincidences to conclude
that all of the reactions are catalyzed on the same mineral and
that each intermediate product is formed in the correct position
and orientation to become the substrate of the next regiospecific
reaction of the cycle. The self-organization of a complex cycle,
such as the reductive citric acid cycle, this time on the surface
of FeS/FeS2 , although logically possible, is very unlikely.’54
In conclusion, it would be most surprising if some, perhaps

many, minerals did not play important roles in catalysing in one way
(e.g. oligomerization) or another (e.g. a simple cycle) some of the first
steps to life, but getting to the next stage seems to be as elusive as ever.
back to the test tube

So such problems as trying to synthesize the sugar ribose (or even a
precursor) or building chains of molecules in a credible prebiotic en-
vironment are fraught with problems. What then of the many other
experiments that have been carried out in the attempt to decipher one
or other stage in the otherwise mysterious journey by which life may
have emerged? It is certainly the case that some do seem, at least at
first sight, to have a ring of reality. Such is often claimed for the famous
experiments by Stanley Miller and Harold Urey, in which electrical
discharges in a primitive atmosphere led to the formation of a strik-
ing panoply of organic molecules. Yet, as we shall see, the elegance of
this set of experiments may conceal less well publicized difficulties.
There is, moreover, a rather general problem, already alluded to in
the context of other experiments: are the conditions, such as those
chosen by Miller and Urey, even remotely relevant to the early Earth?
At first sight, this seems to be a harsh criticism. It is, of course, un-
derstandable that the investigators will take short cuts. In particular,
they will use reagents that are conveniently free of impurities, ones
that are made available by processes that in turn rely on complex in-
dustrial pathways, and are often used in concentrations that in any
prebiotic milieu seem improbable. The methods employed are care-
fully explained, not least to guide others who might wish to duplicate
or extend the procedure, but it is very seldom that more than a passing
remark is offered as to whether any environment on the early Earth
(or Mars; see Chapter 5) even approximates to what is going on in the
laboratory. There are rather few moments of candour, such as those
connected to a series of experiments investigating oligomerization of
nucleic acid-like compounds, where the investigators remark, ‘How-
ever, it is unlikely that chemically pure substrates of this kind, for
example 2-MeImpG . . . [the chemical 2-methylimidazolide guanosine
used as the template for oligomerization] could have accumulated on
the primitive earth.’55
This question of how far chemical analogies can plausibly be ex-
trapolated in studying the origin of life certainly makes it difficult to
back to the test tube 59
judge when experimental convenience is given too much precedence.

Certainly, one is entitled to wonder whether, for all its elegance, a
study of the self-assembly of nucleotides in the form of monolayers
via the catalytic agency of molybdenum sulphide on highly orientated
pyrolytic graphite56 has a direct relevance to the question of how DNA
was polymerized. Indeed the authors are candid when they write that
the ‘substrate surfaces were used because of their well-defined crys-
tallographic orientations and their convenience in electrochemical
and scanning probe microscopy studies.’57 They go on to remark that
‘The prebiotic relevance of self-assembled monolayers of purine and
pyrimidine bases is one of conjecture,’58 although nothing daunted
they extend their hypothesis by commenting that ‘It is attractive to
suggest an interaction between the purine and pyrimidine monolayers
and amino acids as the physicochemical hypothesis for origin of the
genetic code.’59
It is indeed difficult to distinguish what will develop into an at-
tractive hypothesis rather than mere wish-fulfilment. There are many
other examples where the question of using a convenient substrate or
procedure seems difficult to reconcile with any prebiotic plausibil-
ity. To the example of molybdenum sulphide one could add, for ex-
ample, the use of minute droplet-like structures, known as micelles,
composed of such chemicals as sodium caprylate,60 cetyltrimoethyl
ammonium bromide (CTAB),61 or lithium chloride62 as possible mod-
els for the process of self-replication or oligomerization. Concerning
the last of these substrates, Leslie Orgel shows characteristic caution
when he writes, ‘Clearly this particular replication system is unlikely
to have operated on the primitive Earth. But the model is very sug-
gestive, and may encourage work along the same lines using more
plausible prebiotic substrates.’63 Orgel goes on to remark that the
process of self-copying may be a rather general phenomenon, but he
acknowledges that designing a primitive system of replication is diffi-
cult. Thus he continues, ‘The subunits of the self-replicating polymer
must be potentially prebiotic molecules, that is, molecules that could
have accumulated on the Earth without the intervention of organic
chemists. Different authors will draw the line between plausibly and
implausibly prebiotic molecules in different ways . . . However gener-
ously “prebiotic” is interpreted, it is obvious that it will be harder to
realize a replicating system if we are restricted to prebiotic molecules
and cannot use molecules specially designed for replication.’64 With
this distant whisper of teleology, the perennial problem of the chicken

and egg returns. In evolution we know the egg came first, but in the
origin of life?
Those familiar with investigations into the origin of life may
be surprised that no detailed mention has been made yet of the now-
classic experiments by Stanley Miller and Harold Urey,65 which are
still routinely cited (especially in that zoo of good intentions, the un-
dergraduate textbook) and seem to be dramatic evidence of the ease
with which simple organic compounds can be synthesized under con-
ditions closely similar to those found on the early Earth. All the ap-
propriate ingredients were there: from the spotless glassware handled
by the dedicated and intelligent student of a Nobel Laureate, to the
seething organic broth emerging beneath the flicker of artificial light-
ning. Yet it is now clear that the elegance of these and subsequent
experiments66 may have little, if anything, to do with what may have
happened on the early Earth. The difficulties fall into several cate-
gories (notes 16 and 66), and I want to re-emphasize how pervasive
the difficulties really are.
First is the general assumption that these experiments form
some sort of continuum with extraterrestrial organic chemistry, with
the implicit assumption (at least in some quarters) that the emergence
of life is a cosmic process (Chapter 3). Thus, it is said that the pro-
file of the amino acids produced in these experiments closely mirrors
those found in carbonaceous meteorites. But this is a rather selective
reading of the evidence. First, even though some amino acids, includ-
ing not surprisingly simple ones such as glycine, formed in the exper-
iments also occur in meteorites, the similarities are not that precise.
More important, perhaps, is the evidence (reviewed in Chapter 3) that
at least some of the extraterrestrial amino acids formed under condi-
tions of interstellar chemistry are very different from those of Miller’s
experiments. Another well-known feature of these experiments is
that the enantiomorphs (see pp. 39–40, Chapter 3) of the amino
acids occur in equal proportions. Indeed, this racemic equivalence was
used by Miller to refute the possibility of biological contamination of
the system. This is not to contest the fact that how, when, and where
amino-acid chirality was broken in favour of the left-handed enan-
tiomorphs used by all life are, as we saw in the last chapter, all very
conjectural questions. Even so the most optimistic interpretation of
Miller’s results is that if his experiments really do have a bearing on
back to the test tube 61
the early synthesis of organic molecules, including amino acids, then

somehow the left-handed choice was arrived at a later stage of prebi-
otic history.
Having reviewed a few of the experiments on the origin of life
earlier in this chapter, it comes as little surprise that notwithstand-
ing their classic status the success of individual experiments of the
Miller–Urey type was highly variable. Even the best results, or at
least those giving the highest yields of amino acids, were remark-
ably inefficient.67 These revealed a whole series of complexities, such
as variably declining carbon yields after no more than a few days,
meaning that deciding what is the most propitious ‘atmosphere’ is
fraught with problems. Perhaps more surprisingly the experiments
also depended critically on the physical arrangement of the labora-
tory apparatus. Thus, the simple interchange of a condenser (a piece
of equipment for condensing gases to the liquid state) and the en-
ergy source (a sparker) resulted in hardly any amino acids being pro-
duced, although quite substantial quantities of hydrocarbons were
formed. Moreover, the choice of a spark discharge was determined
not so much to represent the effects of lightning flickering across a
lifeless landscape but because, as Miller and his colleagues note,
‘The ease of handling and high efficiency . . . are factors favoring
its use.’68 Thus sparking is very efficient at synthesizing hydrogen
cyanide and cyanoacetylene, two simple but key chemicals in the
pathway to the subsequent synthesis of the amino acids. Yet the use
of what were effectively spark plugs may be an imperfect analogy to
lightning because experiments in which the electric charge was al-
lowed to build up and suddenly flash curiously produced very few
organic compounds. No doubt, then, thunderstorms rumbled across
the lifeless seas of the ancient Earth, but as even a potential energy
source to drive the prebiotic reactions the importance of lightning
would have been far outweighed by the ultraviolet (UV) radiation from
the Sun. Unfortunately, yields of amino acids seem to be much lower
when UV radiation is used, and to complicate matters it has also been
pointed out that if this were the source of energy then many of the
chemical reactions would have taken place high in the atmosphere,
with the risk of decomposition as these compounds floated down to
the surface of the Earth.
The key criticism, however, that has been levelled at the Miller–
Urey experiments concerns whether the gases put into the flasks were
in any way equivalent to the atmosphere of the early Earth. At the time
of these investigations it was widely assumed that this early atmo-
sphere was devoid of oxygen. This was a reasonable enough assump-
tion given that life, specifically photosynthesis by algae and plants, is
the main source of this toxic and corrosive gas.69 It was also assumed
that the early atmosphere was quite rich in two key gases: ammonia
(NH3 ) and methane (CH4 ). Only a small amount of ammonia is re-
quired (at 25 ◦ C about a millionth of an atmosphere), but as Miller
himself remarked, ‘The details of the ammonia balance on the prim-
itive earth remain to be worked out.’70 In the experiments the levels
of methane were much higher, typically up to 0.2 of an atmosphere.
Again Miller noted, ‘This pressure is used for convenience, and it is
likely, but never demonstrated, that organic synthesis would work
at much lower partial pressure of methane.’71 Nobody disputes that
establishing the nature of the early atmosphere is problematic, but
the general consensus now is that so far as the Earth was concerned
methane, and even more so ammonia, were present only in small,
perhaps negligible, quantities. Rather than being strongly reducing,
the air was probably only slightly so, being composed of a mixture of
carbon dioxide, carbon monoxide, water vapour and nitrogen. In these
conditions attempts to produce not only amino acids but also com-
pounds such as the key precursor hydrogen cyanide (see note 8) are
effectively failures. It is hardly surprising, therefore, that other ways of
generating amino acids abiotically have been sought. As we saw in the
previous chapter, one hypothetical route is simply by cometary impact
(‘extraterrestrial infall’) and suchlike. A popular alternative, already
mentioned, is to consider hydrothermal systems, such as mid-oceanic
ridges, where amino acids can be generated at surprisingly high tem-
peratures, at least in experimental analogues,72 although with some
important provisos.73
Miller himself shows considerable candour in discussing the re-
search programme he has inspired. Despite the iconographic status of
his early experiments in sparking a methane-rich gas mixture, he and
his colleagues have no illusions that they are in any stronger position
to arrive at that long-sought-after ‘one-pot reaction’: chemicals in at
one end, cells out the other. The origin of some of the hydrocarbons
is a case in point. Here the usual invocation is to draw attention to a
process known as the Fischer–Tropsch reaction, where at quite high
temperatures (c. 200–400 ◦ C) high yields of hydrocarbons are obtained
a sceptic’s charter 63
when a mixture of carbon monoxide and hydrogen flows over a cata-

lyst such as iron and nickel. This is a well-known industrial process
(the Mond process), but in the natural world it is likely to occur only
in a volcanic environment. The reaction can, however, make straight
hydrocarbon chains up to only a length of 16 carbon atoms. The longer
chains, which are critical for building the hydrocarbons used to con-
struct a cell wall must be made by some additional process, the de-
tails of which are yet to be established. One can see, therefore, the
attraction of deriving at least some of the organic material needed for
life from an extraterrestrial source (Chapter 3). Although most of the
interest has been in the dumping of raw materials, the synthesis of
so-called amphiphilic compounds, with carbon chains and a propen-
sity to organize themselves into tiny vesicles in which prebiotic
chemical reactions could continue,74 has an obvious relevance to this
problem.
a sceptic’s charter
The description of this bringing together of the basic ingredients for
life has been reduced to near parody by Robert Shapiro,75 but, how-
ever amusing, he is making a serious scientific point. The nub of his
argument, which I paraphrase below, is that even ignoring the com-
plex chemical reactions with their plethora of irrelevant by-products,
the varying stabilities of the resultant chemical compounds, the fre-
quently pathetically low yields, not to mention the various laboratory
protocols that necessarily insist on convenience of chemical purity
and efficiency of human-directed reactions, the range of molecules
necessary for life to emerge can be synthesized only in widely, if not
wildly, different environments. To be sure, in certain cases impor-
tant molecules, such as the simpler amino acids, can be generated in
a series of quite different settings. In other examples, however, syn-
thesis depends on rather specific pathways, while the stability of a
particular molecule is often precarious and sometimes dependent on
a particular environment. The sugar ribose, as we have seen, is de-
composed in only a few weeks, but most notorious in this respect
are the nucleotides, such as adenine, which for long-term survival
ideally require conditions well below freezing.76 Inconveniently, the
reverse holds for other molecules, at least in some high-temperature
hydrothermal settings. So let me give the gist of Shapiro’s scenario
for what might be needed to bring the prebiotic ingredients together
to ignite the vital spark. He asks us to imagine a volcanic island,

large enough to generate a series of distinct climatic zones. In this
volcanic environment not only are there various minerals and muds,
but also hot thermal springs. The Big Island of Hawaii might be a
modern analogue. Thus we can envisage a rainy side, with abundant
thunderstorms, and near the summit sufficient snowfall to generate
glaciers. The atmosphere lacks oxygen and in the thunderstorms hy-
drogen cyanide is generated and so accumulates in the glacier ice.
Suppose, Shapiro continues, that one year the glaciers advance and in
doing so displace the water from partly frozen alkaline lakes, located
at the glacier snouts, into a more temperate zone. Here the hydrogen
cyanide (and its reaction products) happens to run into an acidic hot
spring. Behold, adenine is formed. As Shapiro points out, the quan-
tity is small and the adenine would not survive for very long, but
by good fortune the water flows over alkaline ground, so neutraliz-
ing the reaction. Elsewhere formaldehyde is formed, and from this a
small proportion of the sugar ribose is synthesized as the mixture is
swept into a series of hot pools. This time, however, the water is nei-
ther acidic nor alkaline (i.e. it has neutral pH). Happily the necessary
reaction time to make the ribose is just right before it is carried fur-
ther downstream in ice-cold water. This is helpful: the ribose would
otherwise have been destroyed.
Already even the simple compounds of adenine and ribose are
available by several happy chances. But good luck is our constant com-
panion. The separate streams containing these two chemicals now
merge, and, as Shapiro reminds us, the next product, adenosine (the
precursor of one of the nucleotides of DNA), needs another set of
special conditions to form, including heat and salt. No problem: the
water containing the adenine and ribose is carried in a waterfall to sea
level, where hot salty pools, lined with appropriate minerals, receive
the chemicals. One stroke of good fortune follows another. As Shapiro
continues, ‘It was a very hot day. The sun evaporated the remaining
water in the pool and heated the adenine and ribose in the presence of
salt, converting them in part to the nucleoside adenosine . . . the tides
returned to the tidal pool in a rush, sweeping out its contents and
transporting them farther inland to . . . (the) Darwin Pond . . . No sooner
had the adenosine reached Darwin Pond when successive waves, each
flowing from a different direction, brought in supplies of the other nu-
cleosides to make RNA.’ So Shapiro concludes, ‘Had these chemicals
only been human, they would have embraced at the joy of their first
meeting, and in anticipation of the glorious future that lay ahead of
them. They would then have taken turns, each describing the mar-
vellous and different series of events that led to its own creation.’77
Nor does the story end there. Before life itself can emerge in
Shapiro’s tale more fortunate accidents involving phosphate-rich min-
erals, convenient catalysts, and floods are required; but his point has
already been made well enough. It might have been made, but has it
been heard? To show that Shapiro’s thinking is by no means off limits,
consider these remarks by two of the most respected experts, Gerald
Joyce and Leslie Orgel,78 in a section headed ‘ The prebiotic chemist’s
nightmare:
Scientists interested in the origins of life seem to divide neatly

into two classes. The first, usually but not always molecular
biologists, believe that RNA must have been the first replicating
molecule and that chemists are exaggerating the difficulties of
nucleotide synthesis. They believe that a few more striking
chemical ‘surprises’ will establish that a reasonable
approximation to a racemic version of the molecular biologist’s
pool could have formed on the primitive earth, and that further
experiments with different activating groups and minerals will
solve the enantiomeric cross-inhibition problem. [A reference to
an observation that template oligomerization of nucleotides
proceeds efficiently if one enantiomorph, e.g. the left-handed
form, is present, but is severely inhibited if the other handed
molecule is present79 ]. The second group of scientists is much
more pessimistic. They believe that the de novo appearance of
oligonucleotides on the primitive earth would have been a near
miracle. (The authors subscribe to this latter view). Time will tell
which is correct.80
I should emphasize that Joyce and Orgel are equally careful to point
to encouraging lines of research, but near the end of the chapter they
remark, ‘There remains a lingering doubt that we are on the right track
at all.’81
How then can we escape these dilemmas? The sheer unlikeli-
hood of a ‘one-pot reaction’ emerging in any natural and primitive en-
vironment helps to explain the paradox of both frustration at repeated
failure and hope springing eternal that some environment – pyrite?
clouds?82 hydrothermal springs? – will bring the magic ingredients

together. Present ignorance need not be a reason for pessimism. The
general principles of evolution remind us that advantages need only be
slight; systems can act as templates and molecules may be co-opted
for new uses. And that, perhaps, is the central dilemma in biology.
This book is really about evolutionary navigation, be it through the
immensities of protein hyperspace (Chapter 1) or the arrival at intel-
ligence (Chapter 9). Everything we know about biology argues that it
is seeded with inevitabilities; in principle, there is an incomprehen-
sibly enormous universe of possibilities, but in reality the number
of destinations – metaphorically the Easter Island of Chapter 1 – is
a vestigially minuscule fraction of the theoretical possibilities that
can never be visited, not because the journey would never be made,
but because the journey would never be possible in the Universe we
happen to inhabit.
And this should apply with equal force to the steps involved in
the origin of life. As Manfred Eigen wrote many years ago, ‘The evo-
lution of life, if it is based on a derivable physical principle [in this
case selective value] must be considered an inevitable process . . . not
only inevitable “in principle” but also sufficiently probable in a re-
alistic span of time. It requires appropriate environmental conditions
(which are not fulfilled everywhere) and their maintenance. These
conditions have existed on earth and must still exist in many plan-
ets in the universe.’83 The importance of law-like processes in the
study of the origin of life is also emphasized by Iris Fry.84 In partic-
ular she offers an analysis of the extent to which such hypotheses
as those advanced by Manfred Eigen to explain the origins of repli-
cation can be shifted from a supposedly random basis to one that
presupposes ‘a deterministic element in the evolution of molecular
self-organization’,85 and thereby an imposed directionality. Echoing
this, Kenneth James and Andrew Ellington remark, ‘While Stephen
Gould is fond of saying that re-running the tape of life will always
give different outcomes, it is not clear whether this thought exper-
iment is as true for molecules as it is for multicellular organisms
[so far as the latter are concerned, the exact opposite is the principal
argument of this book]. The strictures of prebiotic chemistry may have
narrowly defined the first self-replicating organic molecules [so] . . . it
becomes possible to at least hypothetically chart some of the inter-
mediate forms that life may have taken.’86
Quite so, and one should be optimistic. Despite their difficul-

ties, as already seen the imaginative hypothesis of Wächtershäuser
may be the most encouraging way forward. Somehow, we must as-
sume that the problems of combining those molecules that prefer hot
springs as against glaciers, not to mention the myriad of often un-
wanted chemical by-products all ready to gum up a reaction, will be
surmounted. Perhaps it is impractical to do more than stumble round
the edges of a problem that may have taken millions of years to solve.
Yet, the suspicion, as articulated by James and Ellington, is that there
may be only a narrow way. Not only here, but everywhere. At present,
however, the problem is that the simplest prebiotic systems are pretty
elementary, and there is no obvious or straightforward way that seems
to lead easily to the specificity of chemical reactions that is so charac-
teristic of life. Experiments are contrived and frankly for the most part
unrealistic, and as often as not the end-result is failure or goo, or both.
There seems to be a feeling that there is some general and quite
simple principle that keeps on being overlooked. Leslie Orgel reminds
us87 that practically all the gem-quality rubies in the world come from
one locality near Mogok in Myanmar (formerly Burma), where pre-
sumably the chemical conditions in the host marbles (metamorphosed
limestones) were ‘just right’. In this context, it is surely interesting
that Francis Crick88 can write ‘An honest man, armed with all the
knowledge available to us now, could only state that in some sense,
the origin of life appears at the moment to be almost a miracle, so
many are the conditions which would have had to have been satisfied
to get it going.’89 Crick is careful to continue by pointing out that the
time available (but see note 16, Chapter 5), the diversity of habitats,
and combinatorial possibilities of chemistry do not exclude life orig-
inating ‘by a perfectly reasonable sequence of fairly ordinary chemi-
cal reactions’.90 More than two decades on from Crick’s ruminations,
however, it still remains the case that the notion of an infinitesimally
unlikely series of chemical reactions – that from our perspective can be
described only as a ‘near miracle’ (or ‘happy accident’, a curious term,
given that it echoes traditional Christian teaching on the nature of
the Fall) – remains the unbidden and silent observer at much of the
discussion of how life originated. Yet, as Iris Fry (note 85) reminds
us, such terminology is effectively that of creationism. Put this way,
nearly everyone will ask that the now unwelcome guest should van-
ish through the adjacent wall, and agree that for all their differences
(soup, clay, clouds . . .) they share the common hypothesis that the
steps from inert to vital must be those of an unbroken continuity.
It would be an uncomfortable corollary if the series of meetings,
interactions, and reactions of those few chemicals that led to the ori-
gin of life were little more than a series of fortuitous and happy flukes.
If so, a scientific campaign for understanding the origin of life is not,
to put it mildly, going to be straightforward. It is hardly surprising that
George Wald91 wrote, ‘One has only to contemplate the magnitude of
this task [of making life] to concede that the spontaneous generation of
a living organism is impossible. Yet here we are – as a result, I believe,
of spontaneous generation.’92 This quotation has become justifiably
famous, but it is sometimes forgotten that in its original context there
was a very strong underlying assumption that such a process of spon-
taneous generation could be possible only if there were enough aeons
of time. As we shall now see, there probably were not, and to make
matters worse what obviously did happen in our Solar System may
itself be either a very rare occurrence, or, dangerous thought, possibly
unique.
5 Uniquely lucky? The
strangeness of Earth
Wandering though ruins, along deserted thoroughfares and past top-

pled columns, can be a melancholy, if not disquieting, experience. For
some it is the sense of vast futility of vanished and vanquished king-
doms, of which one might read in Shelley’s Ozymandias. But ruins
open other prospects, not least that sense of majesty and awe such
as the Company felt as they swept along the River Anduin past the
Argonath, the giant sentinels of Númenor. For some today such emo-
tions are either a dangerous or a worthless currency, a price to be
paid for our disenchantment. Yet for innumerable generations men
and women have stood before temples and shrines, now derelict if
not entirely vanished, to scry the Moon, perhaps as a heart-catching
sliver at dusk or an immense harvest orb pulling itself above the hori-
zon. The potency of the Moon needs no emphasis. Yet it has only
recently come to be realized that it, too, is a gigantic and lifeless
ruin, a shattered world, but a satellite whose existence, we now re-
alized, has had a profound effect on the Earth and its four-billion-
year-old cargo of life. Indeed, the peculiarities of the formation of the
Moon help to epitomize the principal theme of this book, the odd
fortuitousness of the world in which we find ourselves, where again
and again matters seem to be remarkably well arranged. To some,
this is just a happy string of coincidences, but the strangeness of the
Moon, which perhaps we take too much for granted, serves to open
a series of vistas into the equally peculiar construction of our Solar
System.1
the shattered orb

To return to the beauty of the Moon: in what way is it a ruin? Even a
glance reminds us that it is far from homogeneous in appearance.2 To
a first approximation, in the reflected sunlight much of the Moon’s
surface is bright, but there are also conspicuous and more-or-less cir-
cular dark areas, once thought to be seas across which sailed the lunar
inhabitants or selenites. These areas are still referred to as the maria
(singular mare). The Moon has no atmosphere and there is thus no
70 uniquely lucky? the strangeness of earth
weathering of its surface, nor is there any crustal recycling akin to

terrestrial plate tectonics. Even the Moon’s volcanism is long extinct.
Today the ancient surface of the Moon is modified only by the im-
pact of meteorites and comets. The large craters, such as Copernicus,
Tycho, and, on the far side of the Moon, Giordano Bruno, exhibit dra-
matic rays that in radiating from the point of impact provide graphic
evidence for collisions. The two craters visible from Earth are, in com-
parison with the age of the Moon, relatively young. Copernicus was
formed about one billion years ago, but Tycho is much younger. It was
formed perhaps 50 or so million years before a similar object hit the
Earth, spelling out the doom of the dinosaurs. Just like the Earth, the
Moon continues to receive high-velocity visitors. One of these, it has
been claimed, was recorded as a dramatic eye-witness account in a me-
dieval manuscript, written by the chronicler Gervase of Canterbury
in 1178. Possibly it was this impact that formed the Giordano Bruno
crater.3
Such impacts are relatively infrequent, and it is agreed that
the surface of the Moon was largely shaped by intense episodes of
bombardment early in the history of the Solar System. The principal
bombardment (Fig. 5.1) literally pulverized the surface with repeated
impacts, churning the surface to rubble and bestrewing the Moon with
craters. As the meteoritic material was swept up by the Moon (and, as
we shall see, the other planets) so the intensity of the impacts dimin-
ished. There was, however, a last, violent interval when the Moon
experienced a series of very energetic impacts.4 The standard view
is that these bodies were presumably also products of continued ac-
cretion, and so would have had diameters of the order of a hundred
kilometres or so.5 It was the impact of these massive bodies6 which
led to the formation of the darker maria, such as the Mare Tranquilli-
tatis, where the astronauts of Apollo 11 first stepped on to the Moon.
The relative smoothness of the maria, which have rather few craters,
shows that they must have been formed after the main bombardment
episode. These maria are effectively huge pools of basalt, now solid
rock, but molten lava that flowed into the circular depressions formed
by the impacts. It appears that the basaltic magma was not formed as a
direct result of the collision, but was formed by melting deeper within
the Moon and then ascended through marginal cracks to fill the giant
craters. In some instances, as in Mare Imbrium, the tops of mountains
formed earlier still protrude above the lava plains.
battering the earth 71
10−1
10−2
Densest cratering
Uplands
Cumulative crater density
10−3
Tranquillitatis
Procellarum
10−4 Copernicus Crisium
Imbrium
Tycho
10−5
10−6
0 1 2 3 4 5
Age (Ga)
figure 5.1 The inferred bombardment history of the Moon in terms of

cumulative density of cratering since the formation of the Moon about
4500 million years ago. Note the intense rate of early cratering, before
about 4000 million years, which then declines with the formation of the
huge maria, of which four are listed here. Thereafter, the rate of impacts
gradually falls off, with a steep decline from about 1000 million years
ago. In this interval infrequent giant collisions include those that led to
the formation of Copernicus and Tycho. (Reprinted from Icarus, vol. 92,
C. F. Chyba, Terrestrial mantle siderophiles and the lunar impact record,
pp. 217–233, fig. 1, copyright (1991), with permission of Elsevier
Science, and also the author.)
battering the earth

So it is that when we look at the Moon riding in the sky we see
inscribed upon it events of almost unimaginable violence, inflicted
during the early history of the Solar System. Even so, both the Moon
and those ancient times may seem remote and scarcely of relevance
to life on Earth. In fact, the opposite is the case. Consider that if the
Moon was subject to so heavy a bombardment, then the much greater
gravitational attraction of the Earth, with a mass almost one hun-
dred times greater than that of the Moon, would almost certainly
have made it subject to an even more severe bombardment. Unlike
the Moon, the Earth now shows no direct trace of this bombardment:
the ceaseless processes of erosion and crustal recycling have oblit-
erated all the evidence. Many of the impacts must have been large,
and a few were cataclysmic. What happens when something very big
hits the Earth?7 The amount of energy released is enormous, so great
in fact that the heat generated was probably sufficient to evaporate
the entire world ocean. The net result is that the surface of the Earth
would then have become an inferno. Initially there would be an at-
mosphere of rock vapour, but this would cool quite quickly. This rock
would rain down in a couple of months to form a layer carpeting the
Earth to a depth of several hundred metres. The Earth would still be
a searing desert, and the atmosphere would remain a boiling mass
of water vapour and other gases: a scene in some ways resembling
the present-day Venus, where the surface temperatures are an uncom-
fortable 480 ◦ C. Venus has become trapped in a permanent furnace,
too close to the Sun and with an atmosphere that acts as a power-
ful and irreversible greenhouse. On the early Earth, however, after
each giant impact the excess heat was radiated into space, the atmo-
sphere cooled, and it began to rain, for perhaps 3000 years as the oceans
refilled.8
It is estimated that during the later bombardments the Earth
may have been pounded by five or so of these monster impacts, each
capable of evaporating the ocean and presumably sterilizing the sur-
face. In the case of the lunar maria, which as noted above, were also
formed in the late bombardment episode, it appears that although
they formed during a fairly substantial interval, the largest impacts
are quite closely clustered in time and, more significantly, terminated
rather abruptly about 3800 Ma ago. Now this happens to be a rather
significant date in Earth history, because it is the age of the oldest rea-
sonably well-preserved sediments.9 These are the famous sediments
from the Isua Supergroup of central – west Greenland. Despite meta-
morphism, a wide variety of lithologies, including waterlain conglom-
erates, are known. Although fossils are lacking10 the presence in these
sediments of carbon, altered by metamorphic heat to the mineral
battering the earth 73
graphite, is possibly, if controversially, indicative of life.11 The main

reason for thinking this is that there is evidence of fractionation in the
graphite in favour of the lighter isotope carbon-12, a process widely
accepted as evidence of ancient photosynthesis.12
Earlier rocks are known,13 but these have been so distorted and
altered by the processes of metamorphism that any evidence for life
will have been destroyed. We may never obtain direct evidence from
the geological record of when life either emerged on Earth or, as some
propose, arrived as unintentional Martian colonists (see pp. 75). In
one sense, however, this ignorance of whether life emerged here or
came from Mars may not matter overmuch. If we assume that the
end of the late stage bombardment episode is correctly dated at about
3800 Ma, then the fact that the first identifiable sediments also date
from this interval is more or less what we would expect: before this the
surface of the Earth was repeatedly harrowed by giant impacts. Before
this time, the impacts may have seriously impeded, if not entirely
frustrated, the emergence or survival of life.14
As already noted, the actual history of the Earth’s bombardment
does not survive, but the scarred surface of our daughter satellite
provides its own mute testimony. One estimate, for example, sug-
gests that the Earth received in total between 17 000 and 22 000 major
impacts,15 of which a few were truly colossal. The fact that the bom-
bardment drew to a close at about the same time as the deposition of
the oldest known sedimentary rocks, containing carbon that is per-
haps of organic origin, leads to three interesting possibilities. In brief,
they are as follows: (a) no sooner were the conditions benign enough
on Earth than life evolved, very quickly; (b) the bombardment episode
was severe in the extreme, but life had evolved appreciably earlier and
hung on; or (c) life evolved, but on Mars, and a regular traffic ensured
that as soon as it was safe the Martian bacteria settled down to enjoy
their new home.
Thus, the first possibility (a) is that the degree of devastation,
with at least several immense collisions, would have released suffi-
cient energy to evaporate the oceans and sterilize the surface of the
Earth. This implies that the time available for life to emerge may have
been remarkably short,16 given the evidence already noted for life be-
ing present more or less at the time this catastrophic interval ended.
One such estimate suggests that the time life needed to emerge was
as little as 20 Ma,17 and some origin-of-life enthusiasts seem quite
content with even 10 Ma.18 This, incidentally, is not to say that life
did not emerge earlier. Quite possibly it did; in principle it could
have evolved several times,19 but was repeatedly snuffed out when the
oceans boiled and billions of tons of rubble descended to a fiery surface.
What of the other two possibilities mentioned above? It is
scarcely surprising that an estimate of a mere 20 Ma (or even 10 Ma)
for life to originate is sometimes greeted with considerable unease, if
not disbelief. As we saw at the end of the previous chapter (see p. 68),
it is a common assumption that life will certainly emerge at least in an
Earth-like milieu, but it needs not only time, but aeon stacked upon
aeon as the chemicals blindly combine and re-associate, with the thin
red thread of vitality worming its way past uncounted blind alleys
and entropic pits. One response is that life is indeed an abrupt pro-
cess of self-assembly, by implication a cosmic inevitability, as natural
as water being wet. Quite possibly so, but the experimental morass
reviewed in the preceding chapter might give us pause for thought.
So can we find a loophole, to allow more time for the emergence of
life?
There are certainly two possibilities. The first is that despite
the violence of the giant impacts, life not only evolved much earlier,
rocked by wave after wave of planetary disaster, but still managed to
cling on. Nobody doubts the tenacity of life, but not until the discov-
ery of deep-sea hydrothermal ecosystems with bacteria that incredibly
are able to function at temperatures close to 100 ◦ C was there any hint
as to how early life might have survived repeated thermal catastro-
phe. It is certainly interesting that many of the bacteria that appear
to fall closest to the root of life are adapted to function at very high
temperatures, and hence are known as the extreme thermophiles.20
This could indicate either the ancestry of life on a hot Earth, or more
probably an ability to survive a series of thermal ‘bottlenecks’, each
imposed by the vast quantities of energy released by a giant impact.
Despite the widespread popularity of these ideas (a sort of bacterial
equivalent of Benvenuto Cellini’s story of the fire-salamander bask-
ing in the flaming coals), some scientists continue to argue that the
most primitive forms of life were adapted to much more moderate
temperatures.21
So perhaps life is even more ancient than we suppose, with its
origins in the great heat of hydrothermal vents and its early survival
set against a backdrop of mega-impacts and searing temperatures far
the mars express 75
beyond any human tolerance. But there is a third way: in the begin-
ning our world was still far too dangerous, and by the time conditions
were stable enough to allow life to begin to evolve, it was far too
late. Hence, the third of the possible explanations for life appearing
no sooner than the great bombardment had ceased: life had indeed
evolved, but elsewhere, on Mars. Terrestrial proto-life didn’t stand a
chance.
the mars express

At first sight the alternative to removing life’s origins from the fiery
heat of the early Earth to the arguably more benign habitat of Mars
seems incredible. After all, Mars is today so inhospitable that few ex-
pect to find living organisms there. But early in the history of the Solar
System, it may have possessed a major advantage over Earth. First, its
smaller size (it has about ten times less mass) means a weaker grav-
itational field, which in turn would have ameliorated the episode of
giant impacts.22 Moreover, it is generally thought that early in its his-
tory the surface of Mars was considerably more benign than it is now:
warmer, wetter, perhaps even with an ocean.23 So if Mars is the cradle
of life, then how did we, or more specifically, our bacterial ancestors,
get here? The recognition of more than twenty Martian meteorites,24
including the now-famous Allan Hills (ALH 84001)25 find and the dog-
destroying Nakhla fall (Chapter 3, p. 34) is clear enough evidence that
pieces of Mars can be blasted from the surface by impacts. Indeed, cal-
culations suggest that the vast majority of Martian meteorites arriving
on Earth escape detection because on average one such fall per month
would be predicted.26 Geologically the transit time from planet to
planet is relatively short: a Martian meteorite typically spends about
a million years in space. But it can be much less, and given a par-
ticular trajectory of launching about one in every ten million would
arrive in only a few years. Evidence from the exposure of these mete-
orites to cosmic rays as they drifted through outer space shows that
their times of arrival on Earth are relatively recent, all of them hav-
ing landed in the past 15 Ma. It is equally conceivable, however, that
about 4 billion years ago some such chunk of rock, but with a cargo
of microbes, landed safely on Earth.27 One obvious problem, whether
life forms such as bacteria could survive both the large accelerations
and sudden changes in velocity as a meteorite was blasted off the sur-
face of a planet like Mars, has been addressed experimentally, with the
conclusion that an abrupt departure from Mars presents few problems

for survival.28 So we may really be Martians.
Meanwhile, on Mars itself the environmental conditions
steadily deteriorated: the ocean dried out, the surface temperatures
plummeted, so that today in the Martian winter the ice caps grow
as carbon dioxide is precipitated. So, too, Mars lost its atmosphere
to outer space, with the result that now the air there is far thinner
than it is at the top of Everest. On the Earth, the story was different.
The likelihood is that before too long we shall have information as to
whether life did evolve on Mars. If it did emerge, or just conceivably
still survives, then we should learn whether it had (or has) any bio-
chemical similarity to terrestrial cells. It needs to be remembered, of
course, that if the Earth were fertilized from Mars, the reverse is in
principle possible. It seems, however, that the return journey, while
not impossible, is still considerably more difficult.29 It also remains
possible that evidence for the most ancient crust of the Earth will
be pushed back into the time of major bombardment, with an out-
side chance of finding remains of organic activity. If, by some remote
chance, very ancient terrestrial life were found, this might make a
Martian colonization less likely.
At present, however, we face many other intangibles. Perhaps,
especially if it eventually transpires that Mars was never an abode to
life, we shall simply have to accept that life on Earth did evolve very
rapidly. This begs many questions, but it also offers some intriguing
possibilities. As already noted in Chapter 4, the attempts to delineate
a plausible set of prebiotic pathways have met with limited success.
Laboratory investigations may target, or just as likely, stumble across,
a neglected possibility. Nevertheless, the present situation is ironic.
The fact is that as and when we do make a living organism from a
prebiotic brew, and in principle this cannot be impossible, it will be
constructed by intelligent and directed intervention. In other words, it
will have to be designed or, if you wish to duck teleological terminol-
ogy, at least engineered. From our present perspective, it might be no
more unreasonable to invoke participation by a visiting squad of alien
molecular biologists. Alternatively, and more scientifically, it may be
found that when life evolves it can only do so by one route. There
is a famous remark by George Wald30 that is particularly apposite,
along the lines that so long as you can teach terrestrial biochemistry,
then your students will have no fears as they wander into the Great
making the solar system 77
Examination Hall of Arcturus to sit Biochem. IV. So, too, writing on

‘The universal nature of biochemistry’31 Norman Pace remarks,
it seems likely that the basic building blocks of life anywhere will
be similar to our own, in the generality if not in the detail. Thus,
the 20 common amino acids are the simplest carbon structures
imaginable that can deliver the functional groups used in life . . .
Similarly, the five-carbon sugars used in nucleic acids are likely to
be repeated themes . . . Further, because of the unique abilities of
purines and pyrimidines to interact with one another with
particular specificity, these subunits too, or something very
similar to them, are likely to be common to life wherever it
occurs.32
making the solar system

Let us now imagine that at least the potential for life is more or less
universal. In doing so we shall conveniently neglect, but not entirely
forget, the difficulties in understanding how it might have been as-
sembled. I shall assume, perhaps incorrectly, that life is confined to
planets with liquid water. So far as the search for planets beyond our
Solar System is concerned, until a few years ago, matters were not very
encouraging. Now, with the discovery of tens of what are referred to
as extra-solar planets, everything is changing.33
Although remote planetary systems had long been postulated,
the first clear hint of what would be found came with the discov-
ery of discs of gas and dust surrounding stars. It is widely believed
that planets nucleate from such discs. Of these structures perhaps the
best known is the one associated with the Southern Hemisphere star
β Pictoris, where an immense disc extends into space for a distance
many times greater than the diameter of our Solar System. It is conjec-
tural whether planets are forming in this particular disc,34 but in any
event with the existing telescopes no direct image35 of any extra-solar
planet is obtainable. How then is it possible to detect their presence?
Nearly all the recent series of extraordinary discoveries depends on a
method known as high resolution stellar spectroscopy. The principle
of the method is based on the fact that although the star is much more
massive than the planet, the latter also has a mass and so exerts a grav-
itational pull that slightly affects the star. Not surprisingly, given the
enormously disparate masses of star and planet, the effect on the star
is very small. Nevertheless, as the planet moves round in its orbit so

the star is pulled in different directions. As a result, the wavelengths of
the light from the star as seen by observers on Earth are very slightly
shifted because of the relative motion between the source and the
observer (the Doppler effect). The shift can be measured by interfer-
ometry, and periodic shifts in the spectral lines indicate that a distant
planet has been detected.
The measurements are very sensitive, and only massive planets
capable of imposing measurable perturbations on their stars can be
detected. In addition, the measurements need to be made over a pro-
tracted period, sometimes several years, before the evidence is suf-
ficiently compelling. Even so, the sensitivity of the technique and
the precision of the fits are both quite remarkable (Fig. 5.2). About
90 extra-solar planetary systems have so far been detected (Fig. 5.3).
In most instances the available data indicate that there is one planet
orbiting each star. This is not to say there are no other planets in each
system. Quite possibly there are, but they are too small to exert a
measurable influence on the star. There are, however, a number of
observations where the data indicate a two-planet system (e.g. stars
HD 12661, HD 37124, HD 38529), and in a few cases even a three-
planet system.36 In Upsilon Andromedae, for example, the spec-
troscopy of the star indicated a planet orbiting the star every 4.6 days,
but there was a residuum of data; when analysed further this revealed
the presence of two more planets (Fig. 5.4), respectively taking about
240 days and 3.5 years to orbit. So, too, 55 Cancri is now known to pos-
sess three planets: two close to their sun, and the third, substantially
larger, in an orbit somewhat beyond the equivalent position of Jupiter.
The results to date are spectacular, but they are also deeply
sobering.These remote planetary systems are patently very different
from our Solar System. So far as the search for extraterrestrial life is
concerned, this seems discouraging, but some qualification is needed.
First, the limitations of the detection method mean that only very
large planets can be discerned. These are typically somewhat larger
than Jupiter, which itself is about 300 times more massive than the
Earth. A number of the extra-solar planets are truly gigantic (and it
is important to remember that the estimate is a minimum because it
is the mass times the sine of the inclination of the orbital plane rela-
tive to the observer). The planets circling HD 162020 and HD 202206,
for example, are each estimated to have a mass equivalent to about
figure 5.2 Detection of an extra-solar system, in this case a planet

orbiting the 55 Cancri. The dots represent measurements of the Doppler
velocities that arise as a result of minute shifts in the spectral
absorption of the starlight as the gravitational tug of the orbiting planet
distorts very slightly the shape of its star. The sinusoidal line is the
predicted pattern; notice the goodness of fit. These data suggest the
planet has a mass slightly less than Jupiter (0.84), orbits the star in about
a fortnight (14.64 days), and is located ten times closer to the star than is
the Earth (at 0.11 Astronomical Units). Note that subsequently evidence
has emerged of 55 Cancri possessing two more planets; see Fig. 5.3.
(Redrawn with permission from fig. 1 of R. P. Butler et al. (1997) Three
new ‘51 Pegasi-type’ planets. Astrophysical Journal, vol. 475, pp. L115–
L118, copyright of the authors and the American Astronomical Society.)
14 Jupiters, while one of the planets in HD 168443 approaches

17 Jupiter masses. On such planets gravity would, by terrestrial stan-
dards, be immensely powerful,37 and outside any oceans (whose ex-
istence, as we shall see, may in any event be problematic) life would
probably be only microscopic and present as thin films: a difficult
place for a human to visit. A shower of rain would be like stand-
ing under flying gravel, and walking down steps would probably re-
sult in multiple fractures. In reality many of these enormous planets
may be more similar to gas giants like Jupiter,38 effectively without
a solid surface and where any sort of terrestrial ecology is probably
(5.2 AU)
Sun
HD187123
HD75289
TauBoo
51Peg
Ups And
HD76700
HD49674
HD217107
HD162020
HD38529 (3.5 AU)
55Cnc (5.9 AU)
GL86
HD195019
GL876
Rho CrB
HD168443
HD114762
70Vir
HD80606
HD37124
HD12661
HD210277 KEY
HD128311
>1 Jupiter mass
16CygB
47UMa (3.7 AU) >1-- 4 Jupiter masses
14Her >4--11 Jupiter masses
HD216437 >11 Jupiter masses
0 1 2 3
Astronomical units
figure 5.3 A sample of extra-solar planets, scaled to our Solar System and the size relative to Jupiter.
(An astronomical unit is the distance from the Earth to the Sun, and is equivalent to 1.49 × 1011 m.)
Data taken from websites given in note 33.
unimaginable. As we shall see later (Chapter 6), this has not prevented
speculation about possible forms of life, although such hypotheses are
unlikely to be compatible with the extreme turbulence of such giant
worlds.
The real surprise, however, is not so much the enormous size
of these newly discovered planets but how closely some of them orbit
their stars (hence their nickname of ‘hot Jupiters’). Some are substan-
tially nearer to their stars than Mercury is to the Sun, whirling around
their stars in a few days. The searing surface temperatures that are in-
evitable so close to the star almost certainly preclude the possibility
of life. To complicate matters, many of these planets have eccentric
orbits. Even on the Earth, with a rather mild eccentricity (0.0167) the
variation in distance from the Sun is clearly discernible in climatic
changes. For extra-solar planets with eccentricities up to 0.93 (as for
HD 80606), the thermal regime at the surface will be much more
variable.
Although with present-day techniques we can detect only giant
planets, the continuing torrent of discoveries strongly suggests a trend
towards the majority of planets being relatively small. For example,
the planet orbiting the star HD 76700 has a mass approximately one-
fifth that of Jupiter, while that of HD 49674 is even smaller. As more
observations are made, planets more distant from their stars, which
take longer to complete an orbit, will also be detected. As noted above,
the third planet of 55 Cancri is remote from its star, taking more than
14 years to complete its orbit. It is important to stress that much
smaller planets, similar in size to the Earth, are still beyond the lim-
its of detection. There is so far no reason to think that such planets
cannot exist, somewhere. Telescopes suitable for their detection are
already being designed. And there is some cause for optimism. Some
earlier computer models of solar system formation were able to pro-
duce terrestrial planets, but there were difficulties. For example, both
the eccentricity and the inclination of the orbits were substantially
greater than those of the Earth, and there were also problems in ‘mak-
ing’ smaller planets, comparable to Mars and Mercury.39 Still, refine-
ments continue to be made, both as a result of new knowledge of
extra-solar systems and faster methods of computation. The results
obtained by John Chambers are impressive,40 inasmuch as it is now
possible to produce a range of hypothetical solar systems (Fig. 5.5)
equipped with a variety of terrestrial planets which ‘are reasonable
100
50
Velocity (m s−1)
−50
−100
Ups And
50 100 150 200

Days
150
Ups And
100
Velocity residuals (ms−1)
50
−50
−100
4.6-day sinusoid
removed
−150
1992 1994 1996 1998 2000
Time (years)
removed
analogues for Earth and Venus’,41 although conjuring up Mars and

Mercury is still proving more difficult. In addition, in these models the
gas giants Jupiter and Saturn are already in place, and their presence
is integral to the evolving story of planetary formation closer to the
Sun. Not surprisingly, there is a diversity of systems, and it seems
possible that up to six planets could be formed, although they were
not produced in these simulations. That would be a fine sight on a
starry evening. Whether there would be anybody to appreciate them
is another question.
These simulations demonstrate that while broadly analogous
systems can develop, an exact counterpart to our Solar System may
be much less likely. How important this might be in constraining the
emergence of life, let alone of intelligence, is rather conjectural, but
the answers depend in part on the position of the so-called Circum-
stellar Habitable Zone (CHZ) (see below). It is worth noting, however,
Chambers’s closing remarks on these simulations, to the effect that
‘the high concentration of mass in the region occupied by Venus and
Earth is not reproduced in any of the simulations, and this discrep-
ancy shows no sign of disappearing as a result of using different initial
conditions. This represents perhaps the most important outstanding
problem for theories of the final stage of terrestrial planet formation.’42
It is probably the models that will require further tweaking; after all
the reality entails starting with billions of orbiting and colliding bod-
ies, rather than about 150 planetary embryos, and time intervals of
millions of years, not a month or so of computer time. But the alter-
native is also possible: solar systems with terrestrial planets occur,
perhaps with some frequency, but our Solar System is still very atyp-
ical. Time will tell.
figure 5.4 (opposite) Detection of an extra-solar system in Upsilon

Andromeda. Upper panel shows detection of the principal planet, which
orbits the star in less than a week (4.61 days). Once the 4.6-day
sinusoidal pattern was removed, there was a residuum of Doppler
velocities, and these are shown in the middle panel. The lower panel
shows the best fit, consistent with the presence of two more planets.
(Redrawn with permission from fig. 4 of R. P. Butler et al. (1997) Three
new ‘51 Pegasi-type’ planets, Astrophysical Journal, vol. 475, pp. L115–
L118 (upper panel), fig. 2 of R. P. Butler et al. (1999), Evidence for
multiple companions to υ Andromedae. Astrophysical Journal, vol. 526,
pp. 916–927 (middle and lower panels), copyright of authors and the
American Astronomical Society.)
Making more terrestrial planets
Start 3 Ma
0.8
0.6
0.4
0.2
0.0
10 Ma 30 Ma
0.8
Eccentricity
0.6
0.4
0.2
0.0
60 Ma 200 Ma
0.8
0.6
0.4
0.2
0.0
0.5 1.0 1.5 2.0 0.5 1.0 1.5 2.0
Semi-major axis (AU)
figure 5.5 Making terrestrial planets. Panel above shows a simulation

with 154 planetary embryos with a bimodal distribution of masses: half
the mass is housed in 14 larger embryos, and the other half in the
remaining 140 miniature planets. In this simulation the basic structure
of the solar system with four main planets has emerged by about 30 Ma,
although about 25 smaller planetismals remain, of which many plunge
into the star. The panel opposite shows the end-product of sixteen of the
simulations, in comparison with our Solar System (top). Simulation 23
is the end result of the panel above. Relative sizes of planets are shown;
also shown are their perihelion and aphelion distances (the horizontal
line), i.e. the nearest and furthest distances to the star that thus define
the eccentricity of the orbit. Spin axes of each planet are depicted with
an arrow. Reprinted from Icarus, vol. 152, J. E. Chambers, Making more
terrestrial planets, pp. 205–224, figs 3 and 10, copyright (2001), with
permission from Elsevier Science, and also the author.
Such models give tantalizing glimpses of what might be the re-

ality of other worlds, but at present they remain computed constructs,
virtual possibilities that for one reason or another might never be man-
ifest. On the other hand, there is, as we have already seen, a known
reality of extra-solar systems (within the limits of how best to inter-
pret minute fluctuations of distant starlight). What needs particular
emphasis is that the configurations of these newly discovered systems
figure 5.5 (cont.)
were to a large extent unpredicted. These new results are forcing rad-
ical reconsiderations of how the nebular discs collapse into planets.
What to us is a familiar arrangement in our Solar System of an inner
series of small rocky planets and an outer group of gas and ice giants
may be very much less common than was once thought, and may be
very rare indeed.
Perhaps the most dramatic implication is the strong likelihood,
indeed near certainty, that the giant planets detected so far in the
ninety-odd extra-solar systems actually formed much further away
from their stars, and only subsequently were moved into their present
position of proximity.43 The reason for thinking this is that it is
effectively impossible for a planet to accrete so close to the star, and
mechanisms exist to explain why the planet should spiral inwards. It
will be obvious that even if life evolves on such planetary giants it will
be doomed as its home migrates closer and closer to its sun. Indeed,
in many instances it is likely that the planet’s ultimate destination
is the star itself, to be swallowed in the thermonuclear maelstrom.
To observe such a catastrophe is beyond our present technology,
but evidence for such an engulfment already exists.44 If the planet
continues to circle the star, the earlier process of migration explains

the otherwise puzzling eccentricity of the orbit. There is also some
evidence that over time this orbit may become more circular. The
inward migration of a giant planet has, however, another ominous
consequence: should the system contain two or more of these giant
planets, then the likelihood is that at least one will be ejected into
the wastes of interstellar space.45 As already noted, the existence
of smaller planets, perhaps Earth-like, in such extra-solar systems
is still conjectural. If they turn out to be much more scarce than is
hoped, this may be because they were slung-shot into the interstellar
wilderness early in the history of the solar system when the giant
planet spiralled inwards, rather than because smaller planets were
not formed in the first place. On a lost planet the prospects for life
surviving seem grim as its parent sun slowly recedes to become ulti-
mately just another twinkling and hopelessly remote star. As Stuart
Weidenschilling and Francesco Marzari remark, ‘Many potential
systems of terrestrial-type planets may be disrupted by a gas giant in
their midst; we would owe our existence to the possibly fortuitous
circumstance that only two cores in our own system, Jupiter and
Saturn, grew large enough to accrete gas’.46
This is not to say that we should take too Earth-bound a view
of where life may be allowed to evolve. In some quarters hope seems
to spring eternal. Thus, no sooner had the extra-solar super-Jupiter
planets been discovered than the possibility of life inhabiting smaller
moons circling these monsters was raised.47 The wish-list for such a
world is interesting. Apart from the problem of whether giant planets
can retain any moons during their inward migration, the desiderata
include a body large enough to retain an atmosphere, an effective mag-
netic field if embedded in a powerful magnetosphere (as possessed by
Jupiter), and association with a giant planet that does not have an
eccentric orbit around the star. There is another interesting thought
connected to the possibility that before inward migrations the moon
might have been very well endowed with volatiles if it accreted near
the ‘snow-line’. These volatiles might serve to form a global ocean
when moved closer to the star. So life is a possibility in such a regime,
although this idea is not free of difficulties.48 In part, the inspiration
for this proposal was news from Europa, which rivals volcanic Io as
the most remarkable of Jupiter’s satellites. Images from the spacecraft
Galileo revealed an icy crust, with clear evidence of dynamic motion.
rare moon 87
Geophysical measurements indicate that beneath the ice, which may

be more than a hundred kilometres in thickness, there is probably
an ocean,49 and it is speculated that life may have evolved within its
depths. So, if the extra-solar planets are generally uninhabitable, intel-
ligent life might conceivably emerge on such moons, with their coun-
terparts to Kepler and Galileo struggling to make sense of planetary
dynamics on a moon slung beneath the huge orb of their parent planet.
rare moon
At present the search for extraterrestrial life would receive immense
encouragement if a solar system at least similar to ours were to be
detected. Building a broad facsimile, however, may only be a first step,
because on closer examination our Solar System possesses a series of
apparent oddities (note 1) that may play crucial roles in the emergence
of life. Here our Moon, distant Jupiter, and the rare spectacle of a
comet all indicate that once again the Solar System may represent a
very special arrangement, and Earth an equally special abode.
Let us return to our most familiar planetary companion, the
Moon. As we have seen, it is scarred with a record of the early and
cataclysmic history of the Solar System, mutely reminding us of the
cataclysmic beginnings to the Earth’s history and the precarious, and
perhaps repeatedly frustrated, hold that life then had. Although there
have been a number of theories as to how the Moon came to be asso-
ciated with the Earth, it is now widely agreed that our satellite was
itself a product of a violent impact between the newly formed Earth
and another body, approximately the size of Mars.50 The reasons for
thinking that the Moon is a genuine daughter rather than a chance
capture are various; they include certain geochemical coincidences
and the nature of the dynamical relationship between the two bod-
ies. The Moon itself, however, is very old; it records the early bom-
bardment episodes, and rocks collected by the Apollo missions date
back close to the dawn of the Solar System.51 Given that the Earth
had formed by about 4500 Ma ago, perhaps 100 Ma after the initi-
ation of the Solar System,52 then the mega-impact that led to the
Moon could have been only a little later.53 Some evidence suggests
that at this early stage the Earth may have been somewhat smaller
than at present, and that the impact added substantially to its mass.54
Alternatively, the collision could have occurred somewhat later as the
Earth approached its existing size.55
The collision must have been truly cataclysmic, and has been
analysed by the application of some remarkable computer models.
These give a glimpse of just how catastrophic the event must have
been, but more importantly they reveal that making the Moon may
be far from a straightforward process. Not surprisingly, modelling this
impact is computationally intensive, and delineating the exact na-
ture of the collision depends on a number of variables that in turn
are difficult to constrain. These include the mass of the collider, its
composition (for example, how iron-rich it was), the amount of an-
gular momentum available, and whether the collision was glancing
or head-on. In one such simulation56 the impactor plunges through
the mantle of the proto-Earth, settling on the core. As it does so the
massive release of energy ejects the existing atmosphere and replaces
it with one made of vaporized rock. As these investigators dryly note
the hemisphere which received the impact was ‘devastated’, while the
opposite hemisphere fared little better when a wall of molten rock
flowed across it. Other simulations show the early Earth deforming
as the result of the impact, like a high-speed photograph of a bouncing
tennis ball.
A key result of this giant impact, however, was the detachment
of sufficient material to form the Moon. It is here that we may begin
to run into difficulties. The simulations tend to suggest that building
a single large Moon from the material blasted off by the giant impact
is difficult.57 The alternatives might be either a series of small moon-
lets or even a ring of material, the latter analogous to those encircling
Saturn (and also, as recently discovered, Jupiter, Neptune, and
Uranus), being more likely to end up orbiting the devastated Earth.
At first sight a ring of debris around the Earth would seem a promis-
ing start towards the aggregation of a larger moon. Some of the models
are not, however, very encouraging: either the single body is too small
or the disc inconveniently aggregates into several moons.58 This sim-
ulation of the possible methods of coalescence of impact debris into
larger bodies is necessarily simplistic and looks at the behaviour of
only 37 ‘particles’, each with a radius of approximately 500 km. Even
this was computationally extremely demanding, but, as the investi-
gators remark, they ‘identify only a few scenarios in which a single
Moon results’.59 Further analyses60 are somewhat more optimistic,
although a recurrent difficulty is that a single large body tends to
form near to the Earth, close to the Roche limit, the point at which
rare moon 89
the gravitational field of the Earth would tear the new Moon to pieces.
Other simulations tended to form several smaller moons, and the con-
ditions necessary to form our Moon seem to remain very sensitive to
starting conditions. The problem of propelling a large mass of ma-
terial clear of the Roche limit is also stressed by A. G. W. Cameron
(see note 57). One way to achieve this is a collision with high angular
momentum. The question then arises as to how this angular momen-
tum was subsequently lost. It is important to stress that increasingly
sophisticated simulations are serving to constrain some of the bound-
ary conditions from which the Moon could form. Various sensitivities
still remain, and the work done so far suggests that a wide variety of
possible end-points different from the actual configuration are equally
feasible. A comprehensive simulation may explain much of the his-
tory of the actual impact, but apparently minor variations could have
led to a profoundly different end-point.
As Jack Lissauer has remarked, ‘It’s not easy to make the
Moon’.61 This is brought into sharp focus by some earlier remarks
by A. G. W. Cameron and W. Benz.62 Thus, concerning the origin of
the Moon they wrote:
We deal with a unique event. In the absence of the Giant Impact,

it seems plausible that the Earth would have developed with a
massive atmosphere comparable to that possessed by Venus. Just
as in the case of Venus, it seems unlikely that life would have
developed in such an environment. Thus, there may be a very
large selection effect which means that reasoning creatures will
speculate about the probability of a Giant Impact only on
Earth-like planets which have suffered a Giant Impact and lost
most of their primordial atmosphere.
They continue, ‘This issue lies well beyond the present reach of
science.’63
Although not everyone would agree that the nature of the Earth’s
atmosphere was critically controlled by the Giant Impact, it still
remains the case that even given such an immense collision, forming
the Moon was difficult, possibly extraordinarily difficult. Yet, without
our daughter satellite this world would certainly be a very different
place, and arguably much less benign.64 One advantage of having a
Moon is that it confers considerable stability to the axis of rotation of
the Earth, making it difficult to shift it appreciably from the existing
◦
inclination of about 23 .65 To be sure there are controversial hypothe-
ses proposing that in the geological past there have been major shifts
in obliquity, with the Earth falling in effect on its side, to explain
climatic anomalies such as glaciations in equatorial regions.66 The
notion of profound changes in obliquity has not been well received,
but there is little doubt that without the Moon the Earth would have
been just as subject to chaotic shifts in obliquity as have occurred in
the other planets.67 At the least such shifts in obliquity would have
had major, possibly catastrophic, climatic consequences (Fig. 5.6).
But having the Moon has other advantages. Neil Comins,68 for
example, reminds us that the coupling between the Earth and its large
Moon leads to the transfer of angular momentum from the larger to the
smaller body, because of such factors as tidal friction. This has the re-
sult during geological time of both increasing the separation between
the Earth and Moon and, more importantly, decreasing the Earth’s
rate of rotation. The most obvious result of this slowing in rotation
is an increasing day length and a corresponding smaller number of
days in the year, for which the fossil record shows some confirmatory
evidence.69 If, however, the rate of rotation had not decreased, then
Comins suggests, by analogy with other rapidly rotating planets, that
the wind speeds on the Earth would be persistently higher, with many
more violent and protracted storms. A planet with life, perhaps, but
with winds tearing across the world at an average speed of 200 km per
hour it would be an uncomfortable place to exist. Marine life would
face continuously stormy seas, while on the continents vegetation
would be tough and ground-hugging and most animals would perhaps
be burrowers.
Having the Moon therefore seems to confer significant advan-
tages. Without it life would presumably still be possible, but it might
be more stressed and probably very different; and the emergence of
intelligence (Chapter 9) might well be compromised. Early in the his-
tory of the Solar System other giant impacts on Mars and Venus might
have spalled off material to make moons. The disproportionately large
size of the metallic core of Mercury has been used as an argument for
another giant impact that smashed away most of that planet’s outer
layer.70 Some scientists71 have consequently tended to argue that the
Earth’s Moon is not so exceptional: perhaps, they say, Venus lost any
satellites it had by the action of solar tides, while Mars was too small
and gravitationally weak to retain any such moons.72 The evidence
23.5° obliquity
305 +5°
Surface temperature (K)

285
+85°
265
−85°
245
90° obliquity
−85°
340
Surface temperature (K)
+85°
300
+5°
260
0 90 180 270 360
Longitude, with respect to vernal equinox
figure 5.6 A model of the climatic effects of an Earth without a Moon,

so allowing massive changes in obliquity, perhaps from the existing and
stable value of 23.5◦ to 90◦ . Shown here are the respective temperature
cycles for near-equatorial (5◦ ) and near-polar (85◦ ) latitudes. Solid lines
are for northern latitudes; broken lines represent southern latitudes.
Note the contrast between the relatively benign seasonal swings of our
seasons, as against the wild oscillations in an Earth with an obliquity of
90◦ . In particular, Antarctica would vary in a year from freezing to a
temperature of 350 K (equivalent to 80 ◦ C), beyond the lethal limit of
most organisms. (Figure from ‘Earth–Moon interactions: Implications
for terrestrial climate and life’ (fig. 4) by D. M. Williams and D. Pollard
in Origin of the Earth and Moon, edited by R. M. Canup and K. Richter,
c 2000 The Arizona Board of Regents. Reprinted by permission of the
University of Arizona Press and with the permission of the authors.)
from the modelling of the Giant Impact that is thought to have pro-
duced the Moon, however, suggests that the impactor may have been
as large as Mars, and possibly substantially bigger. Even in the early
history of the Solar System, such a collision would have been very
rare, and as we saw above even size alone may not guarantee a Moon.
Also, it may well be that had this Giant Impact not occurred then
the Earth would have been a substantially smaller body. Is that so
important? It turns out that it is.
just the right size

In a fascinating discussion of alternative scenarios Jack Lissauer73 in-
vites us to consider Earth-like planets, but somewhat smaller and
larger than our familiar home (Fig. 5.7). In neither case is life neces-
sarily excluded, but if it were present the biosphere would probably be
impoverished as compared with what we enjoy, or at least should en-
joy. On a smaller planet gravity is weaker, the consequences of which
are much more precipitous mountains, a thinner atmosphere, and sub-
stantially lower surface temperatures. The internal heat budget of this
smaller planet would be reduced, and its more rigid crust might rule
out plate tectonics. There would probably not be large oceans, and pos-
sibly only shallow seas. On this rather undynamic planet, life might
be constrained, not least because levels of atmospheric oxygen would
be substantially lower than on Earth.
What of a planet substantially larger than the Earth? It would
have stronger gravity to be sure, leading to a much more subdued
topography and probably a globe mostly, if not entirely, covered in
oceans. As we saw earlier, such a possibility was also raised for an ex-
cessively water-rich moon that might orbit an extra-solar giant planet.
While aquatic life might develop in either situation it would be a prob-
lematic habitat because without continents there would be a greatly
reduced influx of nutrients and minerals derived from weathering, al-
though volcanic and hydrothermal activity might provide some com-
pensation. Even if intelligent life, perhaps dolphin-like, emerged, the
transition to a technology based on metallurgy and controlled com-
bustion would probably be impossible. So, if we wish to meet con-
ceptualizing beings we might concentrate on a planet pretty much
the size of the Earth with a large Moon. It would be a further curious
coincidence if the Moon were of a size and at a distance to eclipse ex-
actly the nearby sun during the time a sentient species was becoming
jupiter and the comets 93
figure 5.7 A comparison of Earth-like planets, with our Earth in the

middle flanked by smaller (left) and larger (right) versions. Smaller
means drier and precipitous landscape, while bigger means an ocean-
world with little, if any, emergent land. (Reproduced from fig. 13.9 of
Planetary sciences, by I. de Pater and J. J. Lissauer (Cambridge
University Press, Cambridge, 2001) with the permission of the authors
and Cambridge University Press.)
interested in astronomy.74 Perhaps that is not asking too much – until,

that is, we look at the Solar System’s goalkeeper, mighty Jupiter.
jupiter and the comets

Notwithstanding the dramatic images of its turbulent cloud belts,
transmitted back to Earth by the Galileo spacecraft, to most of us
Jupiter remains remote and very far removed from terrestrial circum-
stances. Probably at about the time Hitler was climbing to power in
the 1930s Jupiter happened to capture a comet. This was Shoemaker–
Levy 9, now famous because on one of its approaches to Jupiter, it
disintegrated into a string of fragments that ultimately fell on the
planet in July 1994.75 Dramatic pictures revealed the violence of the
impact, but still in a world far away. Yet this capture and destruc-
tion actually appear to have a direct relevance to our own existence.
As George Wetherill76 reminds us, the powerful gravitational ‘well’
provided by Jupiter serves to deflect and capture innumerable comets
that would otherwise continue to hurtle towards the inner Solar Sys-
tem. Not only that, but other estimates77 suggest that about half the
comets entering the planetary Solar System from the more remote
Oort Clouds are then flung into interstellar space by the slingshot
effect of Jupiter’s gravity. The importance of this observation will be
returned to below. This still leaves plenty of comets roaming around
the Solar System in a variety of orbits, but Wetherill estimates that
without the shielding influence of Jupiter the comet flux to the Earth
would be about a thousand times greater. That would be bad news for
life on Earth and, as Wetherill observes, ‘this could frustrate the evolu-
tion of organisms that observe and seek to understand their planetary
system.’78
The danger of impacts has been brought into sharp focus, of
course, by the famous end-Cretaceous extinctions, the so-called K/T
event.79 Given the protracted time that ecosystems require to recover
from such disasters, typically a couple of million years,80 it is difficult
to believe that a repeatedly traumatized biosphere, suffering recurrent
and frequent impacts, would be similar to that of our own Earth. More
than a hint of this grim conclusion comes not only from the mayhem
associated with the extinctions, but also from some rather startling
evidence concerning the aftermath of perhaps the two greatest catas-
trophes, specifically the end-Permian (c. 250 Ma) and end-Cretaceous
(c. 65 Ma) events. Here the ecological disruption seems to have been
so severe that it almost seems as if the evolutionary clock had been
wound back to a time when there were no animals; a ‘return’ to Pre-
cambrian ecosystems dominated by microbial activities, with bacteria
and various mostly single-celled eukaryotes. Such a world, before an-
imals literally imprinted themselves on the biosphere, was evidently
very different from the one familiar to us. Thus, following the Permian
debacle there is some intriguing evidence for the wide-scale return
both of stromatolites81 and other microbially bound sediments.82
Analogous evidence for profound ecological disruption after the
end-Cretaceous catastrophe comes from a study of the sedimentary
carbon. The aftermath of this catastrophe has been something of
a paradox because although the marine algae, and thereby primary
productivity in the oceans, recover fairly quickly as a whole, the

marine ecosystems remain in a state of protracted crisis for much
longer. The resolution of this problem appears to be the continued
near-absence of those animals that can graze or otherwise harvest
this plankton.83 The net result of an absence of such grazers is readily
detectable because of a distinctive change in the geochemical imprint
of the organic carbon. The recycling of the carbon is then effectively
dependent on bacterial metabolisms. This indicates a state of affairs
in the post-catastrophe ocean that is reminiscent of the Precambrian
oceans,84 a time, of course, before grazers had evolved. These vari-
ous lines of evidence, indicating how sluggish were the responses of
post-catastrophe biotas, thus suggest that on a comparable planet with
greatly elevated rates of extraterrestrial bombardment the ecosystems
would remain in a state of near-permanent turmoil. Even if a sentient
and cultural species emerged, the fragility of its civilization and the
delicate nature of the associated infrastructures would be such that
frequent major collisions would be, to put it mildly, challenging.85
Paradoxically, the occasional giant impact may be beneficial, at
least in the longer term. First, it needs to be remembered that even
if Jupiter acts as the ‘goal-keeper’ so far as comets are concerned, it
is indirectly responsible for steering meteorites towards the Earth.
This is because the great majority of these rocks are derived from
the so-called asteroid belt, which in effect represents a ‘failed’ planet,
the coalescence of which was prevented by the massive presence of
Jupiter. Moreover, it is the gravitational resonances set up by the plan-
ets, especially, as it happens, the one connected with Saturn, that are
responsible for perturbing the asteroid orbits so that there is a contin-
uous supply of bodies directed towards the inner Solar System. Their
time there is relatively short-lived, and most will ultimately plunge
into the Sun or be captured by Jupiter, but before that there is the
ever-present possibility that one of them may intersect the path of
the Earth.86 How then might such impacts be beneficial? In princi-
ple, by wiping the biological slate largely clean, so that new groups
are given fresh ecological opportunities. The example most usually
quoted is the demise of the dinosaurs following the end-Cretaceous
(K/T) event, which made possible the rapid radiation of the mammals.
No bolide, the argument runs, no adaptive burst among the mammals
and so ultimately no humans. This is an apparently cogent supposi-
tion, especially if your world-view is that history is driven as much
by accident and blunder as by any inherent tendencies that in a bio-

logical realm might be driven by adaptation and competition.87 But
does the demise of the dinosaurs as a portal to mammalian success
really pass muster? Possibly not, and for two reasons, one particular
and one general, which are returned to in more detail below. First,
unbeknown to the survivors of the K/T catastrophe, in due course the
world was to head into a time of global refrigeration, culminating in
the present-day ice ages. With the dramatic cooling of the temperate
and high-altitude zones, it is likely that the warm-blooded groups, no-
tably the mammals as well as the birds, would have been at a selective
advantage. Thus it seems likely that the mammalian radiations would
have been delayed, rather than never happening; in other words, post-
poned rather than cancelled. The second and more general reason is
the main theme of this book, that of evolutionary convergence. As
all the principal biological properties that characterize humans are
convergent, then sooner or later, and we still have a billion years of
terrestrial viability in prospect,88 ‘we’ as a biological property will
emerge.
Barring a daring technology in the future, human civilization
will be hard-pressed to destroy or deflect incoming bolides and so
avoid impact-driven catastrophes. Perhaps a solar system unencum-
bered with roaming comets and asteroids would be a safer place? Possi-
bly so, but paradoxically it is arguable that without these bolides there
might not have been any terrestrial observers to admire the beauty
of a passing comet, let alone know the fear that one day an asteroid
might repeat the K/T catastrophe. The reason for this has already been
touched on earlier (Chapter 3, pp. 42–43). It is only necessary to remind
ourselves that very early in the history of a solar system, as the star
begins to shine and the volatiles, such as water, are driven outwards to
the ‘snowline’ (approximately where Jupiter presently orbits), the in-
ner rocky planets are in serious danger of remaining dry and, in the
absence of water, presumably sterile. It is generally thought that the
fact that Earth has an ocean is due to the delivery of large quantities
of volatiles early in the history of the planet. (Mars and Venus prob-
ably originally also had oceans, but for different reasons things there
went badly wrong.) It is in this context that once again the presence
of a Jupiter may become very important for the future habitability
of the Earth. This is because early in the history of the Solar Sys-
tem the volatiles the Earth would ultimately ‘need’ for such things
as dolphins, ants, and mystics, were far removed, out at the ‘snow-
line’. This, as mentioned, is adjacent to the orbital position of Jupiter
and it was Jupiter that was in a position to displace the water-rich
asteroids89 back towards the inner Solar System. If, however, Jupiter
had formed somewhat inward of its present position, then the Earth
might have remained sterile and dry. This is because Jupiter would
have still efficiently scattered the volatile-rich asteroids, but in the
wrong direction, outwards and away from the inner rocky planets.90
But what of the comets? Given that they appear to be largely
composed of ice, would they not be an ideal source of the Earth’s
volatiles?91 So far as water is concerned, however, it seems that the
comets may have made only a minor contribution. The reason for
thinking this is that the ratio in water between the isotope of ‘heavy’
hydrogen, known as deuterium, and the ‘light’ hydrogen is, in the few
comets yet examined, very different from that of the Earth’s oceans.
Hence the preferred source from the asteroids (note 89). This does
not mean, however, that comets played no part in ensuring the hab-
itability of the Earth; far from it. The principal source of the organic
compounds, the raw material for the origin of life, was probably de-
livered to the Earth at a late stage of its accretion, by a cometary
bombardment.
So, if Jupiter had been in the wrong position, perhaps there
would have been no oceans, and perhaps if the comets had not ar-
rived, no life? In this context it is therefore intriguing to learn that
cometary systems around other stars may be less common than might
be expected, and perhaps even rare. There are several reasons for
thinking this.92 In the case of our Solar System the estimated pop-
ulation of comets runs into the trillions, perhaps about 1013 . Broadly
these fall into two categories: those located somewhat beyond the
orbit of Neptune (the Kuiper Belt); and much further out a vast num-
ber, occupying an immense volume of space, far beyond the plan-
ets, in what is known as the Oort Cloud (see note 77). It is thought
that the Kuiper Belt and Oort Cloud supply respectively the so-called
short- and long-term comets. So what does this have to do with comets
from other solar systems? The explanation is that on the edges of our
Solar System the gravitational pull of the Sun is so weak that any
nearby disturbance, perhaps a passing star (not in itself an unlikely
occurrence given that the outer Oort Cloud extends about half-way
to the nearest star), will exert its own gravitational field and so disrupt
the position of many comets. Some will be displaced to fall inwards

towards the planets and the Sun, but others will be hurled outwards
into deep space. Now the curious thing to note is that it appears that
all the comets that have been observed by astronomers derive from
either the Kuiper Belt or the Oort Cloud. Not one seems to be a vis-
itor displaced from an extra-solar system. This is slightly puzzling,
because calculations suggest that if equivalents to the Oort Cloud oc-
cur elsewhere then there should have been a small number of such
cometary visitors to our Solar System. These interlopers would be
recognized, because they would swing in on hyperbolic orbits rather
than the customary tighter parabolic trajectories.93
To be sure, there is some evidence of what may be analogous to
cometary infall occurring in the giant disc around β Pictoris.94 These
bodies are inferred from variable and sporadic observation of spectral
absorption lines, especially those of magnesium, calcium, and alu-
minium. They also appear to be larger than the Solar System comets,
and therefore might be better classified as infalling planetismals that
evaporate as they approach the star. In a comparable way extra-solar
equivalents of the Oort Cloud might be detected by evidence of va-
porization of engulfed comets as an ageing star swells towards its
red giant stage.95 Preliminary searches have not been encouraging,96
but recently the presence of Kuiper-like comets has been inferred
around an ancient star whose increasing luminosity is evidently va-
porizing the icy comets, which is revealed by the detection of the
resulting water vapour.97 The comets of the equivalent to the Oort
Cloud, if it exists, are too remote from this star to suffer the same
fate, and at the moment the number of extra-solar systems with a
large reservoir of volatile-rich comets (or similar bodies) is necessarily
conjectural.
It would certainly be surprising if no equivalents of a cometary
Oort Cloud existed elsewhere.98 The possibility, however, is that even
if extra-solar systems are abundant the differences from our Solar Sys-
tem may include not only the planetary configurations, but also an
absence of volatile-rich bodies. Even if such bodies exist, with their
latent reservoirs of water and other light elements, there still needs to
be a method of delivering them to any suitable planet where life can
arise. But in one sense that is only the beginning of our problems. Life
may conceivably arise in many Solar Systems but then, because of one
exigency or another, it is snuffed out. The evolution of intelligence
just the right place 99
may be an inevitability, but on Earth at least this process has taken bil-
lions of years. For this to happen demands the persistence of a zone of
safety, that is a Circumstellar Habitable Zone (CHZ). To us humans, of
course, the concept of a habitable zone is exceedingly narrow: without
technology we can neither fly nor burrow, and swimming beyond very
modest depths requires elaborate paraphernalia. Even the biosphere,
from the deepest bacteria to the migrating storks drifting past turbo-
props at 20 000 feet (6000 metres), occupies an exceedingly thin skin,
wedged between the intense heat and pressure of the Earth’s interior
and the sterile vacuum of outer space. Yet despite these flanking hos-
tilities beneath our feet and above our heads, the Earth itself occupies
a zone of comfort, a planetary Habitable Zone, and it is to this topic
we now turn.
just the right place

What we now know of the infernos of Venus and the super-arctic
wastes of Mars is a clear indication that being either too near or too
far from the Sun may have perilous consequences. The geological
record of the Earth, on the other hand, has shown that whatever the
vicissitudes of the surface environment, perhaps most spectacularly
manifested in terms of the episodes of major glaciations (‘snowball’
earths99 ) and super-torrid equatorial regions,100 the surface tempera-
ture has been remarkably constant. Whether or not this is a reflection
of self-regulating homeostasis mediated by biotically driven feedback
(as most influentially promoted by James Lovelock101 ) is less mate-
rial to this chapter than the wider question that needs to be posed:
how likely are planets to find themselves in a zone which ensures
long-term habitability?
The question of how to define the Circumstellar Habitable Zone
(CHZ), how wide (or narrow) it might be, and how it might evolve,
was addressed some years ago by Michael Hart.102 The fate of our near-
est planetary neighbours is a clear enough indication that this habit-
able zone must be relatively narrow (somewhere between the orbits of
Mars and Venus), and most calculations suggest that it is substantially
narrower. The principal problem in arriving at precise parameters is
that the Sun’s luminosity is increasing steadily, but various feedback
processes on the planet, such as those linked to the amount of at-
mospheric carbon dioxide and water vapour (both greenhouse gases),
mean that the calculations are not simple. Assuming that a planet’s
orbit is fixed, which may not always be a reliable supposition,103 the

general conclusion has been that the inner boundary of the CHZ will
change little, but the outer boundary will extend quite markedly.104
There are further complications in these estimates that need to be
taken into account: the type of star, and hence its brightness; whether
it produces a drenching of ultra-violet radiation; its lifespan; and its
size. For example, if the star is two-and-a-half times as large as our
Sun, then it will burn very quickly, probably too fast for life to get
much beyond the first stage.105
Nor do the complexities end there, because planets like the
Earth are not only dynamic systems but they also change through
geological time. The net result of all these factors is a complex set
of possibilities, in terms of stellar and terrestrial histories, that mean
that both the definition and the evolution of the habitable zone are
not easy to constrain and must depend on various assumptions. An
accurate estimate, however, will have implications not only for the
long-term habitability of our Earth, but also for the likelihood of find-
ing near-equivalents elsewhere in the Galaxy. One recent analysis is
of particular interest. Siegfried Franck and his colleagues suggest that
not only has the habitable zone contracted significantly in the past
billion years, but that in another billion years it will have moved
beyond the Earth (and also as it happens will have narrowed signifi-
cantly) (Fig. 5.8), so that higher life at least will have disappeared under
the increasingly intolerable temperatures (Fig. 5.9).106 Although the
long-term prognosis for the Earth is not good, indeed Franck and his
colleagues simply remark, ‘In about 500 Myr [500 Ma] – the biosphere
ceases to exist’,107 their calculations of both the evolution of the hab-
itable zone and the number of Earth-like planets are used to suggest
that in our Galaxy the total of such planets might be 500 000 (give or
take). How many are suitable for the emergence of life, let alone civ-
ilizations, is much more of an imponderable. The strong hint, on the
basis of what I have already said, is that of the half-million, 499 999
are not suitable.
A further indication that this assessment, which might be re-
garded as vaguely discouraging, might be realistic comes from a rather
different line of evidence. This indicates that even if earth-like plan-
ets are relatively common, most of them will be significantly older
than the Earth, on average by a rather remarkable 1.8 Ga.108 Plenty
of time for civilizations to evolve and develop a taste for interstellar
Time (billion years)
0 0.5 1 1.5 2
R (AU)
−1
Sun Venus Earth Mars
figure 5.8 Evolution of the circumstellar habitable zone (CHZ; shaded)

from 1 Ga (1 billion years) ago, through today to 1 Ga in the future,
relative to the orbits (distance in Astronomical Units (AU)) of Earth,
Venus, and Mars. Note that 1 Ga ago Mars was well within the HZ, but
1 Ga in the future the CHZ has shrunk dramatically and the Earth’s
oceans are boiling. (Reproduced with permission from
Naturwissenschaften, from the article Planetary habitability: is Earth
commonplace in the Milky Way?, by S. Franck, A. Block, W. van Bloh,
C. Bounama, I. Garrido, and H.-J. Schellenhuber, vol. 88, pp. 416–426,
fig. 5; 2001, copyright Springer-Verlag, and also with the permission of
the authors.)
travel? The argument for this conclusion is important because it has

an immediate bearing on the detection of the extra-solar planets and
the so-called ‘hot Jupiters’. The key observation is that the stars that
are known to possess planetary systems are what astronomers call
metal-rich.109 By this it is meant that they contain elements heavier
than helium, which is the first product of the thermonuclear burning
of hydrogen within the stellar interior. Metallicity, in an astronom-
ical context, is, therefore, a wider concept than the more familiar
notion of metals such as nickel and iron. The fact that metal-rich
stars also possess planets is probably no coincidence, and such ele-
ments as silicon, oxygen, and aluminium are essential, not only for
the rocky planets like the Earth, but also for the cores of the gas- and ice
figure 5.9 Surface temperatures of the Earth, from 2 Ga ago and

looking forward 1.5 Ga; from a billion years on things are not looking so
good. (Reprinted from Chemical Geology, vol. 159, S. Franck,
K. Kossacki, and C. Bounama, ‘Modelling the global carbon cycle for the
past and future evolution of the earth system’, pp. 305–317, fig. 9,
copyright 1999, with permission of Elsevier Science.)
giants. The converse is also true: stars of low metallicity will gener-
ally lack planets.110 It also seems likely, however, that stars with a
proportionally higher metallicity are more likely to be accompanied
by ‘hot Jupiters’. But, as we have already seen, as such a giant migrates
inwards from its original orbit to position very close to its star so it
will either destroy or eject any earth-like planets. There is therefore
a fine balance in a solar system between having a star with sufficient
metallicity to ensure the existence of earth-like planets and having
a star with too high a metallicity that will lead to the spawning of
destructive ‘hot Jupiters’. These observations can then be combined
with calculations of the rate of star formation (and thus, in suitable
circumstances, the formation of planets) and the assumption of a

steady build-up of metallicity. This last assumption is based on the
fact that the metals (in the astronomers’ sense) are synthesized within
the star, and may subsequently be blasted into interstellar space dur-
ing a supernova (where more metals are made). Remnants of such
supernovae eventually reaggregate as new stars, which will tend to
be more metal-rich. Hence the overall conclusion is that, given these
trends, the majority of planets similar to the Earth will have been
formed at a substantially earlier stage. Plenty of time not only for life,
but also for intelligence to evolve? So where are these planets?
The irony that the destructive cataclysm of a supernova is a
necessary prerequisite for life, because only thus can the elements
essential to biology be broadcast into space and ultimately gathered
together in a new solar system, has often been remarked upon. Yet any
biosphere unlucky enough to be relatively close to a neighbouring star
that does explode will face a traumatic interval. Nor is this the only
risk. Sudden and extremely energetic bursts of gamma-ray radiation
are regularly detected by Earth-based instruments, although their ori-
gin is still rather mysterious. Perhaps they result from the collision
of two neutron stars. Such events release some 100 000 times more
energy than a supernova, a colossal 1053 ergs,111 so placing biospheres
thousands of light years away at risk. Now it is curious that calcu-
lations by Ray Norris112 suggest that such events should afflict the
Earth roughly every 200 million years, yet here we are. Even if this
risk factor (as well as that of more proximal supernovae) is decreased it
does not easily explain the apparent immunity of our biosphere. One
explanation (if that is the word) is that if ‘we have already survived
for some 20 times the mean interval between catastrophes . . . [then
we are] very lucky indeed.’113 If that is correct then one conclusion
might be that ‘we are alone in the galaxy.’ James Annis,114 however,
offers an interesting twist to this prospect, suggesting that while these
gamma-ray bursts would reset the evolutionary clocks across much of
any galaxy that experienced such an intense pulse of radiation, over
time the frequency of these catastrophic events would decline, and
ultimately the emergence of intelligence would become possible. On
Earth the fossil record indicates that this process took about 108 years.
What applies here should, on the Copernican principle of mediocrity
(that is, there is nothing special about the Earth) apply elsewhere.
Thus Annis suggests that as the Galaxy becomes relatively safe, so
‘A previously forbidden configuration is now allowed. It is likely that

intelligent life has recently sprouted up at many places in the Galaxy
and that at least a few are busily engaged in spreading. In another
108 years, a new equilibrium state will emerge, where the galaxy is
completely filled with intelligent life.’115
There is another turn to the story of habitable zones that might
also narrow the likelihood of finding a real counterpart to Earth. This
is not in the context of distance from a star, but within the galaxy as a
whole. Guillermo Gonzalez and colleagues suggest that galaxies like
our Milky Way also have relatively restricted zones of habitability.116
There are several reasons for this that range from the potentially vio-
lent nature of the galactic centre with its great density of stars, some
of which will inevitably explode, not to mention the resident black
hole, to possibly more frequent perturbations affecting the equivalents
of the cometary Oort Cloud, so greatly enhancing the bombardment
rates of the inner planets. In addition, as we have already seen, the
relative metallicity of a star is correlated to the likelihood of its pos-
sessing both planets and destructive ‘hot Jupiters’. Metallicity of stars,
however, declines away from the galactic centre, and this also helps
to define, at least approximately, the habitable zone of a galaxy.
That certainly still leaves millions of candidate stars, but, as
Gonzalez and his colleagues argue, this leaves the Sun (and thereby
the Earth) occupying ‘an especially comfortable region of the Milky
Way’. Recalling the paradox of the Fermi question (‘Where are they?’),
they continue, ‘Any civilization seeking a new world would, no doubt,
place our solar system on their home-shopping list.’117 Not only are
we probably alone in our Galaxy, but looking further afield may sug-
gest that any neighbours are very, very remote. As Gonzalez and his
colleagues also remark, ‘The broader universe looks even less invit-
ing than our galaxy. About 80 per cent of stars in the local universe
reside in galaxies that are less luminous than the Milky Way. Because
the average metallicity of a galaxy correlates with its luminosity, en-
tire galaxies could be deficient in Earth-size planets. Another effect
concerns the dynamics of stars in a galaxy. Like bees flying around a
hive, stars in elliptical galaxies have randomized orbits and are there-
fore more likely to frequent their more dangerous central regions. In
many ways, the Milky Way is unusually hospitable: a disk galaxy with
orderly orbits, comparatively little dangerous activity, and plenty of
metals.’118
a cosmic fluke? 105
a cosmic fluke?
Earlier I remarked on the almost gleeful abasement of humans, not
least to inform us that we are insignificant worms in the cosmic
drama. One powerful ingredient of this dreary world picture is the
Copernican triumph, of Earth the Insignificant. Perhaps so, but it
could be that our planet and its Solar System are both very much
odder than is realized. Life may be a universal principle, but we can
still be alone. Suppose, at least for the sake of the argument, if not hu-
mility, we really are. Suppose also that the Earth is genuinely a cosmic
accident, a chance fluke arising from spinning clouds of dust and gas.
Such a view is now widely accepted, but so, too, are such principles
applied with equal conviction to the history of life. From its start-
ing point, however and wherever that strange event might have been,
orthodoxy states that evolution can potentially explore a million dif-
ferent trajectories. Even if somewhere there is a planet like the Earth,
so the argument continues, there may well be life but assuredly no
biped writing lines similar to these. In the next few chapters I shall
try to persuade you to take another view.
6 Converging on the extreme
One of the boasts of the physicist is that if she is standing on an Earth-

like planet on another galaxy, then the apple she tosses to her male
colleague will follow a trajectory whose course may be followed by
precise equations. The biologist, however, could not even begin to
address the probability of there being sexes, let alone hands and ap-
ples (not to mention serpents) on other worlds.1 That is certainly the
widely accepted view: rewind the tape of life, as S. J. Gould repeatedly
claimed, and let it replay: assuredly next time round the world will be
a very different place, with a vanishingly small prospect of anything
like a human emerging. I have already argued forcibly against such a
position,2 and the purpose of much of the rest of this book is to de-
velop in more detail why the trajectories of evolution are much more
severely constrained than is sometimes supposed. Nevertheless, this
is not the popular view. The present consensus is that, first, this world
is only one of many similar ones, although as we have already seen
(Chapter 5) such optimism is open to some doubt. Second, and more
significantly, it is widely agreed that notwithstanding the unremitting
processes of biological diversification even the most trivial differences
in the starting conditions would have led to entirely different evolu-
tionary histories, each with a radically different outcome. Nothing
like a butterfly, a daisy, or a dolphin and certainly, as we are repeat-
edly informed, nothing like a human. Not only do these alternative
histories debar the emergence elsewhere of anything that is familiar
from this world, but even, it is argued, of more general biological prop-
erties. Thus, it has been repeatedly proposed that even intelligence
itself may be just one more chance end-product of the innumerable
possibilities thrown up by evolution. In due course I shall attempt to
show the exact opposite.
universal chlorophyll?
At first sight the problem of addressing the question of evolutionary
alternatives seems almost intractable. How can we begin to compare
something of which we know nothing with that with which we are
universal chlorophyll? 107
familiar? Self-evidently, but also trivially, the history of life is unique,3

and perhaps the absence of extraterrestrial communication4 is sim-
ply because intelligence is just another evolutionary quirk. Yet con-
straints on the alternative possibilities and trajectories through time,
at least in evolution, can be demonstrated by at least three avenues.
These are respectively: (a) to make a genuine attempt to consider the
alternatives, (b) to devise experiments that at least in a restricted way
rerun evolution, and (c) to look again at the history of life and enquire
whether the ubiquity of convergence, be it in the anatomy, behaviour,
or molecules, is anything more than a biological curiosity.
In the first chapter I mentioned the very surprising degree of
efficiency of the genetic code. In principle the combinatorial immen-
sity of the ‘code hyperspace’ means that there may be many other
equally effective genetic codes, but as also explained one can impose
a number of additional restrictions that suggest that the code adopted
on Earth really is not just adequate, but quite remarkably good. There
are, moreover, a number of other ways in which we can consider bi-
ological alternatives. An appropriate place to start is at the base of
the planet’s life-support system: that is, photosynthesis, and specif-
ically the molecule that serves to trap the energy of the sunlight:
chlorophyll. Chlorophyll must be a very ancient invention,5 because
it occurs not only in the familiar plants but also in various bacte-
ria. Of these, one group, usually referred to as the cyanobacteria (or
sometimes blue-green algae), is of particular note. This is because,
unlike the other photosynthetic bacteria, the oxygen it produces, as
a by-product of the photosynthetic process, was (and, of course, still
is) released into the surrounding water and thence to the atmosphere.
Green land plants and eukaryotic algae, the latter most familiar as
the various sorts of seaweed, also release free oxygen. This is, how-
ever, a consequence of the photosynthetic organelles within the cell,
known as the chloroplasts. They were once free-living gram-positive
bacteria, similar to the living cyanobacteria, but which merged symbi-
otically with eukaryotic cells, ultimately to lose all independence.6 In
this way, billions of years ago, the atmosphere of the Earth, originally
free of oxygen, slowly became oxygenated, and as a result increasingly
corrosive. Thus, the planet started to rust.7 The principal episode of
oxygenation is usually identified as occurring at about 1.9 Ga ago.
This is inferred principally on sedimentary evidence, notably the dis-
appearance of mineral grains susceptible to oxidation, notably those
108 converging on the extreme
of pyrite and uranium oxide (uraninite), and the development of red

beds, whose colour signifies the onset of atmospheric oxygenation.
As oxygen levels climbed, so organisms had to adapt to this toxic
and reactive molecule. But the process of oxygenation started much
earlier. Just how geologically ancient photosynthesis is likely to be
is hinted at in the oldest known well-preserved sedimentary rocks.
These are from west Greenland (Chapter 5, p. 72) and as already noted
contain carbon. The carbon has, however, been heated by temperature
and pressure within the Earth and is now in the form of graphite,8
but it may still retain a signature of very ancient photosynthesis. To
understand how this is possible, one needs first to recall that photo-
synthesis is a process whereby carbon in the atmosphere, in the form
of carbon dioxide, is combined with water and ultimately transmuted
to sugars, using sunlight as the source of energy. The carbon itself oc-
curs in the form of two isotopes, one slightly heavier than the other:
carbon-13 and carbon-12.9 The processes of photosynthesis ‘prefer’ to
use the slightly lighter carbon-12, and hence the carbon stored in the
plant tissue is slightly enriched in the lighter isotope. Photosynthesis
thus imprints an isotopic signature, and provided that the carbon sur-
vives in the sedimentary record one can infer the ancient activity of
chlorophyll, even if no other evidence for photosynthesis survives.10
And although altered to graphite, the carbon in the most ancient sed-
iments, dated at about 3.8 Ga, looks as if it, too, might once have
passed through the photosynthetic process.
The chemical processes of photosynthesis are very complex. Ex-
actly how the Sun’s photons are captured to yield the necessary energy
within the photosynthetic ‘factory’, located within the chloroplasts of
the green leaf, is still not completely understood. Despite the apparent
miracle of sunlight pouring down on orchards and lagoons to produce
apples and sea grass, it is clear that the process of photosynthesis is far
from perfect. There are many chemical steps, and if an engineer had
been in charge of the design process, the accountants and administra-
tors would by now have been asking awkward questions. In particular,
the activity of one key enzyme (known as d-ribulose-1,5-bisphosphate
carboxylase, or RuBisCO for short) is severely hampered in the pres-
ence of oxygen, which ironically it also produces by catalysing another
chemical reaction that accordingly competes with the process of pho-
tosynthesis. The net result is that valuable energy is consumed in an
unavoidable metabolic process known as photorespiration.11 It is as if
universal chlorophyll? 109
the crew on a sinking ship are ordered into the hold to stem the flood,
and are all issued with sieves.
Nor is this the only problem the photosynthesizers face. An
essential prerequisite of the process, of course, is carbon dioxide. Sup-
pose, however, that the amount of this gas in the atmosphere were
to plummet drastically. This might be due to a number of natural
processes, such as episodes of massive mountain-building that expose
vast areas of rock and scree where the carbon dioxide can be ‘soaked’ up
by the enhanced rates of rock weathering. Plants would be faced with
an extremely serious problem, and it was one that actually started
about 15 million years ago. At this time the levels of atmospheric
carbon dioxide began to decline precipitously, in part probably be-
cause of the uplift of a series of huge mountain belts, most notably
the Himalayas. The plants’ solution was to modify the steps in the
photosynthetic process, leading to a transition from so-called C3 pho-
tosynthesis to a C4 mode (the numbers refer to the number of carbon
atoms in the first compound to be formed). The details of this C4 pho-
tosynthesis are considered in Chapter 10, but what is worth noting
here is that its evolution is rampantly convergent.
The processes of photosynthesis are, therefore, hedged in with
many constraints, but even so chlorophyll is a remarkable molecule
that effectively underpins the entire biosphere. It is thus rather sur-
prising to learn that not only does chlorophyll fall short in such mat-
ters as the effectiveness of its RuBisCO, but it also seems to suffer
from rather more general difficulties. George Wald, in particular, has
pointed out that if one compares the absorption spectra of the vari-
ous types of chlorophyll to the available visible light spectrum of our
Sun the match is, to put it mildly, disappointing (Fig. 6.1). Clearly
the chlorophyll has to absorb some sunlight or it simply would not
work. In addition, different types of chlorophyll vary somewhat in
their absorption spectra: chlorophyll d, for example, shows a rather
remarkable shift towards absorption of red light.12 Accordingly chloro-
phyll has some latitude in the wavelengths of light it is best adapted
to absorb, but it is all the more remarkable that the lion’s share of
the Sun’s radiant energy remains largely untapped. And from this ap-
parent anomaly Wald draws a very interesting inference. He argues
that, however desirable a ‘perfect’ chlorophyll might be, it is simply
not attainable: as the initial supplies of prebiotic ‘soup’ run out, so
life must move to the situation where it can synthesize its own food
figure 6.1 A comparison between the spectrum of the Sun and the
absorption spectra of three types of chlorophyll. Note how the spectra of
the chlorophylls differ, but none is able to intercept the maximum
output of the Sun. An ‘ideal’ chlorophyll would mirror the outline of the
Sun’s spectral profile, leading George Wald to suggest that not only is
terrestrial chlorophyll the best available, but anywhere else in the
Universe on a planet the same chlorophyll will occur. (Redrawn from
figure by Emai Kasai on p. 96 of G. Wald’s article ‘Life and light’,
Scientific American, vol. 201 (4), pp. 92–108. Copyright c 1959 by
Scientific American, Inc. All rights reserved.)
from the light of its sun. Accordingly the necessary machinery, in-
cluding chlorophyll, must evolve. On this basis Wald draws a bold
conclusion, remarking that ‘When that time comes, it seems to me
likely that the same factors that governed the exclusive choice of
the chlorophylls for photosynthesis on the Earth might prove equally
compelling elsewhere.’13
We have already seen that the chances of encountering habit-
able planets may be much lower than is generally supposed, but what
if we do locate an alien biosphere? Wald’s arguments suggest that
however many light years we may be from the Earth, if we discover a
planet and wander along its remote shores with their entanglements of
the wheels of life? 111
seaweeds or explore its immense forests, sure enough there will be the
same and all-too-familiar chlorophyll soaking up the light of an alien
star.14
the wheels of life?

Nor is chlorophyll the only way to look at alternatives in terms of
evolutionary possibilities. Another approach is to assess the diver-
sity of life and ask whether there is anything one might reasonably
expect to see, but seemingly has failed to evolve. The most popular
of these thought-experiments is to ask why no organism with wheels
has ever evolved. While this is certainly a biomechanically demanding
concept, it does not seem to be intrinsically impossible. The nearest
approach in nature to our familiar technology of bicycles and prams
seems to be the rotary mechanism associated with the base of the
whip-like process or flagellum that serves to propel certain bacteria
through a watery medium.15 This organic motor is built of a series of
proteins, and the arrangement of rings and sockets that holds the base
of the rotating flagellum is extraordinarily machine-like. The speeds
of rotation are astonishingly high: some bacterial flagella turn at more
than 1000 revolutions per second.16 Not surprisingly the complexity
of assembly17 has attracted the attention of so-called ‘intelligent de-
sign’ theorists, who argue that in such cases of irreducible complexity
(i.e. remove any component and the mechanism falls to bits) one must
conclude that the theory of evolution ultimately fails, so necessitat-
ing a teleology controlled by either a designer or Designer.18 While
we should not underestimate the difficulty in explaining how such a
flagellar motor might have evolved, everything else we know about
evolution indicates that the pathway to construction will involve the
twin processes of cobbling together and co-option, with at least some
of the proteins being recruited in quite surprising ways from some
other function elsewhere in the cell. This is not to deny that the ques-
tion as to how the flagellar motor was assembled is still unsolved, but
in principle its origin should ultimately be no more inscrutable than
explaining any other complex organic structure. Interestingly, this
gyrating bacterial flagellum is not the only type of molecular motor;
almost unbelievably even a part of a single molecule is able to rotate
freely.19 Even so, and fascinating as these molecular motors may be,
there seems to be no way in which they could simply be scaled up to
provide the equivalent of a wheeled giraffe, or whatever.
The only other way in which at least some animals can loco-
mote in this fashion seems to be to turn the entire body into a wheel.
Perhaps the most remarkable of such examples are certain shrimps
(stomatopods) that can execute up to 40 consecutive rolls,20 and a
pangolin21 on the island of Siberut, west of Sumatra, which has been
observed to roll itself up and then go crashing through the under-
growth of a steep slope.22 Other examples, such as cartwheeling
spiders and sea anemones,23 not to mention the so-called circumro-
tatory corals,24 are interesting in their own right, but scarcely qualify
as wheeled organisms. Organic wheels, therefore, seem to be an ex-
cluded possibility, and it is not difficult to see why. In a thoughtful
study Mike LaBarbera points out that, among other factors, any ani-
mal wheels of reasonable size would presuppose effectively flat and
continuous surfaces.25 Hence our need for roads as well as, more lo-
cally, ramps for wheelchairs. In the natural world as often as not, and
especially on sea floors, this means acres of mud and other soft, sticky,
substrates, ideal for getting bogged down. On this planet, and perhaps
everywhere, any wheeled organisms are going to be bipeds cycling to
the pub.26
fortean bladders
Is there any other neglected possibility? R. A. Fortey, for example, has
pondered whether any major ecological niche remains untenanted,
and suggests that one such niche might exist.27 Thus he writes that
‘High in the atmosphere there is a stream of air which transports in-
sects and spiders, like some plankton of the ether. Could an aerial
‘whale’ have evolved to harvest this stratospheric protein: a light,
flying animal with a wide feeding gape, an animal that could cast
a shadow across the sky?’28 Fortey is careful to note that there might
be good explanations why such an animal could not exist, but for
whatever reason he does not choose to address them.
In principle, such a Fortean bladder,29 named in honour of that
master of intangible aerial phenomena, Charles Fort (see also note 49),
should be biologically unproblematic.30 Perhaps the best analogy is
the swim-bladder found in many fish, a gas-filled organ that confers
buoyancy.31 The ever-present problem of leakage of gas is controlled
by deposition in the swim-bladder wall of guanine crystals32 that
evidently confer impermeability. In looking at the swim-bladder of
the conger eel Eric Denton remarked that removal of the silvery layer,
fortean bladders 113
containing the guanine, reduced the permeability by about one hun-

dred times.33 This arrangement is augmented by an ingenious net-
like arrangement of blood vessels (known as the rete) that provides a
counter-current system so that the gas dissolving into the blood as it
exits from the gas gland of the swim-bladder is transferred across the
rete and so returns to the bladder.34 Nor is a gas-filled bladder unique
to the fish; the females of the octopus Ocythoe possess an equiva-
lent structure.35 A somewhat similar arrangement also occurs in the
complex ‘jellyfish’ known as the siphonophores,36 most familiar per-
haps as the Portuguese man-of-war. Here, too, buoyancy is conferred
by a gas-filled bladder, with an associated gas gland that releases car-
bon monoxide.37 The versatility of these siphonophore bladders is
quite remarkable, given that one type, Thermopalia,38 secretes gas
despite living in a hydrothermal ridge habitat that lies at an ocean
depth of about 2.5 km, where the hydrostatic pressure is equivalent
to a crushing 260 kg per cm2 . Mention should also be made of the
gas-filled vesicles found in numerous species of bacteria, especially
aquatic cyanobacteria, where they evidently confer buoyancy.39
The gas employed in the fish swim-bladders is mostly a mixture
of oxygen and nitrogen, and is perfectly suitable for conferring buoy-
ancy. In principle, however, the biological generation of the lighter-
than-air gas hydrogen should be possible.40 Hydrogen, after all, is
extremely abundant, albeit combined chemically with oxygen in the
form of water. It is known, for example, that some microbes41 (and mi-
crobial mats42 ) can generate hydrogen as a side effect of the fixation of
nitrogen. More specifically, in trying to assess how realistic a Fortean
bladder might be it is worth considering some specialized structures
found in the cells of some microbes. These are a type of organelle43
known as the hydrogenosome which, as the name suggests, gener-
ates hydrogen.44 The process is connected with the generation of the
molecule ATP (which is involved with energy metabolism), and en-
tails the reduction of protons (i.e. hydrogen ion, H+ ) to form molecular
hydrogen. In most other circumstances involving ATP the electron ac-
ceptor is oxygen, but the hydrogenosomes are found in organisms liv-
ing in anoxic conditions, for example, stinking muds or the rumens
of cows, where oxygen is excluded. The key enzyme involved with
hydrogen production is, unsurprisingly, referred to as hydrogenase.45
Not only is it central to the operation of hydrogenosomes, but it
has also been found in organisms that are normally aerobic. In this
instance the enzyme evidently kicks in during times of oxygen

crisis.46
There seems, therefore, to be no biological reason why a Fortean
bladder could not employ hydrogen-producing microbes, recruited by
symbiosis, and store them in an area guarded against any influx of de-
structive oxygen. Such mass packaging of bacterial symbionts is by no
means unusual. One can think, for instance, of the deep-sea hydrother-
mal rift-worms (known as the vestimentiferans), which possess a spec-
tacular organ, the trophosome, that is packed with sulphide-oxidizing
bacteria. These bacteria are integral to the metabolism of this worm,
not least because it lacks a gut. This example is also of interest in
the context of Fortean bladders because the worms (like practically
all animals) require oxygen, that is, they are aerobic, yet they have
an ingenious metabolism that is able to transport the normally toxic
reduced sulphur to the symbionts without poisoning themselves.47
Another example, in a less extreme environment, is the so-called bac-
teriome, such as occurs in the aphids and other insects.48 In the aphids
and their relatives the bacteriome is a specialized organ, which as the
name suggests is packed with bacteria. Like the trophosome it is a
symbiotic association, but here the insect augments its diet with bac-
terially derived metabolites, because the food otherwise consists of
nutrient-poor plant sap. After considering these various contrivances
and the steps necessary to envisage an animal with a gas-filled blad-
der, rendered impermeable by a guanine lining and containing a pro-
tected hydrogenosome-packed organ pumping out hydrogen, we might
conclude that such a structure does not seem beyond the realms of
possibility.
There are, however, cogent biological reasons why we are no
more likely to see a Fortean bladder gliding over the Alps than a flotilla
of flying pigs.49 There seem to be two problems. First, as Steve Vogel50
has pointed out, there is a simple area-to-volume problem. The incip-
ient evolutionary stage of the Fortean bladder would almost certainly
be rather small, and here the restricted volume in comparison to a
relatively large surface area would make lift-off problematic, espe-
cially since the bladder wall would be very thin and prone to leak
the small hydrogen molecule. There is, moreover, another problem,
which is that any such organism that ever got afloat would soon starve
to death. While small animals may be swept high into the air, the idea
of some sort of ‘plankton of the ether’ is a fantasy because above the
a silken convergence 115
ground the concentration of organisms drops off very rapidly with in-
creasing height.51 This is, of course, why on land and in the air there is
no equivalent to the aquatic suspension feeders that strain the much
more viscous water for tiny particles; apart from birds such as swal-
lows, the nearest approach to terrestrial suspension feeders are the
web-spinning spiders.
a silken convergence
Spiders, of course, are celebrated for their ability to extrude liquid pro-
tein through the narrow nozzles of their spinnerets, and thereby draw
out a slender thread of silk.52 Although it is generally agreed that these
spinnerets are modified appendages the origins of the silk glands are
still rather mysterious.53 Even so, spider’s silk itself is probably most
familiar from the orb-webs, even though it is also employed in a num-
ber of other ways. The ogre-faced spiders, for example, capture their
prey with a gladiator-like net of sticky silk (see also Chapter 7, p. 156,
for a discussion of the huge eyes of these spiders). The mechanical
properties of spider silk, not least its strength, are quite remarkable.
It is also well known that the orb-weavers (the araneoids) can secrete
several types of silk,54 each with its own combination of physical prop-
erties, including mechanical properties (e.g. tensile strength, tough-
ness), and degree of stickiness, that make it suitable for different parts
of the web. These physical properties arise from the behaviour of the
silk proteins,55 known as fibroins, and the key role of a few amino
acids, notably alanine and glycine. The proteins are, of course, coded
for by specific genes,56 and some of the structural units in the fibroins
are remarkably conservative, apparently maintaining the same organi-
zation for more than 100 Ma. Clearly, therefore, functionality cannot
be jeopardized and this is a good example of what is called stabilizing
selection.57 It is a curious thought that an insect blundering into a spi-
der’s web is encountering a molecular predicament that is unchanged
since the time of the dinosaurs. Not only that, but, as Catherine Craig
and her colleagues remark, the evolution of these proteins has been
by ‘selection for functional properties specific to the ecological pur-
poses for which they are used.’58 These comments were made in the
context of discussing such properties as ultraviolet reflectance. This
determines how visible the silk is to a passing insect and thereby
appears to explain the dramatic evolutionary success of the spiders,
especially the familiar orb-web spinners. Craig et al. suggest that these
innovations have led to a 37-fold increase in their species numbers,

as well as invasion of new habitats.
Silk production is integral to a spider’s existence, and it might be
thought to be an invention unique to these remarkable animals. Not
a bit of it. Many other examples of silk production in a wide variety of
other arthropods59 are also known. These include, of course, the silk
moth whose cocoon is the source of raw silk for fabrics.60 This and the
other arthropod silks are all convergent on that of spiders.61 In some
ways the differences among these silks, including their extrusion from
a wide variety of secretory glands and their various functions (e.g. as
lifelines62 ) are as interesting as the similarities. Spiders alone produce
large and well-organized entrapment webs, but many insects use silk
to trap their prey. In New Zealand and Australia, for example, the larva
of a fungus-gnat (Arachnocampa) is luminescent (i.e. it is a type of
glow-worm) and so attracts its prey towards vertical threads of sticky
silk,63 a feature also found in some other flies.64 Silken nets are also
employed by larval insects in aquatic circumstances, in such groups
as the midges (chironomids)65 and caddis-flies (trichopterans).66 Typi-
cally these provide an extremely fine silk mesh, suitable for trapping
tiny food particles. Certain types of caddis-fly (e.g. Plectrocnemia),
however, construct nets that are not only, for the size of the larva,
surprisingly large,67 but also aim to trap prey in a spider-like ‘sit-and-
wait’ strategy.68 Interestingly, this mode of capture occurs where the
density of suspended food is too low for suspension feeding, thus echo-
ing the improbability of the Fortean bladders straining after gnats and
other such aerial ‘plankton’.
Other insects are also adept weavers, and some spin a silk whose
molecular organization (a β-parallel sheet) is closely parallel to that
of the spiders. Of all the examples of arthropod weaving, perhaps the
most remarkable is the ant Oecophylla.69 This, the weaver ant, de-
serves a mention, not only because of its intrinsic fascination, but
also because as a social species it provides a prelude to the complex
organizations seen in the other ants, notably fungus farming in attine
ants and carnivorous raids by army ants, that in their different ways
provide a series of insights into evolutionary convergence (Chapter 8).
In the weaver ants the nest is constructed by drawing together leaves
that are sufficiently flexible to provide a tent-like structure.70 While
the leaf is being held in position other workers fetch larvae, the heads
of which are equipped with silk glands. Gripped by the mandibles of
matrices and skeletons 117
the worker, the larva keeping itself rigid, a tiny patch of silk is laid
down on the leaf surface before the larva is carried across to the other
side of the leaf where the silk thread, spun out during its transport, is
reattached. Not only do ants construct tents; so also do some tropical
bats,71 although the latter do not of course employ silk.72 And the
purpose of these tents? Protection of a sort, but apparently for males
guarding their harems.
matrices and skeletons

So silken webs and tents full of expectant females may be the order of
the day, but wheeled animals steadily progressing across the plains of
Africa while admiring a fleet of Fortean bladders seem to be excluded:
it appears as if the possibilities of life have been quite thoroughly ex-
plored. That we can have some confidence in this conclusion can also
be gleaned from various attempts to identify all possible combina-
tions of a particular biological feature. In this instance the approach is
to define a matrix that provides a reasonable proxy for combinatorial
space, and thus allows one to see whether the same solution to an
evolutionary problem has been found repeatedly as against evolving
rarely, if not only once. Most interesting in this context are the com-
binations that ‘ought’ to exist but do not. A pioneering analysis in this
area was by the famous botanist George Ledyard Stebbins, who looked
at character combinations in flower structure.73 He found that only
about a third of the possible combinations were actually realized, but
of the unoccupied majority many were either structurally impossible
or evidently poorly adapted. Stebbins also noted that common solu-
tions were due to several factors, including ease of construction and
what might be called phylogenetic facilitation; that is, closely related
groups of flowering plants are predisposed to arrive at certain floral
arrangements. In the latter case it is important to stress that since
Stebbins published his paper in 1951 there have been very extensive
revisions in plant systematics, especially on the basis of molecular
data. It would be an interesting exercise to see in the light of this
new evidence how much of the matrix of flower structure is actually
occupied by convergent forms.
A similar approach to using a matrix was adopted more recently
by Roger Thomas and Wolfgang-Ernst Reif74 in an ambitious attempt
to encapsulate all types of animal skeleton, in a so-called ‘skeleton
space’ (Fig. 6.2). This ‘space’ was defined by such features as shape,
figure 6.2 A plot of ‘skeleton space’, a matrix that divides the skeleton
into seven major categories (and 21 variables), starting with whether
they are internal (as in us and other vertebrates) or external (as in
arthropods). This gives a total of 186 pair-combinations, of which a
handful are effectively impossible. Of the remainder, the majority are
either abundant or common, having evolved many times; in rare
instances evolution of such an arrangement usually has occurred at least
twice. (Redrawn from fig. 12 of Thomas and Reif (1993; citation is in
note 74), with the permission of the Society for the Study of Evolution
and the authors.)
matrices and skeletons 119
style of growth and assembly, and degree of interplay. The matrix

was defined by a total of 186 possible paired combinations, and sig-
nificantly it showed that practically no combination, other than the
structurally improbable, had not been invented at least once. Indeed,
in many cases a position on the matrix showed multiple occupancies:
in skeletons convergence is frequent. This leads to a general point, to
which I shall return later (Chapter 10) concerning the link between the
exploration of biological ‘space’ (however defined) and convergence,
to the effect that barring the physically impossible and adaptation-
ally compromised, it appears that as a general rule all evolutionary
possibilities in a given ‘space’ will inevitably be ‘discovered’.75
In the context of skeletons it is also worth remarking on some
further instances, both specific and general. An interesting example
concerns that remarkable biomineral, bone. This shows an immense
versatility, particularly in its so-called trabecular form in which thin
struts of bone (the trabeculae) orientate themselves according to the
prevailing stress field, thereby combining strength with lightness. The
precursor of bone, both embryologically, and probably in evolution, is
cartilage. In such groups as the sharks and rays cartilage continues to
serve as the principal skeleton. Yet clearly this material has design
limitations, as is evident from the jaw of the cow-nose ray where a
sort of ‘trabecular cartilage’ has developed. This is structurally and
functionally convergent on trabecular bone, presumably evolving in
response to the stresses imposed on the jaw as the creature grinds
its molluscan prey.76 Bone typifies the vertebrates, and while other
groups of animals possess many and wonderfully variable skeletal con-
figurations, it might be thought that bone is a unique biological inven-
tion. What could be more different from a vertebrate than a barnacle?
These are familiar on seashore rocks and are actually sessile crus-
taceans, that is, close relatives of the shrimp and lobster. Barnacles
encase themselves in a ‘box’ of calcareous plates, yet remarkably one
species (Ibla cumingi) has evolved phosphatic plates with a distinctive
and complex structure convergent on lamellate bone.77
Skeletons are usually thought of as tough and resistant, or at
least until the snail meets the thrush. Yet biologically the concept of
the skeleton can be cast much wider, to soft organs filled with fluid.
Probably the most familiar example is penile tumescence when en-
gorged with blood, a useful trick that has evolved several times.78
Moreover, hydrostatic skeletons are an integral arrangement both of
entire animals, such as the earthworm, and of specific organs, such as

the eye stalks of snails and legs of spiders. In each case the pressure of
the effectively incompressible fluid acts as an antagonist to the con-
traction of a muscular sheath. Hydrostatic skeletons are widespread
and in activities such as burrowing often play the key role.79 This is
not, however, the case in the various limbless vertebrates, such as the
snakes, which, as might be expected, rely on the contraction of longi-
tudinal muscles in association with the backbone. There is, however, a
very striking exception in the form of a group of limbless amphibians,
relatives of the frogs and newts, which are known as the caecilians.80
It has long been recognized that they are much more efficient burrow-
ers than other limbless vertebrates, and it now transpires that this is
because they have effectively transformed themselves into worms.81
By allowing the backbone to move independently of the skin, as well
as employing a system of helically arranged tendons, the overall style
of locomotion is strongly convergent on various invertebrate burrow-
ers: a vertebrate body plan has been remarkably transformed into a
hydrostatic skeleton.
play it again!
But what of counterfactuals, the replaying of history that exerts such
an eerie fascination? Just consider the ‘what ifs’ of European history: a
successful Spanish armada? A Catholic Europe, with science emerging
in a society that presupposed that the Universe was both rational and
had a moral architecture?82 A Charles the First who engaged and de-
feated the Scots in 1639? No English Civil War? No Lord Protector? An
eighteenth century without the brilliant chemist Antoine Lavoisier?
No reform of the French gunpowder industry, a vital component
of the armament shipments to the American colonies in rebellion
against the British Crown?83 Napoleon Bonaparte triumphant at
Waterloo and Europe’s first police state fully operational by 1820,
rather than three more attempts (1870, 1914, 1940) – and as for the
the fourth? German hegemony and the Euro by 1916 as the British
Cabinet avoids conflict by persuading the Germans not to invade
‘gallant, little Belgium’? England invaded by the Nazis in 1940? The
Bloomsbury group taking their cyanide, while in Cambridge a very se-
nior figure glances at King’s College chapel and then carefully adjusts
his swastika armband?84 Looking for trends in history is now deeply
unfashionable, yet so far as the biological equivalent is concerned
play it again! 121
we find that here at least trends have a reality, and are not only
self-evident in the history of life but are even testable in the
laboratory.
The organism of choice is, unsurprisingly, a bacterium, specif-
ically the microbiologist’s equivalent of the fruit-fly and the mouse,
that is, Escherichia coli, more usually referred to as E. coli. This has
some substantial advantages for the study of evolution.85 The genet-
ics of this bacterium are well understood, and population growth is
very fast: in favourable conditions the cells divide as frequently as
once every 20 minutes. Moreover, even as the descendant lineages are
allowed to mutate along a myriad of evolutionary trajectories the an-
cestral ‘race’ can be preserved by the simple expedient of freezing, to
be reawoken when a comparison needs to be made between the start-
ing point and a population that might have been evolving for literally
thousands of generations.
And it is on this basis that Richard Lenski, Michael Travisano,
and their colleagues have undertaken a series of elegant experiments
to see how evolution replays itself. In one experiment these inves-
tigators set up 12 separate populations of E. coli.86 Each was then
allowed to diversify – effectively to evolve – independently for
2000 generations. At this stage each of these populations was di-
vided into three, and the resultant 36 populations were then left to
evolve yet further, for another 1000 generations. At this stage, 36 dif-
ferent evolutionary trajectories have been followed. So far, so good,
but now these populations are introduced to a novel substrate, the
sugar maltose. E. coli much prefers glucose, and maltose is biochem-
ically challenging. For each population a new ‘history’, the struggle
with maltose, will now unfold. The investigators specifically wanted
to see what would happen with the bacteria living on the maltose
substrate after 1000 generations had elapsed. The real objective of the
experiment was to weigh up statistically the relative importance of
three factors: those of chance (life’s grand lottery), adaptation (with the
implications of trends and optimization), and history (play it again!).
Which is the most important? If chance played the main role then the
36 populations would be expected to evolve in all sorts of different
directions, each caught in a contingent net of circumstances. Con-
versely, if adaptation by selection was important then the populations
should converge on one (or more) stable end-points. And how on earth
can we rerun history multiple times? That is the point of arranging
the 36 replicates, each of which, having been allowed previously to

diversify, now represents a different starting point when transferred
to the ‘maltose world’. Clearly enough at the beginning of the rerun,
that is, transfer to the maltose substrate, history has to be highly sig-
nificant, simply because each population represents a different start-
ing condition. After the 1000 generations of maltose-based existence
the historical component was strongly reduced, and correspondingly
there was a strong degree of convergence; that is, different routes led
to the same end-point (Fig. 6.3). Such convergence implies, of course,
a corresponding increase in adaptation as E. coli ‘learns’ to get to grips
with maltose. And the role of chance? In this evolutionary experi-
ment, at least, it was of negligible importance. The simple conclusion
is that history does not go away, but is swamped by the effects of con-
vergence. There are many other experiments involving the rerunning
of evolutionary histories of this bacterium. As noted, E. coli depends
on sugars, but what happens if independent lineages are sometimes
deprived of food and then glutted? To cope with this recurrent ‘feast
and famine’ the separate lineages arrived at different genetic solu-
tions, but even so the overall direction of evolutionary change was
again strongly parallel.87 And the tape of evolution can be rerun with
respect not only to food substrates of E. coli but also to adaptation
to changing temperatures: once again chance and history are not the
main players.88
There are certainly some important qualifications to these ele-
gant experiments,89 but it is also important to emphasize that simply
because convergence overrides history this does not mean that evo-
lution comes to a dead end.90 Even with a similar end result, that
is convergence, the different populations of bacteria remain genomi-
cally very diverse, and also, despite adapting to the new substrate of
maltose, the genomes do not ‘fossilize’ but remain very dynamic.91
Hence stabilization in the form of convergence need not imply the
cessation of evolution. And the roles of chance and history are not al-
ways negligible; in these experiments cell size was determined mainly
by these factors.92 Moreover, in further experiments,93 again compar-
ing evolution on glucose and maltose substrates, the populations left
to replicate on their ‘home’ substrates rapidly adapted in much the
same way to make the best of the food source, but when they were
swapped to the other substrate their behaviour in either case was very
different.94 Travisano concluded that ‘Apparently small differences in
play it again! 123
A
1.4
Derived fitness in maltose
1.2
A
A
1.0
0.8
0.6
0.6 0.8 1.0 1.2
Ancestral fitness in maltose
B
0.30
Derived fitness in maltose
1000 generations later
0.20 Before
0.10
0.00
Ad Ch Hi
ap an sto
tat ce ry
ion
figure 6.3 Rerunning the history of evolution, with the bacterium
Escherichia coli. The upper panel shows the relative increase of fitness
of 12 replicate populations (A→L) of E. coli, each subdivided three
times, when placed on a maltose substrate. The dotted line would be the
case of no increase in fitness, and clearly all 36 sub-populations have
increased their relative fitness, even those that ancestrally had a high
fitness to maltose. The lower panel disentangles the relative roles of
adaptation, chance and history during this process. Note chance has a
very limited role, that of history declines substantially, while that of
adaptation increases markedly. (Reprinted (excerpted) with permission
from M. Travisano, J. A. Mongold, A. F. Bennett and R. E. Lenski,
Science, vol. 267, pp. 87–90. Fig. 2A and 2B. Copyright 1995. American
Association for the Advancement of Science.)
the selective environment can greatly alter patterns of adaptation and

divergence.’95 These experiments are important, both because they
can quantify to some extent the role of history, and also because de-
spite the apparent simplicity of the system they reveal in their various
ways a number of unexpected complexities. Particular responses in a
given lineage of E. coli, when faced with a metabolic or environmental
novelty, are often not predictable, presumably because of our relative
ignorance of either physiological or genetic factors. Even so, these ex-
periments also demonstrate recurrent examples of parallelism: despite
different starting points, different lineages often converge on particu-
lar destinations.
A particular strength of the E. coli experiments was that the
quantitative data could be handled statistically by analysis of the vari-
ances. Nevertheless, it is natural to ask whether what applies to a flask
full of E. coli is of any relevance to the wider world and the likelihood
of evolutionary convergence. Amitabh Joshi reminds us of some of the
principal problems that might accompany any attempt to use bacteria
as a guide as to how the world works, not least in the case of animal
evolution. The main points he makes are: (a) the only observations
made are at the beginning and end of the experiment, e.g. 1000 gener-
ations of adaptation, so ‘making it difficult to study the time course
of changes in the relative contributions of adaptation, history and
chance’; (b) being asexual, bacteria evolve largely by mutation rather
than sexual exchange of genes; (c) the relative importance of ‘ancestry
and past selection history’ are not easy to separate; and (d) bacterial
populations are usually orders of magnitude larger than most eukary-
otes, so what is termed ‘genetic drift’ may be less likely.96 So what
happens when we attempt an analogous experiment, rerunning the
tape of evolution on laboratory animals rather than E. coli? Almost
inevitably the animal in question has to be the fruit-fly (Drosophila).
In the context of one such set of experiments, looking at the adap-
tation of larval feeding rates according to the degree of crowding,97
Joshi’s concluding remarks are telling: ‘Thus, the footprints of his-
tory, at least in sexually reproducing species, would, metaphorically
speaking, appear to be no more than transient impressions on genomic
dust’.98
In other cases, however, the footprints of history may not be so
easily erased, especially if the clock of evolution runs backwards. This
was the aim of a series of experiments, again using Drosophila, to see
play it again! 125
how easy it might be to reverse evolution by returning the fruit-flies to

the ancestral environment from which they had diverged by artificial
selection, e.g. for timing of reproduction, over many generations.99
In most cases there was either a complete or partial convergence to
the ancestral state: evolution had run backwards. In the relatively
few other cases a new evolutionary trajectory was followed or, rarely,
evolution effectively stood still (no change). Here, therefore, there are
various histories, with particular genetic architectures predisposing
some (but not all) lineages to convergence. Even so, when evolutionary
reversal to the ancestral state occurs it is patently due to ‘selection
for adaptation to the ancestral environment’.100 Given the ubiquity of
selection so we can expect a ubiquity of convergence.
Adaptation is ubiquitous and convergence is an inevitable con-
sequence of the constraints on life, although placing convergences
into any sort of quantitative framework is still in its infancy. The
work on E. coli and Drosophila certainly shows one way forward, and
there are some interesting parallels to these investigations. One such,
albeit on a very different scale, is Dolph Schluter’s investigation of
different communities of finches.101 By studying these birds and the
communities102 they formed he was able to show that in terms of
variance there were significant convergences in various traits, such
as the number of species and mean size. Only in the case of mean
shape, which was controlled by phylogenetic factors, could the role
of history be identified as important. Much the same story emerges
elsewhere. One particularly striking example of convergence in which
the historical contingencies are pushed to one side concerns a group
of Caribbean lizards known as the anolids. Jonathan Losos concluded
that their convergence is identifiable, not only in terms of striking
similarities between various species representing specific anolid eco-
morphs on different islands, but also in overall community profile.103
One important inference of his work is that the ecological structure
has evolved by competitive interactions that determine how the habi-
tats came to be occupied. This analysis of anolid convergence is jus-
tifiably well known, but it is by no means the only such example:
among the mammals the rodents are similarly instructive.104
The investigation of these vertebrate communities and their
convergences is a sort of natural experiment. Even if it were thought
desirable to arrange a rerun this would probably require thousands
of years, and possibly much longer. That, of course, is a particular
advantage of using a rapidly evolving and relatively simple organism

like the bacterium E. coli. A number of similar experiments have
already been undertaken using other microorganisms,105 and with ad-
vances in technology there are new possibilities for experimental in-
vestigation at even the molecular level. This topic will be discussed
subsequently (Chapter 10), but here it is worth noting one particular
piece of work, by Amelie Karlstrom and her colleagues, because like
the E. coli story it is based on a rerun of ‘history’.106 Karlstrom et al.
noted that in the field of biochemistry there are often multiple solu-
tions to a given problem, but even so particular motifs do reoccur. This
they investigated as a kind of repeated evolutionary trial, by design-
ing a particular protein (a catalytic antibody aldolase). This showed a
similarity in efficiency and mechanism to the naturally formed equiv-
alent, but in the two trials the investigators found that a particular
amino acid (lysine) at a key site (number 93 in the chain) was a recur-
rent necessity if the enzyme was to function properly.
attacking convergence
Such multiple evolutionary trials suggest that life is far from some sort
of contingent muddle, yet the criticism might be levelled that they
are either rather general, such as the example of the matrix of skeleton
space, or very specific, good for an anecdote over a pint of beer in the
Green Dragon but hardly of any real evolutionary significance. Some
examples do indeed seem to be little more than curiosities, such as the
Jurassic crustacean that is reminiscent of the trilobites,107 or a fossil
bryozoan that adopts a form of spiralling fans that as a bryozoan design
had disappeared more than 200 million years earlier.108 Occasionally
these reappearances seem more like zombies.109 Such is the case in
certain fossil snails found in sediments of Triassic age, that is, the
geological interval that followed the great end-Permian catastrophe.
They are remarkably similar to the pre-catastrophe snails, but it is not
clear whether this is because they are in fact survivors of the debacle
(Lazarus taxa) or actually represent new species, but evolving along
the same old tracks (Elvis taxa).110
What convergence is, and what it might imply, I hope will be-
come clearer as this book unfolds. A few preliminary remarks, how-
ever, may be apposite. First and most obviously, biological conver-
gence can mean many things and operate at many levels. As I shall
argue, however, there are some common implications, despite the
attacking convergence 127
apparently bewildering range of examples. Second, the branching na-

ture of evolution means that although the identification of conver-
gence presupposes a known starting point, as often as not the reality
is an almost infinite regress of investigation. Nevertheless, I believe
the topic of convergence to be important for two main reasons. One is
widely acknowledged, if as often subject to procrustean procedures of
accommodation. It concerns phylogeny, with the obvious circularity
of two questions: do we trust our phylogeny and thereby define conver-
gence (which everyone does), or do we trust our characters to be con-
vergent (for whatever reason) and define our phylogeny? As phylogeny
depends on characters, the two questions are inseparable. In reality
there are, of course, independent data sets and sometimes an histori-
cal record, so the enterprise is not crippled. Even so, no phylogeny is
free of its convergences, and it is often the case that a biologist believes
a phylogeny because in his or her view certain convergences would
be too incredible to be true. This is mostly correct, but not always.
The second reason for the importance of convergence has certainly
been touched upon by others, but, I suggest, is not only neglected but
might provide the nucleus of an interesting research programme (see
Chapter 10, pp. 308–310). In brief, convergence offers a metaphor as
to how evolution navigates the combinatorial immensities of biolog-
ical ‘hyperspace’, a topic already touched upon in Chapter 1. Conver-
gence occurs because of ‘islands’ of stability, analogous to ‘attractors’
in chaos theory. This view of life, however, begs two interesting ques-
tions. First, given the immense dimensions of this ‘hyperspace’, how
is it that navigation is ever successfully achieved? Variants of this
question are, of course, repeatedly proposed by opponents of evolu-
tion. Convergence gives us a clue as to how the metaphorical Easter
Island is located, but just as this Pacific Island is surrounded by oceanic
wastes, so too perhaps are the ‘islands’ of biological habitability. I sus-
pect that not only will the bulk of biological ‘space’ never be occupied,
but it never can be either.
It is therefore my strong belief that the topic of convergence is
very far from being just another example of the frustrating impreci-
sions of biology where apparent rules and laws melt into exceptions
and counter-examples. To echo the idea that convergence might give
some surprising insights into a deeper structure of biology, there is
some evidence that might at first seem to be simply anecdotal. Dur-
ing my time in the libraries I have been particularly struck by the
adjectives that accompany descriptions of evolutionary convergence.

Words like ‘remarkable’, ‘striking’, ‘extraordinary’, or even ‘astonish-
ing’ and ‘uncanny’ are commonplace. It is well appreciated that sel-
dom are the similarities precise,111 and this in itself is as concrete a
piece of evidence for the reality of evolution as can be provided. Even
so, the frequency of adjectival surprise associated with descriptions of
convergence suggests to me that there is almost a feeling of unease in
these similarities. Indeed, I strongly suspect that some of these biol-
ogists sense the ghost of teleology looking over their shoulders. Nor
is this an unworthy sentiment. The eeriness of convergence is central
to how evolution navigates across the combinatorial immensities of
biological ‘hyperspace’. And when we look at some of these examples
of convergence, it is not difficult to see why.
Take, for example, the remarkably close correspondence be-
tween the first walking leg of the praying mantis and another insect,
known as Mantispa.112 As is self-evident, both are modified from a
generalized insect leg for grasping (Figure 6.4), yet they belong to dis-
tantly related groups within the insects, respectively the mantids and
neuropterans.113 The similarity of the raptorial appendage is striking
enough, yet there is evidence that such an arrangement evolved in-
dependently a third time, in another group of insects, the possibly
unfamiliar rhachiberothidids.114 This case exemplifies a more gen-
eral point because some investigators had earlier allied these insects
with the mantispids, whereas a more extensive phylogenetic analysis
points to convergence. Each insect therefore seems to have arrived at
effectively the same raptorial solution. Note also how in the mantids
and mantispids the similarities extend to the large eye and prehensile
‘neck’. Incidentally the power, rapidity, and accuracy of the predatory
strike, at least in the praying mantids, is quite remarkable.115 The
striking action takes two-tenths of a second, and hits home 90% of
the time. A standard method of releasing a sudden surge of power is
by the elastic release of stored energy, a feature found convergently in
many animals and also employed in the catapult. Despite its advan-
tages the mantids do not employ such a system, presumably because
of the delays imposed with reloading the elastic. Their remarkable
striking ability includes a muscle whose contraction rate equals that
of the fastest mammalian muscle.
While on the topic of insect legs it is also worth mentioning that
pad-like attachment structures, of which there are two alternative
figure 6.4 A remarkable case of convergent evolution, between the

raptorial fore-limb of the praying mantis and the neuropteran Mantispa,
both arriving at the same solution independently from the generalized
insect leg. (Redrawn from figs. 1–3 of Ulrich (1965; citation is in
note 112) with the permission of the author, Natur und Museum and
Senckenbergischen Naturforschenden Gesellschaft.)
designs, are also convergent within the flying insects.116 Insects are
a rich source of insights into many other convergences, some of
which (e.g. compound eyes (Chapter 7), eusociality (Chapter 8), hal-
teres (Chapter 7), silk (see above, p. 115), and trachea (Chapter 10) are
mentioned elsewhere. So, too, are the related crustaceans. In the im-
mediate context of attack and defence the repeated evolution of the
pectinate claw is a striking example.117 Yet more intriguing is the
repeated emergence of a crab-like form118 from among the decapod
crustaceans, otherwise familiar in the form of the lobster. This exam-
ple is of particular interest because, in contrast to many, if not most,
examples of convergence where the arrangement has a clear function-
ality in response to a particular adaptive challenge, in the recurrent
emergence (five times) of a crab-like form there is no apparent associ-
ation with a specific mode of life or environment.119
Let me give you another example, again from the world of at-
tack and mayhem. This is in the form of the independent evolution
of dagger-like canines in both placental cats (the sabre-tooth felids)120
and a group of South American marsupials known as the thylacos-
milids (Figure 6.5).121 In fact, the evidence suggests that even within
the placental felids the sabre-tooth habit evolved several times.122 The
dinosaur enthusiast Bob Bakker has also suggested that an analogy to
these sabre-toothed mammals can be found in the allosaurids of the
Jurassic.123 Although as a group the marsupials, best known in the
guise of the kangaroo and wombat, tend to be regarded in some gen-
eralized sense as inferior to the placentals, this is too simplistic.124
So, too, the rich, but now largely extinct, diversity of South American
marsupials is widely regarded as having been competitively inferior
to the onslaught of the placental mammals that surged south when
the linkage to North America was secured via the newly emergent
Panamanian isthmus several million years ago.125 In fact, in the mar-
supial thylacosmilids the sabre showed a number of what appear to
be design advantages when compared with the placental equivalents,
including the possession of a protective flange (the placental sabre-
tooth known as Barbourofelis shows the nearest equivalent), a self-
sharpening mechanism (presumed to act against some sort of horny
pad), and a deeper insertion into the skull that presumably afforded
a more secure housing for the canine. It need hardly be stressed that
despite this manifest convergence neither group escapes its imprint
of phylogenetic history, which is marked, for example, in the spe-
cific structure of the teeth.126 It is also worth remarking that the
figure 6.5 Convergence in the sabre-tooth, between the marsupial
thylacosmilid (upper) and placental cat (lower). (Copyright Marlene Hill
Donnelly (the Field Museum, Chicago), with permission, reproduced
from fig. 10 of L. G. Marshall’s The great American interchange – an
invasion induced crisis for South American mammals (pp. 133–229), in
Biotic crises in ecological and evolutionary time, edited by M. H.
Nitecki (Academic Press, Orlando, Florida).)
131
development of these massive canines, which it appears were more

probably used to shear through the flesh of the prey rather than to
engage in stiletto-like activities, should not be taken to imply a uni-
formity of hunting adaptations.127
This instance of sabre-teeth is just one of the many convergences
between marsupial and placental mammals. Textbooks tend to em-
phasize the various similarities between the marsupials of Australia
and placentals living in other regions of the world. John Kirsch
(note 124) also remarks on the more general convergences between
placentals and marsupials, and rightly remarks that such examples
are only of real interest if the resemblance turns out to be more than
superficial. In particular he draws attention to interesting examples
of convergence in the brain structure of marsupials and placentals,
notably those involving sensory input through stereoscopic vision
and whiskers, while John Johnson draws attention to striking con-
vergences in at least three regional brain specializations,128 examples
that are ultimately of considerable significance in the context of the
rise of intelligences (Chapter 9).
It is also worth mentioning that a number of striking conver-
gences can be found within the placental mammals themselves,129
such as between the rodents and the group known as the artio-
dactyls (hippos, cows, and so forth), as well as among the artiodactyls
themselves.130 So, too, broader comparisons between the adaptive
radiations of placental mammals that originated in two broad re-
gions, defined as Laurasia (Asia, Europe, and North America) and
Africa, define numerous similarities. Thus, in the component super-
groups, referred to respectively as the Laurasiatheria and Afrotheria,
the many parallels include the emergence of ungulate-like, aquatic,
and semi-aquatic (‘otter’), insectivore and burrowing, and even ant-
eater (note 21) forms.131 Yet more intriguing examples of convergence
in mammals come from comparisons between the evolutionary path-
ways of ungulates and the predators that pursue them. The palaeon-
tologist Bob Bakker132 notes that in six separate evolutionary lines a
whole series of anatomical features (e.g. reduction of side toes, elon-
gation of long bones (metapodials) of paws, but shortening of fingers
(phalanges)) follow ‘the same morphological pathway’. He concludes
that ‘this striking case of iterative parallelism and convergence . . . is
a powerful argument that observed long-term changes in the fossil
record are the result of directional natural selection, not random walk
through genetic drift’.133
figure 6.6 Convergence in

the pike morphology, i.e.
fusiform, sit-and-wait/stealth
fish-hunters, arrived at
independently (top to bottom)
by the classic pike (Esox
lucius, Esocidae: Alaska),
Belonesox belizanus
(Poecillidae: Central America),
Gobiomorus dormitor
(Eleotridae: Central America),
Acestrorhynchus microlepis
(Characidae: South America),
Hoplias malabaricus
(Erythrinidae: South America),
and Hepsetus odoe
(Hepsetidae: Africa). (Redrawn
from fig. 11D of Winemiller
(1991; citation is in note 134)
with permission of the author
and the Ecological Society of
America.)
In their different ways mantids and sabre-tooths are power-

ful examples of convergence in the context of overpowering prey.
Hunting style has, moreover, wider fields of similarity. In an assess-
ment of diversification among freshwater fish Kirk Winemiller drew
attention to recurrent patterns of what he called ecomorphological
convergences.134 One example is the repeated evolution of an eel-like
morphology from phylogenetically divergent sources, as seen in the
North American brook lamprey, neotropical swamp eels, and African
spiny eel. In the context of hunting, and especially lunging preda-
tors, there are again striking convergences on different continents,
this time towards a pike-like morphology (Fig. 6.6) and again from
ancestors that are only distantly related to each other. Given these
convergences among the fish, it is not surprising to see other exam-
ples of recurrent evolution. These include similar patterns of teeth,
as in sea-breams,135 or related trophic specializations within the en-
demic cichlids of central Africa.136 Fish, therefore, in their various
ways are a rich source of insight into evolutionary convergence, but
as we shall see (Chapter 7) there are yet more remarkable examples,
such as with the generation of electrical fields and endothermy, that
in the wider scheme of things are arguably even more instructive.
figure 6.7 Convergence in desert plants, specifically the stems of the

New World (Sonoran Desert) cactus Peniocereus striatus (left) and
African (Kenya) spurge Euphorbia cryptospinosa (right). Scale bar is
1 mm. (Reproduced from fig. 3b, g of Felger and Henrickson (1997;
citation is in note 137), with permission of the authors and Cactus and
Succulent Society of America.)
convergence: on the ground, above the

ground, under the ground
Convergence is pervasive, and moving from the animals let us take
a particularly striking example from the plants. Consider two desert
plants, one a cactus (Peniocereus striatus) that grows in the Sonoran
Desert of Mexico, the other a type of spurge (Euphorbia cryptospinosa)
from east Africa. Both are rather unprepossessing plants, straggly,
and to the uninitiated, which the plant presumably hopes includes
ravenous herbivores, look as if they are dead. Not only are the
plants quite similar in general habit, but the stems are astonish-
ingly convergent,137 with a characteristic cross-section that shows flat
ribs and intervening recesses (Fig. 6.7). This arrangement represents a
clever adaptation, because in times of severe drought and water stress
the ribs fold tightly together, thereby shielding the furrows where
the pores (stomata) for gas exchange are located, but from which wa-
ter can also readily escape. The interiors of the stems are also very
similar, with the interior cells storing water and a central pith rich
in food (starch). The similarities between these plants even extend to
the reddish pigment, located just beneath the surface, which confers
the moribund appearance to the plant. Of course, there are differences
such as the details of the stomata, which in the cactus can be sealed
convergence: on, above, under the ground 135
with some sort of water-soluble substance, whereas in the spurge there

are short spines that, it is hypothesized, restrict air circulation and
thus water loss.
Cacti and spurges are only distantly related, and their common
convergence is because of the rigours of living in an arid environment,
although the degree of similarity is, to put it mildly, noteworthy. To
a considerable extent evolutionary convergences have been placed in
the context of adaptation to extreme environments. Not surprisingly,
this specific and striking correspondence between the stems of a cac-
tus and a spurge is only one of a whole swath of similarities that
almost inevitably unite plants living in conditions of extreme arid-
ity, the so-called xerophytes.138 Similar remarks apply to the various
Mediterranean-style floras in areas as remote from the Mediterranean
itself as West Australia, South Africa, and Chile,139 and extend to
plant–animal interactions such as seed dispersal by ants.140 The flo-
ras are not, it needs to be stressed, identical and the various differences
also have evolutionary significance. Even so, in a survey of how plants
function from the tropics to the tundra it is clear that convergence,
for example in leaf structure,141 is a pervasive phenomenon.142
Defying gravity and exposed to wind, not to mention such loads
as a fruit crop or a passing arboreal animal, it is not surprising that
the anatomy of plants reveals many structural constraints, nor is
it odd that the number of solutions to a particular biological prob-
lem may be severely limited. One such example is the convergence
seen among the girder-like structures, known as petioles, which sup-
port leaf systems.143 In itself this is to be expected: buildings are not
made of blotting paper and neither are plants. There are, however,
more specific organizations seen in plants that would seem to be
unique. What could be a better example of evolutionary uniqueness,
of the sheer quirkiness of life, than a flower? Other worlds we may
presume will have their equivalents to plants. Most probably they will
build their cell walls with a carbohydrate similar to cellulose, and they
will almost certainly power their photosynthesis by using chlorophyll
(p. 110). So, too, in form they will range from ground-creeping herbs to
lianas and trees.144 All these are rather generalized predictions, based
either on the universal nature of biochemistry (see p. 77) or reflecting
the broad constraints of life.
On the other hand the specifics, such as flowers, let alone any-
thing remotely similar to say a tulip or a daisy, are quite out of the
question. They are surely a contingent happenstance of life on Earth,

not life on Threga IX. From a historical perspective this must in one
sense be true given that the flowering plants (the angiosperms) evi-
dently evolved from a single ancestor, probably at some time in the
Jurassic. Matters may, however, be rather more complicated. Whether
the smiling biped on Threga IX offers us a tulip is not really relevant;
remember we are in pursuit of general biological properties. Thus, in
the case of the flowers, they display several specific characters. Most
obvious, and appreciated, are the two whorls of leaves modified to
form the often brightly coloured petals and the associated bract-like
sepals. The reproductive arrangement follows the well-trodden path
of male (pollen from the stamen) goes to female (via the stigma and
ultimately the contained ovule), the happy conclusion of which is the
fertilized seed.145 To be sure, the structure of the flower is specialized,
sometimes remarkably so as in the orchids, but the basic principle
of plant sex (pollen meets ovule) is from an evolutionary perspec-
tive much more ancient than the flowering plants themselves. Yet no
other group of plants has a reproductive organ exactly like those of
the flowering plants. Not only is the flower highly distinctive, but its
evolutionary origins are still rather mysterious: seldom can a review
paper avoid parroting Darwin’s remark to the effect that the origin of
the flowering plants remains ‘an abominable mystery’.146
Yet, as is well known to botanists, there exists an otherwise
rather obscure assemblage of plants known as the Gnetales.147 They
in turn belong to a larger group, known as the gymnosperms, a group
that is likely to be more familiar in the form of the pine tree and other
conifers. As with other gymnosperms, the reproductive organs of the
Gnetales are located on cones, but in this case they are remarkably
flower-like.148 Even more intriguing is a rather strange phenomenon
known as double fertilization. In both the flowering plants and the
Gnetales the pollen grain carries two nuclei, which having landed
on the female stigma are delivered to the ovule by the agency of the
pollen tube. One of the male nuclei has, of course, the heroic task
of fertilization, which leads to the formation of the seed. The second
nucleus, however, is not the runner-up because it fuses with another
female nucleus. Hence the term double fertilization. In the case of
the Gnetales naturally enough this second fertilization provides an
extra, so-called supernumerary, embryro.149 In the flowering plants,
however, there is a further twist to the story because typically yet
another female nucleus joins in to produce a triploid cell (i.e. three

haploid sets of chromosomes combine, one set from each of the three
nuclei). This cell, equivalent to the supernumerary embryo, then di-
vides to form a mass of tissue known as the endosperm. Triple sex is
not just a baroque ornamentation: the endosperm has a specific func-
tion, that is, to provide nourishment to the adjacent and growing seed.
This is widely regarded as one of the key innovations to explain the
overwhelming success of the flowering plants, whose evolutionary ra-
diation has many interesting parallels to the Cambrian ‘explosion’ of
marine animals.
So distinctive is this arrangement of double fertilization that,
when it is taken together with the flower-like structures of the
Gnetales, a strong supposition has arisen that the origins of the flow-
ering plants are to be found in a gnetalean-like group that flourished
perhaps about 150 Ma ago.150 Nor are flowers and double fertiliza-
tion the only similarities that exist between the two groups. Among
the long list of features in common between the flowering plants and
Gnetales, particular mention has been made of the similarities in ves-
sels (that is the xylem, where the water-conducting tubes are open
rather than consisting of a series of discrete cells (tracheids))151 used
to transport water through the plant, and in the genus Gnetum the
venation of the leaf. Taken together, these similarities have been used
to support a close evolutionary connection between the two groups
and thereby explain the unique origin of flowers.
Yet all this turns out to be convergent. Despite some early re-
search from molecular biology supporting a link between flowering
plants and Gnetales, more extensive analyses both decisively reject
this relationship and propose that Gnetales is surprisingly close to
the conifers.152 So on this basis it is argued that the cardinal features
of flower-like structures and double fertilization153 have been arrived
at independently, as indeed have a number of other characters.154 So,
too, other supposed similarities, such as the structure of the xylem,
are now shown to be exaggerated.155 All this, of course, presupposes
that the phylogenetic conclusions are correct. It needs to be pointed
out that there are a number of extinct plants, notably a group of
interest to the herbivorous dinosaurs and known as the bennetti-
taleans, which also have some flowering plant-like characteristics.
Accordingly, given current phylogenetic uncertainties, it remains pos-
sible that some features actually evolved earlier in the history of
plants and so were respectively inherited by both the Gnetales and

the flowering plants.156 Most of the latest research (note 152), how-
ever, continues to place the Gnetales very close to the pine trees
and their relatives,157 so making the argument for convergence the
more likely. That such features as specialized sex organs (a.k.a. flow-
ers) open to insect pollination,158 not to mention the nutrition of the
embryo,159 should emerge in more than one lineage of advanced plants
is surely not that surprising.160 Perhaps flowers are an evolutionary
inevitability; and, moreover, even here the options are not unlimited.
Once again, convergence of floral anatomies is widespread,161 reflect-
ing such constraints as methods of pollination by both insects and
birds.162 So we might speculate that not only does Threga IX have
plants and flowers, but perhaps something not so different from a bird
pollinating a flower.
Birds themselves provide further compelling examples of evolu-
tionary convergence. There are, for example, repeated trends: towards
dark plumage in tropical seabirds;163 in the many independent ex-
amples of long-distance migrants the development of more pointed
wings164 as well as parallel changes in the skeletal structure;165 a
cluster of characters (for example, in wing and leg structure) that con-
verge in at least some grain- and insect-eating birds living at very
high altitudes;166 convergences in the aerially feeding insectivorous
birds167 and nectar feeders;168 and a striking convergence between
certain aquatic birds, specifically the grebe and the loon.169 Indeed,
birds as a whole are a rich source of insights into the prevalence of
evolutionary convergence,170 as well as having some striking sim-
ilarities with other groups. A well-known example is the parallels
between humming birds and the sphinx moths, both hovering above
flowers as they draw out the nectar.171 As we shall see in Chapter 8,
the birds are even more instructive in a wider context, as when we
encounter that ‘honorary mammal’ the kiwi, consider the physiology
of warm-bloodedness, or compare their powers of vocalization: once
again striking convergences emerge.
There are, moreover, other sorts of convergence that are equally
intriguing. Birds form a popular part of many peoples’ diet, but in
New Guinea one bird, known as Pitohui, is avoided because of its un-
pleasant taste. The bird’s repellent nature stems from a chemical (a
type of steroidal alkaloid) that interferes with the sodium channels
of the nervous system. The chemical itself is most probably se-

questered from the bird’s diet. Yet this chemical defence mechanism
has evolved independently in the Neotropical poison-dart frogs,172 and
has also been found in another New Guinea bird, probably unrelated
to Pitohui.173 Nor is this the only example of chemical convergence;
a repellent is also found in the crested auklet. This bird, a denizen of
the subarctic Pacific and Bering Sea, has evolved a chemical conver-
gency with a group of insects known as the heteropterans, in which
it occurs in repugnatorial glands.174 So, too, in plants themselves the
extensive employment of defensive alkaloids, cyanogenic glycosides,
and proteinase inhibitors also appears to be rampantly convergent.175
Herbivores, of course, take steps to avoid noxious plants, and in the
case of the lizards, which are predominantly insectivorous carnivores,
the several independent shifts to plant-eating have led to a correlated
evolution of chemosensory abilities that represents a type of conver-
gence in response to a common selective pressure.176
Within the mammals also there are compelling examples of con-
vergence. One of the best-documented case histories concerns the so-
called fossorial species, that is, animals such as the mole which dig
their way through light soils. Just as there is a parallel between the
sabre-tooth cats and marsupial thylacosmilids, so the example of con-
vergence of the Australian marsupial mole to its placental equivalent
is a well-known and striking example. Somewhat less widely appre-
ciated is that among the extinct marsupial faunas of South America
(home of course to the sabre-toothed thylacosmilids) there is another
striking parallel to the placental moles, this time in the form of a re-
markable group known as the necrolestids.177 But the convergences
do not stop there, and among the extant burrowing mammals the sim-
ilarities extend across a much wider range of groups than just the mole
and its various marsupial equivalents. In total this evolutionary con-
vergence encompasses no less than three orders, 11 separate families,
and at least 150 genera (Fig. 6.8).178 Convergences are extensive, in-
volving not only anatomy, but extending to physiology, behaviour, and
even aspects of genetics. Some of the most obvious similarities, such
as powerful digging forelimbs with prominent claws,179 rudimentary
eyes and small size (including small testicles, sensibly enough for
the males dragging their way along tunnels), are rather predictable.
Living in burrows, where levels of CO2 from soil respiration are often
Scalopus Rhizomys, Urotrichus
Geomys Scapanus Talpa Spalax Ellobius Tachyoryctes Scaptonyx
Spalacopus Ctenomys Clyomys Amblysomus Cryptomys Notoryctes

Chrysochloris Georychus
figure 6.8 Convergent evolution in the burrowing (fossorial) mammals, including the familiar mole
(Talpa), mole-rat (Spalax), and marsupial mole (Notoryctes). (Reproduced from fig. 1 of E. Nevo (1995)
Mammalian evolution underground. The ecological–genetic–phenetic interfaces, Acta Theriologica,
Supplement 3 (Ecological genetics in mammals II, eds. G. B. Hartl and J. Markowski), pp. 9–31, with
permission of author and Acta Theriologica.)
elevated, places physiological burdens on these digging mammals, and

here too we see convergences. In the following chapter I explore some
of the other fascinating examples of convergence in terms of sen-
sory systems. The subterranean mammals provide some outstanding
examples, most notably in the case of hearing (pp. 190–191). Unsur-
prisingly, visual systems are correspondingly usually highly reduced,
but as we shall see (pp. 175–177) in the case of the bizarre star-nosed
mole there is another even more remarkable convergence between its
nose and the eye.
As I shall argue, these convergences involving sensory systems
are of considerable importance in assessing the likelihood of recur-
rent emergence of such biological properties as intelligence. Linked
to this is the generally rather neglected evidence for convergence
in behaviour (see also Chapter 10). Study of the burrowing mam-
mals also throws up apparently quirky examples of behavioural con-
vergence, such as their response when confronted with a carrot.180
In a survey of more than 20 small mammals, it was found that
without exception the subterranean species much preferred to start
eating at the lower tip. Rather remarkably, one type, the Zambian
common mole rat, can identify the polarity of the carrot even if
both ends of the root have been cut off. This uniform preference
for attacking the carrot at one end makes sense, of course, be-
cause a burrowing mammal is much more likely to encounter the
lower end of a plant root. In the case of the burrowing animals
here, too, there are other striking parallels in behaviour; typically they
are solitary, communicate seismically,181 and are generally highly
aggressive.182 Such convergences of fossorial activity are, of course,
easiest to document in extant faunas, but it is also worth drawing
attention to some striking convergences in fossil taxa. The controver-
sial status of the possible Cretaceous bird Mononykus is returned to
below, but equally remarkable are highly specialized palaeanodont
mammals (Epoicotherium and Xenocranium) from the Oligocene
of North America. These animals, equipped with powerful frontal
claws, a prominent snout, and lacking sight, are strikingly con-
vergent in many respects on other fossorial mammals, especially
the group of insectivores known as the African golden moles (or
chrysochlorids).183
Perhaps the most remarkable example of behavioural con-
vergence involves the evolution of eusociality, most famously in
figure 6.9 The naked mole-rat, Heterocephalus. (Photograph courtesy

of H. Burda, University of Essen.)
Heterocephalus, better known as the naked mole-rat (Fig. 6.9). The

preface of the main book184 on naked mole rats has a telling story:
In the mid-1970s R. D. Alexander [one of the book’s editors] . . .

lectured on the evolution of eusociality . . . In an effort to explain
why vertebrates had apparently not evolved eusociality, he
hypothesized a fictitious mammal that, if it existed, would be
eusocial. This hypothetical creature had certain features that
patterned its social evolution after that of termites (e.g., the
potential for heroic acts that assisted collateral relatives, the
existence of an ultrasafe but expansible nest, and an ample supply
of food requiring minimal risk to obtain it). Alexander
hypothesized that this mythical beast would probably be a
completely subterranean rodent that fed on large tubers and lived
in burrows inaccessible to most but not all predators, in a xeric
[dry] tropical region with heavy clay soil (p. viii).
The mythical beast emerged as the naked mole rat.

Eusociality refers to a colonial system whereby typically only
one female is reproductive, and the remaining individuals are divided
into several castes, e.g. workers, that operate in a cohesive and coor-
dinated way, notably in food collection and care of the young. Despite
figure 6.10 Another social

mole-rat, a bathyergid.
(Photograph courtesy of
H. Burda, University of Essen.)
its complexity, eusociality is patently convergent. One notable feature

of this convergence is not only the social structure, but also the fact
that the colony makes strenuous efforts to protect the breeding fe-
male/queen from danger. Eusociality has evolved several times in
the mammals, including another group of mole-rats (the bathyergids,
Fig. 6.10)185 as well as some social voles.186 The overall convergence,
however, extends much further because there are striking parallels
with various groups of insects (most famously in the hymenopter-
ans (ants, bees, and wasps)), but also the termites and various other
groups such as the gall thrips187 and certain beetles.188 In the insects,
where of course it is best known, not only is eusociality rampantly
convergent, but also remarkably instructive in terms of parallels with
human organization and activities (agriculture, warfare, and competi-
tive exclusion; see Chapter 8). The often striking differences between
the various castes of a eusocial insect, most familiar in the contrast
between the bloated and near-immobile queen versus beweaponed
and agile soldiers, has another parallel in the naked mole rats because
rather remarkably (for a vertebrate) the colony also shows morpho-
logical castes.189 Nor are insects the only eusocial arthropods; this
characteristic has also evolved in some crustaceans, in the form of a
coral-reef shrimp (Synalpheus),190 where eusociality has evolved in-

dependently three times.191
A recurrent theme of this book is not only the implications of
convergence for evolution, but also the problems it can pose for its
resolution. Not, I hasten to add, in terms of the reality of evolution.
This is emphatically not in question; rather the reverse: the details
of convergence actually reveal many of the twists and turns of evo-
lutionary change as different starting points are transformed towards
common solutions via a variety of well-trodden paths. Rather, conver-
gence, if not identified, may lead to blatantly erroneous phylogenetic
reconstructions. The constraints imposed by the burrowing mode of
life offer a potentially telling example of this problem. There is an
extraordinary fossil reptile, Mononykus, from the Upper Cretaceous
of Mongolia. It has attracted considerable attention, because of both
its strange morphology and its proposed relationships. Bipedal, and
with a fore-limb equipped with a powerful claw, Mononykus has
been interpreted as a flightless bird on the basis of various skeletal
characteristics.192 As such it is regarded as a key discovery in the
evolution of early birds, even though Mononykus certainly remained
firmly on (or in) the ground, and its age means that it post-dates a
number of more advanced birds that have been unearthed in strata of
Lower Cretaceous age in north China.193 When originally described
the strange arm of this animal, with its powerful claw, was recog-
nized as similar to that of digging mammals, although other features of
Mononykus appeared to preclude this mode of life.194 Zhonghe Zhou,
however, has suggested that the adaptation to burrowing is the correct
interpretation,195 and by neglecting this and other convergent features
the role of Mononykus as a very strange bird needs rethinking.
Moles, other fossorial mammals, and possibly even Mononykus
offer, therefore, an excellent example of numerous separate evolution
trajectories converging towards a common solution, best adapted for
occupation of this demanding habitat. It is hardly surprising, there-
fore, that if we turn our attention to a somewhat analogous ecol-
ogy, that of burrowing frogs196 and reptiles, especially the limbless
snakes and the sand-dwelling (and sand-diving) lizards, we again find
instructive examples. Among the snakes, the convergences include
those of brain structure, reflecting the various adaptations to a fos-
sorial existence, including diminished vision and the ability to com-
municate seismically. So far as the burrowing reptiles are concerned,
once again there are certainly differences but also convergences.197

Among the lizards such a fossorial mode of life has evidently been
adopted independently at least eight times, and the mechanical prob-
lems imposed by entering and moving in the aeolian dunes have led
to many parallels.198 Even so, historical antecedents play their part.
As Nick Arnold cogently remarks, with specific respect to ear reduc-
tion in iguanian lizards, ‘Different lineages of organisms often evolve
a number of similar traits independently, but the order in which these
are assembled may often be different, even in ecological analogues,
especially if the taxa concerned are not closely related . . . This [is the]
phenomenon of equipotentiality, where more or less the same overall
condition is reached by different routes’.199 These comments are of
key importance to the overall theme of this book: the evolutionary
routes are many, but the destinations are limited. In a related context
it is also worth remarking that many lizards have toes equipped with
fringes, which assist not only with movement across loose sand, but
also other substrates, including running across water.200 Not surpris-
ingly, the evolution of toe fringes is both convergent and correlated
to the nature of the substrate upon which the lizard moves.201 In a re-
lated fashion the evolutionary trend in lizards that become limbless,
and thereby adopt a snake-like habit, has manifested itself several
times. Significantly, despite this convergence, there are only two ef-
fective destinations represented by distinct types of ecomorph: small
and short-tailed forms that burrow and larger ‘grass-swimmers’ with
long tails.202
If we reconsider deserts, but this time in the context of the
mammals living largely above the surface, again we find a richness of
convergences.203 This is especially true among the rodents,204 many of
which are bipedal hoppers, nocturnal and possessing enormous ears:
the desert kangaroo rats are well named. Here, too, there are some
interesting details. Deserts in South America, notably the Monte of
Argentina, and Australia are effectively unoccupied by these special-
ized rodents, yet, as Michael Mares notes, we can be sufficiently
confident from the general prevalence of convergence that in these
deserts the same story so far as it concerns rodents will probably
unfold.205 If evolution is not inevitable, it is at the very least highly
predictable. South America is of particular interest because it appears
that an extinct group of marsupials, known as the argyrolagids,206
were strongly convergent on the placental desert rodents, especially
the kangaroo rats and a convergent equivalent known as the jerboas.

G. G. Simpson simply remarks, ‘the argyrolagids present one of
the most striking known examples of evolutionary convergence’.207
Mares208 proposes an interesting explanation for the relative fail-
ure of the incoming placentals to diversify in this desert habitat in
South America. He suggests that when they flooded south from North
America as the Panamanian isthmus emerged to separate the Atlantic
from the Pacific (see p. 130) this niche had already been occupied by
the argyrolagids, which had then only recently become extinct. One
of the immigrants, a rodent known as Eligmodontia, does, however,
seem to be evolving in the direction of turning into something like a
kangaroo rat.
To conclude: it does indeed appear to be the case that on the
extremities of life the options narrow, choices diminish, and conver-
gence is the norm. Let us imagine some suitably remote planet, inhos-
pitable, with vast deserts, and a thin atmosphere. Life, if it existed,
might be strangely reminiscent of similar regions on Earth. Tough
plants, burrowing animals, and in the sky flying organisms with re-
markably pointed wings: a sort of mirror image of our world when the
going gets tough. But surely, so the argument continues, in the more
benign areas where the mainstream of life enjoys an exuberance of di-
versity, matters are far less constrained. In this riot of forms, we might
suppose, nearly anything is possible. Evolution is little more than a
chaos of contingency, free of trends and with a myriad of equally likely
destinations. Here the problems of convergence melt away. Nothing
could be further from the truth.
7 Seeing convergence
If I enter a dark wood and happen to hear a woodpecker hammering at

a tree, I at least have the potential satisfaction of knowing that here in-
deed are two unique products of evolution – human and woodpecker –
or so it would appear. The example of the woodpecker is often pre-
sented as one of the best examples of evolutionary uniqueness, with
its specific adaptations that enable it to make percussive attacks on
trees. The corollary that follows is that the evolutionary process will
lead to other unique end points:1 on this planet, as it happens, wood-
peckers and humans; somewhere else, life, but no woodpeckers, no
humans. It may be centuries, if ever, before we can test the latter
supposition, but this may not be necessary. Here on Earth the his-
tory of life can unfold many times, especially on isolated islands and
microcontinents. In Madagascar, for example, there are no woodpeck-
ers, but another group of birds, the vangids, is identified as a ‘true
substitute’.2 Nor might this be the only example of convergence on
a woodpecker. To be woodpecker-like does not automatically mean
you have to be a bird, and three groups of mammals have, in a number
of different ways, converged on a woodpecker-like habit.3 To be sure,
nothing precisely like a woodpecker has evolved independently, but
broad biological generalities still arise. So, too, in an analogous way,
can we say humans are really unique? As we shall see in Chapter 9,
again much is convergent. Yet surveys of life also rightly emphasize
its richness, fecundity, and beauty, if not its strangeness. Time spent
in at least some woods is conducive to reflection: think of all the cu-
riosities of adaptation leading not only to those woodpeckers, but also
to acid-squirting beetles4 or swimming animals encased in the equiva-
lent of newspaper.5 All very intriguing, if not odd; perhaps some might
say even bizarre? Is this not evidence of an almost infinite number of
evolutionary pathways?
Yet now, still in my metaphorical wood, I see a cat. On its left
ear sits a mosquito. Both animals can see me: the cat through the
dilated pupils of its camera-eyes, the mosquito, at least after a fashion,
through its compound eyes.6 But again, at least I know that not only
148 seeing convergence
have both these types of eye – camera and compound – evolved several
times, but even the very mechanisms of seeing in terms of the neural
architecture have also shown multiple convergences. Have I missed
the right road? I give a sceptical cough; both animals can hear me,
one through its triangular, furry ears and the other using the so-called
Johnston’s organs. What could be more different? Yet the auditory
mechanisms are convergent. The cat and the mosquito can smell me,
but again the differences between the nostrils of this feline mammal
and the ‘nose’ of that blood-sucking insect (actually located on its
antennae) are in reality superficial. This is because the mechanisms of
olfaction are also strongly convergent. Even the way in which the cat
and the mosquito walk is convergent in terms of their neural circuits,
reflexes, and motor control: four legs or six, it really does not make
very much difference.7
a balancing act
So, if convergence is going to be a guiding principle in understanding
evolution, then of all the areas worth investigating one of the most in-
teresting must surely be to look at what constraints, if any, accompany
the development of sensory organs. It is here, if anywhere, that we
can approach the wider problem of the evolution of nervous systems,
brains, and perhaps ultimately sentience. And this in turn might give
some clues as to whether indeed intelligence is some quirky end point
of the evolutionary process or whether in reality it is more-or-less
inevitable, an emergent property that is wired into the biosphere.
A similar question has been succinctly posed by David
Sandeman.8 He writes, ‘it may be of interest to ask whether there are
sensory organs that are unique and not convergent . . . One of the most
bizarre and ingenious equilibrium systems that have arisen is without
doubt the halteres of the dipteran flies’ (his emphasis).9 The halteres,
it may need to be explained, are the organs for balance in the flies.
They are located on the back of the fly and resemble tiny drumsticks.
In flight they swing continuously to act as gyroscopes that serve to
detect angular accelerations. It has long been realized that the hal-
teres are derived from the pair of hind-wings of the fly (most insects
have two pairs of wings), and are indeed, in Sandeman’s terminology,
‘bizarre and ingenious’. Surely these qualify as a unique biological
invention? Not a bit of it. Sandeman’s question was rhetorical. He
reminds us that very similar structures are found in another group
a balancing act 149
figure 7.1 Electron micrograph of a strepsipteran insect, in the form of

the male emerging from its host. Note the drumstick-like halteres
arising from behind the head, as well as the compound eye, the
arrangement of which is convergent on that of the trilobites.
(Photograph courtesy of Dr J. Kathirithamby, University of Oxford.)
of insects known as the strepsipterans (Fig. 7.1), which on molecu-

lar evidence do not seem to be closely related to the dipteran flies.10
Strepsipteran halteres are different, however, because they are de-
rived from the pair of fore-wings. They are found only in the males,
whose wings are put to use for a short mating flight that they under-
take once they emerge from their host, such as a wasp, in search of
the females that (with few exceptions) remain permanently within the
host.11 The parasitic strepsipterans differ, therefore, quite markedly
from the dipterans, yet it is clear that their halteres act in the same
gyroscopic fashion.12 However bizarre they may seem, halteres are
not a unique invention. Indeed, there is evidence for yet other conver-
gences. One is in an African beetle that has modified its wing-cases
(known as the elytra) into haltere-like organs;13 and a fourth exam-
ple can perhaps be found in a group of hemipterans known as the
coccoideans.14 The coccoideans are probably most familiar for the
swollen galls the larvae form in plants, but it has been observed that
in some of the adult males the hind-wings are ‘reduced to haltere-like

structures’.15
Swinging tiny drumstick-like structures in the air is one way
of keeping a stable course, and man-made gyroscopes adopt a some-
what similar principle. For animals, a much more widespread method
of achieving balance is by means of structures known as statocysts,
which typically consist of some grains of a mineral attached to fine
hairs. This arrangement is sensitive to gravity, the nervous connec-
tions to the hairs being stimulated according to how the heavier grains
fall. Statocysts have evolved repeatedly, and are found in groups as di-
verse as the jellyfish, crabs, and cephalopods.16 Indeed, a comparable
structure is even known in the group of single-celled organisms, the
ciliates.17
In humans and other vertebrates balance is achieved by an exten-
sion of the statocyst principle, employing mineral grains known as the
otoconia. These are located within sac-like structures housed within
the inner ear. Immediately adjacent to these sacs are the so-called
semicircular canals, which are also involved in maintaining a sense
of balance. Such a sophisticated system might at first be thought to be
without a parallel. Primitively, that is in the jawless fish, the lamprey,
there are two canals, but in the more advanced vertebrates there are
three canals, set at right angles to each other. The canals are filled
with fluid and contain sensory hairs that by registering angular accel-
erations transmit information, albeit in a way not fully understood,
to the brain to permit effective and controlled balance. Unique? No,
because a convergent arrangement is found in the swimming crabs,
although here the arrangement is directly derived from the more prim-
itive arrangement of the statocyst.18 In comparison, however, these
statocysts are strongly modified, and within the canals hair-like recep-
tors monitor the fluid flow induced by the crab’s movement. Even so,
why should a crab in this respect imitate a fish? As so often when dis-
cussing convergence, the mode of life gives the necessary clue. These
crabs are very agile swimmers, capable of rapid manoeuvre in pursuit
of their prey. Balance is clearly at a premium, and it is not so surpris-
ing that they have arrived at a system strongly reminiscent of the one
seen in the more primitive vertebrates with their two semicircular
canals. As might also be expected, these canals are filled with fluid.
In contrast to the vertebrates, however, this fluid is composed of sea
water and is incorporated each time the crab moults its skeleton. At
returning the gaze 151
this juncture it has to secrete a new skeleton, including a pair of the

balancing organs which, once filled with sea water, are then sealed
to the outside world. In conjunction with a sophisticated visual sys-
tem these balancers allow these active and fast-swimming crabs to
perform their adroit manoeuvres.
The senses, five or otherwise, may be separately identifiable,
but the various inputs received by the animal through the nervous
system will require appropriate integration. In the case of balance in
the swimming crabs an acute visual control is an unsurprising corol-
lary. In reviewing this linked system in the crabs and comparing it
with the vertebrates Sandeman remarks that they have both ‘achieved
the same end using similarly constructed sense organs’,19 albeit from
very different origins. Sandeman continues, however, by noting that
notwithstanding the sophisticated sense of balance the overriding fac-
tor in this system is the effective capture of light, and so it is to the
topic of the eye that we now turn.
returning the gaze

Yet, concerning the eyes there is an obvious difference.20 Vertebrates,
including humans, use a camera-like arrangement, whereas the crabs
employ the compound eye that is typical of most arthropods. Has
the principal of convergence broken down? Not at all, because it
is among eyes that we find some of the most compelling examples.
The existence of camera- and compound eyes reminds us, of course,
that solutions to biological evolution need not be unique, but are sim-
ply very strongly constrained.21 Even so, in at least the case of eyes it
is possible to argue that the camera-eye is inherently superior to the
compound arrangement: both are viable, but when it comes to asso-
ciation with sentience the camera-eye may prove to be the winner.
When considering convergences between camera-eyes it is al-
most inevitable that comparisons will be drawn between the camera-
eyes of vertebrates and those of the advanced cephalopods,22 notably
the squid and octopus, which are one group of the molluscs (Fig. 7.2).
This is the stuff of textbooks, but there are well-known differences.
Most notable are those between the relative position and detailed
structure of the light-sensitive layer, the retina, which arise as a
result of the different embryologies in vertebrates and molluscs. In the
molluscs the retina is derived from the outer layer (known as the
ectoderm, and effectively equivalent to our skin). As it infolds to make
Annelid Cephalopod Vertebrate
Optic nerve
Retina
Pigmented layer
Nuclear layer
figure 7.2 Convergence of the camera-eye, including the classic comparison between the octopus (cephalopod) and human
(vertebrate), as well as the alciopid polychaete (annelid). (Redrawn from various sources.)
the eye-cup, the associated nerve cells extend into the body to make
their connection with the brain. In contrast, the vertebrate retina is
effectively an outgrowth from the central nervous system. The net
result of this process is that, in contrast to the molluscs, in the ver-
tebrates the nerve cells come to overlie the retina. The exit point of
these nerves to the optic tract, which then leads to the brain, results in
a ‘hole’ in the retina, better known as the ‘blind spot’. Accordingly the
arrangement of nerve cells and retina in molluscs and vertebrates is
reversed, the cephalopods having arguably the more ‘sensible’ design
of nervous layer beneath sensory retina.
The cephalopod lens is also formed in a rather different way.23 It
has a rather remarkable layered structure,24 and being inflexible can-
not change its shape to assist in the process of focusing.25 It is still
not clear exactly how the cephalopods manage to focus, but most
probably it is by moving the entire lens backwards or forwards.26
The style of construction of the lens may also explain why at least
some cephalopods are able to correct for spherical aberration. (This is
a phenomenon in which the light rays, on passing through the lens,
fail to focus at a single point, resulting of course in a blurred image.)
Even so, correction for such aberration seems to be less refined than
in the fish. In the fish eye the refractive index (which is a measure
of how much the light ray is bent as it passes from one medium to
another) varies through the thickness of the lens. In cephalopods, how-
ever, spherical aberration may be dealt with by using non-spherical
lenses, which in turn is possible because of their different mode of
formation.27 These differences perhaps represent relative disadvan-
tages of cephalopods as compared with fish.
Control of eye movements is, of course, critical. This is achieved
by the so-called extra-ocular muscles that are attached to the exterior
of the eyeball. Through what is known as the oculomotor control sys-
tem, routed via the brain, this arrangement is linked to the organs
of balance and so provides a coordinated arrangement, referred to as
the vestibulo-oculomotor reflex.28 As already noted there are inter-
esting convergences in the systems of balance, which in cephalopods
are represented by the statocysts and in vertebrates by part of the
inner ear. Given the nature of the oculomotor system it is not sur-
prising to find convergences in the brain pathways between some
cephalopods, such as the octopus, and the vertebrates.29 The degree of
similarity between these systems is, in one sense, hardly surprising,
but there are differences: for example, comparison of the oculomotor

reactions, which allow the eye to swivel in its socket by the contrac-
tion of the extra-ocular muscles, shows the vertebrate arrangement to
be superior.30
All these differences are important, but it is still the case that the
similarities between our eyes and those of a cephalopod are very strik-
ing. It is rather less well known that a similar camera-eye has evolved
independently in several other groups. The most notable example is
in a group of marine annelids, close relatives of the more familiar
earthworms.31 The group in question is known as the alciopids.32
These eyes are strikingly similar to those of the vertebrates and
cephalopods (Fig. 7.2), and because the annelids are also relatively
closely related to the molluscs the retina has the same arrangement as
in the latter group.33 There is, moreover, another and striking conver-
gence in the alciopid eye. This is in the form of the so-called accessory
retinas, which are light-sensitive patches located nearer to the front
of the eye. These, too, are strongly convergent on similar structures
found in some deep-sea fish and cephalopods.34 Returning to the mol-
luscs, but this time considering the gastropods (or snails) we find that
in this group a camera-like eye seems to have evolved independently at
least three times.35 Specifically these are in the pelagic heteropods,36
the familiar shore-snail known as the winkle (Littorina),37 and a large
herbivorous tropical snail called Strombus.38
Nor does the list of convergently evolved camera-eyes quite end
there: there are two more examples, each in their own way quite
surprising. The first example comes from the much more primitive
cubozoans. These are a type of jellyfish (hence the alternative name
of cubomedusae) renowned both for their highly toxic stings and for
their remarkable eyes. The eyes are similar in construction to other
camera-eyes, with a large lens located in front of the retina39 but sepa-
rated by a layer of cells that may help in focusing.40 Cubozoan jellyfish
belong to a primitive group of animals, the cnidarians, which also in-
cludes the sea anemones and corals. While primitive eye-spots are
known in other cnidarians,41 at first sight the sophistication of the
cubozoan eyes, which typically total eight arranged around the mar-
gin of the swimming bell, is quite surprising. Cubozoans, however, are
active and highly agile swimmers, have obvious visual acuity,42 and
uniquely for cnidarians engage in bouts of copulation.43 What is par-
ticularly interesting is the relative simplicity of the nervous system,
which consists of a nerve net linked to a series of four pacemakers,

a neural architecture that is effectively imposed by the jellyfish body
plan.44 Thus there is no brain,45 yet complex eyes and sophisticated
behaviour.
Seeing without a brain has certainly attracted notice, although
as we shall see below (p. 165) other organisms have an eye that ev-
idently can focus an image without even the benefit of a nervous
system. But the case of the cubozoans deserves wider scrutiny. As one
worker has remarked, ‘If cubomedusae have developed unique image
forming eyes, is it possible that they have also developed a unique
method of processing the information?’46 In this context it is also
worth noting that nerve nets are by no means confined to primitive
animals. They also occur in the echinoderms (the group represented
by the starfish, brittle-stars, and sea urchins) and their relatives, the
somewhat more obscure hemichordates (which, as it happens, are also
quite closely related to the chordates, the group to which we belong).
Interestingly, there is now evidence for unexpected complexities of
function in the echinoderm nerve net.47 Not only that, but there is
now evidence that part of the brittle-star skeleton has been converted
into a sort of compound eye, convergent on the arrangement seen in
the trilobites (p. 159). Whether anything like an image is seen by the
brittle-star is very conjectural, but here too the suspicion is that these
brainless animals possess hidden levels of sophistication.
More specifically, knowing how it is that cubozoans see might
even have a bearing on extraterrestrial alternatives. What is evident
in these animals is that the nervous system shows levels of auton-
omy, so that at times it can act as an integrated network yet in other
circumstances show a greater degree of independence and be capable
of dealing with directional inputs. As Richard Satterlie and Thomas
Nolen remark,48 ‘The presence of four pace-makers in cubomedusae,
as well as their semi-independent coupling, allows significant modula-
tion of swimming by complex sensory systems and fast, directionally
accurate behavioral responses in a radially arranged, non-cephalized
nervous system. Functionally, one could view [this] as a form of con-
densation of neural networks, analogous to cephalization in bilateral
animals’.49 What they are saying, albeit in somewhat technical lan-
guage, is that coordinated and rapid activity may not automatically
necessitate a brain with an attendant battery of sensory organs. Just
as there are several ways of building eyes, each also convergent, so
perhaps arising from a common substratum there is the possibility of

several sophisticated neural arrangements evolving, and perhaps even
distinct pathways to advanced intelligence marked by such features
as plasticity of behaviour, memory, temperament, and even sleep?50
These possibilities are explored in subsequent chapters, but we can
at this point note that it is surely an exciting prospect to explore
both the nature of the building blocks (e.g. proteins) and the levels of
complexity that lead both to obvious differences and to striking sim-
ilarities in the intelligences of cephalopods, insects, and vertebrates.
And does an underlying neural commonality undermine the principle
of convergence? Not at all. In the next chapter we shall look at the
extraordinary sophistication of the social insects: here, if anywhere, is
the nearest analogy to an extraterrestrial intelligence. There may well
be several types of intelligence, but the number will be restricted and
the pathways to each type strangely similar. And arguably there will
be deeper, if not universal, similarities: in principle all will understand
algebra.
So camera-eyes are remarkably diverse in their distribution, and
they form an obvious contrast to the compound eyes that are so typ-
ical of the arthropods. In due course we shall see that in exceptional
circumstances a compound eye can be transformed into a sort of
camera-eye. Yet this is not the only route: another example of a sort of
camera-eye is encountered in some of the spiders, notably those adept
at net-casting, including a type known as Dinopis. This is a hunter of
a special sort, suspending itself from a web but itself clasping a net of
sticky and elastic silk. When prey is detected, the spider extends its
net, rapidly drops and then pushes the net against its prey. Then, as it
ascends, the net folds around the hapless victim.51 The spider is noc-
turnal, and like other members of this group has eight eyes. In Dinopis
two of the eyes, located on the posterior median part of the head, are
strikingly large.52 (This explains why these spiders are often called
the ogre-faced spiders.) In cross-section each eye consists of a cup-like
retina and overlying lens, the lens being embedded in an iris. Inter-
estingly, the texture of the lens changes in consistency through its
thickness, from jelly-like at the front to being harder further back. Evi-
dently this, in combination with the retina, acts to correct for spher-
ical aberration, in a manner similar to that of the fish lens (see above).
The spectacular eye of this spider does not provide a good image; its
function is rather to collect light, which is a useful attribute given the
spider’s crepuscular activities. A somewhat similar camera-like ar-

rangement is found in the antero-median eyes of jumping spiders, but
there the emphasis is very much on image formation,53 to the extent
that they can recognize video pictures of their prey (and mates).54 In
addition, in these spiders the retina has a quite extraordinary stacked
structure, which even incorporates a telephoto system of a kind.55 It
has also been proposed that this multi-layered retina makes colour
vision possible, and evidence exists for spectral sensitivity.56 What is
not in doubt is that for their size these spiders have a quite remarkable
visual acuity. The occurrence of simple eyes, effectively camera-like,
in these spiders is curious. This is because the majority of arthropods,
the phylum to which the spiders belong, employ compound eyes, as
evidently did the ancestors of the spiders. As we shall see (pp. 160–162)
the nearest approach to a simple eye in the other arthropods is based
on a radical reconfiguration of the standard compound eye. It remains
an open question as to why this spider route was not taken more
often.
The hallmark of the animals that evolved camera-eyes, be they
squid, vertebrate, heteropod snail, alciopid polychaete, cubozoan
jellyfish, or even spider, seems to be that all are active, mobile, and
typically predatory. At first sight, the two other examples I have al-
ready briefly mentioned, those of the snails Strombus and Littorina,
seem to be marked exceptions. Strombus is generally slow-moving
and herbivorous. In remarking on the strombid eye Mike Land makes
no attempt to conceal his puzzlement when he writes, ‘What the eye is
used for, other than simple taxes [i.e. responses to stimuli] for which it
seems over-adequate, remains a mystery’.57 Yet there is considerable
evidence for electrical, and by implication visual, activity in the eye of
Strombus (note 38). Perhaps we have underestimated this seemingly
unassuming gastropod? Strombids are a very successful group in the
tropics, and appear to have displaced similar forms such as the pelican-
shells (aporrhaids) into colder waters.58 Not surprisingly, members of
both groups are anxious to avoid attack, especially by fish, crabs, and
drilling snails, but the strombids are particularly adept. As Kaustuv
Roy notes, when escaping they ‘generate a series of very rapid leaps
away from the predator . . . [showing] some of most specialized and
effective escape responses known among gastropods’.59 Perhaps a so-
phisticated camera-eye is more useful to the strombids than might at
first be thought?
What of the other snail, Littorina? There is little doubt that

the winkles can see better than most other gastropods. This acuity
is evident from a number of tests that show, for example, that vari-
ous species of Littorina can orientate themselves at night, perhaps by
recognizing silhouettes,60 and are adept at navigating both by day and
beneath starlight.61 It is perhaps this snail’s habitat, which is typically
a gently sloping tidal flat,62 that provides the explanation, inasmuch
as there is a premium in being able to recognize particular shapes,
especially plant stems. This is because as the tide rises, and the preda-
tors move in, Littorina can climb the stems to relative safety.63 Such
optical acuity is probably an important component in explaining the
immense success of the humble winkle.
eyes of an alien?
It is beginning to look, therefore, as if active, fast-moving, and at least
in some cases intelligent animals opt for the camera-eye.64 So, too,
we need to remind ourselves that commonalities of eye structure may
lead to convergences in brain structure and the evolution of visual cen-
tres, which is at least the case within the mammals.65 Yet, so far as
visiting aliens go the story more often looks to the immense cylinder,
embedded in the cricket pitch, the hatch slowly opening, a hideous
scrabbling before the bug-eyed monster levers itself out and begins
its obligatory, laser-cannon-powered destruction of the pavilion. The
inspiration for this abrupt interruption of the smooth running of the
Home Counties is presumably the alien-like appearance of insects,
robotic limbs, and empathy-free compound eyes. Clearly, given the
immense success of the arthropods, the compound eye has its advan-
tages. For example, the eyes of some deep-sea crustaceans are remark-
ably adept at collecting the minute quantities of light available at
depth, and do so by employing a sort of fibre-optic system.66 This spe-
cialized arrangement has arisen independently in two pelagic groups,
specifically a galatheid crab and the euphausiids.67 Not only that,
but it now seems likely that the compound eye itself is convergent
in the arthropods, evolving independently in a group of crustaceans,
the ostracods.68 It is interesting also to note convergent evolution of
the compound eye in the sabellid annelids69 and bivalve molluscs70
(Fig. 7.3). In the sabellid annelids this type of eye evidently has evolved
independently several times.71 In both these groups the shared con-
vergence seems to reflect, however, a rather different adaptation from
eyes of an alien? 159
figure 7.3 Convergence in

the compound eye, in the case
of the mantle eye of the
bivalve Arca. (Photograph
courtesy of Dr T. R. Waller,
Department of Paleobiology,
National Museum of Natural
History, Smithsonian
Institution.)
that found in the arthropods; here the compound eyes appear to act as
a sophisticated optical alarm.72
Yet it is in the arthropods that the compound eyes are most
familiar, ranging as they do from the familiar example of the insects
to the calcite lens of the extinct trilobites.73 Trilobite eyes are often
claimed to be unique, but in fact they show two interesting conver-
gences. The first is in terms of general organization, and entails a
comparison with the eyes of the strepsipteran insects,74 a group that
we encountered earlier in the context of their convergently derived
gyroscopic halteres. The overall resemblance to trilobite eyes, espe-
cially the more advanced type known as schizochroal (see below) is
quite striking, but strepsipteran eyes (Fig. 7.1) are not calcitic. An
equivalent arrangement, albeit on a smaller scale, is found, however,
in a group of echinoderms mentioned above, the brittle-stars. Here,
on the upper side of the five arms, there are the characteristic calcitic
ossicles that make up the flexible skeleton, but evidently modified
for the purposes of vision.75 As in the trilobite lens, the optic axis is
parallel to the line of sight, an important feature because the mineral
calcite shows strong double refraction. This means that light passing
through the crystal at an angle to the optic axis is split into two rays,
resulting in the formation of a double image. Along the optic axis
of the crystal, however, the rays remain together, so that the trans-
mitted image remains single.76 This convergence in calcitic eyes may
extend further because each brittle-star ossicle is composed of a series
of lenses, each forming a dome-like structure, which is strongly rem-
iniscent of a trilobite compound eye. Just what the brittle-star ‘sees’
is rather conjectural because the nervous system is a diffuse net with-

out a central brain (see note 47), and is therefore reminiscent of the
cubozoan jellyfish (see above). Even if images are not perceived, the
brittle-star’s sensitivity to external stimuli is evident in its ability to
avoid predators rapidly and in the striking changes in its colour on a
daily cycle.77
Trilobite eyes have also attracted considerable attention because
the internal structure of the more advanced type of lens, which is re-
ferred to as schizochroal,78 appears to have a double component. Such
a configuration, it is claimed, acts to minimize the spherical aberra-
tion of the perceived image. Accordingly this lens structure evidently
anticipated, by hundreds of millions of years, human technology with
its grinding of lenses with the same aim of minimizing such optical
distortion of the image.79 Interestingly, a somewhat similar arrange-
ment, again to correct spherical aberration, is found in the calcitic
lenses of the brittle-star (note 75). So far as the trilobites are con-
cerned, not everyone, however, is entirely convinced by this elegant
story. At least in some cases post-mortem changes, such as recrystal-
lization in the eye structure, may provide artefacts of structure that
were not present in the living trilobite.80
The magic of being able still to see through the lens of a
trilobite81 does not mean that one can necessarily avoid a distorted
vision of these extinct animals, but let us return to the relative merits
of camera- and compound eyes. Will the alien monsters be bug-eyed,
or should we prepare ourselves to return the steady gaze of sentience
through an eye like ours? As it happens, more probably the latter. This
is because, despite its versatility, the compound eye simply does not
match the light-collecting powers of the camera-eye. Apart from the
ogre-faced and jumping spiders discussed earlier (pp. 156–157) there
seems to be only one exception to the rule that arthropods typically
have compound eyes (see also note 73). This concerns the unique eyes
of a shrimp belonging to a group known as the mysids.82 As in arthro-
pods, the principal eyes are compound, but embedded in the posterior
region of each eye there is an extraordinary addition in the form of a
giant lens (Fig. 7.4). In effect the shrimp has evolved a simple eye, con-
sisting of a single enormous lens that overlies a giant crystalline cone.
This serves to channel the light to a retina composed of more than a
hundred of the individual optical units that are referred to as the rhab-
doms. Effectively these rhabdoms, which in the normal compound eye
figure 7.4 The compound eye of a mysid shrimp. Above, entire eye
with simple eye located on posterior side. Below, transverse section of
eyes. Simple eye is on lower side. (Photographs courtesy of Professor
D-E. Nilsson, University of Lund.)
each underlie a separate lens, have been diverted to supply the gi-
ant lens. In life the eyes of the shrimp can move freely, and used
together the two giant lenses evidently act as a binocular. Looking
at this animal changes our perception; as Dan Nilsson and Richard
Modlin remark, ‘Viewing the live animal from behind . . . gives an
almost uncanny feeling of being observed’.83
So here in one particular arthropod we have a design that ap-
proaches that of our own eyes; yet from the point of view of evolution
it appears to be a dead end. As the investigators remark,
The ultimate development is . . . a large single facet and crystalline

cone supplying light to a common retina of numerous rhabdoms.
There can be no further improvement along this line and [the
giant eye of this shrimp] has reached the ideal design for a
compound eye. Ironically, the ideal compound eye is not really a
compound eye at all. It has been turned into a simple eye in order
to exploit the advantages of imaging through a single aperture.
There is thus little reason for surprise over the unique design in
[this shrimp]. It would be more appropriate to wonder why [it]
does not share its ingenious solution with other crustaceans or
with insects.84
The answer may lie in the difficulties in reorganizing the neural cir-
cuitry and, as it happens, with the notable and already noted exception
of some spiders (see p. 157), nearly all other arthropods remain stuck
with their compound eyes.
In the final analysis camera-eyes are not only different from
compound eyes; they are better. This is because in the compound
eye the visual array can be increased both by expanding the overall
area of the eye and also by the increasingly dense packing of the lens.
There are, however, limits to both, and in particular as the lenses
shrink in size so their ability to collect enough light to function is
compromised. Calculations suggest that if humans were to rely on
a compound eye to achieve equivalent vision it would have to be
at least a metre across, and more realistically up to 12 metres wide
(Fig. 7.5).85 On distant planets there might (or might not) be bug-eyed
aliens, but any astronomers are almost certainly going to be equipped
with camera-eyes. Perhaps, too, they will have binocular vision, a rea-
sonable enough assumption given its convergent acquisition in birds
and mammals.86 Yet, it is also worth remarking here that in a certain
figure 7.5 Why we don’t have

compound eyes. A conservative
estimate of the size of eye a human
would require to provide even a
near-equivalent of the camera-eye.
(Reproduced with permission from
Neural principles in vision, edited
by F. Zettler and R. Weiler, in the
chapter (pp. 354–370) by
K. Kirschfeld, fig. 7c; 1996,
copyright Springer-Verlag and also
with the permission of the author.)
sense the differences between compound and camera-eyes are, at one

level, rather superficial. For example, the process of visual tracking in
which an animal’s coordination of vision and locomotion ensures suc-
cessful interception of prey (or mate) shows that between humans and
such insects as the mantids and hoverflies87 there are ‘some remark-
able similarities’.88 Not surprisingly, given these similarities, binocu-
lar vision has also evolved independently in a number of insects.89 As
we shall see later, there are also convergences in the molecular archi-
tecture of vision, notably among the proteins known as crystallins and
rhodopsins, both of which are essential for any animal eye to work.
Moreover, the convergences discussed below in such sensory modali-
ties as olfaction and hearing make it less surprising that there may be
parallels in the neural processing of, say, insect and vertebrate eyes.90
The limitations of biological form and the reality of the adap-
tive framework provide a realistic guide to what there will be ‘out
there’. Even let us suppose that our extraterrestrial has the inevitable
camera-eyes, but these are mounted in little turrets and able to swivel
independently. Again we can find a striking convergence here on
Earth. Such an arrangement is well known in the chameleon lizards. It
entails some remarkable changes to the basic camera-eye. Among the
most striking of these is the suppression of the lens and its replace-
ment for refractive purposes by the cornea, which is duly equipped
with muscles to effect the changes in shape necessary for focusing.
How odd, but this rearrangement is strongly convergent on the eyes
of the sand-lances, a group of fish.91 And despite the different media,
air as against water, there is a common adaptive explanation. Both
chameleons and sand-lances are capable of exceedingly accurate and
rapid strikes at their respective prey: in effect the lunging of the entire
fish is equivalent to the darting of the chameleon tongue.92
In discussing eyes a number of qualifications are important.
First, it is agreed that eyes have evolved independently very many
times. In their classic paper Luitfried v. Salvini-Plawen and Ernst
Mayr93 write, ‘Summing up the different and convergent sequences
towards eye perfection in general, there are about 20 or even more
independent lines of differentiation including at least 15 cases of
independent attainments of photoreceptors with a distinct lens.’94 Al-
though I have emphasized the importance of compound and camera-
eyes and their convergences, unique optical configurations do occur.
Perhaps the best example is the remarkable eye (or eyes, given that
an individual has a whole series distributed along the shell margin)
of the scallop, which although camera-like has a system of inter-
nal mirrors.95 As Michael Land aptly remarked,96 this represents ‘an
almost unique example of a type of eye that should exist, but which
had not been found until relatively recently’.97 Even so, the reflec-
tivity of the eye depends on the mirror-like properties of the material
guanine.98 This compound has already been encountered in the rather
different context of providing thin, flexible sheets that would help to
gas-proof a Fortean bladder (pp. 112–113). In an optical context, how-
ever, not only is guanine found in scallop eyes, but it also forms the
reflective structures, known as the tapetum, that are found in the
eyes of certain vertebrates, notably deep-sea fish, where they act to
concentrate light.99 Nor does its ‘optical’ use stop there, because the
silvery guanine is also widely used as a reflector to channel the light
generated by the luminescent organs known as photophores.100
figure 7.6 The optical

apparatus of two
dinoflagellates,
Erythropsidinium (left) and
Warnowia (right), magnified
respectively ×460 and ×380.
Note the bulbous lens and
underlying cup. (Reproduced
from fig. 1 of Greuet (1968;
citation is in note 106), with
permission of Urban and
Fischer Verlag. (European
Journal of Protistology.)
Second, and even more interestingly, eyes are by no means re-

stricted to animals. Most familiar are the various sorts of eye-spot
(more strictly light antenna or photosensory apparatus) in the protis-
tan eukaryotes such as the single-celled Chlamydomonas or globular
colonial Volvox.101 Here, too, independent evolution is widespread,102
and it also turns out that the visual protein (rhodopsin) employed in
these eye-spots may be convergent on the equivalent protein found in
our eyes and those of other animals (see pp. 171–172). Typically the
protistan eye-spot has a reddish or orange colour; this is due to a layer
of carotenoids that serve to reflect the incoming light back against the
photoreceptor.103 Even so, among the examples of protistan eye-spots,
the optical system in the group known as the dinoflagellates is par-
ticularly remarkable.104 In this group a variety of eye-spots occur,105
but in certain taxa (e.g. Erythropsidinium and Nematodinium), the
organism, which, remember, is a single cell, bears a protruding bulb-
like structure. In appearance, this is almost telescope-like (Fig. 7.6).
Its structure is complex, but in essence it consists of a pigmented
cup (melanosome) surmounted by a crystalline lens (or hyalosome)106
which can refract light so that it focuses on the underlying sensitive
layer.107 Despite the absence of a nervous system, this tiny optical pro-
tuberance, less than a tenth of a millimetre long, is strikingly conver-
gent on the animal eye,108 although most probably it originates from
a chloroplast.109 Somewhat nervously, in describing the behaviour of
these dinoflagellates, Ester Piccinni and Pietro Omodeo110 remark,
‘It is unthinkable that an apparatus of this sort, in spite of the so-
phistication of its design, can function like an image-forming eye’,111
but they speculate that these organisms may detect their prey and
then entangle it with explosively discharged thread-like structures,112

which are also convergent.113
clarity and colour vision

The existence of these primitive eye-spots exemplifies once again the
important, but I believe relatively neglected, principle of inherency,
whereby the basic building blocks of complex structures are available
long before they are recruited for new and more sophisticated tasks.
This observation is important for two reasons. First, if such building
blocks, necessary for the assembly of complex structures, are readily
available – as seems to be the case – then, unless there is some ne-
glected and at present mysterious principle of evolution that imposes
an almost insuperable barrier of some sort, there seems no reason why,
in a world subject to adaptive scrutiny and competition, that sophis-
ticated forms, such as eyes, should fail to evolve repeatedly. Second,
and as a consequence of the first point, it is apparent that evolutionary
novelty is often only skin-deep. This is because its emergence relies
more on co-option and redeployment than invention.
Thus, in the case of the eye, what, in some ways, could be more
remarkable than having a window into the head, that is transparent
tissues that not only allow transmission of light but remain func-
tional for decades? The solution is provided by a number of specific
proteins that serve to confer transparency to the lens. For this reason
they are known as the crystallins, although it is important to stress
that they are not specifically crystalline in the way, for example, of a
mineral like quartz. As is now well known, these crystallins clearly
must have been co-opted from serving some previous function. This
is because originally these proteins evolved for completely different
purposes, most usually associated with providing the organism with
resistance to physiological stress.114 Of these stress resistors, the best
examples are the heat-shock proteins that are widespread and found
even in bacteria: in this sense crystallins are inherent in evolution
from billions of years ago. Nor is their recruitment difficult to un-
derstand. Long-term stability in the face of environmental insult is
clearly highly desirable, especially when repair mechanisms that en-
tail a blood system would compromise optical transparency. In addi-
tion, the proteins recruited for this purpose are typically rather small
in size. This favours close packing in regular arrays that, together with
a water-rich environment, confers the necessary transparency.115 And
clarity and colour vision 167
however strange the construction of the eye may be, such as is seen
in the remarkable mirror optics of the scallop eye (which actually has
a direct parallel to some telescopes95 ), the transparency is still con-
ferred by the agency of crystallins.116
The importance of the crystallin proteins in conferring trans-
parency to tissues is nicely shown in what might be called a ‘reverse
eye’. In some squid, for example, the body bears bioluminescent organs
(the photophores) that emit flashes of light. The tissue of this ‘reverse
eye’ is derived from the musculature, yet not only are there sometimes
lenses117 but once again the transparency is conferred by recruitment
of a crystallin.118 More generally, the wide range of co-opted proteins
that provide transparency, be it in eyes or ‘reverse eyes’, is another
striking example, not only of convergence, but also of biochemical
convergence.119 Just as I speculated that alien astronomers (if there
are any) would be searching the skies using camera-eyes, so it would
also seem more than likely that inside their eyes the lenses would be
packed full of the direct equivalents of the crystallin proteins.
The other essential component of any eye is a mechanism for
converting the incoming photons into electrical signals that, some-
how, the brain can interpret as an image. In detail this entails a very
complicated sequence, but the key element is a group of proteins
known as the opsins. Of these the protein rhodopsin is probably the
most familiar. Its basic structure is of some interest in that we shall
encounter the arrangement of seven α-helices, each of course made of
a string of amino acids, spanning two sides of a lipid membrane in an-
other sensory context (pp. 180–181). Embedded within the rhodopsin
molecule, and attached to two of the α-helices, is a molecular unit
known as a chromophore. This is usually a derivative of vitamin A,
known as retinal120 and hence explains why such a vitamin disorder,
sometimes genetic, can lead to difficulties with sight.121
Rhodopsin is, therefore, as integral to vision as are the crys-
tallins. This protein also offers further insights into the nature of
convergence in at least two ways. The first is rather specific, and
concerns the – to us – familiar property of colour vision. The second is
more general and also controversial: how many times has rhodopsin
evolved? To humans the glory of a rainbow is appreciated because
our colour vision is trichromatic (blue, green, red), a relatively un-
usual condition among the animals. In mammals, for example, it is
otherwise found only in the apes and their close relatives, the Old
World monkeys, and convergently in the howler monkeys of the

New World122 (as well as possibly in some of the more primitive
prosimians123 ) and some Australian marsupials.124 That we are able
to see in colour is generally thought to be because our tree-dwelling
ancestors had strong incentives to locate either appropriately coloured
fruit125 or leaves.126 In some of the New World monkeys, such as the
marmoset, there is a particularly intriguing mixture of colour discrim-
ination whereby all the males see dichromatically, whereas the fe-
males are either trichromatic or dichromatic according to whether the
individual is respectively heterozygous or homozygous.127 Here, too,
there is some evidence to suggest that this peculiar arrangement might
have evolved independently among the New World monkeys,128 and
might in fact confer an adaptive advantage in the co-operative search-
ing for fruit by males and females, given that each is attuned to differ-
ent visual signals.129 Some support for this notion might come from
humans who happen to be dichromatic. Their apparent deficiency
is offset by a remarkable ability to penetrate the disguises of cam-
ouflage, be it military130 or natural, as in detecting otherwise cryptic
moths.131 Many other mammals132 and some invertebrates are dichro-
matic, while other animals can see beyond our range of visual sensibil-
ity into the region of the ultraviolet.133 As is well known, on account
of the relative commonness of red–green colour deficiency in male hu-
mans, colour vision has a genetic basis,134 and it is also evident that
the origins of trichromacy are related to episodes of gene duplication.
The changes in the structure of the rhodopsin molecule that
make possible the absorption of particular wavelengths of light, and
so the perception of colour, depend on substitutions of one amino acid
by another at key sites along the length of the protein.135 In red–green
vision the replacement typically involves the insertion of a hydroxyl-
bearing (i.e. OH) amino acid in place of a non-polar one (i.e. one that
‘dislikes’ water), although changes that involve the substitution of a
different-sized amino acid can also play a role.136 Quite why a par-
ticular substitution should alter the wavelength of light absorbed has
only recently become clear.137 The sites of substitution are highly
specific and are limited to a mere handful, of which even fewer are
really important. This has led to the identification of the so-called
‘site rules’, which for red–green vision involves five sites. It is not
so surprising to learn, therefore, that not only has red–green vision
evolved independently, but at least in some cases its molecular basis
clarity and colour vision 169
is strongly convergent. One well-known example is the convergence

in red vision between a fish and mammals where two, and possibly
three, sites show identical substitutions.138 Indeed, within the verte-
brates the evidence continues to accumulate that the ‘five-site rule’
is very widespread,139 and although there are some variations these
are much more minor.140 So, too, with the registering of blue wave-
lengths of light, where in a number of vertebrates there are a number
of specific substitutions of the rhodopsin molecule.141 Rather remark-
ably, in at least some toothed whales and seals there has been a loss of
blue-sensitive visual pigments, an absence that is evidently conver-
gent in these aquatic mammals.142 For animals living in deep water a
shift in the opposite direction, towards the red end of the spectrum,
is to be expected, and here, too, at least in a series of distantly re-
lated teleost fish, the spectral tuning is again controlled by a series of
specific substitutions.143 For ultraviolet (and violet) vision in various
vertebrates, the story of site substitutions is slightly more compli-
cated. In the ancestral form, ultraviolet vision evidently depended
on amino acid replacements at about eight specific sites, and for the
most part only limited substitution is possible. The birds, however,
regained sensitivity to these short wavelengths by the replacement of
an amino acid at a completely different site.144
Rather less is known about the molecular controls on spectral
absorption, and by implication colour perception, among the inver-
tebrates, but broadly the same picture emerges.145 The case of the
cephalopods, such as the squid, is particularly important because of
the striking convergence in terms of the overall arrangement of the
camera-like eye (Fig. 7.2). The octopus is evidently colour-blind, but
in one species of squid, where the rhodopsin146 shows a spectral shift,
there is a key substitution at one site that is evidently the equiva-
lent of one of the mammalian sites involved with colour vision.147
What of those animals, notably the arthropods, with compound eyes?
They certainly have colour vision,148 and in at least some instances
there are interesting convergences with the vertebrates. Thus in some
butterflies both blue- and red-shifts have exact parallels with mam-
mals, while among the butterflies themselves other convergences
exist.149 So, too, among the crayfish, in those regions of the rhodopsin
molecule that are similar to those for the vertebrates several sites
may be involved with a shift in spectral absorption.150 As a group,
insects have a wide range of spectral sensitivities. In the fruit-fly
(Drosophila), for example, the principal window of vision is in the

blue and ultraviolet range, and this too appears to depend on strongly
constrained amino acid sequences. Certain other site substitutions,
however, have parallels in the vertebrates.151
It should also be noted that many animals, such as fish and birds,
incorporate colour filters in their retinas by employing oil droplets to
absorb particular parts of the light spectrum.152 This also appears to
be convergent. So, too, many animals have independently developed
the capability of detecting polarized light, perhaps most famously the
honeybee, which uses differences across the sky to assist it to navi-
gate. In certain instances the sensitivity to polarized light is such that
it is analogous to colour vision, and hence is referred to as polariza-
tion vision. This is known to occur in the stomatopods,153 a group of
crustaceans long known for their extraordinary eyes154 and a lifestyle
that encompasses effective hunting and communication. It is perhaps
less surprising that polarization vision has evolved convergently in
the cephalopods such as the octopus and squid.155
universal rhodopsin?
As it happens, a number of other proteins can act as light receptors,
and they too provide some illuminating cases of convergence. Yet
the rhodopsin molecule seems to offer various advantages: as Russell
Fernald156 has remarked, in the context of vision although an opsin
‘is not the only way to detect light . . . it has proven irresistible for use
in eyes’.157 Rhodopsin is remarkably widespread, and apart from its
near-universal adoption in the various visual systems this molecule
has long been known to occur in the bacteria. These organisms exist
as two kingdoms, the Archaea and the Eubacteria, of which a typical
representative is E. coli. The most notable examples of rhodopsin are
in those Archaea referred to as halophiles. As their name indicates,
they live in salt-ponds, where their immense numbers frequently
colour the lagoons red or purple, which is the colour of the pigment
associated with their rhodopsin. The molecular architecture of this
so-called bacteriorhodopsin,158 with seven α-helices spanning a lipid
membrane, is the same as the molecule in our retina. So, too, is its
basic dependence on light, but in the bacterial variety the energy of
the Sun’s photons is used to drive a proton pump159 to transfer hy-
drogen ions (i.e. protons, H+ ) as the basis of the cell’s energy system,
specifically by the synthesis of ATP.
universal rhodopsin? 171
Bacteriorhodopsin has received a vast amount of attention. This

is principally because of its now-classic role as an ion pump and the de-
sire to elucidate its detailed structure as a protein.160 Yet a remarkable
fact and an intriguing inference concerning bacteriorhodopsin have
emerged. The former concerns its widespread distribution. Its occur-
rence in archaeal bacteria is well known, but only recently has it been
appreciated that in oceanic ecosystems this molecule occurs in the
eubacteria (in which organisms it is referred to as proteorhodopsin),
where it appears to play a key role in the harvesting of the sunlight in a
way analogous to photosynthesis.161 Nor does the distribution of
bacteriorhodopsin stop there. As we have seen, various single-
celled protistans, including some green algae, have eye-spots. It is
hardly surprising, therefore, to find rhodopsin in the single-celled
Chlamydomonas and colonial Volvox.162 This story has had its con-
volutions, especially with the identification of a membrane-bound
protein, referred to as chlamyrhodopsin, in the eye-spot. This, how-
ever, has a molecular structure that is markedly different from that
of genuine rhodopsin.163 It is now clear that not only does rhodopsin
itself occur, but it evidently has a bacteriorhodopsin-like structure.164
Bacteriorhodopsin, however, is even more widespread, and it
also occurs in the fungi. One instance is in the parasitic chytrid-
iomycete fungi. Their reproductive cells, known as zoospores, are
equipped with a rhodopsin-bearing photoreceptor that guides this
swimming stage towards the light.165 In many, and perhaps all, these
fungi the rhodopsin employed is equivalent to the bacteriorhodopsin.
The intriguing question, therefore, is whether this type of rhodopsin
is strictly the same as that found in our eyes, or is in fact convergent.
The latter seems to be more likely. First, it has long been appreciated
that there is no discernible similarity in the amino acid sequences
when the two types of rhodopsin are compared.166 Furthermore, de-
spite the overall similarities of molecular architecture there are some
specific differences.167 In their review of rhodopsins, John Spudich
and colleagues168 conclude,
the genome sequence data presently available and the three-

dimensional structures of the molecules themselves argue that
nature discovered retinal [i.e. the vitamin A derivative] twice, and
both times found it useful, when solvated with seven helices, for
photosensory signalling as well as other phototransduction
functions. The two-progenitor hypothesis would require that

archaeal sensory rhodopsins and mammalian rhodopsins have
converged on remarkably similar mechanisms of receptor
photoactivation. Such closely similar mechanisms could result
from ‘likely reinvention’ determined by the inherent properties of
retinal as a chromophore.169
The remarkable similarity of the molecular architecture of the two

rhodopsins may, therefore, well be convergent. If this is the case it is
perhaps not surprising that bacteriorhodopsin is also very sensitive to
substitutions at key sites. Thus, in a way analogous to substitution at
key sites controlling colour vision, so meddling with site 85, for ex-
ample, can completely destroy the ability of the bacteriorhodopsin to
pump protons.170 And the parallels may be even closer, because it ap-
pears that some types of bacterial rhodopsin are also capable of a colour
discrimination,171 suggesting once again the possibility of a universal
property. Wherever life sees light it will be through the agency of a
rhodopsin, and where it sees red it will depend on key amino acid
substitutions. So, too, the sentient inhabitants of Threga IX will also
enjoy those spectacular red sunsets, by camera-eyes, through trans-
parent crystallins, and by means of rhodopsin.
Light reception by organisms is therefore very widespread.
Rhodopsin may have evolved twice, but other proteins can also act
as light-receptors. Opsins may well be best, but the other examples
are also instructive in the context of both light sensitivity and also
convergence. For us, the role of light is effectively synonymous with
vision, yet, as is widely appreciated, there are also perceived cycles
of light and dark. Most notable, perhaps, are the circadian rhythms
of the 24-hour solar cycles, of day and night.172 Also important is
the longer-term seasonality, with the most familiar effects in higher
latitudes of autumnal colours in plants, as well as hibernation or
migration in animals. Such rhythms and changes are very widespread
and are monitored by light-sensitive molecules. Of central importance
are the so-called cryptochromes, cryptic neither in being difficult to
find nor in establishing a photoreceptor function, but because until
quite recently their true structure was problematic. Cryptochromes
are proteins with a flavin component (and linked to a vitamin B2 -based
pigment) that are sensitive to blue light.173 They occur, for instance,
in the retinas of mammal eyes, where they play an important role
smelling convergence 173
in the setting of the circadian clock, effectively the discrimination

between night and day.174 The importance of this clock is familiar
from the disorientating effects of jet lag; more seriously, disruption
of the circadian clock is implicated in major malfunctions during the
night shifts of nuclear reactors and chemical plants.175 Danger might
also lie elsewhere. For example, disruption of circadian rhythms by
the widespread use of artificial light may also be implicated in the
increased prevalence of breast cancers.176
Given their role in light sensitivity, it is unsurprising to learn
that cryptochromes occur also in algae177 and plants.178 What is more
interesting is that the plant cryptochromes appear to be very an-
cient, whereas those of animals are more recent and an independent
invention.179 Such a convergence is less surprising in one sense, be-
cause in both cases the precursors of these cryptochromes lie with
a group of proteins known as the photolyases whose function is to
repair DNA damaged by light, especially by ultraviolet radiation.180
Curiously enough, although humans have cryptochromes most proba-
bly we lack photolyases, thereby increasing our vulnerability to ultra-
violet radiation and, by implication, skin cancers.181
smelling convergence
Given that key molecules required for vision, such as rhodopsin and
the crystallin proteins, evolved in single-celled organisms this sug-
gests that given time182 and the adaptive value of light discrimination
then the evolution of the eye seems to be a near inevitability. And so
fascinating is the story of the evolution of the eye, not to mention our
particular dependence on vision, that we tend to forget other types of
sensory world, even when walking the dog at night with the air full of
bats: to us other senses, such as those of olfaction and echolocation, are
almost closed books. Yet the principle of evolutionary convergences
should encourage us to think, not only of recurrent re-emergences
of given senses, but also about deeper similarities of sensory input,
such as those of hearing, generation of electrical fields, and perhaps
even the sense of smell. In this last case there have been some curi-
ous experiments, such as those by the remarkable Victorian polymath
Francis Galton. Today he is probably best remembered for his inter-
est in hereditary principles and the related application of eugenics,183
the monstrous nature of which was articulated with characteristic
prescience by G. K. Chesterton.184 Galton’s autobiography185 is still
entertaining, even if it reveals an individual to whom self-doubt was

an almost complete stranger.186 Perhaps that may explain some of the
less appealing aspects of Galton. Thus, in his absorbing account of
the recognition of the importance of fingerprints in forensic investiga-
tion, Colin Beavan187 argues that for all his contributions to this area,
driven in part by the hope that fingerprints would yield eugenic in-
sights, Galton was less than fair to the other pioneers, not least to one
Henry Faulds. Indeed, Beavan shows that behind the façade something
unpleasant, but not so rare, lurked: ‘Other people’s success aroused a
venomous jealousy in Galton.’188 Well, well.
In his memoirs Galton mentions a paper he had written, entitled
‘Arithmetic by smell’.189 Galton’s interest was evidently sparked by
what is called non-verbal representation, crudely whether we (and
by implication other species) can think and handle abstract con-
cepts without recourse to words. So Galton decided to investigate
one of the further reaches of mental representation. Here is what he
wrote:
Leaving aside Colour, Touch and Taste, I determined to try

Smells. The scents chiefly used were peppermint, camphor,
carbolic acid, ammonia, and aniseed. Each scent was poured
profusely on cotton wool loosely packed in a brass tube, with a
nozzle at one end . . . A squeeze of the tube caused a whiff of
scented air to pass through the nozzle. When the squeeze was
relaxed, fresh air was sucked in and became scented by the way. I
taught myself to associate two whiffs of peppermint with one of
camphor, three of peppermint with one of carbolic acid, and so on.
Next, I practised small sums in addition with the scents
themselves, afterwards with the mere imagination of them. I
banished without difficulty all visual and auditory associations,
and finally succeeded perfectly. Thus I fully convinced myself of
the possibility of doing sums in simple addition with considerable
speed and accuracy, solely by imagined scents. I did not care to
give further time to this, as I only wanted to prove a possibility,
but did make a few experiments with Taste, that promised equally
well, using salt, sugar, quinine, and citric acid.190
It is easy to smile at Galton’s equivalent to an olfactory abacus,191 yet

these convergences of sensory input may have more subtle evolution-
ary implications. Consider, for example, the star-nosed mole (Fig. 7.7).
figure 7.7 The star-nose mole, a typical mole in terms of its powerful
fore-limbs, but remarkable for its tentacular nose, with its nostrils
clearly visible. (Photograph courtesy of Professor K. C. Catania
(Vanderbilt University, Tennessee).)
We have already met the moles as a splendid example of fossorial con-

vergence among the burrowing mammals. The star-nosed variety is
unusual in constructing its burrows in wet and boggy ground, and it
also spends a considerable amount of time swimming.192 This mole
inhabits the eastern part of North America, but it may be a recent
immigrant because fossils have been found in Poland.193 It is almost
blind, but by way of compensation has an extraordinary nose consist-
ing of 22 mobile and fleshy appendages, somewhat like a star (Fig. 7.8).
This nose, however, is not used for smelling, nor do the tentacles
serve to capture food directly. Rather, as Ken Catania has so elegantly
shown, this remarkable structure retains its sensory powers for the
detection of prey, but is modified in a most interesting fashion.194
The surface of the nose is densely studded with button-like sensory
structures, known as Eimer’s organ (Fig. 7.8).195 It is estimated that
the nose carries about 25 000 of these mechanoreceptors. Not surpris-
ingly, the nose is exceedingly sensitive, and although it is only about
a centimetre across it is supplied with five times as many nerves as
run into the human hand. But the real surprise is that not only is this
nose ultra-sensitive, but more importantly it shows some remarkable
parallels to the eye. In particular, one region of this nose towards the
figure 7.8 Details of the nose of the star-nosed mole. Above, detail of a
tentacle, studded with the sensory Eimer organs. Opposite, electron
micrograph of the nose. The analogue of the fovea is represented by the
two central lobes beneath the nostrils. (Photographs courtesy of
Professor K. C. Catania (Vanderbilt University, Tennessee).)
figure 7.8 (cont.)
central area is especially well supplied with nerve endings. As such it

can be compared directly to that region of the eye known as the fovea,
where the retina possesses exceptional sensitivity. Both this part of
the mole’s nose and the fovea show a comparable degree of nervous
acuity.196 In making comparisons between nose and eye Catania is un-
equivocal about their similarities when he writes, ‘The many surpris-
ing parallels between the somatosensory system of the mole, and the
visual systems of other mammals, suggest a convergent and perhaps
common organization for highly developed sensory systems.’197 Inter-
estingly, in another underground mammal, the naked mole-rat, which
we met earlier (Chapter 6, p. 142) in the context of the convergent
evolution of eusociality, nearly a third of that part of the brain allo-
cated to touch perception (the somatosensory cortex) is rededicated to
the prominent incisor teeth.198 Effectively these teeth take the place
of the usual rat whiskers, and given that the naked mole-rat is for all
intents and purposes blind, it is not surprising that the area of the
brain normally employed in vision appears to have been almost en-
tirely taken over by the somatosensory cortex. So, in a way analogous
to that of the star-nosed mole, these naked mole-rats probably ‘see’

with their teeth.
‘Seeing’ with a nose (or even teeth) allows for some intriguing
speculations. Imagine a planet ejected early in its history from its par-
ent solar system, which as we saw earlier may be an all-too-common
occurrence (Chapter 5, p. 86). A position whirling through the Stygian
gloom of interstellar space might seem to make such a lonely world
an improbable abode for life,199 especially at an advanced level. Prob-
ably so, but let us imagine also an abundance of radioactive elements
in the planetary crust, whose continuing breakdown by decay serves
to maintain subterranean ecosystems at viable temperatures. Perhaps
in caverns, which on Earth extend at least 1.5 km beneath the surface,
any animals would be ‘seeing’ in much the same way as the terrestrial
star-nosed moles and naked mole rats. Some hint of how adept such
hypothetical animals might be also comes from studies of how the
blind mole-rats navigate their complex burrow systems in a way that
would do justice to a rat in a maze.200 Sensory cues may possibly come
from the Earth’s magnetic field, but it also seems conceivable that the
mole-rat is able to build a cognitive map: one more hint of mental-
ity in ‘dumb’ animals (see Chapter 8).201 There are even instances
at the surface of the Earth in which eyes are lost, or more strictly
eaten, without deleterious effects. One particular example concerns
a population of tiger snakes living on Carnac Island, off the western
coast of Australia. Silver gulls attack these snakes and remove the
eyes, but the snakes are in a sense ‘over-designed’ and they are able to
survive202 using other sensory mechanisms.203 It is also worth noting
that while snakes cannot re-grow eyes, where regeneration is possi-
ble, as in arthropods, amputation of a sensory appendage may lead to
its replacement by other appendages with their own neurological (and
sensory) characteristics.204
The remarkable proboscis of the star-nosed mole provides a
number of other useful and relevant insights into the tension between
novelty and constraint in evolution. In one sense the origins of this
mechanosensory nose are not so peculiar; the related talpids, that is
the other true moles, also have highly sensory noses, again studded
with Eimer’s organs. More specifically, comparisons with moles such
as Townsend’s mole (a species of Scapanus) give a good idea of how
the star-nose might have arisen.205 Even so, in its evolved form the
star very much represents a biological novelty. Many animals have
fleshy appendages of various kinds, which routinely employ a partic-

ular developmental pathway that depends on specific developmental
genes (notably one called distal-less). In the star-nosed mole, how-
ever, the appendages develop in a unique manner, without parallel
elsewhere.206 Why this mole should pursue such a novel develop-
mental route when the genetic template for appendage formation is
readily available, and indeed presumably coded for in the limbs of
the mole, may seem rather odd. There are, however, particular con-
straints. Catania and his colleagues suggest that the need to keep the
neural connections, without which of course the nose is functionless,
means that the standard method of ‘let’s build an appendage’ simply
is not available. Finally, remembering the topic of convergence, it is
worth mentioning that the development of the elaborate nose is not
without consequences for the architecture of the skull. In particular,
the chewing apparatus is quite modified (e.g. weaker dentition and
more fragile mandible), and as such is convergent on other animals
that are either vermivores or termite eaters and thus tend to swallow
their prey.207
As noted above, the star-nosed mole does not employ its spectac-
ular proboscis for olfaction. But here, too, those animals that do have
a sense of smell show some striking similarities. So it is that despite
major anatomical differences the basic mechanism of olfaction seen
in the insects (usually located in the third segment of each antenna)
and vertebrates (the nose, of course) depends on nervous structures
known as the glomeruli.208 These act to connect and integrate the
olfactory messages that are then conducted as electrical signals via
the nervous system to the brain, where the discrimination of odours
takes place. To be sure, the proteins (and thereby genes) that serve
to bind the odours, as the first step in olfaction, are different,209 but
as John Hildebrand and his colleagues have emphasized, despite the
molecular genetics being quite different, our sense of smell and that
of the insects (and other arthropods) works on the same principles,210
and so encompasses such specific examples as the sex pheromones.211
Not only do the properties of the insect glomeruli ‘have precise coun-
terparts in vertebrate olfactory bulbs’,212 but Nick Strausfeld and John
Hildebrand go on to remark, ‘Basic types of olfactory information
coding are remarkably similar in both vertebrates and insects’ with
various excitatory responses to odours strongly indicating ‘common
strategies for processing odor information at the level of glomerular
relay neurons’. These writers are fully aware of the implications when
they continue, ‘The primary olfactory centers in the brains of verte-
brates and insects thus appear to share canonical cell arrangements
and common physiological properties . . . Furthermore, vertebrates
and insects show remarkable parallels in events underlying the de-
velopment of glomeruli . . . Do such commonalities suggest that the
glomerulus, as an olfactory functional unit . . . originated in a common
bilaterian ancestor [i.e. to insects and vertebrates]? Or, is common de-
sign and physiology the consequence of convergent evolution? Is there
a uniquely logical response to common selective pressures such that
to construct a glomerulus requires the same set of rules?’213
A definitive answer is not yet possible, but given the differences
in the molecular genetics of olfactory receptor coding, the evidence
points strongly to convergence. This is because the predecessors of
the vertebrates, the amphioxus animal (see Chapter 1), and the in-
sects, that is the aquatic crustaceans,214 do not possess such glomeruli,
which indicates that the method of effectively identical olfaction
arose independently.
Once again it seems that the options are limited. There may
be only a few ways to smell, and, as we have seen in the star-nosed
mole, if a nose is not used for olfaction it can end up as the sen-
sory equivalent of an eye. And again there may be deeper similar-
ities pointing towards the inevitability of evolutionary end results.
The processing of the olfactory signals, and especially the recogni-
tion of particular odours, involves a synchronized neuronal activity
that leads to an oscillatory firing of the neurons. The associated net-
works are not only similar in both the insects and the vertebrates,
but they in turn may also find parallels to what occurs in the process-
ing of visual stimuli in the mammals, at least.215 Further evidence for
more profound similarities between the processes of olfaction and vi-
sion also comes from an investigation of proteins known as arrestins
which, as their name suggests, are involved with the termination of a
physiological process, such as the quenching of phototransduction in
rhodopsin.216 Given that the proteins involved with olfaction have a
basic similarity to such visual proteins as rhodopsin,217 it would not
in principle be surprising to find arrestins also involved with the olfac-
tory transduction pathway. What, therefore, is the more remarkable
is evidence for a specific family of arrestins that is involved in both
visual (eye) and olfactory (antenna) transduction in insects.218 This is
the echo of convergence 181
not the only example of such overlap in molecular transduction in

the two systems,219 and it is further evidence for a deeper, and largely
unappreciated, commonality of these sensory systems.
the echo of convergence

The richness and versatility of the sensory worlds does not end here.
Many animals, in addition to the star-nosed mole, inhabit gloomy and
crepuscular worlds, but generate elaborate and sophisticated fields of
perception that arguably are as refined and sensitive as any visual or
olfactory system. Most familiar in this regard, perhaps, are the ex-
traordinary powers of echolocation possessed by the bats,220 a group
that offers other insights into convergence.221 The basic principles
of echolocation are fairly self-evident, and are of course employed in
a somewhat similar way by many people who are blind.222 So, too,
with cats.223 In the case of the fast-moving bats, however, what is par-
ticularly remarkable is that even when the animal is very close to its
insect prey, where the neural responses can no longer keep track of the
returning echoes, the bat still keeps a lock on its prey. Evidently the
bat must employ a filtering mechanism of some sort, but how this
signal is actually interpreted by the brain is still mysterious.224
Another intriguing feature is that not only will the bat navigate
through crowded vegetation in pursuit of its prey, but the hunted
moths have acute hearing and are capable of evasive action.225 More
extraordinary are those moths that also produce ultrasonic clicks, the
purpose of which appears to be to jam the bat’s sonar.226
Echolocation is not confined to the nocturnal sonar of the
bats. It is also well known in various marine mammals, such as the
dolphin,227 but in the context of considering convergence we shall
postpone a visit to these wonderful animals because of their involve-
ment in the yet more intriguing area of intelligence (Chapter 10).
Concerning the techniques of echolocation there is, however, an-
other fascinating example of convergence that is perhaps rather less
known. This involves those birds that inhabit the deep, dark recesses
of caves, where they often share their domiciles with the bats. No-
table in this regard are the South American oilbirds and the Asian
swiftlets, whose saliva-bound nests are eagerly sought for those who
find bird’s nest soup exceptionally tasty. As their name suggests, the
oilbirds have extensive fat deposits, which if rendered provide an oil
of exceptional quality and purity. These birds attracted the notice of
the great traveller and naturalist Alexander von Humboldt, and sub-
sequently were engagingly described by Donald Griffin.228 Oilbirds
are nocturnal. They are also noisy animals, but can they echolocate
when travelling to or from their perches in the caves? Griffin listened
to them as they poured out of their cave at twilight, remarking, ‘Per-
haps the clicks were call notes, perhaps they were something analo-
gous to profanity, or perhaps they were symptoms of some other avian
emotion whose nature we could not guess.’229 His subsequent exper-
iments, however, showed that the ability to echolocate was genuine,
with the series of sharp clicks clearly audible to humans. Continu-
ing work230 has shown that the resolution of oilbirds’ echolocation
is rather crude, at least when it comes to avoiding discs deliberately
suspended in their flight path.
Interestingly, despite having well-developed eyes the oilbirds
appear to lack binocular vision,231 which is otherwise a general char-
acteristic of birds and convergent on other vertebrates (p. 162). In some
ways this makes the behaviour of the Asian swiftlets all the more ex-
traordinary, since their echolocatory abilities enable them to avoid
much smaller objects.232 Although their echolocation does not rival
that of bats,233 nevertheless in their natural habitat the pitch-dark
caves are filled with flying swiftlets, seldom if ever colliding and read-
ily finding their respective nests. Bats, dolphins, and even some birds,
therefore, have all entered a sensory world that is apparently almost
completely unknown to us (note 227). It is fascinating to speculate
what sort of echolocatory image these animals ‘see’. Although it is
now time to turn to the even stranger world of electrical perception, it
is worth remembering that as the commonalities of convergence con-
tinue to emerge so we may find that apparently alien perceptions, and
even mentalities, may not be as remote as has sometimes been imag-
ined. Even to enter the mind of a bat may not be as difficult as is some-
times supposed. We return to this topic in Chapter 9 (pp. 265–266).
shocking convergence
Of the sensory systems it is perhaps those involving the generation of
electrical fields that are the most remote from human experience. Per-
haps it is not surprising that it has been suggested that the exquisitely
sensitive nose of the star-nosed mole possesses an electrical sensi-
tivity, but this proposal is viewed with considerable caution.234 Even
so, such sensitivity is certainly possessed by some other mammals,
shocking convergence 183
notably the primitive monotremes.235 It is among the fish, however,

that we find not only some of the most remarkable examples of elec-
tric generation (and perception), but also – and by this stage are we
really surprised? – yet more striking examples of convergence. Thus,
electrogeneration has evolved independently at least six times and in
each case entails the modification of muscle cells, albeit from a vari-
ety of locations.236 The phenomenon of electrical generation by fish
must have been shockingly apparent to incautious waders and fisher-
men for a very long time, and certainly at least as far back as classical
times.237 A famous physician, close to the Imperial family of the first
century ad, one Scribonius Largus, wrote of the use of the electrical
discharges of the torpedo ray, a familiar denizen of the Mediterranean
and Atlantic coasts (although it is found in all oceans), for the re-
lief of intractable headaches.238 Such a therapy was also endorsed by
Dioscorides, a celebrated medical authority active at about the same
time as Scribonius Largus, who, despite a reputation for travelling,
spent at least some time in Tarsus.239 He also knew of this treatment,
but more alarmingly employed the torpedo’s electrical capacity for the
treatment of the prolapsed anus.240 Interestingly, the greatest of the
ancient physicians, Galen, disputed the efficacy of this galvanic treat-
ment not only for headaches but also for more fundamental matters.241
Nor was familiarity with the electrical properties of certain fish re-
stricted to the classical world. African tribes, for example, were also
familiar with the therapeutic powers of such fish, and according to
reports provided by Jesuit missionaries and other travellers the inhab-
itants of Ethiopia employed the discharges for a variety of purposes,
ranging from the expulsion of evil spirits to the control of fevers.
The properties of electrical fish also attracted the attention of
both doctors and scientists in nineteenth-century Europe. Among the
former were those intrigued by the surge of enthusiasm associated
with mesmerism and other examples of what was termed ‘animal
magnetism’.242 Of the scientists, the most famous was the physicist
Michael Faraday, who undertook an extensive series of experiments in
the late 1830s. These, not surprisingly, focused on the measurements
of the electrical current but also included observations on the killing
of prey and its ingestion by the fish. Faraday was a thoroughly hands-
on, or in this case, hands-in, scientist. In one such experiment he and
two colleagues had their hands in the tank when ‘suddenly [the fish]
gave a shock which startled us all and was perfectly satisfactory as to
figure 7.9 A mormyrid, or elephant-nose, fish (Campylomormyrus).

(Photograph courtesy of Professor C. D. Hopkins (Cornell University).)
the generality of the discharge. Mr Gassiot evidently felt it least’, but

Faraday dryly continues, ‘. . . I daresay Mr Bradley [the] most.’243
The fish Faraday employed was an electric eel from South
America,244 which belongs to a freshwater group known as the gym-
notids. They show a series of quite remarkable convergences with
an effectively unrelated group, the mormyrid fish (Fig. 7.9),245 which
inhabits the lakes and rivers of Africa.246 In both gymnotids and
mormyrids the twin processes of electrogeneration and electrorecep-
tion are strongly convergent, although it should be noted that the
underlying neurological mechanisms are certainly different.247 These
fish have been the subject of intensive experimentation, and the func-
tional significance of their electric signalling and responsiveness has
given glimpses into an extraordinary electrical world, which might,
however, have analogies with our more familiar sensory modalities.
Experiments typically involve the recording of electrical activity and
observation of behaviour, but while an important strand of investi-
gation concerns studying how individuals react, it is also possible to
‘fool’ the fish with an artificial equivalent, consisting of an electric
dipole connected to a computer that can respond to the electrical sig-
nalling of the fish.248
Almost without exception, the electric organ is derived from
muscle, not from nervous tissue,249 but unsurprisingly the cells
(or electrocytes) are specialized and commonly form a structure equiv-
alent to a stack of gelatinous discs with a rather remarkable series of
interpenetrating stalks.250 As might also be expected, the electric or-

gan itself has evolved. Primitive and advanced arrangements can thus
be recognized, but there is evidence in this organ for both reversion as
well as convergence.251 Control of the electric organ is from the brain,
via particular nerves known as the electromotor nerve axons. The re-
sultant electric organ discharge (or EOD) that is transmitted into the
water varies remarkably in duration, frequency, number of peaks, and
polarity according to the species concerned.252 There is, moreover,
one important difference between the gymnotids and mormyrids
inasmuch as with one exception the mormyrids produce the elec-
trical signal as discrete pulses, whereas the gymnotids produce an
effectively continuous signal as a wave form. In the case of the gym-
notids there is also some evidence that the signals associated with
communication, especially important in the sexual context, may be
generated in a different region of the fish from those concerned with
the imaging of the surrounding environment.253
An interesting possibility is that one of the driving forces to-
wards an increasing complexity of the signals may have arisen as a
defence against those predators that are capable of detecting the elec-
trical impulses. Thus, by shifting the frequency of part of the signal
the fish become more difficult to detect,254 although it certainly does
not render them immune to attack.255 But this is not the end of the
matter, because the enhanced electrical signal can then be utilized in
sexual communication.256 The gymnotid and mormyrid fish therefore
live in what is effectively an electrical world in which the signals are
received by the electroreceptors and then transmitted to the brain,
which allows the animal to perceive inanimate objects,257 especially
when encountered in a novel context,258 and to recognize or commu-
nicate with other inhabitants of the murk, be they friends or foe. It
is also interesting that the two types of electroreceptor found in the
skin, to detect direct and alternating currents respectively, are again
very similar in both gymnotids and mormyrids.259
Because the environment in which these fish swim is effectively
one of electrical uproar arising from the competing signals of other
fish, it is hardly surprising that each species produces a highly specific
electric signal. Such call signals, analogous to those issued by individ-
ual aeroplanes, are doubly important because the general electrical
racket produced by dozens, if not hundreds, of electric fish nearby is
further augmented by the frequent tropical thunderstorms with the
associated lightning, adding to the ‘electrical cacophony [and] . . . back-

ground roar’.260 Not only is there distinction between species, but in
some cases either sex or the juveniles of a particular species also pro-
duce a characteristic and distinct waveform. For the most part, how-
ever, the signals of individuals of a given species appear to be identical.
This may, however, be a simplification because in one species of gym-
notid the signatures of individuals are discernibly different, a factor of
considerable importance in social contexts such as defence of home
territory.261 But there are exceptions, effectively a convergence, where
up to four species produce a similar electrical signal. It comes as little
surprise to learn that in such cases the fish aggregate in schools.262
In this electrical Babel separation of signals is, therefore, essen-
tial. What happens, however, if two fish of the same species produce
their respective signals simultaneously? Potentially the result would
be problematic, for the two signals would lead to destructive interfer-
ence, and the whole point of signalling would be lost. To circumvent
this problem the fish have evolved a mechanism that is referred to
as the jamming avoidance response (JAR) in which the fish changes
its signal frequency.263 Sensible enough, but two things are remark-
able about this avoidance of jamming. First, the fish are astonishingly
sensitive to potential interference. They will respond to signal mod-
ulations with an amplitude difference of 0.1% and a timing disparity
of 400 nanoseconds or less, each fish shifting its signal pattern in a
few microseconds.264 Second, and even more remarkably, the algo-
rithm used by the gymnotids and mormyrids to shift the signal has, of
course, evolved independently but it is identical. Nor do the conver-
gences end there, because computationally similar neural algorithms
also occur in the owl, where acuity of sensory perception is acoustic
rather than electric.265 In the owl,266 not surprisingly, neither the neu-
ral mechanisms nor the implementations are identical to those of the
electric fish, but this does not prevent the emergence of deeper sim-
ilarities. So when we look at the general arrangement in gymnotids
and mormyrids, it is clear that even though different parts of their
respective brains are employed for the interpretation of the electrical
signals there are also a number of significant similarities in the neural
circuitry.
Given the degree of convergence between the gymnotids and
mormyrids267 in terms of electrical activity it comes as less of a sur-
prise that they occupy similar environments, living in waters that are
either muddy or otherwise opaque ‘blackwater’, and typically showing

nocturnal activity.268 In such a setting the powers of vision are decid-
edly restricted269 and, at least so far as the mormyrids are concerned,
an early investigator, M. R. McEwan, remarked, that ‘The sense of
sight of these fishes is not at all keen.’270 This observation was con-
sistent with her study of the mormyrid retina, which demonstrated
a number of peculiar structures that probably represent adaptations
to the low levels of light.271 Aspects of their life cycles and reproduc-
tion are also strikingly similar, including the embryological stage.272
Although the body form of each group is rather different, in terms of
their ecology, and especially feeding, there are important parallels in
these fish. Thus there is a recurrent tendency to evolve elongate and
tubular mouths (Fig. 7.9) with weak jaws and a feeble dentition. In
arriving at this arrangement, these fish have opened a new resource
in the form of a rich bottom-dwelling fauna of worms and larvae that
more orthodox arrangements of mouth and jaw would find difficult to
exploit.273 Moreover, in at least some instances the prey is detected by
the generation of the electrical field.274 Indeed, Tyson Roberts iden-
tified an even closer link between feeding and electrical production
when he wrote, ‘I would go even further, and suggest that the interre-
lation between electrical faculties and feeding played a decisive role
in the initial divergence of the gymnotoids and mormyroids from non-
electrically specialized ancestors,’275 as they entered the twilight zone
of the Congo and Orinoco. Nor should this mud-grubbing life lead
us to underestimate the extraordinary nature of these electric fish.
As already indicated, the jamming avoidance mechanism is clearly
a complex response involving a sophisticated set of adaptations. By
now it should be clear that these fish occupy a rich world of electri-
cal signals and social communication and, relative to at least some
other fish, one of enhanced perception. This is consistent with a neu-
ral plasticity that in filtering out the predictable indicates that these
fish occupy a contextural environment in which learning and memory
are the norm.276
So it appears that, in a way analogous to the star-nosed mole,
these fascinating fish ‘see’, not tactilely but in an electrical world.277
So, too, it seems reasonable to think of other electrically sensi-
tive animals, such as the sharks, also constructing an electrosen-
sory landscape.278 Nor need this parallel, however unfamiliar it is
to humans, surprise us, because when we look at the brain activity
12.0
Elect. sim.
Visual sim.
7.0
Position along fish
2.0
−3.0
0.0 60.0 120.0 180.0
Position in brain
figure 7.10 Seeing in the mormyrid fish, in terms of visual stimulation
(dots) and electrical stimulation (asterisks) with respect to the response
of a moving image or electrical stimulation along the body length versus
the corresponding response position in the brain, specifically tailward of
the anterior margin of the optic tectum. The overlap and parallelism of
either stimulation suggests that the sensory modalities are in one way
equivalent. (Redrawn from fig. 7A of W. Heiligenberg and J. Bastian
(1984). The electric sense of weakly electric fish. Annual Review of
Physiology, vol. 46, pp. 561–83, 1984. With permission, from Annual
Review of Physiology, vol. 46, c 1984 by Annual Reviews, www.
annualreviews.org, and the authors)
associated with both the electrical and visual systems of a mormyrid

there are obvious similarities (Fig. 7.10). Despite their weak vision
(note 271) these fish are not actually blind, and in at least some of
the gymnotids there is evidence that the electrical and visual signals
are integrated.279 Indeed, some have argued that the parallels between
electroreception and sight can be pursued to the point where elec-
troreception yields some sort of analogy to colour.280
It is a fascinating, and largely untouched, problem to begin to
speculate just how different are the sensory modalities of animals. A
relatively familiar example of this is the ability of some people to use

one sense perception to appreciate or understand another, such as see-
ing music in terms of colours, or more rarely as taste or shape.281 So,
too, Galton’s experiments (note 189) using olfactory stimuli for math-
ematical calculations indicate the existence of equivalent modalities.
So, too, experimental procedures in which animals are either deprived
of one modality, e.g. sight,282 or subject to transplants,283 reveal ev-
idence for neural plasticity. Thus, a part of the brain that in normal
circumstances would be involved with vision is now able to process
auditory information. So, too, transplantation, involving transsexual
exchange of appendages (specifically feeding claws) in fiddler crabs,
induces new sensory modalities that linked to the central nervous
system invoke novel behaviours. More specifically in the case of the
electric fish, it has been suggested that the organization of nervous
centres involved with electrosensory perception (the electrosensory
line lobe) has intriguing similarities to mammalian hearing (in the
dorsal cochlear nucleus).284 What is important in this context is that
the mammalian system is an evolutionary innovation, since equiv-
alents to this neuronal arrangement do not exist in the reptiles or
amphibians. Similarities of neuronal arrangement may not guaran-
tee, of course, equivalence of perception. There is more than a hint,
however, not only of emergent systems in advanced organisms (yes,
evolutionary progress is real), but also of an underlying commonal-
ity that might presuppose that extraterrestrials with nervous systems
will hear, see, and smell in very much the same way as we do, and if
that is so will also possibly have similar mental processes.
It should be pointed out that electric fish have not abandoned
other sensory mechanisms. Mormyrids, for example, also have an
acute sense of hearing.285 These remarkable fish, however, show an-
other peculiar feature that is probably instrumental in sound recep-
tion. In a number of fish the sense of hearing is evidently improved by
connecting the swim-bladder to the skull through a series of bones (the
weberian ossicles), which presumably serve to amplify the sound. This
arrangement is found in various fish, such as the minnows, as well
as the gymnotids. The mormyrids, in a sense, go one better because
their embryological development leads to parts of the swim-bladder
coming to lie, as isolated sacs, against each ear region, and these sacs
are probably instrumental in their exceptionally acute hearing.286 It
is hardly surprising to learn that at least some mormyrids are adept at
producing sounds, which arise from the action of a drumming mus-

cle connected to the swim-bladder. In the mormyrid Pollimyrus the
courtship sounds, described as grunts, moans, and growls, are broad-
cast by the males and evidently form an important complement to
their electrical signalling.287 It appears, therefore, that even if an an-
imal experiences a rich sensory world, be it visual or electric, other
modalities are by no means neglected and may show intriguing paral-
lels. So, too, with hearing, to which we now turn.
hearing convergence
Hearing, of course, depends on the transmission of vibrations, and
their transduction into electrical signals that in ways analogous to
sight and smell are interpreted as sound. For most of us hearing is
almost as important as sight, and since the invasion of land by the
tetrapod vertebrates, effectively beginning with an amphibian-like
creature,288 this sensory modality to detect and track airborne sounds
has been extensively developed.289 Yet the evolutionary evidence sug-
gests that the evolution of the tetrapod ear, with its tympanic mem-
brane and associated features, such as pressure-relief mechanisms
and differential ossification to provide an otic capsule that helps to
refine the auditory mechanism, has evolved independently several
times among the vertebrates.290 More specific convergences are also
found. For example, in the vertebrates impinging sound waves are in-
geniously amplified via the hairs located within the spiral-shaped part
of the inner ear known as the cochlea.291 This structure contains spe-
cialized hair cells that are involved with auditory perception. While
the differences between the arrangement found in birds as against
mammals have long been appreciated, it is also evident that there are
important convergences.292 So, too, within the mammals themselves
there are some interesting examples of convergence, perhaps most
notably among the many subterranean species that, as already noted
(pp. 139–141), provide many other insights into convergence. In their
fossorial environment the sensitivities of hearing, not surprisingly, are
shifted towards the lower sound frequencies. Even so, vocalizations
and hearing carry only for about five metres.293 Across longer dis-
tances communication is by seismic pulses, e.g. head-banging or foot-
drumming, and does not appear to utilize the ears.294 The auditory
systems of the underground mammals nevertheless show some strik-
ing convergences linked to the shared mode of life. Unsurprisingly, the
hearing convergence 191
external ear lobes (or pinnae) are greatly reduced, if not absent. Within
the middle ear there are a number of parallel changes to meet the need
for detecting low-frequency sounds; far from this part of the ear being
vestigial it is actually highly adapted.295 So, too, within the inner ear
there are further convergences, such as those concerning the arrange-
ment of the semicircular canals that probably play a significant role
in navigation through the maze-like world of these mammals.296
Other animals hear, but in groups such as the insects this has
long been assumed to be a passive process, lacking the sophisti-
cated physiological features found in mammals, in which acuity of
hearing is sensitive to the animal’s metabolism and also shows a
feature known as autonomous vibration, which is important in the
process of sound amplification. This now appears to be incorrect. In
such dipteran insects as the mosquito the equivalent of the ear is a
complex antennal structure, including the so-called Johnston’s organ.
This structure is richly endowed with nerve cells, is correspondingly
very sensitive to sounds,297 and now it transpires that the hearing
of mosquitoes is much more similar to that of mammals than was
realized. As Martin Göpfert and Daniel Robert298 remark,
Like vertebrates, mosquitoes actively improve hearing by

enhancing sound-induced vibrations in a frequency- and
intensity-dependent way. Considering the vastly different
anatomies of mosquito hearing organs and vertebrate ears, such
similarity in the mechanical processing of sound is surprising . . .
it seems reasonable to assume that active auditory mechanics in
vertebrates and insects evolved independently and in parallel.299
Nor is the versatility of the Johnston’s organ confined to land. In wa-

ter it is capable of detecting pressure variations, as in the whirligig
beetles. When diving through water these insects are evidently capa-
ble of sensing pressure waves produced during their swimming that
are reflected off adjacent objects. Not only does this help to prevent
collisions, but it has also been noted that even in the densest flocks
of gyrating whirligigs the individuals never crash.300
Other insects have hearing but typically this depends on a tym-
panal arrangement.301 This is analogous to our eardrum, being com-
posed of a thin sheet of cuticle. Interior to this is an air-filled cavity
and a complex structure known as the chordotonal organ.302 Such an
arrangement is found in many insects, such as the moths,303 as well as
the group (the orthopterans) that includes the grasshoppers, cicadas,

and crickets. The songs of the orthopterans are familiar to us, and
serve them for sexual communication. For the crickets and their rel-
atives, however, such broadcasts carry risks, not least the unwanted
attention of parasitoids. These include a particular group of dipteran
flies known as the ormiinids (e.g. Ormia), the females of which seek
out hosts to lay maggots on or near the luckless cricket.304 Now, as we
have seen, the standard dipteran ‘ear’ is in the form of the Johnston’s
organ. This is, however, ill-equipped to register the intense and high-
frequency sounds generated by the crickets. One can almost guess
the solution, because the ormiinids have evolved a tympanic ear, lo-
cated behind the head, which is strongly convergent on that of the
crickets.305 As the investigators aptly remarked, ‘for a fly to act like a
cricket, it must hear like one.’306
This example is also interesting in several other respects. First,
it is a rare example of acoustical sexual dimorphism; after all, it is the
female carrying her maggots that needs the more acute hearing. Why
the males also have good audition is less obvious, but may result from
the common need to avoid predatory bats.307 In addition, despite their
tiny size and relative simplicity, the acuity of these ears and their di-
rectional ability are little short of astounding.308 The ormiinids belong
to a larger group of dipterans, the tachinids. There are thousands of
related species in this group, and all are parasitoids. In contrast to the
ormiinids, however, they rely on visual and olfactory cues in search
of their hosts. Why then did the orminiids evolve this ear? Presum-
ably it was because the nocturnally singing crickets would otherwise
be impossible to locate; so their pinpoint accuracy in hearing opens
up a whole new resource, made possible by the ‘invention’ of an ear.
So far as the orminiids are concerned, this is an evolutionary novelty,
but it is necessarily convergent on its host. Nor is this some fluke-like
event. As we have already seen, the tympanal ear of insects is blatantly
convergent and is an evolutionary inevitability given the co-option of
sensory stretch receptors that are inherent to this sort of body plan,
that is one consisting of an articulated exoskeleton. More specifically,
the orminiid ‘solution’, which is how to hear its host, has evolved in
another group of parasitic flies known as the sarcophagids, who home
in to the sound of cicadas.309
These examples of convergence, interesting as they are, are
of more general importance because they lead to a consideration of
hearing convergence 193
deeper similarities of sensory perceptions. In particular, the process

known as transduction, whereby a mechanical force (e.g. sound) is
converted into an electrical signal, is evidently the same in insects
and vertebrates.310 This, in turn, however, begs the question of how
many molecular alternatives, in this case the ion channels, necessary
to generate the electrical signal are really available, and the extent
to which such channels and the component proteins that form them
have been recruited for this purpose from some pre-existing function.
Similarly, as Martin Göpfert and Daniel Robert remark (note 298), the
genes (notably Math 1 and atonal)311 involved in vertebrate hair-cells
and the insect Johnston’s organ are the same. This commonality, how-
ever, may be less informative than might first appear. Thus, the atonal
gene is involved both with the development of stretch receptors (i.e.
the chordotonal organ) from which the insect tympanal ear is derived
(notes 302 and 305) and photoreceptors,312 as well as the olfactory ner-
vous system.313 A parallel, and more famous example, is the control of
eye development by the Pax-6 gene.314 It would be a mistake to think
that this gene ‘makes’ an eye, any more than atonal ‘makes’ ears:
they are really switches, and in the case of Pax-6 its primitive func-
tion may have been involved with the differentiation and patterning
of nervous systems.315 This would explain its role, not only in eye for-
mation, but also its involvement in olfaction,316 the brain,317 and the
pituitary.318 As for many developmental genes, a role in the nervous
system need not preclude co-option, and this might explain its appear-
ance even in the pancreas (see also p. 240, Chapter 9).319 To be sure, the
compound eyes in arthropods (and almost certainly their convergent
equivalents in annelids and molluscs) and camera-eyes of vertebrates
and cephalopods (and almost certainly their convergent equivalents
in heteropod gastropods and alciopid polychaetes) are all controlled320
in one way or another by Pax-6, but the eye structures themselves,
in terms of macroscopic arrangements, nevertheless, have evolved by
convergence. So, too, with ears.
I argued above that any alien astronomers would use a camera-
eye, and when they ask for a drink we can be pretty sure the
mechanism of hearing will be strangely familiar. So, too, as they sniff
the gin and tonic they will register the same olfactory signal of ju-
niper berries, while convergence of taste should lead them to close
momentarily those camera- eyes. And as already indicated, there will
be yet deeper similarities, not only in transduction mechanisms using
recurrent protein designs, but also the way in which meaning is con-
veyed and recognized. Human speech, and thus hearing, is well known
for what is referred to as ‘categorical perception’. That is, despite a con-
tinuum of sound we divide the signal into discrete categories. So, too,
with other vertebrates, such as birds,321 and also insects.322 As Robert
Wyttenbach and his colleagues remark, ‘categorical perception may
be a ubiquitous feature of perceptual systems in many animals, in-
vertebrate as well as vertebrate.’323 Star-nosed moles and stridulating
crickets, echolocating bats, and electric mormyrids are all reminders,
not only of a rich sensory universe, but also of one in which emerging
mentalities may find fertile ground for evolutionary innovations but
ultimately shared experience.
thinking convergence?
When we reconsider this range of animals, embedded in their respec-
tive environments of complex sensory perceptions, it seems safe to
conclude that, despite what to us are largely alien means of percep-
tion, the many analogies that exist with our sense of vision are a pow-
erful argument for the importance of convergence in evolution. But
is this so important? At one level the conclusion might seem rather
trivial. After all, it only reminds us that the world is subject to con-
straints; nervous systems interpret the sensory input, be they from
noses, eyes, ears, or electric organs, and they all have a fundamen-
tal similarity of operation. Hence the trajectories towards any style
of perception in any possible evolutionary history are indeed greatly
restricted. Convergences are inevitable, but again, so what?
There is, however, something else worth knowing. Let us re-
turn to the mormyrid fish, but this time to look at their brain.324 Not
only does it have a strange, almost crystalline, internal structure, but
also a well-developed series of lobes (Fig. 7.11). Most remarkable in
the latter context is the enormous enlargement of the cerebellum (no-
tably the region known as the valvula cerebelli), although a number of
the other regions (e.g. the acoustico-lateral area, mesencephalon, and
corpus cerebelli) also contribute to its extraordinary size. Consider for
example the remarks by R. Nieuwenhuys and C. Nicholson when they
write, ‘Opened the skull of a mormyrid fish reveals, however, a totally
different picture [from the standard teleost brain, say of a trout]. It is
immediately apparent that the brain is of a relatively enormous size
and that none of the structures mentioned above [in the basic brain]
thinking convergence? 195
C3
C4 Va
CL
ELL
C2 C1
Cr TL
figure 7.11 Longitudinal section through the brain of a mormyrid fish.

Note, in particular, the immense size of the cerebellum, which
comprises all parts of the brain above the dotted line. The various parts
of this hypertrophied structure include the remarkable valvula cerebelli
(Va), the central cerebellar lobes (C1 to C4), the crista cerebellaris (Cr),
the electrosensory lateral line lobe (ELL), the transitional cerebellar lobe
(TL), and the caudal cerebellar lobe (CL). Total length of the brain is
about 12 mm. (Reproduced from fig. 7 of Nieuwenhuys and Nicholson
(1969; citation is in note 324), with permission of R. Nieuwenhuys.
Photograph kindly made available by Professor C. C. Bell, Oregon
Health Sciences University.)
can be readily seen.’325 In Fig. 7.11, note, in particular, how the cerebel-
lum ‘flows’ over much of the rest of the brain, in a manner reminiscent
of the massive neocortical expansion of the human cerebral lobes. Not
only are these lobes enormously enlarged, but in addition to this hy-
pertrophy the total area of the brain is further increased by both a series
of deep folds and a series of fine ridges. It is all the more intriguing to
note that in relation to its body size the mormyrid brain is compara-
ble to the disproportionately enormous brain within our skulls. Not
only that, but more than half the consumption of body oxygen in the
mormyrid is accounted for by the brain, a figure three times greater
than for humans.326 And this is hardly surprising; as already noted,
these fish live in a rich electrical environment, requiring sophisticated
neural processing involving recognition of other fish and navigation in
near-darkness through a cluttered and potentially lethal environment.
And the gymnotids, with their convergent electrical system; what of
their brain sizes? Certainly their enlargement is less dramatic than in

the mormyrids, but even so they too have – in comparison with most
other fish – larger than average brains.327 The enlargement of the gym-
notid brain is located in a region known as the torus semicircularis,328
and this appears to be the principal area of electroreception. Even so,
the overall similarities of the cerebellum-like structures in both gym-
notids and mormyrids are strikingly equivalent in terms of sensory
processing and thus reinforce the degree of convergence in these two
groups of electrosensory fish.329
If brains can get big independently and provide a neural machine
capable of handling a highly complex environment, then perhaps there
are other parallels, other convergences that drive some groups towards
complexity. Could the story of sensory perception be one clue that,
given time, evolution will inevitably lead not only to the emergence of
such properties as intelligence, but also to other complexities, such as,
say, agriculture and culture, that we tend to regard as the prerogative
of the human? We may be unique, but paradoxically those properties
that define our uniqueness can still be inherent in the evolutionary
process. In other words, if we humans had not evolved then something
more-or-less identical would have emerged sooner or later. Such an
idea is dismissed out of hand by nearly all evolutionary biologists, but
that is exactly what the next two chapters aim to show.
8 Alien convergences?
I have just woken from a hideous nightmare, the sort of dream that
haunts you for days afterwards. I have visited, or been sent to, a terri-
ble world, and to make things worse it had strange parallels to ours,
only it was alien and mechanical. I am standing on a wide highway,
with robot-like workers rushing in either direction. On either side of
this stream miniature versions scurry everywhere, patrolling the mar-
gins, ceaselessly on guard. Such is the flow of traffic that the earth
has long since been beaten flat; not a shred of vegetation obstructs
the road. In one direction the army of helots marches empty-handed,
but towards a truly immense forest which I can see only dimly as
there is something not quite right with my vision. My sense of smell
is, however, overwhelmingly acute, and wave after wave of olfactory
messages urge me to obey the twin commands of unceasing work and
obedience. Still looking towards the distant forest, I see that from
this direction the army returns, each worker scurrying forward with
a parcel of food. New olfactory orders are issued, and I am impelled
to follow the returning stream. After walking for hours I am driven
into a set of huge subterranean chambers. Here the activity is at a
frenzy, yet there is also an underlying order. My senses reel: from the
ceilings hang huge clumps of what look like some sort of fungus, but
it is being tended with the utmost care and attention. Now I can see
workers busy with weeding; others seem to be applying herbicides,
and yet more are either collecting pieces of the fungus or extending
the zones of cultivation. Yet for all its strangeness this world is a lit-
tle too familiar for comfort. The workers rushing past me are indeed
literally bug-eyed monsters, sinister antennae constantly in motion,
with hardened carapaces, and in some cases carrying miniature ver-
sions of themselves. But it is also clearly an organized society, a vast
underground city entirely dependent for its well-being on the cease-
lessly maintained fungus farms. As I awake from this dry and soulless
world, I realize suddenly that I have not been among the extraterres-
trials, but have been mingling with the leaf-cutting ants who quite
independently of humans have invented agriculture.
198 alien convergences?
figure 8.1 Attine ants at work. Left: with mandibles poised above a
leaf. Right: with a minima. (Photographs courtesy of Cameron Currie,
University of Kansas, Lawrence.)
down in the farm

What, at first sight, could be more specific than agriculture? To be-
come a farmer entails a series of familiar processes, from maintenance
of gardens, transport, weeding, application of herbicides, manuring,
cropping, to the exchange of cultures. That is effectively how we pur-
sue our agriculture. So, too, and convergently, do the leaf-cutting ants
(Acromyrmex and Atta)1 that flourish in Central and South America.2
It was they who were the subject of my Vorstellung at the beginning
of this chapter. As their common name suggests, these remarkable
insects (Fig. 8.1) use their sharp mandibles to excise pieces of leaf3
that are then transported back to the communal nest. Although typi-
cally associated with forests, these leaf-cutting ants have managed to
extend into desert environments.4 In some species there is a remark-
able division of labour, in which a rather small number of worker ants
leave the nest early, climb the trees, cut away the leaf pieces, and then
drop them to the ground. There another group, arriving slightly later,
find the pieces, cut them smaller, and carry them to the ‘road’, where
a third group of ants transports the vegetation back to the nest.5 An-
other example of such division is either the ‘bucket brigade’ whereby
some species of ant hand over plant material at junctions of the ‘road’
system6 or the formation of caches when supply exceeds the rate of
down in the farm 199
processing the leaves within the nest.7 Travel is often via well-defined
and more-or-less permanent paths,8 which may stretch for consider-
able distances, sometimes 100 metres. On the main thoroughfares the
ground may be cleared to form a smooth highway that in turn maxi-
mizes walking speeds. In these social insects three distinct castes are
recognized, and in addition a miniaturized variety (the minims) may
also occur. Some of these minims spend time working in the gardens,
as well as riding on the larger ants, where they engage in grooming
and cleaning activities. Another important role, only recently appre-
ciated, is to patrol the edges of the trails along which the foragers
march, on the lookout for danger and when appropriate issuing an
alarm.9 Yet others ‘hitch-hike’ on the leaf fragments as they are being
carried back to the nest, in part probably to defend the leaf-carriers
from attack by parasitoid flies (known as phorids) and possibly also
to begin the preparation of the leaf prior to its arrival at the nest.10 In
the nest itself, the soldier caste is vigorous in defence.
The collected leaves are not eaten directly11 but are used to pro-
vide a mulch for fungus gardens that are located within the nest. The
leaves, of course, are fresh and the initial preparation, which as al-
ready noted may start during transport, includes a stripping away of
the outer waxy layer. This process, achieved by a sort of licking, also
appears to inhibit the activity of associated microorganisms, the con-
trol of which is a central necessity to the health of the fungus farm.
Thereafter the leaves are shredded and pulped, and at this stage are
ready for fungal innoculation.12 The saprophytic activities of the fungi
break down the plant material, especially the resistant cellulose, and
so provide an edible crop for the ants.13 The gardens are subject to
careful and ceaseless maintenance. Weeding,14 especially of infected
areas, is undertaken principally by tiny ants (the minims). In weed-
ing several minima typically loosen the offending item before it is
removed by larger ants. In addition to weeding, the minima also en-
gage in grooming, that is the removal of alien spores. If necessary the
ants will also transfer the fungi to parts of the nest with more suitable
humidity15 and temperature.16 In addition to weeding there is also the
activity of pruning, again to encourage the harvest.17 The application
of fertilizers is in the form of a manure, an excrement rich in nitrogen
as well as an enzyme supplement.18 The harvest is ready and, in the
more advanced types of cultivation, cropping involves the cutting off
of the knob-like ends of the fungus, which are rich in protein, sugars,
and other compounds.19
As on our own farms there is, however, a recurrent danger of
invasion by pathogens. For the ants’ fungal gardens the principal risk
comes from a virulent parasite, also a fungus, Escovopsis. If it is not
controlled, the garden is soon converted into a mouldering and black-
ened ruin.20 How do the ants avert disaster? They apply the equiva-
lent of a herbicide, specifically in the form of an antibiotic.21 This is
derived from streptomycetean bacteria, the filaments of which grow
attached to the bodies of the ants.22 In addition to these fungicides
it is likely that other chemicals, including again antibiotics, secreted
by the ants themselves, also help to inhibit the growth of unwanted
bacteria and fungi.23 Both provide defences against infestation, but in-
terestingly the ants show a trade-off whereby newly imported leaves
are protected by the ants (mostly minims) using their own secretions.
This helps to provide an all-round microbial defence. However, in the
older parts of the fungus garden, where the fungal biomass is higher
and presumably more vulnerable to attack by the parasitic Escovopsis,
the larger ants responsible for this area of maintenance rely on the
bacterial defences.24 Nor are these the only risks faced by the ant-
gardeners: on occasion a specialized predator aggressively sweeps in,
expels the attine ants, and usurps the garden.25 These so-called ‘agro-
predators’ are a particular species of myrmicine ant, unrelated to the
leaf-cutters and related attine ants. In addition to usurping the gardens
these myrmicines have been observed to place their larvae adjacent to
those abandoned by the vanquished attines, where most probably the
latter larvae act as a convenient source of protein. Yet more terrible
fates await other attine nests when their citadel falls, after stiff resis-
tance and appalling casualties on both sides, to army ants. Here part of
the resistance entails plugging nest entrances and building barricades
with leaf fragments.26
military convergence
In parentheses, one should note that the army-ant syndrome has
evolved independently several times.27 This particular example of
convergence is important for two reasons. The first is simply
connected to the intrinsic fascination of a sophisticated biological
organization, replete with militaristic metaphors and an associated
military convergence 201
folklore of marauding columns of invincible ants sweeping all be-

fore them. In fact, in some parts of the world an invasion of a house
is welcomed as it sweeps out pests and other unwelcome visitors,
and for the most part the army ants offer no danger to humans.28
Arguably of greater importance is the emergence of a complex sys-
tem that encapsulates a remarkable social system and thereby de-
fines a different sort of collective intelligence.29 As such the army
ant adaptive syndrome has several noteworthy and recurrent features.
The most celebrated aspect is the relentless mobile columns, capable
among other things of building bridges constructed of their own bod-
ies (see also note 32), and self-organized into an advancing front that
can span as much as 20 metres of tropical forest floor. Animals flee,30
but those either trapped or overwhelmed are rapidly dismembered,
the pieces being despatched back to the nest by exceptionally effi-
cient teamwork.31 Such transport of the dismembered pieces of prey
is referred to as super-efficient, inasmuch as the total transported to
the nest typically involves cooperative efforts and exceeds in bulk
what the individual ants could achieve, however the piece was di-
vided. Such super-efficient transport is one of the convergent features
of army-ant behaviour. It is also important to note that recruitment to
the team is highly dynamic, with individuals ‘choosing’ involvement.
The nest itself provides protection for the young, yet it is built of the
living bodies of the ants.32 Interestingly, in these bivouacs (Fig. 8.2),
which are temporary because the colony is highly nomadic, there is a
careful thermoregulation of nest temperature.33
In the context of this book, however, the principal topic is not
convergence per se, but the attempt to establish the likelihood of the
repeated emergence of complex biological systems. Thus it comes as
little surprise to learn that population sizes of some leaf-cutter ant
colonies are enormous, with the inference that such numbers can be
sustained only by the ants’ careful system of fungal agriculture. Yet,
these ‘cities’ are overtopped in numbers by some species of army ant,
notably Dorylus, where the population of a single colony can reach
20 million workers. Such enormous numbers seem to be a result of
both growth and aggression, and as such emerge as a result of an arms
race.34 Even though genuine arms races in natural biological systems
are probably rather uncommon,35 such examples as the army ants
awake uneasy echoes of human organization.
figure 8.2 The bivouac of the army ant Eciton. (Photograph courtesy of
David Kistner, California State University, Chico.)
Not only is this true for the army ants, but study of other social
insects may reveal other interesting evolutionary parallels. Consider,
for example, the advanced eusocial bees,36 with their familiar division
of labour, notably the queen and pollen-gathering workers. These an-
imals have fascinated humans for millennia. In England it is still the
custom, when the beekeeper has died, for a friend to go to the hive and
tell the news to his or her bees. In a different way, in past centuries the
industry and organization of the hive invited obvious political analo-
gies. Now this fascination extends to evolution with the realization
that bees have cognitive capabilities and a plasticity of behaviours
that otherwise are known only in the vertebrates,37 and, no less sur-
prisingly, other features such as exhibiting sleep-like states.38 Or is it
surprising? Sleep may have several functions, but one widely agreed
purpose is the need to consolidate memories.39
In any event, the complexity of bees not only has some intrigu-
ing parallels to vertebrate mental processes, but has been arrived at
military convergence 203
by an independent evolutionary trajectory that may have an unnerv-

ing parallel in our hominid history. This is because there is evidence
that a once-wide diversity of eusocial bee groups has been whittled
down through competition to only two groups (the apinid honey40 and
meliponinid stingless bees41 ) over geological time.42 Interestingly, al-
though the evolution of advanced eusociality in these two groups of
bees has generally been regarded as stemming from a common an-
cestor, recent work suggests that they arrived at this arrangement by
convergence.43 Moreover, Michael Engel remarks that it appears that
during the
geological expanses of time, aggressively foraging advanced-

eusocial species had a negative impact, not only on non-social
species44 but also and especially on other advanced eusocial
lineages. A similar situation is reported from another highly
competitive, social species (albeit not eusocial) for which the
fossil record indicates the aggressive exclusion of related social
genera and species that may have caused their extinction – namely
Homo sapiens . . . the only survivor of a prior hominid
radiation.45
‘Big, fierce societies’46 may lead not only to the emergence of col-
lective intelligences but to ones that are, from our point of view, so
alien as to be useful in envisaging extraterrestrial societies. In this
context some remarks by Nigel Franks, a leading specialist on army
ants, are intriguing, if also in places self-confessedly speculative. In
reviewing the extraordinary social structure of these insects he ob-
serves how it is among the army ant colonies ‘we see the emergence
of flexible problem solving far exceeding the capacity of the indi-
vidual’, a capacity that finds no rival in other animals, apart from
humans. This, as Franks stresses, presupposes effective communica-
tion, so that ‘intelligence, natural or artificial, is an emergent property
of collective communication . . . This is exactly what happens when
army ants pass information from individual to individual through the
“writing” and “reading” of symbols, often in the form of chemical
messengers or trail pheromones, which act as stimuli for changing
behavior patterns.’ So it is that we can see a transformation of a sys-
tem from being hard-wired to flexible via ‘increasingly sophisticated
patterns of communication’.47 Such a pheromone-based society48 may
be one of the few genuine alternatives to the emergence of human-like

intelligence, which as discussed in the next chapter is itself almost
certainly convergent. In genuinely speculative mode Franks draws at-
tention first to eye reduction (each eye to a single facet in Eciton),
which poses the question as to how the colony can navigate, given
the vestigial eyesight of each worker. He continues,
In my wildest dreams, I imagine that the whole swarm behaves

like a huge compound eye, with each of the ants in the raid zone
front contributing two lenses to a 10- or 20 m wide “eye” with
hundreds of thousands of facets. Each ant has possibly only the
slightest directional preference, but through tactile signals and
trail pheromones these preferences might be collated and
amplified across the swarm. In this way the army ants could
comprise a parallel-processing computer of intriguing yet
awesome simplicity.49
Since writing this, Franks has continued his exploration of the

convergent organization of army ant societies, and has also presented
an absorbing analysis of the self-organization of the swarm raids. As
he points out, the rules for a raiding front can in themselves be quite
simple, even though the net result is uncannily ‘intelligent’ with an
‘active architecture’. His comments on biological self-organization
are of particular interest. It is, he explains, an assumption that such
self-organization is ‘order for free . . . By this they [other authors] seem
to suggest that self-organizing biological patterns occur so automati-
cally that natural selection has had little influence over the evolution
of their structure. I believe that such a viewpoint is extremely mis-
leading’. So when we consider how army ants organize their swarm
raids Franks argues that the processes of natural selection remains in
force, for instance, by
rather precisely [tuning] both the behavioural rules of thumb of

the ants and the chemical properties (e.g. volatility) of their
pheromones . . . Self-organization theory does not suggest that
natural selection has had no role in the creation of certain patterns
in biology – rather it suggests that natural selection has rather less
to do than one might expect given the complexity of the global
structure. As illustrated by the swarm raids of Old World and New
convergent complexities 205
World army ants, natural selection may have had to select for a
surprisingly small and simple set of rules to generate swarm-
raiding patterns. Thus self-organization theory may help to
explain why we observe such a high level of convergent evolution
in certain biological structures.50
convergent complexities
The behaviour of complex social insect societies, and specifically the
army ants, is more than a digression. This is because the emergence
of such biological complexity may be much more constrained than is
sometimes imagined.51 At the least it is a reminder that carnivorous
species can be more successful, at least in terms of colony size, than
the farming ants (and, as we shall see below, termites). Success, as
even the most jaded (if not jealous) scientist well knows, is relative.
Size, we are reassured from the most reliable sources, is not all. In
their relative adaptive contexts social wasps and bees have taken over
their respective worlds. So, too, the attine ants have evolved a remark-
able system of fungal cultivation. In each of these examples we have
evidence of extensive appropriation of resources, capable of maintain-
ing large and complex societies. Life beyond these ‘cities’ and ‘armies’
remains, of course, diverse and marvellous: the biological world is by
no means reduced to a monochrome. Nevertheless, the repeated rise
of such societies, and at least evidence of the displacement and ulti-
mate extinction of less successful equivalents, suggests that such an
arrangement is a biological inevitability.
So, too, in any specific case if a group ‘decides’ to adopt a partic-
ular strategy, say agriculture, then the routes to success will be very
limited. It is time, therefore, to return to the parallels between ant
and human farming, and thereby consider some implications of such
a commonality. As in our mushroom farms, the activities are carried
out underground, since, unlike plant crops, the fungi have no need
for sunlight. The farms are located in elaborate nests, equipped with
ventilation shafts52 and in certain species also dump-pits – some large
enough to house a man53 – which are used for waste disposal (Fig. 8.3).
Just as in human societies, control of the waste is very important for
the health of both the colony and the fungal colonies, and it now
appears that the risky business of waste management54 may be ‘allo-
cated’ to the older workers, nearing the ends of their lives and less valu-
able to the colony. It seems that once assigned to dump management55
figure 8.3 The fungus farm of the attine ants, with large central
chamber containing pendant clumps of the fungus and one worker (!)
Note the tunnel system leading to the outside world, and the capacious
dump-chambers. (Reproduced from fig. 29 of Jonkman (1980; citation is
in note 53), with permission of Blackwell Wissenschafts Verlag GmbH
and the Jonkman family.)
the workers seldom leave, not least because of an aggressive response

by other occupants of the nest.
Human farming has evolved independently a number of times,
and so, too, has agriculture among the attine ants. Present evidence
indicates that the domestication of fungus was first achieved approx-
imately 50 Ma ago, but there have been repeated episodes (at least
five) of domestication of various groups of fungi as distinct cultivars.56
What originally might have triggered this innovation? Here, too, there
could be parallels to the invention of human agriculture, where it has
been suggested that in some instances, as in ancient Mesoamerica, its
development was spurred by significant climatic changes.57 So, too, a
comparable set of upheavals, perhaps even the ecological mayhem that
succeeded the K/T impact event, could have been the crucial factor in
shifting these ants to agriculture.58 This ability to cultivate fungus is
more than just an evolutionary curiosity. Not only is it strongly con-
vergent on our agricultural activities, but as Neal Weber has remarked,
‘the evolution of a unique skill in fungus culturing has freed these ants
from the usual food limitations and has enabled them to expand in
colony size. In the process a few species have developed some of the
largest ant colonies known,’59 with populations estimated to reach
at least seven million,60 and typically exceeded only by the aggres-
sive army ants (p. 201). Gigantic populations dependent on highly
organized societies: does it sound familiar? And, as with the eusocial
bees, there may even be lessons for us. The fungal gardens are usually
monocultures. But as humans know (or should know), monocultures
have their risks, being notoriously liable to invasion by viruses, fungi,
and other pests. The use of the streptomycene antibiotics (note 22)
has, of course, a direct parallel in human medicine. What seems quite
remarkable is that the resistance conferred by the antibiotics seems
to have been maintained by these ants for literally millions of years,
in comparison to our experience where in only a few decades we have
seen the seemingly inexorable rise of ‘superbugs’, resistant to nearly
all treatments. In reality the system of ant farming and its crop protec-
tion is presumably much more dynamic. It appears that the more vir-
ulent strains of pathogenic fungi occur in the more advanced species
of attine ants, which in turn are more reliant on monocultures. It is
hardly surprising that there is evidence for exchange (and stealing) of
fungal cultures between colonies.61
The agricultural activities of the attine ants are surely one acme
of arthropod organization, and further parallels with human agricul-
ture may well emerge. As a leading worker, Ulrich Mueller, has re-
marked: ‘The more I study [the fungal gardens] the more analogies
I find between human agriculture and ant agriculture.’62 Given the
nutritional value of fungi and the ease of propagation, it is not so sur-
prising that other insects have also learnt the art of agriculture. Thus
the ambrosia beetles, a type of weevil, have evolved a mutualistic re-
lationship with a group of ascomycetes known as the ophiostomatoid
fungi.63 The general arrangement is certainly rather different from
that of the attine ants because these beetles bore into tree trunks,
so producing a complex series of galleries. The beetles are responsible
for introducing the fungi. These are carried in containers (mycangia)64
that consist of cuticular pockets with associated secretory glands that
evidently ‘control the growth and form’ of the fungi.65 The mycangia
share, therefore, a basic function, but their form is extremely varied,
and in different species they occur on many parts of the carapace.
As ever in evolution, ‘needs must’: who cares whether the container
(in this case the fungal-ferrying mycangia) comes from, so long as
it works? So far as beetle and fungus are concerned, both sides are
winners. Evidently the advantage to the fungi is to be smuggled into

new tree hosts, incidentally making these beetles a serious economic
pest.66 The beetles benefit because the fungi help to block plant de-
fences, such as the secretion of resins,67 and also provide a source of
food. Yet in other respects this fungal agriculture is reminiscent of
that pursued by the ants. For example, it has become an obligate as-
sociation, shows vertical transmission from generation to generation,
and provides the main food source, and interestingly in these beetles
appears to have ‘evolved at least seven times’.68
It is also possible that a convergent situation has arisen in the
termites, which although often referred to as ‘white ants’, are actually
relatives of the cockroaches, whereas ants are closer to the wasps. In
one group, the macrotermitinids, there is a striking and close sym-
biotic union between fungus and termite.69 In contrast with the at-
tine ants, which are restricted to the New World, this association is
an Old World one, probably originating in Africa and then spread-
ing to the Middle East and southern Asia. Termite nests are among
the more spectacular of animal constructions, with an elaborate sys-
tem of chambers and tunnels.70 The tunnels are also used for foraging
runs outwards from the nest, for up to 50 m; one calculation gave
for a single nest a total tunnel length of almost 6 km. The forag-
ing tunnels repeatedly branch but crossroads are very rare. The walls
of tunnels are smoothly plastered and sharp bends are cambered. These
tunnels enable excursions to be made in relative safety; the termites
ascend to the surface by short access tunnels that slope upwards from
the main tunnels. These access tunnels are less well built, and may
be sealed when not in use. Their diameter is such that they can also
be blocked by the heads of the large soldier termites. The nest itself
is engineered to allow both the circulation of air and the venting of
waste gases. Also present are storage areas and, at least in experimen-
tal set-ups, latrines. As is also the case in other colonies of social
insects (and their analogues), there are other inhabitants, some ben-
eficial, some not. Some termite colonies, for example, play host to
a group of flies known as the phorids, and here, too, convergences
emerge.71
Growing within the termite nest are spectacular combs of the
fungus, aptly named Termitomyces. It belongs to a group known as
the basidomycetes72 and thus is related to the fungi cultivated by the
attine ants, as well as to the more familiar mushrooms and toadstools.
On occasion, Termitomyces emerges from the nest to form on its out-

side a mushroom-like fruiting body (which is edible to humans), but
evidently this fungus cannot survive outside the termite colony. Thus,
if the termites are removed the combs are soon invaded by other sorts
of fungi and bacteria. Maintenance of the combs may involve salivary
secretions and weeding, yet earlier workers such as Roger Heim73 were
dismissive of the term ‘cultivation’.74 Since then views have changed,
and as Johanna Darlington briskly notes: ‘Heim’s views of the adapt-
able fungus forcing itself upon the reluctant termite is at odds with
more recent experience of Termitomyces as a delicate and finnicky
organism requiring care and cosseting. It seems best merely to take
note of Heim’s arguments and pass on.’75 Yet, while the association
seems to be agricultural, the precise nature of this symbiosis is still
somewhat elusive. Again to quote Darlington,
The fungus is a passive partner, absolutely dependent on the

termites for its survival and success. The termites, on the other
hand, invest a lot of time and energy in caring for the fungus. They
gain substantially from the relationship, although it is proving
difficult to discover exactly how. The evidence favors a nutritional
role, with the fungus comb acting as an external stomach,76
breaking down less digestible components of the forage and so
making them available to the termites . . . An advantage of the
system is that the digestive work of the symbiont is carried on
outside the termites’ bodies, unlike the gut symbionts that do a
similar job for the lower termites. Lower termites have to carry
fermentation tanks around with them, while the Macrotermitinae
leave them parked at home.77
Evidently this agriculture is different from that of the attine ants, just
as human agricultures vary.
Other convergences also emerge. Communication among ants
is famous for its pheromone-based system, but in the termites seis-
mic communication is achieved in the soldier caste by drumming
the head against the substrate,78 a method also used by subterranean
mammals (Chapter 6, p. 141). Seismic communication is also used by
the leaf-cutter ants.79 In these ants, although the sound is produced as
a stridulation, it is transmitted through the ground. Typically stridu-
lations are produced only when the individual attine ant is immo-
bilized, such as occurs by collapse of part of the earthworks. These
cries of distress elicit rapid digging by other workers, who can detect
sounds through 5 cm of earth and will begin rescue work if the thick-
ness is no more than 3 cm. In the termites this alarm reaction can
evidently be triggered either by air currents or vibrations. The more
advanced fungus-growing termites are particularly sensitive in this
respect. Rather remarkably, the signal can be propagated by chains
of soldiers, each re-amplifying the signal in a way analogous to the
now-redundant methods of human communication using smoke and
drums,80 as well as the periodic re-amplifications observed in the
transmission of signals along the nervous system.
Although they are in certain respects less sophisticated, it is ap-
propriate to mention in passing the so-called ant-gardens, which are
arboreal earthen structures known as ‘cartons’. In this arrangement
certain species of epiphytic plant are encouraged to grow on the nests,
from seeds that are collected by the ants and planted in the nest wall. In
due course these plants germinate, grow (and thereby provide a source
of extra-floral nectar), extend roots that probably help to strengthen
the nest, eventually flower, and so produce seeds, the fruit of which
is eaten before they are planted . . . and so the cycle continues.81 In
a few instances fungi are employed in the ‘cartons’, probably to help
bind the structure and possibly also to release antibacterial chemicals.
Interestingly, in some cases the fungus is effectively a monoculture,
apparently maintained by weeding and feeding. So far as can be told,
however, the fungal products themselves are not directly cropped.82
Other activities of these garden ants include the pruning of surround-
ing vegetation, probably to create a ‘fire zone’ to reduce the risk of in-
vasion by other ants,83 the collection of vertebrate faeces, presumably
as fertilizer, and perhaps the choice of plants that, like those cartons
with fungi, release chemicals that help to keep pathogens at bay.84
This section on the ants (and termites) has included more than
its fair share of digressions, but I hope that the common thread of ex-
ploring convergences still runs through the narrative. Consider again
the army ants: living in the midst of their mobile and aggressive col-
umn has its self-evident risks, yet various species have managed to
insinuate themselves into the ants’ social system. Most striking in
this instance are probably the various staphylinid beetles. These have
effectively transformed their bodies into an ant shape that readily de-
ceives the actual ants.85 Hiding yourself in an aggressive raiding col-
umn may yield various benefits, not least an uncontested share of the
figure 8.4 A staphylinid beetle in association with the army ant

Eciton. (Photograph courtesy of David Kistner, California State
University, Chico.)
booty and also some protection from predators. Typically the beetles
travel in the centre of the column, and are not found in the vanguard.
In times of excitement, such as during an episode of emigration, the
beetles may ride on the ants themselves and despite continuous an-
tennal interrogation never seem to be recognized as interlopers. The
way in which these beetles have transformed themselves into ant-
mimics is very striking (Fig. 8.4), and the selective pressure to sur-
vive in a jostling mass of aggressive workers can explain how such
a transformation in the staphylinids has occurred independently on
multiple occasions.86 In at least one instance the convergence extends
to a colour matching, whereby a geographical variation in the ant col-
oration is matched by the beetle.87 As David Kistner (Fig. 8.5) points
out, given that the ants are effectively blind, this mimicry supports
the idea that it serves to dupe ‘educable predators which sit by the
raiding columns to pick up insects which are stirred up by the raid’.88
Nor is this the only example of a convergence between a mimic and
an army ant. A common association is the attachment of mites to
the exterior of insects (and other animals such as birds and mammals,
clinging respectively to feathers and hair; pass me the nit comb), but
in the case of the mite Planodiscus ‘the sculpture of the mite and
figure 8.5 David Kistner and his wife in the field in Ecuador, collecting
staphylinid beetles and army ants. (Photograph courtesy of David
Kistner, California State University, Chico.)
the ant’s leg is nearly identical. Also the arrangement and number of
setae on the mite approximates the arrangement and number of setae
on the leg. Thus when the ant grooms its leg, the tactile stimulation
[as it passes over the mite] will be similar to that of the leg itself.’89
In Chapter 6 (pp. 116–117) I briefly introduced the tent-building
capacities of the aptly named weaver ants. Here, too, there are some
striking cases of mimicry, which here involve weaver ants and a crab
spider, Amyciaea forticeps.90 As it happens, the weaver ant in ques-
tion, found in India and known as the Indian Red Ant (Oecophylla
smaragdina), is also associated with another spider (Myrmarachne
plataleoides) that in turn is ‘a perfect mimic of the red ant; so per-
fect is this mimicry that even experienced biologists may pass it by
as an ant, in the field.’91 Despite its close association with the Indian
Red Ants, which can be very aggressive, M. plataleoides takes good
care to avoid the ants. Possibly it employs its mimicry to fool other
ants from whom it steals their young, as well as to avoid predation

by being mistaken for its genuinely aggressive counterpart. The other
spider, Amyciaea, has only a generalized similarity to the Indian Red
Ant, but this is compensated for by a rather remarkable behavioural
convergence (see also p. 285, Chapter 10) which seems to entail its
pretending to resemble an ant in distress. Thus, when a nearby ant
adopts an alarm attitude, thereby disturbing the other ants. This then
gives the spider an opportunity to attack. Once seized, the meal is
concluded as the spider suspends itself from a silk thread, safe from
the other ants.92
The convergence of mimicry of insects and spiders to an ant
morphology has ‘evolved at least 70 times’,93 and as such gives a series
of fascinating insights into both the malleability of biological systems
and the likelihood of establishing adaptive explanations. The trans-
formation of staphylinid beetles to an ant-like form has already been
noted, and given that ants are also insects it is relatively easy to see
how other insects might come to mimic them. Even so, there is a sub-
tlety inasmuch as the closeness of mimicry might be compromised
as the insect changes in size and shape. The solution, of course, is
for the mimic to resemble successively, in a series of moults, two or
more species of ant.94 The case of the spiders, the other principal ant
mimics, is more remarkable, given the more obvious differences in
body plan, and here, too, illusions connected to segmentation, colour,
appendages, eyes, and surface texture have all been arrived at by a se-
ries of ingenious modifications. Not surprisingly there are degrees of
mimesis, but in at least some instances there is evidence that the phy-
logenetically most derived are also the most ant-like,95 and as such
will define evolutionary trends. As already noted for one spider, these
may be supplemented (or complemented) by various behavioural mod-
ifications. Not all insects need look like ants to insinuate themselves
into ant colonies, and many closely associated species evidently rely
on compatible and thereby convergent chemistries or textures.
In passing one should note that the vast area of biological
mimicry, where one species comes to resemble another, is a perfectly
good example of convergence. Its classic manifestation is in Batesian
mimicry, where an animal, usually an insect, comes to resemble
closely an unrelated but noxious species and thereby escapes preda-
tion. Mullerian mimicry, on the other hand, is where two species,
both unpalatable, converge to resemble each other. So well known
are these mimicries and their various manifestations,96 that for the
purposes of this book they can, I hope, largely be taken for granted.97
The complexity of both ant and termite colonies, not to mention their
various guests and symbionts, also provide a rich field within the topic
of convergence, and one that can be analysed at several levels. Pass-
ing comment has already been made on such built structures as trails
and tunnels, farms, and cartons, and it is clear that much remains
to be discovered about these various constructions in an adaptive
context.98
These insects, moreover, are by no means the only arthro-
pods to have evolved constructional abilities. Take, for example, the
beach-dwelling orypodid crabs. These crustaceans are highly territo-
rial and protect their domains from intrusion by a variety of complex
behaviours that include the building of mud fences and also the plug-
ging of neighbouring burrows. Sometimes the occupant is evicted, but
in other cases it is entombed. As might be expected, at least some of
these behavioural repertoires are convergently acquired within this
group.99 Nor is this the only example of such sophistication of be-
haviour. The fiddler crab is famous for its signalling, achieved by
movement of its hypertrophied claw. It now transpires that the sig-
nalling for mates, in a process known as ‘lekking’ in which numer-
ous associated males signal simultaneously to choosy females (a well-
known characteristic of some birds), also occurs in these crabs.100
hearts and minds

Not only do certain invertebrates provide some intriguing parallels to
the behavioural sophistications of the vertebrates, but they also ap-
proach them in terms of activity and intelligence. Earlier I emphasized
the convergent similarities between the camera-eyes of cephalopods
and vertebrates (p. 151).101 In addition, despite the rather different ar-
rangement of the cephalopod brain,102 the size of this organ relative
to the body exceeds that of many fish and reptiles.103 In addition, in at
least the cuttlefish, the blood–brain barrier,104 essential for control-
ling the chemistry of the brain and thereby a prerequisite for high level
integration, ‘is as tight as that of mammals’.105 The cephalopods also
appear to share another characteristic, the possession of molecules
similar to those known in vertebrates as neurotrophins, a type of pro-
tein growth factor. This, as has been suggested, may be a prerequisite
for the emergence of complex brains.106 Cephalopods are fascinating
hearts and minds 215
animals and, notably in the cuttlefish, they can show an astonishing

range of chromatic signals that are under direct nervous control. To
see the colours flowing across a cuttlefish is a quite extraordinary
sight, and although the colour changes are well attested with re-
spect to camouflage and predator avoidance, it is difficult to believe
that it is not in some way ‘emotional’.107 In the case of the octopus,
there is an intelligence that includes a flexibility in behavioural reper-
toires and the ability to learn and remember.108 Evidently the octopus,
far from being a rule-bound machine, is capable of acting in an au-
tonomous fashion.109 In their natural habitat octopuses have a sound
grasp of the seascape: as one report110 commented, the octopuses
‘often zigzagged through multiple substrata and depths, with many
obstacles obscuring their visibility of the horizon. The divers [who
were tracking them] were often surprised when a forage ended and
they realized the octopus had arrived back at its den while the divers
themselves were still disoriented.’111 Some workers have gone so far as
to talk about the individual temperament, if not personalities, of the
octopus.112
Given these characteristics, it is scarcely surprising that the
octopus and the squid are very active animals, with maximal power
outbursts that compare favourably with those of human athletes. This
type of activity presupposes an effective circulatory system. As such
it finds strong parallels to ourselves in two ways: the structure of the
aorta and the pattern of circulation. Although rather different, both
are related to the necessity of supplying oxygen at a fast enough rate.
In such circumstances high blood pressures are essential. Blood leaves
the heart via the arteries, and to deal with the repeated fluctuations in
pressure it is hardly surprising that the arterial walls are rich in spe-
cific proteins that confer an elasticity to the throbbing tube.113 Nor
is it really surprising that the structure of the aorta wall in squid and
human is strongly convergent.114 The other parallel concerns the gen-
eral arrangement of the circulatory system. Elevated blood pressure
is essential for an active (and intelligent) animal, but if directed to
the respiratory organs (lungs or gills) there is a danger that it will rup-
ture the delicate membranes across which the gases are exchanged.
The solution is simple and to some extent convergent. Only in the
cephalopods115 and vertebrates is the circulatory system completely
enclosed in vessels and so capable of operating at high blood pres-
sures. In both there is effectively a dual system, whereby the blood is
first pumped to the lung/gill in order to collect the oxygen (and dis-
pose of carbon dioxide). The blood is then returned to a second set of
chambers in the heart where it is dispatched at full force into these
elastic arteries. Even so, despite the shared principle there are signif-
icant differences. The branchial heart in cephalopods, which as the
name indicates is responsible for feeding blood to the gills, is sepa-
rate from the main pump. Furthermore, it is not very muscular, and
probably has additional excretory functions.116 The main systematic
heart is powerful and muscular, but its structure only approximates
to the vertebrate arrangement.117
Despite their manifest differences, the convergences between
the cephalopods and vertebrates have attracted the attention of many
biologists. Of greatest significance surely are those that pertain to the
camera-like eye (Chapter 7) and brain (note 108). Others are probably
less significant, but still intriguing. Earlier, in discussing the hypo-
thetical Fortean bladders I remarked on the convergence between the
relatively familiar fish swim-bladder and that of an octopus (note 35,
Chapter 6). Another interesting convergence with the vertebrates is
the development of a cartilage-like tissue in the head.118 In fact, the
roster of convergences is still not complete. Consider, for example,
that part of the cephalopod sensory system associated with the skin.
This is strongly analogous to the lateral-line system found in the fish
and aquatic amphibians,119 and it too is sensitive to pressure waves
travelling through the water.120
Having reviewed earlier (Chapter 7) not only convergences in
particular sensory systems but also possible underlying commonali-
ties (which are equally important), it is not surprising that, at least
so far as the fish are concerned, there is evidence that the lateral line
can help the animal to form a ‘hydrodynamic image’.121 What sort
of ‘map’ or ‘image’ this might form in the brain is still a matter for
speculation. It is likely, however, that the input from the lateral line
produces a ‘pressure world’, analogous both to the ‘electrical world’
of the mormyrids and other electrosensitive fish and to one that is
integrated with other sensory inputs, such as vision.
Self-evidently, as an aquatic adaptation the lateral line of the
fish (and amphibians) was lost as the tetrapods clambered on to land.
In the case of the manatees (or dugongs), which are secondarily aquatic
mammals, the arrangement and anatomy of the post-cranial hairs are
something of an evolutionary novelty. Their sensitivity to changes
hearts and minds 217
figure 8.6 A surprising convergence on the lateral line system of fish

and cephalopods, as manifested in the trailing antenna of a penaeid
crustacean. The structure defines a tunnel with spaced bunches of
sensory hairs, and so is a direct analogue of the pressure-sensitive
system of the other two groups. (Reproduced from Pl. 1a of Denton and
Gray (1985; citation is in note 123) with the permission of the authors
and the Royal Society.)
in water pressure, however, offers an intriguing analogy to the lateral

line,122 and is consistent both with the poor eyesight of these creatures
and with the murky waters they inhabit. Because I regard the conver-
gences among sensory systems as being of particular importance, in
the context of lateral lines it is surely necessary to remark on an even
more striking example: that between the lateral-line system of the fish
and the modified antennae of some swimming crustaceans known as
the penaeids.123 These pelagic arthropods have a pair of flagella-like
antennae (Fig. 8.6), which project from the head at right angles before
bending sharply and trailing parallel to the animal as it swims through
the water. As in many other arthropods, the antennae have hair-like
setae, but these are recurved and so define a tunnel-like structure on
the floor of which are located sensory setae. Despite their quite dif-
ferent origin these antennae function in effectively the same way as
the lateral line of a fish; in each case the basic structure is a tube with
sensory hairs sensitive to changes in the water pressure.
honorary mammals
So building complex circulatory systems, peering at the world through
a camera-eye, and employing intelligence with a large brain have all
evolved convergently. We are, however, in pursuit of the humanoid,
and I shall assume that the galactic equivalent is in some sense
mammal-like. This may seem too bold a claim, but let us see. So far
as the Earth is concerned, there is a simple natural experiment. What
we need to do is find a landmass which the mammals have failed to
colonize. There is an excellent example, and it is called New Zealand.
Indeed, Jared Diamond124 went so far as to say ‘New Zealand is as close
as we will get to the opportunity to study life on another planet.’125
For at least 85 million years126 these islands have remained isolated in
the Pacific Ocean, too remote to be colonized by any of the terrestrial
mammals, other than by the bats and much later the boat-travelling
Polynesians who arrived about ad 1000. New Zealand, however, had
plenty of other inhabitants. Particularly extraordinary are the giant
wingless crickets, known as wetas, relatives of the grasshopper and lo-
custs. These are ecologically convergent on mice and rats, ‘resembling
rodents not only in their biomass, but also in nocturnal foraging and
diets, use of diurnal shelters, polygamy and even their droppings’,127
and when surprised leaping across the forest floor.128
Nor are these the only animals in New Zealand to approach
mousedom. Diamond remarks that the ‘Stephens Island wren . . .
was . . . the world’s only known flightless songbird and functioned
as an avian mouse.’129 So, too, one of the endemic bats (Mystacina)
is also convergent on a mouse-like habit, being partially terrestrial
and when walking on the ground folding its wings to protect the del-
icate membranes.130 Presumably like the bats the ancestors of these
birds had originally flown there, but perhaps because of the absence
of other ground-dwelling competitors many of them became flight-
less. These include an extraordinary nocturnal parrot, known as the
kakapo, which grows as big as a turkey.131 The most famous are the
kiwi and the much larger moa.132 These birds are not closely re-
lated and, arrived at flightlessness convergently.133 The moas have
vanished, and their extinction was almost certainly by way of the
stew-pots, roasts, and fricassees of the Maoris.134 The much smaller
kiwis, however, avoided the category of a walking larder. If, as for
the moa, we had only the bones, it is unlikely that we would ever
have realized that in a land without mammals, the kiwi (Fig. 8.7)
honorary mammals 219
figure 8.7 The kiwi, a flightless bird, emblem of New Zealand, but
also an ‘honorary mammal’. (Reproduced from the figure on p. 103 of
Calder (1978; citation is in note 135), with the permission of the artist,
Alan D. Iselin.)
has converged strongly in this direction.135 The feathers are rather

atypical in being shaggy, in fact almost fur-like. In part this must
be because these flightless birds live in burrows. Like many of the
equivalent-sized mammals, kiwis are also nocturnal, are strongly de-
pendent on a sense of smell, and unusually have nostrils at the tip of
the beak. Moreover, the set of feathers around the mouth is modified
into whisker-like structures.136
It is in its reproduction, however, that the kiwi (Fig. 8.7) is the
most mammal-like. Not so far as giving birth to live young; like the
monotreme mammals, such as the duck-billed platypus, the kiwi still
lays eggs. Given the widespread occurrences of ovoviviparity, that is,
where the egg is retained in the female who ultimately gives birth
to live young (and as discussed below is strongly convergent), it is
something of a mystery why, with one dubious exception concern-
ing a budgerigar living in Dorking,137 no bird has managed to become
capable of live births. There has been no shortage of suggestions, in-
cluding potential problems of immunology (leading to rejection of the
embryo), the type of sex determination (which in birds depends on the
female, with a ZZ–ZW system of chromosomes as against our XX–XY
male determination), and perhaps most plausibly the difficulty in dis-

pensing with a shelled, calcareous egg.138 Even so, it is possible to pre-
dict that if ovoviviparity were ever to evolve in the birds the species
concerned would show the following characteristics: flightlessness; a
lower than average body temperature (one reason why internal incu-
bation may be difficult for birds is their elevated body heat, at c. 40 ◦ C);
and a single egg rather than a clutch. And the closest candidate to this
list? Yes, the kiwi.139
So, even though the kiwi is not ovoviviparous, it still remains
the case that its reproduction is otherwise strongly mammal-like and
unlike that of other birds. Thus, both ovaries remain fully functional,
with the ovum released in the alternate fashion. The incubation period
of the egg is exceptionally protracted, equivalent in time to a mammal
of the same body weight. In reviewing this flightless denizen of New
Zealand, William Calder III was in no doubt that this convergence was
more than superficial when he wrote,
When one adds to this list [that is reproduction] the kiwi’s burrow
habitat, its furlike body feathers and its nocturnal foraging, highly
dependent on its sense of smell, the evidence for convergence
[with mammals] seems overpowering. Only half jokingly I would
add to the list the kiwi’s aggressive behavior . . . When I intruded
on his domain at night, [the male kiwi] would run up to me
snarling like a fighting cat . . . and drive his claws repeatedly into
my ankles until I went away. For this behavior and the many other
reasons [given] . . . I award this remarkable bird the status of an
honorary mammal.140
giving birth to convergence

The case of the kiwi is a useful reminder that despite sticking to eggs,
many other non-mammalian vertebrates have adopted the system of
ovoviviparity, in which the egg is retained in the mother prior to a live
birth. The squamate reptiles (lizards and snakes) are particular masters
of ovoviviparity: Richard Shine estimates that in this group this type of
reproduction has evolved independently approximately 100 times,141
while Daniel Blackburn suggests that overall viviparity has evolved
independently approximately 132 times,142 with examples also in the
anuran amphibians143 and fish.144 Not surprisingly, the retention of
the embryo in the female reproductive tract has led to other con-
vergences, most notably the emergence of a placenta.145 Particularly
giving birth to convergence 221
remarkable in this respect is the similarity between the mammalian

placenta and an ovoviviparous Brazilian lizard (Mabuya heathi)
which, as Blackburn and his colleagues remark,146 ‘has strongly
converged upon a reproductive pattern long believed to be unique
to eutherian [i.e. placental, as against the metatherian marsupials]
mammals’.147 The relevant features, in addition to the placenta and
its nutrient supply, include a remarkably small egg (c. 1 mm), com-
parable in size to a typical mammalian ovum. Not surprisingly the
period of gestation, for a reptile, is prolonged. This has the curious side-
effect that females can be impregnated while juvenile and give birth
to their live young when adults. Other species of the lizard Mabuya
show similar features,148 and it is possible that such viviparity orig-
inated at least four times in this group.149 Other viviparous lizards
show interesting features convergent with mammals. In a European
species of Chalcides (C. chalcides) the placentome, a specialized area
involved with absorption, develops to such an extent that at the time
of birth it can be cast off, and thus uniquely for the reptiles is equiva-
lent to the mammalian afterbirth.150 A curious parallel also exists in
the spade-nose shark. Here, too, the ovum is minute (c. 1 mm), and
quickly implants in the uterine wall where it receives nourishment
through a placenta-like structure. The combination of tiny egg and
major maternal input again leads to a massive increase in the weight
of the developing young.151
In reviewing the occurrences in the lizards, Daniel Blackburn
emphasized that ‘In several important respects, matrotrophic Mabuya
and Chalcides have converged strongly with mammals. Features that
have evolved independently among eutherian mammals and in one
or both of these two reptilian groups [total at least seven],’152 and he
concludes that
Classic textbook examples of convergence, such as the evolution

of flight, and fusiform body shapes in aquatic vertebrates, pale
beside those relating to viviparity and matrotrophy. Among
viviparous vertebrates evolutionary convergence has operated at
virtually every level – not only with respect to viviparity itself,
which has originated over 130 times, but at the level of the organ,
the tissue, and the cell. Such convergences are all the more
striking when we consider that the historical sequences through
which viviparity and matrotrophy have evolved apparently differ
from lineage to lineage. Moreover, the independently-derived
specializations for fetal nutrition refute theriocentric [i.e.

mammal-centred] views of mammalian reproduction and
placentation as uniquely specialized.153
Blackburn’s analysis epitomizes many of the important features of

convergence, especially the degrees of similarity and the fact that the
end-result may be arrived at by various historical pathways. Another
important facet of convergence is emphasized in a study of sharks
and rays, where viviparity (see also note 151) is estimated to have
evolved independently at least nine times. In at least this group there
is a very strong polarity towards viviparity, with only a few counter-
examples.154 It is worth noting, however, that the strong phylogenetic
trend to viviparity is not matched by the evolution of matrotrophy
(nutrition provided by the mother in addition to the yolk), which evo-
lutionarily is much more labile. This has evolved independently four
or five times, but also with multiple reversals.
In the context of non-mammalian ovoviviparity, there is one
other item worth mentioning. Earlier (Chapter 5, p. 96) I suggested
that as a contingent episode the asteroid that hit the Earth 65 Ma
ago, and so led to the demise of the great reptiles, which thus facili-
tated the subsequent rise of mammals, may be misconstrued. Not as a
historical episode: there seems little doubt that the mammalian radi-
ations are a direct consequence of the disappearance of the Mesozoic
reptiles. That, however, does not address the question that concerns
the likelihood, if not the inevitability, of the emergence of such bi-
ological properties as ‘mammal-ness’. Viewed in this light a differ-
ent picture emerges, and for two reasons. First, as the Tertiary planet
cooled, culminating of course in the ice ages, the mammals (which,
recall, appeared about 220 Ma ago, at about the same time as the
dinosaurs) would have been at an advantage anyway, well adapted
to temperate and even polar environments. This suggests that the
rise of active, agile, and arboreal ape-like mammals, and ultimately
a hominid-like form, would have been postponed, not cancelled. My
guess is that without the end-Cretaceous asteroid impact and in due
course progressive planetary refrigeration, the appearance of the ho-
minids would have been delayed by approximately 30 Ma. But even
that is a local history. It is interesting that, at least in the lizards, the
evolution of viviparity is associated with colder climates.155 It is not
the only factor, but recalling also that ‘honorary mammal’ the kiwi,
warming to, singing of, chewing convergence 223
it too suggests that ‘mammal-ness’ will not be the exclusive property

of one evolutionary group. If ‘we’ had not emerged, then (as this and
the next chapter aim to show), rest assured that a viviparous, warm-
blooded, vocalizing, and intelligent species would have done so.
Finally, as if it needed any emphasis, it is important to stress,
as Nicholas Dulvy and John Reynolds specifically remark with re-
spect to shark and ray viviparity ‘Our data do not support a linear,
irreversible progression toward a ‘pinnacle’ of maximum maternal
input.’156 Quite so; evolution is labile, it does show reversals, but the
point still remains that the emergence of various biological properties
is in response to adaptation and is governed by selection. Nor does this
mean that the selective pressures operating in an ocean need be the
same as on the savannah, although as we shall see later (p. 250) here,
too, there are some surprising convergences, specifically between
sperm whales and elephants. Convergence simply tells us that the
evolution of various biological properties is certainly highly probable,
and in many cases highly predictable.
warming to convergence, singing of

convergence, chewing convergence
The stories of the kiwi and non-mammalian ovoviviparity are two re-
minders that being a mammal and evolving towards ‘mammal-ness’
need not be the same thing. There are at least two other features
that typify mammals but are in fact also strikingly convergent: warm-
bloodedness and vocalization. Concerning the first, endothermy (that
is, the phenomenon of having a body temperature elevated in com-
parison with the ambient environment), is rampantly convergent and
occurs not only in the birds and mammals, but also in various fish
(teleosts and sharks) and even in insects.157 The example of the fish
might seem somewhat removed from the path of humanoids, but may
in fact have some relevance. In the fish endothermy is clearly conver-
gent, and has probably evolved on at least five separate occasions.
Examples include the remarkable modification of the eye muscle in
some scombrid fish, to form a heater to warm the brain and retina (see
note 118, Chapter 7). Whole-body endothermy in fish is perhaps best
known in some tuna, famous for their fast swimming and wide ranges.
Barbara Block and her colleagues158 stress that this is a sophisticated
adaptation, and entails a concatenation of changes. These include the
development of countercurrent heat exchangers at three locations in
the body, the movement of red muscles towards the interior of the
animal, combined with the body becoming larger and thicker, and a
style of locomotion known as thunniform, in which swimming is
achieved by oscillation of the tail rather than sinuous deformation
of the entire body. Yet this complex arrangement was convergently
arrived at in the lamnid159 and alopiid sharks, where again it prob-
ably evolved independently in each group.160 There are, moreover,
more specific convergences161 such as the ‘Striking parallels in both
structure and function [that] exist between the shark orbital rete [a
fine network of blood vessels] and the mammalian carotid rete which
serves as a brain cooling system in mammals’.162
The adoption of this type of endothermy is significant because it
evidently allowed an expansion of a habitable zone, and in these fish it
made possible the invasion of a vast new realm, specifically the cooler
(and deeper) parts of the ocean. As Barbara Block and her co-workers
note, the multiple origins of endothermy indicate ‘Strong selection
for this energetically costly metabolic strategy’.163 In other words, en-
dothermy is expensive, but well worth the cost by being highly adap-
tive and accompanied by success. Costs, as ever, have risks, and in
the context of endothermy there is another interesting physiological
convergence. This involves a pathological condition known as ma-
lignant hyperthermia, in which the muscles produce excessive heat
that if unchecked leads to severe tissue damage and death. In mam-
mals it has a genetic basis, and may manifest itself in anaesthetized
humans undergoing surgery, as well as in genetically disposed pigs
that panic in the slaughterhouse.164 A strikingly similar syndrome is
found in highly stressed tuna, especially those caught on a hook and
line. This leads to so-called ‘tuna burn’, badly damaged musculature
and subsequent rejection by the fastidious Sashimi enthusiast.165
What then is the significance of the convergent evolution of
endothermy in the birds and mammals? C. G. Farmer has argued that
among the welter of possibilities the key factor is parental care.166 Of
central importance is the need for sustained exertion, to enable the
parent(s) to collect sufficient food not only to maintain their ravenous
fledglings or litter, but also for them to grow as quickly as possible and
thereby decrease their vulnerability. Even in those birds and mammals
that characteristically save energy by entering a state of torpor and
letting their body temperature decline, incubating birds and pregnant
mammals typically retain a higher body temperature. Farmer suggests
that the original trigger for bird and mammal warm-bloodedness was
connected to reproduction and the increased production of hormones,
e.g. from the thyroid, which in turn are also important determinants
of metabolic rates. She continues by noting that
The convergent evolution of . . . parasagittal limb posture . . . a

large pulmonary diffusion capacity . . . high oxygen-carrying
capacity of the blood . . . a large tissue diffusion capacity . . . a
completely divided cardiac ventricle . . . an extensive coronary
circulation and compact myocardium . . . high systematic blood
pressure . . . and a single aorta . . . may be due, at least in part, to
selection for an ability to sustain vigorous exercise which is
requisite for birds and mammals to provision their young.167
Parental care is, of course, a particular hallmark of the birds and

mammals,168 and Farmer concludes that this in turn may have pro-
moted both ‘complex social structures [and] . . . vocal signaling’.169
And that might seem to be the furthest these convergences can be
pushed. Birds are warm-blooded, indeed generally showing signifi-
cantly higher body temperatures, and of course they can also sing.
Is the ability to sing only a vague similarity? Not a bit of it; now it is
time to see just how far vocalizations are also convergently arrived at.
The ability of birds to vocalize needs, of course, little empha-
sis: twittering, booming, squawks, hooting, and melodious song are
all familiar, as are the mimics, such as some parrots and the mynah
birds. One African Grey I knew could produce a remarkable reper-
toire of sounds, including the Greenwich time pips with eerie ac-
curacy and now, her companion Caroline Pond tells me, is well up
with mobile phones and car alarms. And there are more poignant in-
stances. Darwin, for example, in a section on the extinction of races
and tribes remarks that ‘Humboldt saw in South America a parrot
which was the sole living creature that could speak a word of the lan-
guage of a lost tribe.’170 Even so, we tend to regard bird mimics as
simply amusing, and their songs as only incidentally beautiful and
in reality vocal expressions of territorial demarcation and the urgent
priorities of mate attraction. Perhaps so, but it transpires that in real-
ity there are under-appreciated similarities between our vocalizations
and those of the birds, both in terms of songs and musical output171
(and even drumming with a stick172 ) and neurology. Thus, with re-
spect to song, both birds and humans share such features as ‘interval
inversions, simple harmonic relations, and retention of melody with

change of key’, while our ‘simple melodic canon . . . is reminiscent of
the matched countersinging of many bird species.’173 So, too, claims
of human uniqueness with reference to such matters as the voluntary
control of the supralaryngeal tract, a prerequisite for human language,
in reality finds parallels in African Grey parrots.174
So far as neurology is concerned, the behavioural biologists
Allison Doupe and Patricia Kuhl remark,175 concerning birdsong and
human speech, that ‘there are striking similarities in how sensory
experience is internalized and used to shape vocal outputs, and how
learning is enhanced during a critical period of development’, and they
continue, ‘similar neural mechanisms may therefore be involved’.176
Significantly, Doupe and Kuhl use the phrase ‘strikingly similar’ a
number of times. Thus, in referring to the vocal control systems in
songbirds and humans, they comment how continuing study ‘reveals
numerous anatomical and functional similarities in the organization
of neural pathways for vocal production and processing . . . both fields
[of enquiry] grapple with strikingly similar questions about how sen-
sory and motor processes interact in vocal learning and production,’177
while concerning the learning and self-organization of sounds they re-
mark that the ‘perceptual patterns stored in memory serve as guides
for production [and are] strikingly similar’178 in birdsong and human
speech.
In some ways the similarities of these various processes are all
the more surprising, given both the various differences in avian and
mammalian brain structure, e.g. the absence of a multi-layered cortex
in the avian brain, and in some species of bird strong sexual dimor-
phism of song production. Yet the similarities, striking or otherwise,
still emerge. Even so, Doupe and Kuhl are careful to qualify these re-
marks, noting that although the parallels are striking there are also a
number of obvious differences, most notably the human possession
of a grammar (but see Chapter 9). Yet what they rightly call the ‘nu-
merous parallels’179 between my remarking to my companion on the
beauty of a bird’s song and the song itself, suggest that not only warm-
bloodedness and viviparity but also at least some mechanisms of both
vocalization and song may be widespread across the Galaxy (see note
132, Chapter 9). So, too, given the recurrent emphasis on evolutionary
convergence, it is not surprising to learn that in the birds the powers
of song and vocalization have evolved at least twice.180
Song, the birth of live young, and warm-bloodedness, not to

mention that honorary mammal, the New Zealand kiwi, all provide
compelling examples of convergence that show how features we as-
sociate with the mammals are more widespread and may, therefore,
reasonably be expected to emerge elsewhere. Nor do the examples end
here. Mammalian dentition is well known for its complexity. Earlier,
we encountered the convergent evolution of the massive canines in
the sabre-toothed cats and marsupial thylacosmilids. In addition to
the canines, most mammals have nibbling teeth (incisors) and the
familiar battery of molars for grinding or shearing.181 That this denti-
tion makes sophisticated food processing possible has no doubt con-
tributed to the evolutionary success of the mammals. In contrast, the
dentition of the reptiles is much simpler. Typically it consists of a
sharp array of pointed teeth, and if there is any difference along the
length of the jaw it is usually only one of relative size. Exceptions,
however, are known, and they show striking convergences with the
dentition of mammals. Not only that, but this type of dental conver-
gence has arisen at least twice,182 in two distantly related groups of
reptiles, both of which lived in the Cretaceous. The first concerns a
teiid lizard, from sediments close to the Cretaceous–Tertiary (K/T)
boundary in Montana. Here the posterior teeth on both the lower
(dentary) and upper (maxillary) jaws have become molariform, and
resemble the so-called tribosphenic arrangement of mammalian mo-
lars with its series of interlocking cusps.183 In addition, the method
of regular tooth replacement, whereby old and worn teeth are dis-
carded and new ones take their place, which is the norm in reptiles,
is suppressed.184 As in the mammals, after the loss of the deciduous
(or ‘baby’) teeth, the next set is for life. The second example concerns
some crocodilans from China and Malawi, whose dentition has again
ground towards the molariform solution. In these examples chewing
appears to have been somewhat less effective in its action, without
direct occlusion. Unlike the teiid lizard, which probably ate insects,
these crocodiles were vegetarians.185 While considering the some-
what improbable topic of plant-eating crocodiles, it is surely worth
mentioning in passing another example from the fossil record of a
group of crocodiles (the ziphodonts) that became effectively fully ter-
restrial and in doing so developed something rather like hooves.186
And are crocodiles unique? Well, not really, or at least so far as the
skull is concerned. The spinosaurs,187 hitherto regarded as a rather
enigmatic group of theropod dinosaurs, have a skull that is interpreted

as a crocodile mimic.188
By now, I hope it will be clear that many of the evolutionary fea-
tures that help to define the human are convergent. If such features as
warm-bloodedness, vocalization, and even agriculture can evolve in-
dependently, then so, too, on any suitable planet the same will emerge.
Yet at this stage there is surely a dimension missing. To be sure these
and many other convergent features serve to delineate complex biolog-
ical systems, but the scope is still very wide-ranging and encompasses
animals as disparate as ants, tuna, and kiwi. Even if we grant that
‘mammal-ness’ is a biological property rather than a historical con-
tingency, we still seem to be far removed from anything specifically
human with such hallmarks as bipedality, tool-making, culture, and
intelligence. Let warm-bloodedness, vocalization, and even agricul-
ture be convergent, but surely the hallmarks of the human are simply
the quirky results of contingent happenstance. On Threga IX there
will be much that is reminiscent of Earth, but – so it is widely be-
lieved – any consciousness will be submerged in the inarticulate and
any music will be little more than harmonic babbling. The opposite
turns out to be the case.
9 The non-prevalence
of humanoids?
What we know of the social insects, and especially the extraordinary

organization of agriculture and warfare among the ants, is striking
both in terms of their convergence and in the almost alien nature
of these complex societies. Certainly the jointed skeletons, the com-
pound eyes, the miniaturized clones, and apparently robotic social
organization are a familiar staple of science fiction. Suppose that there
are advanced extraterrestrials: will they be like us, at least vaguely hu-
manoid, or so alien as to defy belief and perhaps even recognition, let
alone communication? The majority certainly tends towards the lat-
ter opinion. It probably owes as much as anything to George Gaylord
Simpson, one of the last century’s great evolutionary biologists.1 He
was a prolific writer, and among the 16 papers he published in 1964
was one baldly entitled ‘The nonprevalence of humanoids’.2
Simpson’s article, presented with characteristic force and intel-
ligence, was a sustained protest against what he saw as unwarranted
extrapolation and conjecture. He argued that the history of life as
revealed on Earth could not possibly be taken as a useful guide to
biological events anywhere else in the Universe. Simpson presented
his argument carefully, acknowledging that planets suitable for habi-
tation would probably be in fairly short supply, and he conceded fur-
ther that the likelihood of life itself arising was even lower. But no
matter: as Simpson remarked, the Universe is a big place, so how-
ever uncommon life was, the total number of planets with life must
be quite large. But, he concluded, whatever happened on these re-
mote worlds in terms of evolution would bear little resemblance to
what we see on Earth. Not surprisingly, given Simpson’s life’s work,3
his argument was strongly neo-Darwinian: while no stranger to the
importance of convergence4 he was adamant that the emergence of
anything like a human would be an evolutionary fluke.5 As Irven
DeVore exclaims, ‘Extraterrestrial intelligence? Not likely.’6 As such,
our presence on this planet was a priori unpredictable because the
evolutionary trajectory that led to us was only possible according to
the ‘precise conditions of our actual history.’7
230 the non-prevalence of humanoids?
interstellar nervous systems?

Nor was Simpson the first to argue along such lines. George Beadle,8
for example, also acknowledged the probability of extraterrestrial life,
but dismissed the likelihood of any such organism being hominid-
like, remarking that ‘the a priori probability of evolving man must
have been extremely small – for there were an almost infinite num-
ber of other possibilities.’ He pessimistically continued, ‘Even the
probability of an organism evolving with a nervous system like ours
was, I think, extremely small because of the enormous number of
alternatives.’9 Perhaps so, but what are the constraints? Central to
an effective nervous system is the sodium channel, which by pump-
ing sodium ions across a membrane provides the electrical activity
necessary for the rapid propagation of action potentials (i.e. changes
in voltage) and hence transmission of the signal along a nerve. This
arrangement is sometimes said to be unique to the animals, evolving
first in the jellyfish and their relatives.10 Even if its primitive func-
tion was linked to the evolution of the nervous system, it is worth
noting that the sodium pump has also been adopted for a wide variety
of other physiological processes.11 Sponges, which lack (and perhaps
never had) a nervous system, are regarded as the most primitive group
of animals and show no evidence for electrical activity in the form of
action potentials.12 Even so, there is a sophisticated single-celled or-
ganism, the heliozoan known as Actinocoryne that consists of a ‘head’
with radiating filaments and a stalk that attaches the cell to the sub-
strate. If disturbed, the stalk and filaments are capable of remarkably
rapid contraction, and the action potential that triggers this response
is mainly dependent on sodium.13
Sodium transport14 is also found in the bacteria, where it has
various roles, such as in respiration and the movement of the rotary
flagellum.15 Rather remarkably it now transpires that in at least one
bacterium this transport, which can occur in a number of different
ways, is specifically by a sodium channel.16 The evolution of sodium
channels therefore does not seem to involve some wildly improbable
process. It should also be noted that ion channels dedicated to the
movement of potassium and calcium ions, which are fundamental in
controlling cell physiology and excitability, are effectively universal.17
It seems very likely that the sodium channel evolved from a pre-
existing calcium channel.18 Nor was this necessarily very difficult to
achieve, given that single mutations can change the behaviour of the
the conceptualizing pancake 231
sodium channel so that it acquires the characteristics of a calcium

channel.19 So why not carry on using calcium, since it was already
available? One problem is that calcium has a number of key roles
within the cell, and sudden fluxes, necessary to generate the electri-
cal signal associated with nervous activity, would lead to intracellular
mayhem.20 Safer by far to design a specific mechanism to handle fast
action potentials. The fact that the sodium channel evolved indepen-
dently of the animals in a single-celled heliozoan, capable of very
rapid contractile movements, strongly suggests that at least from this
point of view Beadle’s scepticism about a nervous system being able
to evolve may be over-pessimistic. Not only that, but Milton Saier21
points out that the various transport proteins, including those in-
volved with ion transport,
apparently exhibit a common pattern of six tightly clustered

transmembrane helical segments with both the amino and
carboxy termini localized to the cytoplasmic membrane surface. It
may be that a three-dimensional transmembrane structure
consisting of six nearly parallel α-helices is particularly well suited
to transmembrane channel formation. Such constraints could
account both for the retention of this structural motif during
evolutionary divergence of homologous transport proteins and for
convergence of evolutionarily unrelated transport proteins.22
Thus, as we saw earlier, not only are there good arguments that aliens
would see and smell using very similar proteins to those we use, but
their electrical conductivity would again converge towards the same
solution.
the conceptualizing pancake

It is worth remembering that the scepticism of such people as Simpson
and Beadle was set in the context of the beginnings of the exploration
of outer space. In addition, it coincided with the growing interest
in what came to be known as the SETI (Search for Extraterrestrial
Intelligence) projects, galvanized by Frank Drake and others with the
formulation of the now-famous Drake equation (which attempted
to estimate the number of extraterrestrial civilizations23 ) and the
OZMA Project (which was a preliminary search for signals using a
radio-telescope in Green Bank, West Virginia).24 Since then the ques-
tion of the non-prevalence (or otherwise) of humanoids has remained
an important ingredient in the SETI programmes, not least because

of its potential implications for communication. Leonard Ornstein,25
for example, took a view broadly similar to that already set out by
Beadle and Simpson. His remarks echo the view of Temple Smith
and Harold Morowitz,26 which I introduced earlier (Chapter 1) in the
form of the metaphor of navigating to Easter Island: that is, finding
functional states, be they proteins or brains, in a vast ‘universe’ of al-
ternatives, most of which are probably entirely maladaptive, is prob-
lematic. At first sight the alternatives are so remote from each other
that any end-product, such as the one that we label as humanoid,
would seem to be as fortuitous an outcome as anything else. Thus,
Ornstein wrote: ‘The full set of messages [that is the end-products
of the Darwinian process arrived at by mutation] tested by selection,
from the beginning of life, constitute only a minute and probably un-
representative sample of different possible messages from which the
sample has been “drawn” . . . Therefore, no matter how prevalent life
might turn out to be, biological evolution on earth can easily have gen-
erated many “inventions”, perhaps including intelligence, which are
unique in the universe.’27 Despite this, Ornstein reminds us of how
it is that by convergence ‘evolution [can] easily “rediscover” certain
useful classes of “technical solutions” to problems of survival, more
or less independently of evolutionary starting points.’28 As a possible
proxy for intelligence he considers that favourite topic in convergence,
the similarity between the eyes of advanced cephalopods and verte-
brates. This, of course, was explored in some detail earlier (p. 151). But
having got that far, Ornstein suddenly takes a blind alley, so to speak,
by suggesting that in fact these eyes are not really independent at all,
so saving his argument of potential uniqueness.29
So, should we accept the arguments against the prevalence of hu-
manoids put forward by Beadle, Ornstein, and Simpson? I think not.
In the same year that Simpson published his review, another worker,
Robert Bieri,30 took effectively the diametrically opposite view. He
reminded us of the ‘severe limitations [which] are imposed on the
number of routes available to evolving forms. The number of alterna-
tive possibilities is by no means infinite; on the contrary, the number
is quite limited. This limited number of available routes has led to the
innumerable cases of convergent evolution in plants and animals,’31 a
few of which I have reviewed in the previous chapters. Bieri concluded
by saying: ‘If we ever succeed in communicating with conceptualizing
the conceptualizing pancake 233
beings in outer space, they won’t be spheres, pyramids, cubes, or

pancakes. In all probability they will look an awful lot like us.’32
Similar sentiments are offered by the physicist Philip Morrison in
a discussion33 of extraterrestrial life when he writes, ‘Evolution does
not repeat itself; its paths across the badlands of the adaptive landscape
are too complex. But all the same it frequently converges upon states
that seem quite similar (attractors?) along several distinct paths,34 He
continues with pertinent comments related to the emergence of such
evolutionary oddities as complex social systems and tool manufac-
ture, both as we shall see being convergent.
So are we, as humanoids, in some sense either very probable, or
perhaps even inevitable? Given that much of this book is about con-
vergences, no bets need be placed. With a sample of one, that is the
Earth’s biosphere and its four billion years of history, it can always be
argued that a descendant of Bieri, an astronaut, will land on a remote
planet and strike up a conversation with one of those conceptualizing
pancakes, but even at the most basic levels it now seems increasingly
likely that whatever alternatives there might be they are going to be
highly restricted. How many avenues are available even for the ori-
gin of life? Given our lack of success in this area (Chapter 4), could
there be only one? Similarly, as I explained in Chapter 6, George Wald
has argued that wherever there are planets whose biosphere is depen-
dant on the light of a star, there, too, there will always be chlorophyll
(and the associated antenna proteins) harvesting the radiant energy.35
What of DNA? In a brilliant research campaign Albert Eschenmoser
and his team have been constructing what they call an aetiology of nu-
cleic acids, a library of alternative possibilities in the realm of DNA.36
Some of the alternatives may equal the versatility of DNA; some, per-
haps may even exceed it. Nevertheless, the various peculiarities of
DNA, touched upon in Chapter 2, are undeniable: it, too, might be a
molecule uniquely suitable for biological processes. Even the strange
optimization of the genetic code37 indicates that although there must
be alternatives, they are not going to be abundant.
If life is universal, it also seems likely that it has a universal
basis. The aim of this chapter, however, is to extend the argument
for pervasive convergence and thereby question the thesis concerning
the non-prevalence of humanoids. It may need no further emphasis,
but what matters is the emergence of particular evolutionary proper-
ties. It is presumably immaterial which specific lineages develop the
necessary complexities of, let us say, camera-eyes and intelligence, so

long as some do. On Earth we can test this by reference to conver-
gence. Nor need this have very much to do with SETI: as I have noted,
life may well be a universal principle, but that does not prevent our
being alone. So, too, the potential for humanoids may be universal,
but that does not rule out the Earth being unique.
the bricks and mortar of life

In tackling Simpson’s challenge there are obvious questions: at what
level of organization should we begin? At what stage do humanoids
become inevitable? Perhaps, as just mentioned, we can allow our-
selves some basics, such as proteins and DNA. But then where? So far
as humanoids are concerned, perhaps our starting point has to be the
great apes? But here there are plenty of convergences, ranging from
the tools of some New World monkeys to the social organization of
dolphins. Perhaps we should start with the mammals? But what ex-
actly is a mammal? Given that other groups of vertebrates converge in
various ways on the mammals, we may wish to distinguish between
the evolutionary lineage we call mammals and some wider property of
‘mammal-ness’. Perhaps the potential for humanoids lurks somewhat
deeper in evolution, maybe among the fish? Here surely the sine qua
non of terrestrial existence for vertebrates, the four legs, is a fluke-like
development from the pre-existing fins of the aquatic fish? But in fact
the development of legs seems highly probable, given a series of par-
allel trajectories towards limbs in the Devonian fish.38 Still not deep
enough? Perhaps some sort of worm? But here too there is plenty of
evidence for vermiform convergence, even among the vertebrates.39
To be sure, man is not a worm; the point is simply that at each and
every stage convergence emerges.
Ape, mammal, fish, worm, or even single cell? There is no sim-
ple answer to the question about the stage at which something like
ourselves becomes overwhelmingly probable. To provide a focus, how-
ever, much of this chapter will concentrate on those elements that we
might regard as the hallmarks of the humanoid – large brain, intelli-
gence, tools, and culture – all of which, needless to say, show examples
of convergence. Despite this emphasis it is worth re-emphasizing the
notion of inherency. By this I mean that there is a more fundamental
level, inasmuch as many of the molecules (or for that matter, genes)
that we rightly regard as essential for our existence were in fact already
the bricks and mortar of life 235
in place at much earlier stages in the history of life. Indeed, in some

respects this basic architecture (we might call it a sort of biochemical
scaffolding) is as central to the discussion of evolutionary inevitabili-
ties as the veneers of life, such as sentience.
A good example is the iron-bearing protein haemoglobin. For
us, and other animals, it has a vital role in the transport of oxygen
in the red blood corpuscles. Haemoglobin, however, is much more
widespread. In addition to its occurrence in some plants, most fa-
mously in the root nodules of such leguminous plants as the clover,
haemoglobin is also found in more primitive organisms from single-
celled protistans, such as the ciliates, to bacteria, including the
cyanobacteria.40 To a first approximation, irrespective of whether the
haemoglobin is in the blood of a cow, the clover it is cropping, or in
the cyanobacteria forming the scum of a nearby pond, the protein is
doing the same thing, taking care of oxygen. But there the similarities
cease. In the leguminous clovers and the cyanobacteria nitrogenous
compounds are synthesized, but their formation critically depends on
the exclusion of oxygen. Hence a role for haemoglobin. This protein
is highly effective; David Goodsell has referred to it, perhaps incau-
tiously, as ‘perfectly designed for oxygen transport.’41 Haemoglobin is
not the only protein involved with oxygen transport, but as we shall
see (Chapter 10) in the case of the related myoglobin and the copper-
bearing haemocyanin, as well as that of haemoglobin itself, there is
again evidence for convergent origins.
The proteins involved in oxygen transport (and storage) are im-
portant because, however ancient they are, and from whatever prede-
cessor they were recruited, they are a vital ingredient in the success of
large, metabolically active animals, including humanoids. Muscles,
and even more so brains, require copious quantities of oxygen. But
that is only one component. Another key element in the emergence
of at least animal complexity is those chemicals suitable for nervous
activity. A key molecule in the context of nervous transmission is
acetylcholine. This compound plays a key role at the junction where a
nerve fibre (carrying the electrical information from the brain) abuts
against muscle fibre (the contraction of which will lead to a desired
action, e.g. running away from a charging ungulate). At this meeting
point the signal has to be transmitted across a narrow junction known
as the synapse. All in all it is a sophisticated system, yet this molecule
is very ancient.42 Acetylcholine is found in bacteria and plants, and
quite clearly has been co-opted in the animals for neural functions. We
know this also because molecules characteristic of the process of ner-
vous transmission and activity are found in single-celled protistans,
notably the ciliates.43 In these tiny organisms there are no nerves,
let alone brains; so what are these molecules employed for? Ciliates
show a sexual process known as conjugation in which individual cells
must first confirm their compatibility (recognition) before adhering in
order to allow exchange of genetic material. All this depends on the
molecule acetylcholine and a number of other key receptors.
The molecular architecture of the nervous system is complex
in other ways. For example, certain hormones, the neuropeptides, are
also important. At first sight it is quite surprising that hormones such
as corticotropin, which is important in the pituitary gland, as well as
β-endorphin-like molecules and dopamine, have also been found in
the ciliates.44 Given that these organisms lack any sort of nervous
system, the function of these neuropeptides is somewhat enigmatic.
Nevertheless in their own way the ciliates are sophisticated organ-
isms and use of messenger molecules is to be expected.45 George
Mackie46 has aptly referred to these molecules occurring in protis-
tans as ‘prophetic’, and they underline the likelihood of more complex
structures, say a brain, emerging from the unicellular substrate.
Other types of complexity in both bacteria and various eu-
karyotic microbes have been comparatively well known for many
years, not least the remarkable propensity for certain bacteria
(the myxobacteria) and eukaryotic slime moulds (Dictyostelium) to
aggregate into quasi-multicellular organisms, some of which display
a mobile slug-like behaviour. There is, however, newly emerging
information, across a wide front of enquiry, that is demonstrating
hitherto unappreciated levels of complexity that correspond in a num-
ber of interesting ways to the social behaviour of animals and higher
organisms.47 Thus, aspects of sociality such as foraging and coop-
erative hunting, specialized dispersal forms, genetic altruism, and
(perhaps most interestingly) communication, using various chemi-
cal signals,48 have now been identified. As Bernard Crespi remarks,
‘The social phenomena uncovered so far allow the first direct compar-
isons between microorganisms and macroorganisms [and in the text
he lists seven social attributes], which reveal convergences in behav-
ior that are clearly suggestive of adaptation.’49 Not only that but, as
pointed out by various authors,50 the social interactions, synchronized
genes and networks 237
activity, and communication between microbes confer on them multi-

cellularity of a sort. This makes one wonder whether even on planets
that, in my view implausibly, were to remain stuck in a microbial
cul-de-sac,51 parallels to the complexity of animals and plants might
not still emerge.
genes and networks

There is one escape route from this apparent dilemma of the simple
being equipped with molecules essential for the complex. In the fine
print of life’s contract there is a clause allowing the lateral transfer
of genetic material between distantly related organisms. Could, let
us say, animals ‘invent’ neuropeptides, and then export the code to
all and sundry? There are certainly now a number of well-attested
examples of such transfers, but for the most part they seem to entail
exchange between more primitive organisms.52 At this stage it seems
much more likely that in a way analogous to the crystallin proteins of
our eyes, so, too, long before there were nervous systems, at least some
of the key molecules were there ‘ready and waiting’. So the principle
of inherency, widely recognized, but more seldom articulated, appears
to hold.
There is, however, a more general problem of assessing how the
sort of complexity we associate with humanoids emerges from life’s
molecular substrate. This is because of both the immediate question
as to when and where on any evolutionary timetable the emergence
of complex forms becomes inevitable, and to anticipate in Chapter 11
the extent to which genetic fundamentalism has any bearing on ex-
plaining both the structure of life and its potentialities. Thus, despite
their iconic status, it is not clear that DNA and genes are in much
of a position to provide any sort of helpful metric when it comes to
deciding either what is inherent or what is complex. First, it has long
been appreciated that the overall quantity of DNA an organism pos-
sesses provides effectively no guide to its relative complexity: single-
celled organisms may contain far more DNA than a human.53 The
enormous variation in the size of genomes has led most biologists
to regard the ‘excess’ DNA as a sort of molecular ‘junk’, surplus to
requirements.
Another problem, and equally serious, is the news emerging
from the genome projects in which the DNA is mapped and the to-
tal number of genes tallied. For a complex animal, such as a fruit-fly
(Drosophila) or nematode worm (Caenorhabditis) to function, thou-

sands of genes are required, significantly more than for the bacteria,
for which a total of about 4000 is fairly typical. Thus, in the worm
Caenorhabditis the estimated total is just over 18 000. So far, so good,
but in its own way it is a relatively simple animal. As an adult it
has a fixed number of cells, which is one of the reasons why it is
particularly suitable as an experimental species. This is because each
lineage of dividing cells can be followed exactly from the fertilized
egg until the final number in the adult is reached. The cuticle is cer-
tainly rather complex, but the general organization, e.g. gonads, gut, of
Caenorhabitis is fairly straightforward. No eyes either, and the worm
has a relatively simple brain. So perhaps 18 000 genes are about right.
Now consider the fruit-fly. A sophisticated flier with its gyroscopic
halteres, a complex brain with capability for memory and courtship,
compound eyes, and a well-differentiated body. All in all a complex
animal, and how many genes? If the worm needs 18 000 what about
30 000 for the fly? Not a bit of it: the fly actually has substantially
fewer genes, totalling 13 600.54 So what about ‘the pinnacle of cre-
ation’: us? Until recently the estimates of our gene total were in the
order of 100 000, but they too are now being revised downwards; some
people are suggesting as few as 30 000. But perhaps we should not be so
surprised. Claims for the primacy of the gene have distorted the whole
of biology, and as I shall discuss in Chapter 11 other views are perhaps
worth entertaining.
Neuropeptides in single-celled organisms without a nervous
system, and more complex animals built with fewer genes than simple
animals, might in their different ways be thought to pose a problem
in biology. How much of a complex organism, say a humanoid, has
evolved at a much earlier stage, especially in terms of molecular archi-
tecture? In other words, how much of us is inherent in a single-celled
eukaryote, or even a bacterium? Conversely, we are patently more
than microbes, so how many genuinely evolutionary novelties can
we identify that make us what we are? It has long been recognized
that evolution is a past master at co-option and jury-rigging: rede-
ploying existing structures and cobbling them together in sometimes
quite surprising ways. Indeed, in many ways that is evolution. New
insights that confirm this view are now becoming available from the
rapid development in our understanding of developmental biology and
the associated genes. The examples most often cited concern features
of animal architecture such as the body axis,55 dorso-ventrality,56

wing/limb and other appendages,57 eyes (camera versus compound),58
and heart.59 Most often this is expressed in terms of comparisons be-
tween fly (arthropods) and mouse (vertebrates). The conservation of
genetic instructions between these rather distantly related animals
originally came as a considerable surprise, but is now realized to be
part of a much wider pattern. Even so, there is still a world of differ-
ence between the human advancing with raised newspaper in hand,
camera-eyes glaring versus the fly through its compound eyes gaz-
ing at the London Review of Books descending rapidly. Splat? Missed
again? Tut! Tut!
Clearly other genes, and more importantly the respective net-
works of developmental processes, in a fly ascending to safety and the
frustrated human are different. The tension that has arisen from the
discovery of widespread genomic similarity of organisms that are oth-
erwise widely disparate in terms of anatomy and behaviour is only
now being explored. Differences there must be, but at a molecular
level these may be trivial, sometimes even a matter of a handful of
amino acid substitutions. Alternatively, they may arise from already
well-known evolutionary processes, notably gene duplication. In the
case of the newspaper-waving human and the other vertebrates this
latter process, notably the double duplication of the Hox genes, is
widely interpreted as offering the fish and all their descendants new
evolutionary opportunities.60
The details of what will be discovered along the way, as the
various gene networks and cascades are documented, will in their
various ways offer fascinating insights into the structure of life. Why,
for example, do practically all mammals have a fixed number of neck
(cervical) vertebrae? In giraffes and moles, for example, the lengths of
the respective necks could hardly be more different, but in both the
number of cervical vertebrae is seven. In contrast, in the other verte-
brates this total is much more variable. All things being equal, it would
be more ‘sensible’ for the giraffe to multiply the number of neck ver-
tebrae, rather than being ‘forced’ to elongate each of the seven it has.
Why then the constraint? An intriguing suggestion, made by Frietson
Galis,61 is that in the mammals a presumably fortuitous coupling has
arisen from the involvement of key developmental genes (especially
Hox genes) in both the laying down of the axial skeleton, including
of course the cervical vertebrae, and the process of cell proliferation.
If, owing to some developmental abnormality,62 the patterning of the

axial skeleton is upset so, too, there is a tendency to develop child-
hood cancers. These are examples of uncontrolled cell proliferations,
which in this case originate in the developing embryo. For mammals,
departure from seven spells lethality. Moreover, in mammals some
cancers may owe their initiation to the production of highly reactive
molecules (known as free radicals). In the mammals, at least, the free
radicals are an unavoidable by-product of an active metabolism. It
may be no coincidence that the few exceptions to the rule of seven in
neck vertebrae are in the metabolically sluggish animals, such as the
torpid sloth. In this sense the rule of seven in mammalian necks is a
good example of stabilizing selection, and may be the ‘price’ to pay in
ensuring the effective development of very complex organisms. Such
a constraint has, therefore, its costs, but when we see the diversity
of mammals it seems that a restriction to seven cervical vertebra in
animals as diverse as bats and camels has been more than offset in
other respects.
The difficulties in making simple equations between genetic ar-
chitecture and the body form of either a worm or a humanoid, let alone
something more primitive, are considerable. Given the widespread
conservation of genes among the animals, it is hardly surprising that
some must date back to at least the dawn of their history.63 A further
problem is the evidence for extensive redeployment of genes so that
they end up with multiple functions. There are now many examples,
but the famous Pax-6 will serve well because its activity is sometimes
regarded as almost synonymous with the origin of eyes. True up to a
point, but as already mentioned in Chapter 7 in the mammals Pax-6
is also implicated in the development of other organs, including the
nose,64 brain,65 pituitary gland,66 gut,67 and pancreas.68 To be sure, in
these animals Pax-6 is involved with the formation of eyes, but in the
eyeless nematodes Pax-6 is still expressed.69 There is a further compli-
cation because in at least one group of animals (the flatworms) Pax-6
is, as expected, involved in eye formation, but if it is ‘knocked out’
during the process of regeneration (for which flatworms are famous)
then eyes still form.70 As already noted in Chapter 7 (see note 315), it
is possible that Pax-6 was originally involved with differentiation of
nervous tissue and thereby became involved with sensory structures
at the anterior of primitive animals which had at most rudimentary
eyes. Indeed, the fact that Pax-6 is also involved with crystallins71
and rhodopsin,72 which as proteins far pre-date the eye, is probably a

result of co-option for effective optical development.
There are many more examples of multiple functions and re-
deployment of developmental genes.73 Nor is this at all surprising;
these genes are effectively switches that instruct, presage, or order
other genes essential for the formation of particular tissues. This has
led to a protracted, if not agonizing, discussion about what really
is the same in evolution. If Pax-6 makes all eyes, then despite the
differences between camera- and compound eyes are they not really
equivalent? Not in any useful sense, because despite the remarkable
genetic conservation74 the genetic instructions that encode for the fly
eye differ in other ways from those for the vertebrate eye.75 Rising
above a molecular perspective is essential if we are going to under-
stand evolution. Indeed, there is already an uneasy feeling that pat-
terns of genetic expression may tell us more about the constraints of
building organic architecture than providing insights into particular
origins.76 Co-option of genes for new functions is hardly unexpected,
but the questions of how? when? and why? are taking more time to
sort out. If new recruitments are commonplace, perhaps even ubiq-
uitous, this opens the possibility that, at least in certain instances, if
evolution wants to build something it may not have many choices.
Lisa Nagy,77 for example, remarks ‘Should vertebrate and insect limbs
be considered homologous [that is, descended from a common ances-
tor] because they are patterned by similar gene networks? Or is the
similarity an example of molecular convergence, representing not an
extreme conservation of limb construction throughout metazoa, but
merely a consequence of a limited number of molecular tools that an
organism has available to change its form?’78
In a similar vein, a study of a developmental gene known as
distal-less points to repeated co-option in convergently derived struc-
tures. As its name suggests, this gene is typically expressed at the
tips of embryonic appendages, and it has been extensively studied in
the arthropods. We now find, however, that distal-less is employed
much more widely than just in limb development during the early
developmental stages of various arthropods. As with Pax-6 the orig-
inal function of this gene is not certain, but some evidence suggests
that its primary role was linked to the development in the embryo
of the nervous system, and especially the sensory organs. Now it so
happens that in arthropods many of the sensory organs are located on
the appendages, and accordingly when there was a need for improved
sensory perception so parts of the body protruded to extend the spa-
tial range of the sensory cells. Only later were such outgrowths on
occasion employed for such purposes as locomotion. The widespread
expression of the gene distal-less is, therefore, effectively a reflec-
tion of the recurrent and independent evolution of such limbs: in
a sense distal-less hitchhikes as a sensory protrusion and is subse-
quently transformed to allow an additional function such as a leg or
an antenna.79 So limbs, like eyes, may be underpinned by a similar
genetic architecture, but the end-product is still convergent. Nowhere
may this principle be more important than in the origin of advanced
nervous systems and the role of a gene known as otx (the vertebrate
homologue of the otd (orthodenticle) gene found in the fruit-fly).80 In
vertebrates otx plays an important role in the early stages of brain de-
velopment. So, too, does otx in insects, but the expression patterns are
not identical. Most probably the gene is conserved, but echoing Nagy
the researchers Nic Williams and Peter Holland81 also remark that
while conservation is the most plausible explanation, the alternative
of ‘convergent evolution cannot be ruled out by current molecular
evidence’.82
jack, the railway baboon

So, genes and developmental pathways are a very important part of
the story, but in themselves they cannot provide the entire picture.
Differences there patently are, but despite their sometimes momen-
tous consequences it may be that from a molecular viewpoint the
rearrangement of a gene or protein is trivial.83 Indeed, in some ways
we hardly needed years of frenetic activity in this area of genetics
and developmental biology to learn this. At a rather early stage of
this research programme the then surprising fact emerged that ge-
netically the difference between humans and chimps is negligible,84
even though there is a vast difference between living in the trees and
boarding the much-delayed 08.06 train from Cambridge – or is there?
What is inherent in more primitive mentalities? The evidence
tends to be rather anecdotal, but consider the following account told
by Euan Nisbet.85 It concerns the baboon, Jack, who lived near to Port
Elizabeth in South Africa, shortly before the time of the Boer Wars.
The baboon had been adopted, when young, by a Mr James Edwin Wide
who worked on the railway. In an earlier railway accident Mr Wide
had lost both his lower legs, and subsequently he trained Jack in
giant brains 243
some rather surprising directions. To get to his post at the signal box
Mr Wide travelled on a special trolley, and Jack’s job was to put the
vehicle on the rails and push Mr Wide to work (Fig. 9.1). On the down-
hill sections of the track both human and baboon enjoyed the ride.
Nor was the trolley Jack’s only duty. Other responsibilities included
pumping and carrying water, gardening, and locking doors and, when
necessary, handing over a key to passing train drivers who needed
to unlock the points giving access to the coal yard. On one occasion,
when Mr Wide had injured his arm in a fall, Jack, who had already been
trained to work the levers for the signals, following commands from
his human friend, took over the actual signalling. As Euan Nisbet
continues, on the basis of reliable witnesses: ‘Jack knew every one
of the various signals and which lever to pull – not unnaturally the
railway passengers objected initially but the baboon never failed dur-
ing his many years of work.’ It is scarcely surprising to learn that
the emotional bond between ape and man was very close, and found
expression by mutual grooming. On one occasion offensive remarks
made to Mr Wide by another railwayman led to Jack jostling him off
the platform. One wonders what Jack would have made of the much-
delayed 08.06. After nine years of service, Jack died of tuberculosis.
Mr Wide was broken-hearted.
This story, which is certainly true, and strangely moving, is
meant to be more than a diversion, because in its own way it is another
example of evolutionary inherency. So we return to the question of the
prevalence, or otherwise, of humanoids. With so many of the building
blocks of life in place, even at the time of the Cambrian ‘explosion’
more than half-a-billion years ago, when does the merely likely be-
come the almost inevitable? Here, in our tracking of the humanoid,
let us grant, for the sake of the argument, that we have organisms that
are able to move freely, have well-developed sense-organs and a ner-
vous system capable of interpreting all manner of signals. On Earth
we call them animals. What then among all the millions of species is
special to us? One feature that might come to mind, so to speak, is
intelligence, but as we shall now see we are not the only players.
giant brains
The many convergences documented in the last chapter, from the
agriculture of ants to the vocalization of birds, are strong evidence
that the evolutionary emergence of many complex systems is highly
probable, if not inevitable. Yet the sceptic will still pause in thought.
figure 9.1 Jack standing against the trolley with Mr Wide; note the lever frame in the background.
(Photograph courtesy of Euan Nisbet, Royal Holloway, University of London.)
giant brains 245
Man
4 Brain size
Dolphin
EQ
2 Chimpanzee
Squirrel
monkey
Sheep
Pilot whale
1 Hedgehog
Rat Elephant
Hump whale Cat
Opossum
−10 0 10 20 30
Coefficient
figure 9.2 Relative size, or encephalization quotient (EQ), of

mammalian brains plotted against a coefficient that relates total volume
of the brain against the surface area and volume of the cerebral cortex.
This comparison is less familiar than the EQ calculated on the basis of
brain and body volumes, with an expected exponent of 0.67. (Redrawn
from fig. 2 of M.A. Hofman (1982), Encephalization in mammals in
relation to the size of the cerebral cortex. Brain, Behavior and Evolution,
vol. 20, pp. 84–96, with the permission of the author and S. Karger AG.)
Such similarities are indeed intriguing but do they really address the
central question as to whether or not humans as a biological prop-
erty are inevitable? This is because as well as having a complex social
system, agriculture, placentas and live birth, warm-bloodedness, and
vocalization, we have something else. Even if some reptiles give birth
to live young, and others chewed the equivalent of celery, their mental
powers neither were nor are conducive to rumination. Whatever may
be said in favour of reptiles, their brain size is distinctly disappointing.
Bigger brains are largely the prerogative of the birds and mammals, al-
though as we saw earlier the electrosensory mormyrid fish also weigh
in with a hefty brain. To a first approximation the size of the brain
scales to the body mass (Fig. 9.2).86 Most mammals, the group upon
which I shall now concentrate, have a brain whose size matches the
body, but some have a brain smaller than would otherwise be pre-
dicted. The tenrec, an insectivore, is one such example.87 Conversely,
other mammals have disproportionately large brains. Elephants are
big, but their brains are even more massive than would be predicted.
Humans, of course, are the exception of exceptions, with a brain ap-
proximately seven times bigger than it ‘should’ be. This strange con-
dition was arrived at by an astonishing neural trajectory that began
about four million years ago, with the later australopithecine apes. It
is often thought that this must have been a biologically unique event,
unrepeatable and dependent on a series of peculiar historical factors
without parallel elsewhere. William Calvin,88 for example, regards
attempts to conjure up particular explanations for the evolution of a
large brain as little more than caricatures. In particular, and referring
to an idea that has wide currency, he regards such a feature as intelli-
gence as being an unforeseen consequence of a neural machinery that
has been selected for some other reason. Intelligence, in this scenario,
is not primarily adaptational and as such might be reduced to an evo-
lutionary fluke, unique to humans and their nearest relatives and, to
echo G. G. Simpson (see note 2), unlikely to be found elsewhere in
the Galaxy.
In fact the evidence suggests otherwise, at least on this planet.
To start with, there is strong evidence that among the primates those
with bigger brains show more innovatory behaviours, social learning,
and tool use, while among the birds those with greater behavioural
flexibilities and adventurousness (or fewer neophobias) again have
larger brains.89 This accords with the fact, returned to at various points
below, that some monkeys, parrots, and crows are all markedly intel-
ligent. So, too, of course, are the great apes, but at least in this case
the fact that chimps, for example, have important parallels to human
mentality is hardly surprising. To find striking similarities to human
intelligence that might persuade the disinterested reader that such
represents a general biological property likely to emerge on any suit-
able planet, we need to turn to the toothed whales (the odontocetes)
and especially the dolphins.
It has long been recognized that for their size some of the toothed
whales, of which the dolphins are one group, have large brains.90 A
straightforward comparison with other groups of mammals, of which
the humans and related anthropoid primates are the most relevant,
nevertheless runs into some difficulties. Rather self-evidently the size
giant brains 247
6
Pacific white-sided dolphin
5 Tucuxi dolphin
Common dolphin
EQ.67 4 Bottlenose dolphin
Dall's porpoise
3
3.5 2.0 1.0 0.5 Ma

A. afarensis H. habilis H. erectus H. sapiens
Ma = million years ago
figure 9.3 A comparison of encephalization quotients, calculated as

the ratio between brain and body sizes to the expected component of
0.67, a porpoise (c. 3.5), four species of dolphin (c. 4.0–4.5), and the
increasing values in the hominid lineage from the australopithecines to
modern humans. Note hominid EQ only pulls past that of the biggest-
brained dolphins about 1.5 Ma ago. (Redrawn from fig. 3 of L. Marino
(1996; citation is in note 92) with the permission of the author.)
range of the bigger whales far outstrips the great apes: a typical killer-
whale weighs about seventeen times as much as a gorilla. Scaling of
body and brain masses across these size differences is not necessarily
easy. More importantly, the buoyancy of sea water means that a large
body size can be achieved without the imposition of a crushing gravi-
tational burden, and in addition a large part of the body tissue consists
of blubber. As its primary function is that of insulation, it is neurologi-
cally inert and accordingly requires no investment by the brain tissue.
The net result is that toothed whales have a rather different ratio of
body weight to brain mass when compared to the primates. Appro-
priate corrections therefore have to be made, but once this is done
it can be shown that not only do several types of dolphin (including
the Pacific white-sided, Tucuxi, common, and bottlenose) have large
brains, but in proportion to the brains of our three nearest relatives
(chimp, gorilla, and orang-utan) they are significantly larger.91 Indeed,
as Lori Marino has shown, until about 1.5 Ma ago, these dolphins were
the biggest-brained creatures on the planet. Only then, at about the
time of Homo erectus, did the brain size of hominids overtake that of
the dolphins (Fig. 9.3).92
The questions then arise as to how, when, and why did dolphin
brains get so large, and to what extent are the convergences with the
brains of the great apes actually informative?93 So far as the origins of
the story are concerned, it is not surprising to learn that well-preserved
fossil skulls of toothed whales are only moderately common. Nor has
it been easy to measure the so-called endocranial volume, which is a
fair guide to brain size, at least until the advent of non-invasive medi-
cal techniques, such as computer tomography (CT). In any event, the
fairly limited evidence suggests that the early whales had brains of
unremarkable size, some of which were, indeed below average.94 The
transition by whale ancestors from a terrestrial habitat to an aquatic
existence, the elucidation of which has been one of the triumphs of
palaeontological investigation,95 was evidently not in itself a spur
to bigger brains. A dramatic increase in brain size is apparent when
present-day dolphins are compared with these early ancestors, but the
exact history is still not resolved. The increase probably started in
the Oligocene, about 30 Ma ago, followed perhaps by a further jump in
the Miocene. Nor, too, is it yet clear whether these spurts of encephal-
ization were geologically sudden or more gradual.96 Nevertheless, it
appears that by the Miocene (if not before) brain size had increased
suddenly, especially in the ancestors of the porpoises and dolphins.
It is surely significant that this vast organ, which metabolically is
ruinously expensive,97 has been maintained for a protracted period,
probably far in excess of that for hominids, and possibly for as long as
20 Ma.98 Evidently big brains may be, in at least some circumstances,
adaptively useful, and are not just fickle blips of happenstance that
in due course sink back into the chaotic welter of the evolutionary
crucible.99
What, however, might have initiated one or more upsurges in
dolphin brain size? At present perhaps the best evidence is that the
trigger was environmental, specifically the dramatic cooling of the
Southern Ocean.100 Among the consequences of this event, which was
the harbinger of global refrigeration that culminated in the present
ice age, was increased oceanic productivity as marine upwelling
intensified.101 As Australia pulled away from Antarctica, marking
the final break-up of the once immense supercontinent known as
Gondwana, so the circum-Antarctic current became firmly estab-
lished, the ice caps began to spread down from the mountains of the
Antarctic continent, and in the adjacent oceans the baleen-sievers
giant brains 249
and predatory toothed whales both diversified, the deep waters filling
with their clicks, whistles, and other vocalizations. This environmen-
tal trigger has an interesting parallel, perhaps, with the hominid story
because it has been suggested that the major increases in brain size in
such forms as Homo erectus may have been encouraged by increas-
ing aridity in Africa, imposing new stringencies but also spurring new
possibilities.102
Large brains may therefore be favoured when the environment
offers a special challenge. Their persistence for many millions of years
requires, however, additional explanations. Of these it may be that the
emergence of sophisticated social organizations (with an emotional di-
mension) and the necessary corollary of advanced vocalizations may
be especially significant. The social structure of dolphin groups, espe-
cially in the well-studied bottlenoses, is directly relevant to the theme
of evolutionary convergence.103 At first sight it might seem rather re-
markable that the bottlenose dolphins show ‘striking parallels in so-
cial organization and complexity with that of the chimpanzee’,104 as
well as parallels to the ateline or spider-monkeys,105 a group to which I
return below on account of a series of other instructive convergences.
Such associations fall into a category known as fission–fusion soci-
eties, which as the name suggests are rather fluid in their composition,
with alliances and coalitions of varying durations.106 Societies of this
type are, as Marino notes, ‘extremely complex because they represent
a constantly dynamic social situation involving the movement of dif-
ferent individuals into and out of groups at various times’.107 Even
so, some individuals may stay together for protracted intervals, and
a common feature is a stable alliance of two to three male dolphins.
These males form temporary consortships with females. Mating is
probably promiscuous, but so far as the females are concerned it is
also highly coercive.108
The similarity with chimp societies is another fine example of
convergence within social systems. Nor is this resemblance likely to
be accidental. On the contrary, it is much more likely to be adaptive
and arises because despite the radical ecological differences there is
a deeper constraint imposed by the patchiness in space and time of
food resources in both ocean and jungle.109 The parallels are not exact;
why should they be? Moreover, as studies of the dolphins continue, so
further complexities in the structure of their societies are emerging.
They indicate that, in some ways, despite having a fission–fusion
society, the dolphins have advanced further than our closest relatives,
the chimps. It is now apparent that their societal structure is more
complex than hitherto realized, and that with dolphin groups form
so-called ‘super-alliances’. This clearly implies that their intelligence
is well suited to handling an extended social network comprising at
least a hundred individuals.110 Such intelligence is also well attested
in the examples of cooperation between dolphins and humans, no-
tably in the dolphins’ assistance with fishing by providing clues as to
where to cast the net or in herding the fish.111
This example of convergence in a sophisticated social context
finds another very interesting parallel, but this time between two
larger-than-average mammals, the sperm whales and the elephants.112
Linda Weilgart and her colleagues comment that despite their
self-evident differences in ‘a remarkable number of ways, including
life history and ranging behavior, sperm whales and elephants resem-
ble each other more than they do other animals – even ones that share
similar ancestries, diets, environments and predators. The closest re-
semblance is found in their complex and unusual, but comparable,
social organization.’113 In both these gigantic mammals the females
and young form highly social units, highly communicative with vari-
ous long-distance vocalizations.114 Socialization is intense, and in the
sperm whales, for example, there seems to be a form of ‘babysitting’ in
which the vulnerable young are cared for by other adults when their
mothers are engaged in deep dives in pursuit of food.115 The males,
in contrast, are solitary and wide-ranging, and return to the mating
game only when they are not merely sexually mature but big enough
to win contests. Social complexity, communal care of the young, in-
telligence, and memory, as well as longevity, seem to be the key in-
gredients in driving this remarkable convergence.
In the sophisticated and dynamic milieu of many cetacean so-
cieties, it is not surprising that vocalizations are complex and var-
ied, especially in the dolphins. As has been repeatedly pointed out,
the relative opacity of water means that facial expressions are of very
limited use, and so sound production and acoustics have largely taken
priority. In the specific case of the dolphins their fission–fusion soci-
eties are probably highly dependent on the production of recognizable
whistles and other noises.116 Indeed, there is evidence that the vocal-
izations are far from a cacophony. Dolphins are accomplished mimics,
and the speed and accuracy with which they can imitate given sounds
giant brains 251
is highly impressive.117 Evidently they can make contact across quite

substantial distances, and one group of researchers have identified
what they term ‘whistle matching’, in which the receiving dolphin
returns an effectively identical whistle to the sender.118 This, how-
ever, is a very controversial notion, and it is much more likely that
the whistles are simply contact calls. In particular, Brenda McCowan
and Diana Reiss119 present evidence that groups of about 12 dolphins
have a shared whistle type, but embedded in this there are subtle
variations that in part are individualistic. The extent to which indi-
viduals do or can recognize each other remains moot, but all would
agree that dolphin whistles are by no means fully understood; interest-
ingly the type of dolphin known as Cephalorhynchus does not produce
whistles. Its vocalizations, however, are convergent on those of the
phocoenid porpoises.120 It has been remarked that given their ability
for mimicry and their frequent contact with humans it is rather sur-
prising that dolphins’ powers of imitation have not extended to our
speech.121 As it happens, certain other marine mammals are not so
restricted. A famous example concerns the captive Harbour seal, one
Hoover. Perhaps because of a bout of illness, followed by attention-
seeking, the seven-year-old Hoover started to mimic English words.122
As the investigators reported, ‘One observer wrote in the files, “he
says ‘Hoover’ in plain English.” I have witnesses’.123 Perhaps because
he lived in Maine, Hoover spoke with a Boston accent, although the
tendency to slur words made him sound inebriated. A somewhat sim-
ilar case concerned a beluga whale, who also repeated his name and
evidently enjoyed socializing with humans.124 Lori Marino also tells
me that in the New York aquarium the belugas imitate the sound of
the elevated train that runs near by.
The failure of dolphins to imitate human speech should not,
however, distract from more intriguing similarities. In their case
what is especially significant is the way the infant dolphins learn
to vocalize.125 As has been pointed out by various investigators,
this process has strong parallels with humans: there is a whole
range of sounds that evidently represent babbling, over-production,
and finally an attrition to a more standard repertoire.126 Not only
that, but notwithstanding the debate concerning ‘signature whistles’
(note 118), it is known that dolphin vocalizations are considerably
more complex than was once thought.127 It is tantalizing to think that
such vocalizations, and their parallels in birds (note 175, Chapter 8)
and some primates,128 might provide an analogy as to how humans

learnt to speak. In this context it is also worth recalling some re-
marks by Marc Hauser and Peter Marler, who commented that ‘no
one would, of course, claim homology between [bird] subsong and
[human] babbling.’ Rather, similarities would be viewed as conver-
gences, illustrating ‘a basic set of strategies that any species would
be likely to employ if it embarks on the development of a system
of communication based on learned signals’.129 So, too, Diana Reiss
remarks on the discovery of a ‘surprising complexity and plasticity
in the communication, orientation, and navigation systems of many
species . . . diverse species either use or can learn to use, to different
degrees, symbolic or referential communication for intraspecific or in-
terspecific exchanges. This suggests that there may be a convergence
or continuity in the communication and cognitive abilities in ani-
mals from different evolutionary paths.’130 These comments on the
commonalities of vocalization and thereby the transmission of infor-
mation are potentially of universal significance, as it is now being
realized that dolphins might provide a model of how to achieve com-
munication with extraterrestrials. Nor need this be the only clue.
Readers of Mary Doria Russell’s The Sparrow will recall how the
aliens were first detected in a transmission of their singing; so, too, the
song of the humpback whales131 may be a contribution to a universal
music.132
Nor should these comparisons in communication be taken as
some sort of fanciful whimsy: it is possible to teach dolphins to un-
derstand sentences and take the appropriate action.133 The actual ex-
periments involved two dolphins, each one of which was taught an
artificial language based respectively on computer-generated sounds
(approximating to, but sensibly enough not identical to, their whistles)
and on signs given by the hand and arm movements of a human.
What emerged was a degree of comprehension that not only asso-
ciated ‘words’ as symbols for given objects, e.g. a ball or hoop, but
mastered both syntax (word order) and semantics (meaning), even
when the ‘words’ were presented in a novel order. Such instructions
provided a short cut to learning new tasks, that otherwise would
have involved laborious routines and training. Indeed, this work has
been taken further and there is now good evidence that dolphins
can understand human gestures that refer to different parts of their
bodies.134 Thus a gestural symbol achieves, for the dolphin, a semantic
giant brains 253
meaning. It is difficult to avoid the conclusion that dolphins are capa-

ble of abstract thought via mental representations that entail symbolic
referents.
Beyond humans such an ability is rare indeed, as Marino notes:
‘The fact that artificial “language” studies can only be meaningfully
attempted with these very few species [dolphins, bonobo chimps,
and perhaps African Grey parrots] suggests that these few species
have converged toward a level of cognitive complexity that allows
them to understand a simple but symbolic and rule-based system of
communication.’135 This is not to imply, incidentally, that in the spe-
cific case of dolphins, they may one day develop a language with
grammar136 , but their syntactical competence seems to be firmly
established.137 Human language may, on this planet, be unique, but
waiting in the wings of the theatre of consciousness are other minds
stirring, poised on the threshold of articulation.
In this sense the wide divergences of opinion as to the extent
to which animal communications are indicative of the origins of hu-
man language, let alone to the possession of sentience by non-human
species, are less important than the realization that what we call
language is an evolutionary inevitability. Most of the cognitive sub-
stratum is already firmly in place. Nor should we be fooled by the
routine criticism that because only some primates, cetaceans, or par-
rots achieve syntactic competence then the rest somehow have failed.
That competence may in fact be more widespread than we realize.138
But even if it is not, then the explanation for its evolutionary emer-
gence is contextual, notably in terms of highly complex social worlds
where communication and memory are adaptations to a constantly
changing world.139 It would actually be more surprising if conver-
gences did not occur in such complex contexts. Of course, the belief
still persists that the emergence of human language was a contingent
fluke, but everything else we know about evolutionary convergence
and the exploration of functional ‘spaces’ makes this seem increas-
ingly improbable. And what of grammar and syntax, how the world,
as well as the word, achieves meaning? So far as humans are con-
cerned, a universal grammar may have evolved as a result of natural
selection that optimizes the exploration of ‘language space’ in terms of
rule-based systems.140 These themes resonate, of course, with much
of this book, that is, to address the question as to how life ‘navigates’
to particular functional solutions. This agenda has the potential, if
not the promise, of revealing both a deeper structure to biological

organizations and a predictability to the process.
Notwithstanding decades of interest, the more we learn about
dolphins the more remarkable they seem to be. Not only are they
intensely vocal, but they show other convergences. For example,
their powers of memory, at least so far as lists are concerned, are
strongly reminiscent of the memory processes seen in humans,141 and
it seems likely that at least short-term memory is retained as internal
representations.142 So, too, there are similarities in their response to
uncertainty, the dither factor. Broadly speaking, when faced with a
problem, humans decide either to avoid it (escape) or to collect more
information. The difficulty is that in times of high uncertainty a de-
cision to collect more information may lead to a catastrophic error.
In comparing the uncertainty responses of humans and dolphins an
investigation by David Smith and his colleagues concluded that ‘The
dolphin performed nearly identically . . . Human and dolphin uncer-
tain responses seem to be interesting cognitive analogs.’143
Another observation seems at first just a curiosity. Dolphins
sleep, but they also need to swim continuously lest they drown by
failing to come to the surface. What do they do? Effectively either the
left or the right side of the brain falls asleep, while the other side re-
mains fully conscious with the corresponding eye open. Interestingly,
birds have convergently arrived at this arrangement.144 It seems pos-
sible that at least some birds can engage in such unihemispheric sleep
while flying, but its principal function seems to be as a protection
against attack. For dolphins, however, having only half a brain asleep
is apparently to keep in contact with the rest of the school.145 Dolphin
brains therefore show a high degree of what is referred to as inter-
hemispheric independence. This is presumably a precondition for lat-
erality, which is a well-known feature of the human brain, and may
possibly occur also in the dolphins. As remarked above, dolphin brains
(Fig. 9.4) are large, and the hominids outstripped them only about
1.5 Ma ago. So far as the degree of folding of the neocortex is con-
cerned, it seems that in this respect the dolphin brain is the more
convoluted.146 The same applies to the cerebellum, at least in the
bottlenose and common dolphins, in which the cerebellum is sig-
nificantly larger than in any primate, including humans.147 Marino
and colleagues suggest that this enlargement of the cerebellum is not
only linked to coordination of movement, but has broader functions
figure 9.4 A comparison of the human (left) and dolphin (right) brains, with some of the principal
differences listed. (From http://brainmuseum.org/specimens/cetacea/dolphin/index.html, courtesy
of L. Marino, Emory University.)
connected with the complexities of sensory processing, memory, and

cognition.
These differences, in neocortical folding and cerebellum, under-
score a key fact. This is to the effect that the large brain of the dolphins
may have arisen from a common mammalian substrate,148 but it is
very far from being identical to that of the human.149 The contrasts in-
clude a marked and very different development of particular areas and
lobes, with the temporal and parietal regions emphasized in dolphins
as against the frontal and occipital in humans (and other primates).
Particularly noticeable is the development in the dolphins (and other
cetaceans) of a unique paralimbic lobe, unknown in other mammals,
which evidently has a special role in a variety of functions, including
vision, touch, and motor activities. As Marino notes, ‘The segregation
of the limbic and supralimbic regions by an interposed paralimbic lobe
is a radical departure from the typical terrestrial mammalian pattern
of cortical evolution;’ she continues, ‘The fundamental topography of
cortical sensory projection regions in a cetacean brain also stands in
stark contrast to that of the primates.’150 Nor are these the only differ-
ences: the neocortex is thinner, the internal laminae are less defined,
and the cellular structure is generally simpler.151 To quote Marino
again, ‘The differences in gross morphology and topological arrange-
ment of cortical sensory zones between cetacean and primate brains
persist at an even deeper level in the form of profound differences
in cortical cytoarchitecture.’152 Yet despite all these differences the
degree of cognitive convergence is striking.
The presence of an enlarged cerebellum has already been alluded
to, and it is probably no accident that it is similarly enlarged in the mi-
crochiropteran bats, which also echolocate.153 We should also recall,
at this juncture, the immense cerebellum of the mormyrid fish with
their highly developed capacities for electrogeneration and electrore-
ception (Chapter 7, p. 182). All these animals are masters of sensory
processing, possessing in their different ways remarkable acuities. In
the case of dolphins and humans it will be apparent that despite both
are mammals their brains, as Lori Marino stresses, ‘represent two
fundamentally different cortical organizational themes . . . [and thus]
compels the conclusion that any similar complex cognitive processes
between primates and cetaceans are convergent.’154 The evolution of
the brain cannot therefore be divorced from its adaptational require-
ments and a corresponding moulding of function to those needs.155
giant brains 257
figure 9.5 Dolphins with

their bubble rings. (Photograph
courtesy of Brenda McCowan,
University of California, Davis
and Lori Marino, Emory
University.)
This is important, not only because it confers a predictability to the

evolution of brain structure (see pp. 266–267), but because it also sug-
gests that while the routes to adaptive success, of which one is a
higher intelligence, may be quite strikingly different (as in dolphins
and humans) the end-points converge again and again.
It remains the case that in some ways the dolphin brain is a cu-
rious amalgam of a rather archaic brain, reminiscent of the relatively
primitive condition seen in such mammals as the hedgehog and the
bat,156 combined with the massive enlargement of the deeply con-
voluted lobes that gives it an uncanny similarity to our own brains
(Fig. 9.4). Even so, the complexity of the dolphins’ social life, sophis-
ticated and specific communication, and ability to mimic and learn
clearly indicate that complex intelligence is not the unique preserve of
humans, and that in some respects the convergence seen in dolphins is
more compelling than the classical comparisons with the apes. To this
list should, of course, be added the playfulness of dolphins, perhaps
best exemplified by the blowing of bubble-rings (Fig. 9.5) where both
some degree of forethought in their production and an ability to mani-

pulate the ascending circle of air are again consistent with some degree
of cognition.157 What is often regarded as a key ability, that of self-
recognition (usually tested by using mirrors), has long been suspected
in dolphins,158 and has now been unequivocally demonstrated.159
This finding is of particular importance because it helps to refute
the notion that something like a human can emerge only by some
quirky path of evolutionary happenstance. As the investigators, Diana
Reiss and Lori Marino, remind us,
‘Bottlenose dolphins, great apes, and humans all possess high

degrees of encephalization and neocortical expansion . . . Yet the
brains of dolphins are markedly different from those of primates
on many levels . . . reflecting the fact that the cetacean . . . and
primate ancestral lines diverged at least 65–70 million years ago.
The present findings imply that the emergence of self-recognition
is not a byproduct of factors specific to great apes and humans but
instead may be attributable to more general characteristics such as
a high degree of encephalization and cognitive ability . . . More
generally, these results represent a striking case of cognitive
convergence in the face of profound differences in
neuroanatomical characteristics and evolutionary history.’160
The evidence therefore seems to be increasingly consistent not only

with a cognitive capacity, but with one arrived at independently and
convergently with humans. These abilities, it seems, can encompass
a comprehension both of symmetry and of signs,161 thereby indicating
that, as I have already suggested, the notion that dolphins might well
be capable of abstract thought is not in itself intrinsically absurd.
In the wider context of the toothed whales (such as the killer
whales) and other marine mammals (such as the elephant seals and the
Weddell seals) it has long been recognized that their vocalizations are
not necessarily uniform across their geographical ranges, but in some
instances can be divided into recognizable variants, that is, dialects.162
Some investigators are now going further, and speak of the cultural
transmission of behaviours (that is, a state in which transmission of
information is no longer entirely reliant on the gene). And what is
patently true of humans most probably applies also to various marine
mammals, and perhaps not surprisingly to some birds.163 For various
whales such transmission involves not only vocalization but other
giant brains 259
activities, such as feeding behaviours.164 In the light of such work

it is difficult to escape two conclusions: first, that the emergence of
cultural capabilities represents a continuum, and second that conver-
gences are inevitable. This is not to deny that humans have gone fur-
ther; they have what has been termed a ‘hyperculture’, but it does not
rule out such a phenomenon evolving elsewhere, whether on Earth or
on Threga IX. That cultural attributes are manifested in such features
as vocalization should not make us forget that their contexts are ones
that are firmly embedded in a social matrix. This in turn has intrigu-
ing ramifications, such as, for example, the convergent development
of the female menopause. Far from it being a puzzle why evolution
‘permits’ post-reproductive females, there are in fact probably sound
adaptive reasons why it should in some complex societies.165 More
than an echo of this has also been found in the African elephants,
where the oldest of the females also tends to be the wisest, the best
able to remember elephants from other groups, and thus able to trans-
mit the requisite social knowledge to her own ‘family’.166 As already
emphasized (note 112), elephant and sperm-whale social structures are
strongly convergent,167 and it has been repeatedly pointed out that to
slaughter the largest and oldest in such matrilineal clans is a stupid
course of action if the result is to deprive the group of the accumu-
lated knowledge that enables it to prosper. It is a small irony that as
the Darwinian synthesis was emerging in mid-Victorian Britain the
rooms were lit by the burning of sperm whale oil, and the soothing
music from the piano arose as the fingers moved across the elephant
ivory keys.168
In the case of whale vocalizations it is also evident that a number
of discrete categories are recognizable, but in the context of their pos-
sible cultural transmission what is significant is that only certain call
types change with time, whereas others remain invariant. As certain
call types change in a particular social group (which has a matrilineal
social structure; see note 165), they will typically diverge in struc-
ture from the sounds made by another social group, although it is also
possible that on occasion calls will in fact come to resemble one an-
other. In either case there is a clear implication that within a social
group a component of this change must be by learning, that is, by a
cultural instruction. In their study of a pod of killer whales and their
vocalizations Patrick Miller and David Bain169 inclined to a view that
the changes arose by cultural transmission mediated by learning, and
they were not afraid to suggest that such learning might occur ‘as a
consequence of exposure to the sounds of tutors’. They concluded that
the ‘Vocal similarity [of the pod] seems to be correlated with the mul-
tilayered structure of killer whale society, and ultimately to the social
interactions that contribute to the stability of the social structure.’170
Shifts in sound production might certainly, as the various investi-
gators are careful to point out, be underpinned by a genetic change,
but it seems just as likely that it is by imitation, improvisation, and
experimentation. So, too, in humpback whales the arrival of some
‘foreigners’ led to a rapid adoption of the new song, literally a ‘cultural
revolution’.171 And should we be so surprised? Like humans, these ma-
rine mammals are long-lived, show a high degree of cognition, demon-
strate prolonged parental care, and form cohesive societies.
In their various ways a number of marine mammals, and es-
pecially the dolphins and some of the other toothed whales, are in-
triguing in their social complexity and cognitive abilities. Much is
convergent with other intelligent mammals, including humans, and
it is a clear enough indication that at least in this biosphere if we had
not emerged as the cerebral species then at some point, and probably
sooner rather than later, someone else would. In this sense, humans,
as a biological property, were inherent from at least the Cambrian
period, if not before. The life of a dolphin appears to be rich and com-
plex, but in evolutionary terms these large-brained cetaceans seem to
have reached an impasse. Because they are highly adapted for swim-
ming and living in the sea, they would seem to be unlikely candi-
dates for the emergence of an advanced technology. Even so, despite
their aquatic milieu, when it comes to tool use then, once again, the
flippered dolphins surprise us. At least one group, principally more
solitary females, has learnt to root up conical sponges and stick them
on to their anterior beak, a structure known as the rostrum.172 Sev-
eral alternative explanations can be entertained. Perhaps it is just that
dolphins enjoy fooling around. Another possibility is that the sponges
have medicinal properties. The most likely explanation, however, is
that the sponge acts as a natural glove and as the dolphin rootles
around in the sea bed, so it receives some protection from an alarm-
ing array of venomous animals, such as the stone-fish, scorpion-fish,
sting-rays, sea-snakes, and the occasional blue-ringed octopus. Nor is
this the only example of tool use by cetaceans. We saw earlier evidence
of cognitive capacities in the formation of bubble rings by dolphins,
grasping convergence 261
but the humpback whales produce nets and curtains of bubbles to

trap their prey.173 Perhaps on planets that are entirely oceanic (see
p. 92) all advanced cognition resides in cetacean-like societies. There,
as Michael Denton reminds us,174 access to fire and the skills of smelt-
ing and metallurgy would for ever be denied.
grasping convergence
If the path to the humanoid is characterized by any feature other than
the carrying of a large brain rich in mentalities, it is the tools dropped
on the way. Chosen and then discarded, tools epitomize intelligence
and the purposeful. To be sure, those constructed by animals may
seem primitive in the extreme, but it is clear enough that human
technologies ultimately are based on the same antecedents. More im-
portantly it is also now obvious that such an evolutionary ability may
be rare, but it has emerged independently a number of times: inherent
in evolution is not only intelligence and cognitive sophistication, but
also technology. Tool use, per se, is of course well documented in a
wide variety of birds,175 but a particularly famous example is provided
by the New Caledonian crows.176 This example has excited interest
because of evidence for the manufacture of both particular tool types,
including hooks,177 and their standardization, features that had been
thought to be effectively restricted to the advanced primates. In ad-
dition, Gavin Hunt documents lateralization in tool use and a rule-
based method of construction.178 He also stresses that in a number
of respects the more famous chimpanzee tool cultures do not match
that of these crows, and concludes:
At the least, crows provide an extant species for learning about the
neuropsychology associated with . . . tool-making, such as
handedness, hook use and the shaping of tools to rule systems,
including an opportunity to see whether left-hemisphere
specialization of the brain for the organization of sequential,
manipulatory behaviours in tool-making might indeed be
phylogenetically very ancient. If crows’ tool behaviour involves
cultural transmission, they also offer the opportunity for studying
tool-making by pre-modern humans where cognitive, behavioural
and social processes may have resulted in largely repetitive rather
than innovative tool manufacture, and symbolism and language
were rudimentary or absent.179
Apart from the parrots, already mentioned, and for which considerable
evidence for cognitive sophistication exists,180 many ornithologists
have remarked on the general intelligence of crows. Recent observa-
tions on tool modification and goal-directed activities in the American
crow bear this out.181
Tool use in chimps has, not surprisingly, received extensive at-
tention. Most of the work focuses, reasonably enough, on the use of
sticks and stones.182 The closely related bonobo is also known to en-
gage in tool-making. Remarking on one particular case, Nicholas Toth
and his colleagues183 noted that the bonobo in question, one Kanzi,
showed ‘exceptional progress to date [but] his skill in flaking stone still
contrasts sharply to that of Oldowan hominids’.184 They were unsure
as to whether this was more a reflection of manipulative abilities as
against cognitive constraints. Interestingly, however, chimps also use
twigs for dental care in activities that include cleaning the teeth and
helping to yank out deciduous ones.185 Much of this activity is self-
directed, but examples are also known of one individual performing
elementary dentistry on another chimp.
Nor is tool use among non-human primates confined to chimps.
Significantly in some of the New World monkeys, whose evolution-
ary history has been separated from that of the Old World monkeys
(and their descendants, the great Apes) for about 30 Ma, have conver-
gently acquired tools. In particular, capuchin monkeys (Cebus), and
also some of the callitrichids (specifically the golden lion tamarins186 )
show extensive tool use for a variety of purposes, including a prim-
itive lithic technology when provided with suitable materials in
captivity.187 Suzanne Chevalier-Skolnikoff188 also emphasizes that
tool use is unlikely to have arisen by trial and error (fiddling around
if you will) but is a direct product of advanced sensorimotor abili-
ties: that is, capability combined with motivation. It is difficult to
escape the conclusion that once such abilities are in place, tool use
becomes an inevitability. In reviewing the earlier literature on tool use
by capuchins, Gregory Westergaard and Stephen Suomi provide an ef-
fective counterpoint to Chevalier-Skolnikoff’s comments.189 They re-
mark, ‘We suggest that the ability to make and use simple stone tools
is a primitive behavioral capacity that may have been “discovered”
numerous times and utilized by more than one hominid genus and
species’.190 This idea echoes an earlier suggestion by Sue Parker and
grasping convergence 263
Kathleen Gibson of parallel developments in the capuchin monkeys

and great apes.191 This theme was returned to by G. C. Westergaard
and his colleagues192 in summarizing a discussion of handedness, lo-
comotion, and tool use. They remark, ‘We do not hold that paral-
lels between capuchins, on the one hand, and apes and hominids, on
the other, resulted from homologous processes. Instead, we specu-
late that an array of behavioral similarities, including food-sharing
and tool-use, evolved through convergent processes in Cebus and the
common ancestor of great apes and modern humans.’193 In parallel
to the chimps, however, the capuchin tools do not rival the most
primitive of hominid technologies, known as the Oldowan. Nor is it
entirely clear why the capuchin seems relatively adept at tool use in
captivity, but there are few reports of such activity in the wild.194 As
Elisabetta Visalberghi195 notes, this may be due to lack of observa-
tion or to their tree habitat, whereas primate tool use is very much
associated with ground activities.196 It is also worth remarking that
in capuchin tool use the manipulation has a pronounced bias to right-
hand employment.197 It is significant that this discussion is put in a
selective context by the intriguing possibility of a separation in the
brain ‘between language and object manipulation’.198 In parenthesis,
it is important to stress again that the point of this book is to explore
the likelihood of the emergence of certain biological properties. It is
not my intention to persuade you that capuchins will one day evolve
into humans; they won’t. To start with, their mentality has strengths,
but also limitations.199 Nor need this tool use necessarily be an appro-
priate guide to early hominid activities.200 The simple point is rather
that tool use is far from being some sort of quirky by-product of evo-
lution, and when tools are first employed other things may become
far more probable, at least in some circumstances.
The examples of the capuchin monkey and especially the New
Caledonian crow demonstrate that tool use is patently convergent.
For us tools and hands are almost synonymous, yet the birds (and
even dolphins) are a reminder of valid alternatives. Yet, by and large,
tool use would seem to be a prerogative of the vertebrates. So far as
there are rivals in a terrestrial context, for advanced complexity we
would most probably turn to the insects, some of whose marvels have
already been explored. Consider, however, this account by Samuel
Wendell Williston of the behaviour of a parasitic wasp.201 It is quite
a long quotation, but if at the end you don’t rub your eyes, obviously
you are more immune to surprise than I am. Thus Williston writes:
An insect, alighting, ran about on the smooth, hard surface till it

had found a suitable spot to begin its excavation, which [when
finished] was . . . nearly vertical, and carried to a depth of about
four inches . . . The earth, as removed, was formed into a rounded
pellet and carefully carried to the neighboring grass and dropped . . .
When the excavation had been carried to the required depth,
the wasp, after a survey of the premises, flying away, soon returned
with a large pebble in its mandibles, which it carefully deposited
within the opening; then, standing over the entrance upon her four
posterior feet [so leaving the front two free], she . . . rapidly and
most amusingly scraped the dust with her two front feet, ‘hand
over hand,’ back beneath her, till she had filled the hole above the
stone to the top. The operation so far was remarkable enough, but
the next procedure was more so. When she had heaped up the dirt
to her satisfaction, she again flew away and immediately returned
with a smaller pebble . . . and then standing more nearly erect,
with the front feet folded beneath her, she pressed down the dust
all over and about the opening, smoothing off the surface, and
accompanying the action with a peculiar rasping sound. After all
this was done . . . she laid aside the little pebble and flew away . . .
Soon, however, she comes back [with] . . . the soft green larva
[which] . . . is laid upon the ground, a little to one side, when, going
to the spot where she had industriously labored, by a few rapid
strokes she throws out the dust and withdraws the stone cover,
laying it aside. Next, the larva is dragged down the hole, where the
wasp remains for a few minutes, afterwards returning and closing
up the entrance precisely as before [an action that is repeated four
or five times] . . . The things that struck us as most remarkable
was the unerring judgment in the selection of a pebble of precisely
the right size to fit the entrance, and the use of the small pebble in
smoothing down and packing the soil over the opening.202
converging on the humanoid

This example of stone tool use in wasps underlines my repeated em-
phasis that it is not my intention to suggest that there is only a single
type of intelligence. Nor do I intend to imply that technologies are
converging on the humanoid 265
necessarily dependent on a unique configuration of the brain: dolphins

themselves hint at the opposite possibility. Even within the mammals
it can be shown that there are several trends in terms of the structural
organization of the brain whereby different regions, such as the hip-
pocampus, neocortex,203 or olfactory bulbs, are variously developed
or suffer relative atrophy. One such study, by Willem de Winter and
Charles Oxnard,204 has many points of interest. As its basis it took
three major groups of mammals, specifically the bats, insectivores,
and primates. To a first approximation each group has gone its own
way: bats into the night sky in pursuit of moths, insectivores rootling
around for worms and suchlike during nocturnal forays, and primates
swinging through the jungle. Each inhabits a very different world,
and not surprisingly in terms of neural architecture each of the three
groups occupies a largely separate area of ‘brain space’, as defined on
the basis of the relative importance of the different regions of the
brain.
That different mammals emphasize different sensory modali-
ties has long been known, and writers attempting to enter the mind
of, say, a dog will typically try to transpose the wealth of olfactory
‘images’ into something more amenable to human comprehension.
So, too, we can ask to what extent equivalences might exist between
sensory assimilations dependent, say, on olfaction, electroreception,
or echolocation. The last case is of particular interest because in a fa-
mous essay Thomas Nagel argued, as part of a consideration of what it
is to be conscious, that the mind of a bat was effectively unknowable
to us.205 Nagel’s article has attracted an enormous amount of atten-
tion, but it may be that he underestimated the underlying degrees of
similarity in sensory perceptions. Thus, in referring to bat echoloca-
tion he remarked that ‘bat sonar, while clearly a form of perception,
is not similar in its operation to any sense that we possess, and there
is no reason to suppose that it is subjectively like anything we can
experience or imagine’.206 Nagel continued that even for such experi-
ences as we might share with a bat, for example ‘pain, fear, hunger, and
lust . . . [they] have in each case a specific subjective character, which
it is beyond our ability to conceive. And if there is conscious life else-
where in the universe, it is likely that some of it will not be describ-
able even in the most general terms available to us.’207 But is this too
pessimistic a view? The many examples of convergences in sensory
modalities already addressed, and exemplified by the star-nose mole
that ‘sees’ with its nose, suggest that there are deep commonalities in
both neural systems and the associated transduction proteins. These
may enable wildly different sensory systems, in effectively unrelated
animals, to build similar cognitive maps.208 If that is so, then arguably
the root problem of consciousness, the qualia (e.g. the redness of the
sunset on Threga IX), have deep-seated similarities that require an in-
terpreter rather than blank incomprehension. Incidentally, this is not
necessarily to deny another point made by Nagel, to the effect that
our neural architecture may for ever preclude the comprehension of
certain realities. Thus he writes ‘one might also believe that there are
facts which could not ever be represented or comprehended by human
beings, even if the species lasted forever, simply because our structure
does not permit us to operate with concepts of the requisite type.’209
An interesting thought, especially if all intelligences have a universal
neuronal basis.
Whatever truth there may be in these suppositions, the study
by de Winter and Oxnard (note 204) of how these three groups of
mammals occupied the various zones of ‘brain space’ leads to some
very interesting, if perhaps by now unsurprising, conclusions. As has
been long known, within each of the three groups there are recur-
rent convergences in terms of life, habit, and behaviour. For example,
among the bats carnivory, by which is meant the ability to capture
live vertebrate prey, has evolved several times.210 So, too, among the
insectivores there have been multiple convergences associated with
the transition from the terrestrial environment either to the aquatic
realm or to a burrowing mode of life; and in the primates leaf-eating
has evolved several times.211 Such convergences are, of course, a prin-
cipal theme of this book, but the work of de Winter and Oxnard is of
particular importance because they find that the convergences that
lead to the adoption of similar life-styles are mirrored in the brain
structure. Neurology, and by implication mentality, overrides phy-
logeny. One significant implication of this analysis is that it points
to the pervasive influence of evolutionary selection: how else can we
explain such similarities? The net result, as Winter and Oxnard note,
is that one can ‘infer important ecological and behavioural attributes
of a mammal from . . . knowledge of its brain proportions alone.’212
So confident are these researchers that in one case, those bats that
visit plants, for example, to feed on nectar, they predicted that cer-
tain of the Old World bats would show this ecology even though at
the time the habit of plant visiting was believed to be restricted to the
New World bats. Subsequent observations on the Old World murinine
bats confirmed this prediction.213 Indeed, these and other analyses214
indicate that the general rules of brain organization confer a wider pre-
dictability to a number of evolutionary processes. As Leah Krubitzer
has remarked on the evolution of cortical structure,
it is hypothesized that future changes [in cortical-field evolution]

will be shaped by similar mechanisms. Indeed, while the product
in a given lineage of several million years of further evolution
cannot be predicted exactly, which features are likely to be
retained, the types of modification that are likely to occur, and
what will not happen can be predicted with some certainty.215
Research on the convergence and lability of brain structure, and

its correlations with function and ecology, also has an immediate bear-
ing on the following section, which specifically addresses hominid
origins. One key step in this process is evidently locomotor adapta-
tions. As it happens, primates possess a very distinctive quadrupedal
gait, and this may well owe its origin to the art of climbing along
and between fine branches. Very specialized, and no doubt unique?
No, the same gait has evolved independently in the marsupials, again
meeting the adaptive challenge of living near the tops of trees.216 As-
sociated with this is the separation in function between the princi-
pally propulsive hind-limbs versus the use of the fore-limbs (arms)
for grasping and feeding.217 Such an arrangement is regarded as ‘the
primary hominoid adaptation’,218 and it represents, of course, a prereq-
uisite for such features as mobile hands, arm-swinging, and in some
cases ultimately bipedality. Yet it is also known that this locomotory
style so characteristic of the apes is strongly convergent on the New
World monkeys,219 specifically the atelines, or more familiarly the
spider-monkeys.220
It would hardly be surprising if such locomotion and especially
arboreal agility sowed the seeds for a particular type of intelligence.
And in a way analogous to the bats we see that the occupation of
‘brain-space’, as defined by de Winter and Oxnard (note 204), shows
close parallels between the Old World apes and their New World ana-
logues. This is strongly echoed in the increases in the respective mass
of the neocortex among these (and other) primates. The increase de-
fines a clear trend, parallel and independent, that is consistent with
selective pressure. Moreover, despite the evolutionary trend that can

be defined simplistically as ‘lemurs to humans,’221 the details are more
complicated. In particular, New and Old World monkeys show overlap
because both have complex social structures that in various respects
show convergences.222 The case of the spider monkeys is particularly
interesting because their social organization is ‘remarkably similar’
to that of the chimps,223 whose parallels to the fission–fusion society
of dolphins were addressed earlier.
Remembering that Old World monkeys (known as the cercop-
ithecoids) and hominoids (including the apes) are a sister-group, and
recalling the convergences of locomotion and brain structure between
Old and New World primates, it seems entirely plausible that by one
route or another, be it as an island population (see below) or one spread-
ing across open prairie or pampas, a terrestrial (as against arboreal)
New World monkey would have adopted a bipedal stance following
the same evolutionary path as the transformation of ape to australo-
pithecine. After all, preadaptations of locomotory ability, parallels in
brain structures, and complex social systems all point in this direc-
tion. Are we so surprised when we read what Martin Moynihan has
written about Cebus, the capuchin monkey:
Captive capucinus and apella [two of the species of Cebus] tend to

explore and play with all the objects they can lay their hands on
even when they are not hungry. They are very prone to take things
apart. They seem to have a passionate desire to discover what is
inside or behind anything that can be pulled or plucked or
dismembered. They can also put familiar and strange objects
together, sometimes in quite surprising ways, when the spirit
moves them . . . they are more easily bored than any other
primates with which I am acquainted . . . With all their
limitations, however, capuchins can cope with some classes of
stimuli with remarkable ingenuity, almost unparalleled below the
level of man himself.224
If we hadn’t walked out of Africa then probably sooner, rather than

later, our analogues would have strolled225 out of South America,
holding tools (see notes 187–189), and probably enjoying the taste of
meat.226
Carnivorous, with tools, and a restless intelligence? Perhaps,
but bipedal? In the sense of true bipedality humans are certainly
very unusual.227 To be sure birds, and some of their relatives, the

extinct dinosaurs, are bipedal; so, too, are the kangaroos. Apes are
typically tree-dwellers, and on the ground proceed by what is known
as knuckle-walking, which is forced upon them by having a hand
designed principally for grasping. Why the lineage leading to hu-
mans became bipedal is controversial, but the fossil record – although
fragmentary – indicates that it happened shortly after the divergence
with the common ancestor, presumably a somewhat chimp-like beast.
Of the various hypotheses explaining bipedality and the concomitant
freeing of the hands, those connected with thermoregulation and the
necessity of keeping cool during times of activity in open, unforested
areas are attractive.228 This is, however, only one idea and not surpris-
ingly discussion on the origins of bipedality remains lively, especially
with growing evidence that the first hominids may have lived in forest
environments rather than open savannah.229 As our knowledge of the
hominid ‘bush’ of diversification rapidly expands, there is, moreover,
another dimension to this problem. Thus, in a commentary on one
sensational new find, that of the c. 6-Ma-old Sahelanthropus from
Chad (see note 245), Bernard Wood230 remarks that the continuing
accumulation of data ‘predicts that because of the independent acqui-
sition of similar shared characters (homoplasy), key hominid adapta-
tions such as bipedalism, manual dexterity and a large brain are likely
to have evolved more than once.’231
There is also evidence that bipedality evolved independently,
not only among the hominids, but also in the apes. In what is now
Tuscany, but about 7 million years ago,232 was a series of islands in
the Mediterranean, where Oreopithecus233 flourished, most probably
having migrated out of Africa. For about 3 Ma this ape enjoyed a period
of geographical isolation,234 and then something both remarkable and
convergent happened: bipedality emerged. Even though reconstruc-
tions of this Miocene ape sometimes portray it as swinging through
the trees, detailed analysis235 of the Tuscan fossils strongly indicates
that ‘bipedal activity made up a significant part of the positional be-
havior of this primate. The mosaic pattern of its postcranial morphol-
ogy is to some degree convergent with that of Australopithecus [the
genus that preceded Homo] and functionally intermediate between
apes and early hominids.’236 The arguments for Oreopithecus being
bipedal depend on the analysis of the hip237 and foot bones. Nor do
the convergences stop there: it seems that this ape also independently
evolved a hand with a type of precision grip apparently similar to the

arrangement seen in the more primitive australopithecine humans.238
This, of course, is a vital prerequisite for the use of tools. But why did
Oreopithecus become bipedal? The reasons suggested for getting up
on the hind feet are particularly interesting. The argument centres on
the advantages of insularity.239 Living on islands240 meant that the
large predators were no longer a menace, and in addition food could
be readily collected from low-hanging branches.241 Things were fine
until reconnections were made to the outside world and the carni-
vores returned (note 234). Oreopithecus vanished, and the world had
to wait a couple more millions of years before the next adventure to
involve bipeds with a precision grip got under way.
It is also worth noting that, at least so far as the vertebrates are
concerned, the ability to use the fore-limbs for skilled manipulation
is remarkably widespread and is not restricted to the higher primates,
as is sometimes thought. Examples can be found among the frogs and
many mammals, including some marsupials, lemurs, and rodents. It
seems, therefore, that when bipedality arose hands were very much
ready for use. As it happens, these repeated occurrences of skilled fore-
limb manipulations are generally thought to have arisen convergently,
although others argue that they are basic to the tetrapod vertebrates.242
As we imagine the bipedal Oreopithecus wandering around the
islands of future Tuscany, it is tempting to think that some such cru-
cial step in the evolution of early humans also took place on islands,
protected from the terrors of the jungle and savannah.243 But, now, as
the only surviving species of Homo, should we not count our luck, ad-
miring ourselves in the mirror as a unique and otherwise irreplaceable
end-product of the evolutionary process? If evolutionary history had
gone otherwise, it might have led to a planet filled with grunts, whis-
tles, howls, and other yatterings, but never the music of Don Giovanni
or Parsifal. The evidence suggests otherwise. It certainly needs to be
acknowledged that the study of human evolution is riven with con-
troversy. What is generally agreed, however, is that even though you
and I are, by definition, lineal descendants of a bunch of African apes,
hominid phylogeny can no longer be construed as an evolutionary rail-
way: from heavily browed and small-brained tree-climber to smooth-
headed bicyclist. Despite the limitations of the fossil material and
the taxonomic squabbles, it is clear that we are sole survivors of quite
a substantial ‘bush’ of hominid diversification. It probably needs no
converging on the ultimate 271
emphasis, but spectacular as have been the advances in understanding

the diversification of hominids, there is no doubt that further discov-
eries will be made. Consider, for example, the report of a new form,
Kenyanthropus, a contemporary of the African australopithecines and
dating from about 3.5 Ma.244 Even more remarkable are substantially
older discoveries, notably of a hominid 6 Ma old (Sahelanthropus)
from Chad,245 and considerably more controversially what appears to
be a bipedal hominid (Orrorin) from coeval sediments in the Tugen
Hills in Kenya.246 Now, however, Homo sapiens is alone, having lost
(or killed?) its sister-species H. neanderthalensis about 30 000 years
ago. Much is made of this dramatic reduction in hominid diversity,
a seemingly perilous shrinkage to just one surviving lineage. Implicit
in this view is the notion of human uniqueness and our vulnerabil-
ity: when we go, by nuclear incineration or social mayhem, that’s it.
Evolution gave us the chance, and we blew it. Not quite.
converging on the ultimate

Giant brains, tool use, bipedality, and even a precision grip are not,
therefore, specific to humans. To be sure, in our species they are each
developed to a high order, but their independent emergence suggests
that in principle there is no reason why they could not be similarly
honed. Still, that does not rule out the possibility that the human lin-
eage possesses unique features, which if lost by some chance would
never be able to evolve again in the contingent turmoil of continuing
evolution. The facts, however, point in precisely the opposite direc-
tion. When we peer into that ‘bush’ of hominid diversity (note 244),
encompassing perhaps as many as seven species of Homo and a
similar number of the more primitive australopithecines (including
Australopithecus, Paranthropus, and Ardipithecus), and the more re-
mote Sahelanthropus (and perhaps Orrorin), then a familiar tune is
heard. Much has been made of this sprawling bush of hominid diver-
sification from, among others, those anxious to demolish any sense
of linearity in our history, especially if the implication is that this
evolutionary trajectory might even be progressive. This discussion,
however, misses the point. Recall that we are interested only in the
emergence of particular biological properties: some will be of princi-
pal interest to a restricted group; for example, teeth to dentists. Others
might be of universal interest; for example, intelligence. In either case,
however, the convergences emerge.247
Concerning dentistry, one of the most spectacular examples

involves an offshoot of the australopithecines. They are character-
ized by huge grinding teeth (appropriately called megadontic) and the
associated development of a massive jaw musculature. This evolu-
tionary excursion, into what is generally agreed to be serious vege-
tarianism, actually arose independently at least twice, and perhaps
three times.248 These robust australopithecines are instructive be-
cause of what they tell about the realities of evolution and the rules of
convergence. The first time these powerful teeth were developed (in
Paranthropus aethiopicus) is correlated with a massive accentuation
of two sets of jaw muscles, specifically depending on the deployment
of the posterior temporalis and anterior masseter. On the second oc-
casion when the robust habit emerged (in Paranthropus robustus and
P. boisei), one of these muscles had already become reduced so that
the grinding process largely depended on the action of the masseter
muscle.
As Rob Foley249 noted, in reviewing this case of convergence,
‘Clearly, in evolution there are more ways than one to crack a nut.’ He
then went on to remark that ‘these anatomical differences [between
the robust australopithecines] reflect alternative ways of solving the
problems of heavy chewing, and which one evolved depended upon
what had already happened in the lineage’.250 Quite so, yet Foley then
concluded, ‘More interesting though, is the whole issue of conver-
gence.’ It is to this topic that he returns in a short but stimulating
chapter published in Structure and contingency.251 A central insight
made by Foley, echoed frequently, but equally widely neglected, is
his remark that ‘Convergence is perhaps the strongest evidence for
adaptation.’252 Foley is in no doubt that key features in hominid evolu-
tion, such as ‘bipedalism or increased meat-eating, characteristics that
underlay the hominid adaptive radiations, are adaptive, functional,
and the product of natural selection.’253 He does not deny the roles
of contingency and other stochastic factors – why should he? – but
his concluding remarks are judicious and tally with what we know
about the realities of evolution. Thus he writes, ‘while contingency
plays a part in the timing and location of evolutionary events, the way
these events are played out, the final biological outcomes, are strongly
influenced by selection, adaptation and function,’ and he concludes,
‘evolution is the outcome of both stochastic and deterministic pro-
cesses. As such, should the tape of life be replayed, undoubtedly, there
would be many differences, but there would also be a very significant

number of similarities.’254
Hominids evolve, and unsurprisingly given its ubiquity, conver-
gence is inevitable. The independent development of nut-cracking is
a dramatic example, but there are also more subtle instances. In be-
coming more gracile and showing a reduction in tooth size our own
species Homo sapiens marks a departure from the general hominid ro-
bustness and actually approaches once again the more primitive con-
dition found in our pre-australopithecine ancestors. The differences
between us and the australopithecines are, of course, much more ob-
vious. In some ways the most significant is the general absence in the
australopithecines of any evidence of a stone tool culture. The earliest
such occurrences, the Oldowan Industrial Complex, are from Gona in
Ethiopia and are dated at about 2.5 Ma.255 An array of slightly younger
(c. 2.3 Ma) tools from Kenya256 is important because reassembly of
the fragments (cores and flakes) shows that tool-preparation involved
a high degree of dexterity. Although most are evidently the product of
early Homo it is almost certain that the late-stage australopithecines,
including Paranthropus, were also tool-makers.257
Most probably the nascent skill in tool use was already present
in the common ancestors of Homo and Paranthropus. What is, how-
ever, of particular interest is the evidence during geological time for
brain-size increase in the Paranthropus lineage, independent of but
parallel to that seen in Homo.258 Why it was that the brain sizes in-
creased has, of course, been the subject of protracted debate, but the
convergence in Paranthropus might help to constrain these possibil-
ities. Thus, Sarah Elton and her colleagues reinforce one current idea
when they write, ‘The manufacture and use of stone tools as a means
to access meat eating might have been a crucial factor in hominin
encephalization. However, it is also possible that increased meat eat-
ing and exploitation of patchy resources occurred because the existing
cognitive ability and skills of hominids enabled them to make tools,
and therefore facilitated meat eating.’259
The likelihood that lithic technologies would evolve more than
once has already been touched upon, and so it seems reasonable that
at some time, sooner or later, one or more lineages would rampantly
encephalize. Thus, in a discussion of early hominid evolutionary ho-
mologies and homoplasies Daniel Lieberman and his colleagues260
remark, ‘An additional question raised [by our study] . . . is whether
enlarged brains evolved more than once in hominid evolution . . . there

is in fact no theoretical reason why, if large brains are advantageous,
that this should not have occurred independently among different
Pliocene hominid lineages.’ Acknowledging the many difficulties in
obtaining enough reliable data, Lieberman and his colleagues com-
ment that their analysis raises ‘the interesting possibility that in-
creased encephalization evolved independently in more than one clade
of hominids.’261 In a similar vein, Gerrit van Vark262 has argued, on
the basis of a multivariate statistical analysis of cranial material dat-
ing back to c. 1 Ma, that two separate lineages show independent
trends towards human form. As he remarks, ‘“Hominisation” is not
as unique a process as many may think.’263 Nor does van Vark think
this process of hominization is yet complete: in this case it may well
be legitimate to extrapolate existing trends into the future.
While these ideas will certainly be controversial, it is beyond
dispute that much of the hominid evolutionary ‘bush’ became asso-
ciated with stone tool cultures. A striking and much remarked-upon
feature is the conservatism of these cultures,264 with vast epochs of
time elapsing before there is a noticeable shift in technology. Stone
tools, of course, may give a rather one-sided view of these cultures.
Thus, the discovery of wooden spears 400 000 years old,265 evidently
designed to act as javelins rather than to be employed in thrusting, are
timely reminders, not only of the cognitive range of extinct hominid
species (in this case Homo erectus or a near-relative), but also of how
much of these cultures may have literally rotted away.
Other hints of increasing sophistication are the controlled use
of fire from approximately 400 000 years ago266 (earlier records are
controversial267 ), and pathological conditions that imply protracted
care of the sick. These include what appears to be a case in Homo
erectus of hypervitaminosis, an agonizing condition brought on by
the unwise consumption of carnivore livers containing vast and po-
tentially lethal quantities of vitamin A, with the implication that the
afflicted individual received long-term care rather than being left to
the hyenas.268 Despite these hints, and remembering that much ev-
idence has been lost, it still seems that in general the pace of cul-
tural change was, from our technophilic perspective, torpid in the
extreme. Much has been made, and justifiably so, of the so-called
Upper Palaeolithic Revolution.269 This refers to the dramatic appear-
ance of sophisticated tool-kits,270 ornaments, e.g. beads, and perhaps
musical instruments271 beginning about 50 000 years ago. On the last
item, Patricia Gray and her colleagues comment, ‘Remarkably, many

different types of scales can be played on reconstructed prehistoric
flutes, and the sounds are pure and haunting.’272 This culture is, of
course, a product of our own species and unsurprisingly their anatomy
is indistinguishable from ours. Yet the so-called anatomically mod-
ern humans appeared substantially earlier, about 125 000 years ago,
although prior to the Upper Palaeolithic Revolution for the most part
they churned out a markedly different set of stone tools. These are
referred to as the Mousterian culture. While these implements are
perfectly adequate they are decidedly less complex, and once again
strangely conservative, showing little change over aeons of time.
What may have been a rather abrupt transition from Mousterian
to the more advanced Aurignacian tool-kits in Europe was possi-
bly heralded by substantially earlier developments, such as those re-
ported from central Africa.273 Indeed, the very concept of an Upper
Palaeolithic revolution is now in question.274 Nor should this sur-
prise us: perhaps what happened in ice-age Europe was a telescoping
of events that elsewhere was a more protracted process. Surely the
important point is that over a relatively brief time we see a clear
emergence of advanced cognitive abilities. Nor should we automati-
cally assume that the emergence of sophisticated cultures can occur
only once: evidence for multiple origins for agriculture and domestica-
tion of animals points in exactly the opposite direction. Nevertheless,
the earliest occurrences of such cultures are highly sporadic,275 and
there does seem to be a real shift in the mental landscape when, about
55 000 years ago, the first compelling evidence for symbolic compo-
sition is found.276 Here, too, there are earlier harbingers such as a
fragment of bone 70 000 years old from the Blombos cave in South
Africa, with a set of largely parallel cut marks,277 and even more spec-
tacularly, and from the same site, rare pieces of engraved ochre.278
Here arguably is the key benchmark, in what was to lead to a cul-
tural effloresence made famous by the cave paintings, figurines, and
other sculptures that were in production from about 35 000 years ago.
If humans were inevitable from the Cambrian period, a visit to the
Moon was on the cards when the Palaeolithic painters surveyed the
bare cave walls of Les Chauvet.
So here, at long last, we have found an example of evolution-
ary uniqueness. Other species of hominid might warm their bodies
beside the fire, hunt with spears, and care for their sick, but were
they really capable of abstract thought and symbolic representations?
Who are those figures in shadows, tracking our own history? Step
forth the Neanderthals, a much-researched and on occasion much-
abused group.279 Either way their study is accompanied by continuous
controversy. Nevertheless, the majority opinion is that they repre-
sent a separate species, Homo neanderthalensis, a view established
on differences in skeletal structure280 and development281 and more
recently on the basis of the recovery of ancient DNA.282 Their mas-
sive and powerful construction, and details such as huge noses, are
evidently adaptations to living in hostile, near-tundra conditions, of
dealing with bouts of intense cold, and an erratic food supply that
led to the adoption of hunting patterns rather different from those
of Homo sapiens.283 Our and Neanderthal brain sizes are equivalent,
and it is beyond all reasonable doubt that they controlled fire and
employed the red pigment ochre, possibly for body decoration. Un-
equivocal evidence for cannibalism also exists,284 but the discovery
of skeletons with severe impairments indicates that the infirm and
crippled were at least sometimes tended to. Famous examples come
from the Shanidar caves in Iraq.285 Of the series of Neanderthal skele-
tons, a majority shows evidence of various traumatic injuries. One of
the individuals (Shanidar I) had particularly dramatic injuries, perhaps
caused by a rock fall, that included cranial damage, a fractured foot,
and an arm that had withered. This individual lived, for a Neanderthal,
to a relatively advanced age, and the extent of his disabilities indicates
a society capable, for whatever reason, of care and compassion.286 Even
so, Neanderthal life was evidently robust and demanding. A general
survey of Neanderthal trauma287 confirms the prevalence of injuries,
as well as the very high incidence of neck and head damage. Thomas
Berger and Erik Trinkhaus remind us that although the old and in-
firm might receive care, the general ‘dearth of older Neanderthals’
suggests that ‘these hominids did not sacrifice the survival of the so-
cial group as a whole when it was threatened by an immobile indi-
vidual’. Despite the likely employment of spears in hunting, these
may have been used principally for thrusting rather than throwing.
As Berger and Trinkhaus dryly note, ‘Given the tendency of ungulates
to react strongly to being impaled, the frequency of head and neck, as
well as upper limb, injuries seen in the Neanderthals should not be
surprising.’288 And if the Neanderthal escaped the charging ungulate,
then there was also the risk of attempted murder.289 What is espe-
cially intriguing is the evidence for deliberate burial (Fig. 9.6). This too
figure 9.6 The Kebara Neanderthal burial. Inhumation was probably

deliberate and may have involved decapitation and removal of the lower
jaw. (Reproduced from the figure on p. 229 of B. Arensburg et al. (1985),
Une sépulture néadertalienne dans la Grotte de Kébara (Israel). Comptes
Rendu des séances de l’Academie des Sciences, Paris, ser. 2, vol. 300,
pp. 227–30) with permission of the Academie des Sciences, Paris.)
has attracted scepticism; one paper is entitled Grave shortcomings.290

Certainly the evidence for the laying of flowers with the corpse seems
circumstantial,291 nor is there convincing evidence for grave goods, at
least in the eyes of most investigators.292 Nevertheless, it seems al-
most certain that at least in certain circumstances the Neanderthals
took care of their dead;293 perhaps for hygiene, perhaps for some other
reason?
And yet the consensus is that a vital spark was missing. It seems
inconceivable that Neanderthals with their use of fire and burial of the
dead did not have some kind of language; who knows, perhaps they
sang? However, the only tangible evidence, which depends specifi-
cally on the position of a small bone (the hyoid) in the throat, is only
indicative of some sort of language ability.294 Differences between the
vocal tracts of Neanderthals and modern humans do, however, exist,
and while there is no doubt that the Neanderthals had a facility for
language, there is still a fierce debate about whether their fluency ap-
proached that of modern humans.295 The general consensus, however,
seems to be that to a first approximation Neanderthals and humans
were equally fluent. Richard Kay296 and his colleagues, for example,
use the size of the hypoglossal canal, which supplies the nerves to
the tongue, and which is effectively the same size, to argue that vo-
cal capabilities of Neanderthal and human were equivalent. A similar
conclusion297 is based on the evidence for, and origins of, thoracic
nervous innervation and its links to sound control. Language is also
consistent with the production of sophisticated stone tools, but one
has to note that these tools are functional and conservative. To be
subjective, there is little sense of beauty in their form.298 In fact, the
so-called Mousterian tool-kits, made both by Neanderthals and by hu-
mans, are in a European context similar in range of types and skills
in execution, until about 50 000 years ago when we begin to diverge
in technological complexity and usher in the next cultural stage, the
Aurignacian.299
In contrast, the Neanderthals seem to have plodded on. Judged
from about 40 000 years ago, but with foreknowledge of their coming
doom, one might have predicted the last Neanderthal to have gone to
the grave300 still clutching in his or her massive hand a Mousterian
tool. This, however, did not happen, because about 35 000 years ago
there was a rather extraordinary cultural effloresence. Exemplified
by what is referred to as the Châtelperronian,301 it is marked by
figure 9.7 Châtelperronian culture, in the form of perforated or incised

teeth and bones, presumably to be worn as ornaments. (Reproduced
from fig. 13 of Zilhâo and d’Errico (1999; citation is in note 308), with
permission of the authors and Kluwer Academic/Plenum Publishers.)
the Neanderthals not only making more advanced tools but, even
more remarkably, by shaping bones and teeth to form artefacts. No-
table finds from the excavations at Arcy-sur-Cure302 include necklace
pieces (Fig. 9.7) composed of canine teeth and fossils. This break-
through, which for obscure reasons did not spread to Iberia,303 took
place after tens of thousands of years of seeming cultural stagnation,
and has been interpreted as largely imitative on the Aurignacian cul-
tures of the incoming H. sapiens, if not acquired by trading or even
scavenging abandoned sites.304 A useful analogy, suggested by Paul
Mellars, is with the cargo cults of such islands as New Guinea and the
New Hebrides.305 On some of the islands, such as Tanna in the New
Hebrides where the John Frum cult still exists, the indigenes con-
struct elaborate cult centres which include model aeroplanes, upon
which it is hoped the desired objects will arrive.306 If these artefacts
had been found in an archaeological context, it would not be sensible,
as Mellars reminds us, to conclude that the islanders had any practical
knowledge of aerodynamics. So too, perhaps, with the Neanderthals:

‘Very good, Arthur; not bad at all, now if we can just hold the flint a
bit higher . . . well, never mind, I expect we can use it for something;
now if you would like to bring a flint from that pile over there, we’ll
continue the lesson . . .’307
A radically different interpretation, however, is now being put
forward that denies that the Châtelperronian culture is imitative. In-
stead the evidence is used to indicate that this breakthrough, where a
necklace had meaning, was independent from H. sapiens and thereby
convergent.308 To say that this proposal is controversial is putting
it mildly: battle-lines are being drawn.309 The disagreements revolve
around the dating, which depends on radiocarbon technology, and the
stratigraphic order of the various cave deposits. The cave stratigra-
phy is notorious for the complexity of the infill histories and for
the problems of subsequent disturbance, by hominids, animals, or
even careless excavators. Yet the evidence is intriguing. In one key
locality, the famous Grotte du Renne at Arcy-sur-Cure in Yonne, cen-
tral France, exceptional Châtelperronian skeletal remains and arte-
facts have been excavated. Interestingly, despite the finesse of the
artefacts, the skeletal morphology of these Neanderthals remained
highly robust, although there are some hints of minor changes in the
loading patterns of the forearm and shoulder.310 In any event this
Neanderthal occupation patently underlies and is separate from the
subsequent Auriginacian influx that marks the arrival of H. sapiens.
These Châtelperronian tools were not imported from some trading
post, but were obviously manufactured on site.
In addition, the Châtelperronian itself was not completely ho-
mogenous. Some argue that it is divisible into a series of separate
and advanced cultures (with names like Szeletian (a central and east
European culture) and Uluzzian (in Italy), all doing approximately
the same thing at the same time. Interestingly, Neanderthal culture
has its own characteristics,311 and among the material recovered is
a bear penis-bone (baculum) with carved circular markings.312 This
find, from Croatia, is probably from an upper level of Neanderthal
occupation, but it is possibly derived from a deeper horizon, dated
at c. 42 000 years bp. This is not to say, incidentally, that cultural
contacts did not exist between some Neanderthals and modern hu-
mans, although the evidence for this is conjectural (see note 312). Cer-
tainly the idea that the two species (if they were genuinely separate)
may have interbred is intensely controversial.313 But if they did at least

have cultural contacts, it is also worth remembering that even if the
Châtelperronian cultures arose independently (which seems likely)
then if there was any contact it would almost certainly be a two-
way process, and Homo sapiens would also have picked up useful
tips.314 Concerning this ongoing controversy Steven Churchill and
Fred Smith (note 299) take a measured view, and although noting
that ‘the last Neanderthals seem to have held to a Mousterian way
of life to the bitter end [see note 303] . . . [others] appear to have been
full participants in the evolving Upper Paleolithic . . . [and] suggest
that Neanderthals had cultural capacities on a par with those of early
modern humans.’315
Translating the stratigraphy of the caves and their contained
artefacts into the daily life of a band of humans or Neanderthals,
each with the daily pressures of avoiding hunger and charging ungu-
lates, surviving illness, and on occasion encountering other groups, is
bound to be contentious. Tools were made, but are now silent. When
does innovation and novelty stop being imitation? This problem is
not confined to the origin of the Châtelperronian cultures. For ex-
ample, it has been suggested that similarities in the tools made in
the Upper Palaeolithic of Europe, and specifically the Solutrean cul-
ture (c. 20 000 years ago, effectively the people who painted Lascaux),
and the somewhat younger North American Clovis culture can be ex-
plained only by transAtlantic crossings thousands of years in advance
of either the Vikings or the Spanish. The majority of researchers are,
however, unconvinced, and interpret such resemblances as there are
between Solutrean and Clovis are simply convergent.316
Still, it is sensible to keep an open mind. What, for example,
are we to make of a terracotta head, apparently about 2000 years old,
that if displayed in a museum would be put in the Roman section?
The only problem is that the head is from Mexico, and was found
in a sequence that is unequivocally pre-Columbian.317 A Roman ship
blown off-course from Spain, crossing the Atlantic?318 Let us further
imagine that two of the passengers were a friend of Paul the Apostle
and a relative of Pliny the Younger. The ship moves slowly across the
Caribbean: one of them looking at the New World with radiant hope,
while the pagan sinks into gloomy resignation.319
To conclude: it is emphatically not my purpose to argue
that agricultural ants, New Zealand kiwis, playful dolphins, or
tool-wielding capuchin monkeys should have inherited the world as

the most sentient species. Rather it is the simple observation that
the ubiquity of convergence will lead inevitably to the emergence of
recurrent biological properties that define the fabric of the biosphere.
Rerun the tape of life as often as you like, and the end result will be
much the same. On Earth it happens to be humans, just as the au-
thor of this book happens to be an academic in Cambridge. So what?
Self-evidently we humans are now utterly different. We have new con-
cerns, new priorities and questions, and, most important of all, new
possibilities.
10 Evolution bound: the ubiquity
of convergence
It was a cold spring day, and the little boy had been out for too long,
playing in the pond. The other members of the family were not pay-
ing much attention, and it was only with a start that the father first
realized that the boy was no longer to be seen; then with a surge of
anguished alarm he saw that not far from where the boy had been wad-
ing something was now floating. All attempts at resuscitation were
futile. With great sadness they left the ambulance as the metal doors
closed. Swiftly they made their way home, already late to prepare for
the Sabbath. Young Adolf was dead.
To judge from the recurrent, if not increasing interest in counter-
factual histories, be it the untimely demise of either tyrant or genius,
the what-ifs of history seem to loom ever larger as the tapestry of
events weaves folly compounded with stupidity, the witless leading
the insane, to the soft applause of the flaccid and the time-server: ‘All
loyal, my lord’. So many missed opportunities, such corruption and
malice, laced with vaunting pride. All seems to be dependent on the
twists and turns of fate, epitomized by the loose nail of the horse’s
shoe by which a kingdom is lost.
So, too, it is now widely thought that the history of life is little
more than a contingent muddle punctuated by disastrous mass ex-
tinctions that in spelling the doom of one group so open the doors of
opportunity to some other mob of lucky-chancers. The innumerable
accidents of history and the endless concatenation of whirling cir-
cumstances make any attempt to find a pattern to the evolutionary
process a ludicrous exercise. Rerun the tape of the history of life, as
S. J. Gould would have us believe, and the end result will be an ut-
terly different biosphere. Most notably there will be nothing remotely
like a human, so reinforcing the notion that any other biosphere,
across the galaxy and beyond, must be as different as any other: per-
haps things slithering across crepuscular mudflats, but certainly never
the prospect of music, nor sounds of laughter. Yet, what we know of
evolution suggests the exact reverse: convergence1 is ubiquitous and
the constraints of life make the emergence of the various biological
284 evolution bound: the ubiquity of convergence
properties very probable, if not inevitable. Arguments that the equiv-

alent of Homo sapiens cannot appear on some distant planet miss
the point: what is at issue is not the precise pathway by which we
evolved, but the various and successive likelihoods of the evolution-
ary steps that culminated in our humanness. To remind ourselves of
what Robert Bieri noted: ‘If we ever succeed in communicating with
conceptualizing beings in outer space, they won’t be spheres, pyra-
mids, cubes, or pancakes. In all probability they will look an awful lot
like us.’2
ubiquitous convergence
Before moving to some examples specifically in the areas of behaviour
and molecular biology, where convergence a priori would be surpris-
ing, let me try to show you how ubiquitous is the phenomenon of con-
vergence. Science fiction is replete with examples of the ‘insectoid’,
vaguely modelled on the apparently robotic scrabblings of a terres-
trial counterpart. Evolutionary orthodoxy, of course, is that such a
creature is a contingent accident, assembled by chance histories and
circumstances. Insects are interesting; insects are monophyletic; but
in the final analysis that is all there is to say. If, however, we consider
‘insectoids’ as a biological property then perhaps something more gen-
eral emerges. So what is the design specification? Among the defining
features of the insects are the following: an articulated exoskeleton
arising from the process of arthropodization; compound eyes; a hexa-
pod gait whereby three of the six walking legs are always on the ground
and thereby define a triangle (two legs on one side, one leg on the other)
that keeps the animal stable; respiratory tubes known as tracheae that
serve to bring oxygen into the interior animal via special openings
(spiracles) on the side of the body; and, to complete this list of evolu-
tionary peculiarities, the development of complex eusocial colonies,
as in the honey bees. All pretty strange, all one-offs in the great lot-
tery of life? On the contrary, all are convergent. Arthropodization may
have evolved as many as four times independently;3 the convergent
evolution of compound eyes was addressed in Chapter 7; a hexapod
gait is inferred in eurypterids (an extinct group related to the scor-
pions) such as Hibbertopterus that moved out of water on to land;4
tracheae evolved independently at least four times5 (as has the method
of gas exchange6 ); and a eusocial organization has not only evolved in-
dependently in many groups of insects (ants, bees, termites, wasps),
ubiquitous convergence 285
but is also found in the shrimps and various mammals, most famously
the naked mole-rats (Chapter 8).
Yet despite the textbook examples of convergence, such as the
camera-eye of vertebrates and cephalopod molluscs, along with the
seemingly arcane, for example, birds that can echolocate (p. 181) or
reptiles with a dentition similar to that of mammals (p. 227), my the-
sis is that both the extent and the importance of convergence have
been consistently underestimated. By this I do not mean so much
the innumerable examples. These are mostly known only to the spe-
cialists, and even among them the convergences tend to be treated
as simple curiosities, with the customary adjectives of ‘remarkable’
and ‘surprising’ (see p. 128). Rather it is my suggestion that by un-
derstanding convergence we can constrain what sort of biosphere we
might expect, especially in an extraterrestrial context. To be sure, for
the most part any discussion concerning convergence tends to revolve
around the immediate realities of the physical world. In this sense the
repeated evolution of features as distinct as stabbing canines (p. 130)
or electroreception in fish (p. 182) are, on a closer analysis, hardly sur-
prising. There are, however, other more complex convergences, such
as the cognitive functions of dolphins (p. 258). Nor in this context
should we be surprised to see the repeated re-emergence of advanced
and sophisticated social organizations. The comparison between the
elephant and the sperm-whale (p. 250) is a fine example, as are the
similarities between the fission–fusion societies of chimp and dol-
phin (p. 249). Because of the general emphasis on physical constraints
in evolution, these examples of societal convergence surely deserve
to be more widely known.
Indeed, I rather suspect that the role of convergence in the be-
havioural realm will produce quite a few more surprises. Earlier I
referred to the way in which electric fish and moths, generating re-
spectively electrical and sound signals, show a convergence in the way
they confuse potential predators. At present, however, the number of
examples of convergences in the behavioural repertoire is rather lim-
ited. Indeed, at first sight such convergences would seem to be rather
unexpected given the complexity and range of responses. Neverthe-
less the examples now available, including the courtship behaviour
of houseflies,7 lacewings,8 crickets,9 and bowerbirds,10 give some in-
dication of the many other examples of behavioural convergence that
probably await recognition. The case of the bowerbirds is, of course, of
particular interest because of the complexity of the decorated bowers

constructed by some species.11 Mike Hansell (note 11) also discusses
whether the bowerbirds have any sense of beauty, a view made pop-
ular by Jared Diamond.12 Hansell is clearly sympathetic to this idea,
alluding to evidence of male apprenticeships and the possibility of
cultural transmission. So it is that once again the birds, this time in
the context of complex behaviours, hint at biological universals.
In this context the famous displays by the fiddler crabs are also
relevant, both because there is indeed evidence for convergent acqui-
sition of complex behaviour, and because of the likelihood that these
are linked to the transition towards a more terrestrial life. In entering
this new adaptive zone, key features, such as rapid locomotion and
excellent vision, are linked to new behavioural capabilities.13 Nor
should this be a surprise. What we already know of convergence in
terms of cognitive sophistication (Chapter 9, p. 258) and sensory per-
ceptions (Chapter 7) indicates that behavioural similarities are al-
most inevitable. A nice example of this is in acridid grasshoppers,
whose song is usually produced the rubbing of the hind-leg. In one
species, however, the same song is produced by rubbing together the
mandibles.14 Even so, the examples of behavioural convergence tend
to elicit the customary cries of surprise. How can we blame the inves-
tigators? Until one appreciates the background these convergences do
seem to be remarkable.
It is, however, important not to adopt too uncritical a view
of such convergences. Identifying examples of behavioural conver-
gence, or for that matter any other type of convergence, presupposes,
of course, a reliable phylogeny. A recurrent feature of phylogenetic
reconstruction is that one set of data, say morphology, produces a
phylogeny that is at odds with another set, such as behaviour.15 Some-
body has got to be wrong, and in being wrong by implication will have
failed to identify certain convergences, either in the morphology or in
the behaviour. What is needed, of course, is an independent data set
such as molecular sequences that may enable the more reliable phy-
logeny to be identified. To a first approximation this works, but even
molecular databases usually produce contradictory results. Phylo-
genies therefore remain in a constant state of flux, and this means,
of course, that in some cases identification of a given convergence is
provisional. So far as the thesis of this book is concerned, however,
this is not a problem, inasmuch as the important comparisons involve
respiratory convergence 287
distantly related groups whose common ancestor lived long ago in the
geological past, and which very probably did not possess the relevant
character we identify as convergent.
respiratory convergence
Despite the apparent ubiquity of convergence, there are some areas
of evolution where the role of convergence would be expected to be
minimal. Of these molecular biology is surely the prime candidate. It
is not difficult to see why. As was discussed earlier, the combinatorial
‘universe’ of protein space is unimaginably vast, which suggests that
the likelihood of navigating to the same ‘destination’ is exceedingly
remote. In itself this may not be terribly important. The success of
proteins depends on their overall structure and, to a first approxima-
tion, the precise sequence of amino acids may not make that much
difference. To be sure, the structure at particular sites may be exceed-
ingly sensitive to the substitution of one amino acid by another, and as
we have already seen in such examples as trichromatic colour vision
(p. 167), these substitutions also give insights into a particular sort of
molecular convergence. Such substitutions, however, presuppose an
existing protein that is already functional. However, we also discover
that, despite the combinatorial immensity of protein ‘hyperspace’, in
other ways the landscape of possibilities is again somewhat limited.16
This is simply because proteins are constructed from a limited num-
ber of ‘building blocks’. Furthermore, in at least the case of certain
structures such as the so-called α/β barrels17 , there is evidence that
the arrangement may have been arrived at convergently. It is useful,
therefore it evolves. Even so, it must be acknowledged that there is
considerable tension, metaphorically speaking, in the study of pro-
tein evolution in deciding whether a structure has evolved because
of a common ancestry, even when no sequence similarity remains, or
because of general constraints on protein structure and function.18
As such, this suggests that molecular convergence at the level
of proteins need not be unexpected. For example, the general view
is that the haemoglobins found in organisms as different as humans
and cyanobacteria are similar because they all derive from a common
ancestral protein. Yet it is also possible that the haemoglobins found
in the protistan ciliates and cyanobacteria actually have a separate
origin,19 albeit with the complication that the cyanobacteria may have
acquired them from the ciliates by a lateral transfer.20 In the context
of these distinctive haemoglobins the possibility of ‘an independent

evolutionary origin’ remains a real possibility.21 So perhaps those ex-
traterrestrials that inhabit oxygen-rich planets will also employ a re-
markably similar respiratory protein. And if not haemoglobin, what
else? On Earth the number of proteins is rather limited. The next most
important, after haemoglobin, is the copper-bearing protein known as
haemocyanin. This respiratory compound is found in many molluscs
and in the arthropods,22 but the molecular structure of the haemo-
cyanin in either group is very different. Quite clearly haemocyanin is
convergent.23
Before leaving the exhilarating topic of the respiratory proteins
it is well worth drawing attention to yet two other examples of molec-
ular convergence. Of these, the second is to my mind even more re-
markable, because it shows how proteins can be different, but end
up with effectively identical functions. In each case the story con-
cerns the molecule myoglobin.24 It is related to haemoglobin, but is
somewhat simpler. It, too, has its bacterial antecedents,25 but as with
haemoglobin there is a strong possibility that the myoglobin that oc-
curs in the cyanobacteria (and the protistan ciliates) evolved inde-
pendently of that found in mammals (and bacteria).26 In any event,
myoglobin is best known for its role in animals as the molecule used
to store oxygen, especially in the muscles. It is no coincidence that
whale muscle is also conspicuously red in colour and that here the
abundance of myoglobin is connected to the need to store substantial
quantities of oxygen during deep dives. So, too, we see convergently
high concentrations of myoglobin in other animals that experience
oxygen stress, notably the burrowing mammals.27 Even more inter-
estingly there is evidence for a more specific molecular convergence
in the myoglobin of diving and burrowing mammals in terms of an
increased content of the amino acid arginine.28
The second example of myoglobin convergence concerns the
gastropods, or snails. In a number of these animals the buccal mass,
which helps to ingest food, is not only very muscular but is by any
measure exceedingly rich in myoglobin. But it is not the usual myo-
globin. It is a substantially larger protein, and is evidently derived
from an enzyme (indoleamine 2,3-dioxygenase (in short IDO)), which
is involved with the degradation of tryptophan.29 So how did it end
up as a respiratory protein, convergent on true myoglobin? There are
several clues. A key fact is that the enzyme carries iron,30 and as
respiratory convergence 289
might be expected this structural characteristic means that it has an

absorption spectrum very similar to those of other proteins, including
myoglobin.31 But in itself such a structural similarity is not enough.
This is because at the stage where the oxygen is being carried, which
forms part of the catalytic process, this intermediary form is very un-
stable. If the protein is to act in any sort of respiratory mode a key
necessity, therefore, is that it must have been able to mutate that part
of the protein structure close to the iron and so confer a new stability.
Only then can the protein serve as a respiratory protein. And this is
evidently what has happened in the abalone (Sulculus), in which this
convergence was first noted,32 and in a number of other snails,33 all
of which belong to the group known as the archaeogastropods. Other
members of this group, however, possess normal myoglobin.
How and in what circumstances, one wonders, did the ances-
tor of these particular snails come to lose its ability to synthesize
myoglobin? A clue of some sort might come from animals rather far
removed from the snails. In the heart of a bullfrog there is no myo-
globin, but an evidently altered and very small haemoglobin has taken
over this essential role.34 But what counts as essential? Rather oddly,
in genetic experiments mice can also be deliberately deprived of myo-
globlin, but seemingly without obvious deleterious effects.35 Such was
evidently not the case for the snails mentioned above. Presumably
once the original myoglobin was lost (for whatever reason), the alter-
native copper-bearing haemocyanin (the usual molluscan respiratory
pigment) was not available as a substitute, and some other protein
which happened to contain iron had to be recruited. That presumably
had to be a rapid process, and in any event once again convergence
emerges.
So, too, with the arthropod silks. As noted in Chapter 6, they are
blatantly convergent. There is, however, a deeper convergence in the
type of molecular structure that characterizes these silk proteins. As
John Gatesy and his colleagues36 point out, there are a variety of high-
strength proteins. These include not only the arthropod silks, but also
certain proteins found in bivalve molluscs (such as the edible mussel
and oyster), where they have such structural roles as conferring
toughness to the shell or the byssus threads that in some instances
anchor the animal to the seafloor. All share the same motifs, suggest-
ing that similar proteins on Thega IX will be less alien than might
be expected. These examples, if anything, exemplify convergence:
whether the journey is by bicycle or airship is less important than the

common destination. One suspects, for example, that if elephants and
sperm-whales could somehow communicate about their strikingly
convergent social systems (p. 250) it would lead to an animated con-
versation about how much they had in common. And in an analogous
way molecular convergences might also point to deeper patterns in
evolution. It is to this controversial topic that we now turn.
freezing convergence, photosynthetic

convergence
Animals faced with the threat of freezing can meet this challenge in
various ways. A common strategy is to alter the osmotic balance of
the internal fluids to produce a natural equivalent of an antifreeze:
adding glycerol or glucose to the blood or equivalent is a standard
biological remedy.37 Another route is to synthesize particular pro-
teins that act to prevent the formation of ice crystals, the growth of
which would rupture the cells and so lead to death. A variety of pro-
teins are able to confer protection against ice growth, but there is
at least one striking example of convergence. It involves two groups
of fish, the Arctic cod and (in freezing waters around Antarctica) a
group known as the notothenioids.38 These groups are not closely re-
lated, but both produce effectively identical protein antifreezes that
are based on a sequence of tri-peptide repeats. Several lines of evi-
dence show, however, that the genes responsible for the production
of this protein are completely different.39 In the notothenioid fish the
gene40 is evidently closely connected to trypsin production41 in the
pancreas (so representing co-option), whereas although not yet specif-
ically identified the equivalent antifreeze gene in the Arctic cod is
evidently entirely unrelated to any pancreatic trypsinogens. It will be
interesting to learn the derivation of this gene in the Arctic cod, and to
see whether convergences exist among some of the other antifreeze
proteins. Thus fish have developed a variety of antifreeze proteins,
but their distribution makes little phylogenetic sense and probably
reflects an ad hoc recruitment by the various groups in the face of
global cooling.42 So, too, proteins have been recruited by plants and
insects to prevent ice formation, and in the insects, at least, there is
evidence for convergence.43 Perhaps less surprisingly, given the na-
ture of the convergence, at least some of these antifreeze proteins
have been under positive Darwinian selection,44 indicating that if ice
freezing convergence, photosynthetic convergence 291
is universal so, too, will be the adaptations to combat its growth in

tissues.
The reason for the polar fish having such proteins is, as noted, to
combat the problem of inhabiting near-freezing waters, especially ad-
jacent to the Antarctica ice sheets. Despite their relative familiarity,
if only from films showing the antics of penguins or the cavorting of
polar bears, from a geological perspective such icy waters (and there-
fore polar ice caps) are rather unusual. For millions upon millions of
years the weather forecast for most of the planet has been ‘Dry and
sunny, with afternoons in the low 40s, with thunderstorms after sun-
set’. Periodically, however, the planet refrigerates. Then once again
the ice sheets smother the poles and so construct huge ice caps. That
the Earth now possesses approximately 30 million km3 of ice is be-
cause of a long period of global cooling. This began about 20 Ma ago,
with the initiation in the Southern Hemisphere of the first major ice
caps. As we saw earlier (p. 248), in the Southern Ocean this cooling
event may have been a trigger for the diversification of the ocean-
going whales, one consequence of which was the rise of dolphin brain
power. Despite the growth of the Antarctic ice sheets, the world only
entered the fully fledged ice age about 2 Ma ago, when the northern
continents and polar ocean became respectively glaciated and covered
with sea ice.
The net result is our present climatic predicament, because al-
though we live in a warm interval there is no reason to think that the
ice age is over. What we regard as ‘normal’ is but an interlude of mod-
erate climatic stability (if not safety) in a longer-term pattern in which
the global climate not only oscillates wildly but may plunge from one
state to another in only a few decades. These oscillations are of widely
varying duration – think for example of the so-called ‘little ice ages’ –
but are most obviously expressed in times of major glacial advance
alternating with warmer interglacial periods. Self-evidently, we live
in one such climatic respite. Nor is the present day the warmest inter-
val. About 400 000 years ago hippopotamuses wallowed a few miles
from my office in Cambridge, and were no doubt the occasional ob-
ject of scrutiny by the equivalent of Boxgrove Man,45 members of
Homo erectus (or a similar species). By 40 000 years ago, however, my
view would have been replaced by bleak tundra, occasionally traversed
by hunting parties of Neanderthals tracking the spoor of the woolly
mammoth.
Knowing that where I work was once home to the near-naked

Boxgrove Man or fur-covered Neanderthal depends, of course, on many
lines of evidence. Records of ancient pollen and the study of such an-
imals as the beetles are particularly important, as are the much more
occasional stone tools left by the hominids. But some of the clearest
evidence for these climatic fluctuations comes from the drilling of
cores through the ice caps. In this way it is possible to read the his-
tory of glaciation to a remarkable degree of accuracy. As the terms
glacial and interglacial imply, the glaciers sometimes advanced and
then subsequently retreated as the world swung into the next warmer
interval; but in the polar regions, and especially on the great ice caps
of Antarctica and Greenland, the snow continued to fall, and it is here
that the critical climatic evidence is encoded in the ice caps.46 In these
polar regions, as the snow turned into ice, air bubbles were trapped and
retained tiny samples of the atmosphere at that time. Not surprisingly,
during any given interval the proportions of the major gases, such as
nitrogen and oxygen, remain constant. There are, however, other so-
called greenhouse gases, notably carbon dioxide and methane, which
show major fluctuations in abundance through the thickness of ac-
cumulated ice. Specifically, as the amounts of these gases increase so
the planet warms and the ice begins to melt. Although much talk of
global warming revolves around carbon dioxide, methane is probably
also of critical importance. At present huge quantities are locked up in
ice and similar hydrates, especially in the tundra and within deep-sea
sediments. Once released, the methane is quite rapidly oxidized. This
means that it has a short residence time in the atmosphere. However,
the substantial release that is expected as a result of already rising
global temperatures will in its turn accelerate the process of global
warming. It is interesting that the geological record has compelling
evidence for the release of massive volumes of methane resulting from
the destabilization of buried hydrates, notably during the so-called late
Paleocene thermal maximum (LPTM).47 So the world is no stranger
to an event of this sort, which is, of course, exactly what we are pro-
ducing today by burning fossil fuels.
Present-day global warming, at least in part the result of the
addition of massive amounts of carbon dioxide to the atmosphere, is
likely to have perilous consequences. However, from the point of view
of the last ten million years or so of geological history, the carbon diox-
ide being released from belching chimneys and fuming car exhausts is
freezing convergence, photosynthetic convergence 293
actually reversing a long-term decline in its atmospheric abundance.

That decline, although natural, is also potentially perilous, at least
for the plants. The reason is rather straightforward. Photosynthesis,
of course, requires carbon dioxide and the fact is that for the plants the
pre-industrial levels of this gas in the atmosphere had reached danger-
ously low levels. This predicament is made the more serious because,
as we have already seen, the process of photosynthesis depends on the
RuBisCO enzyme, which itself shows serious deficiencies, not least
its inefficiency in the (inevitable) presence of oxygen (p. 108).
These difficulties are effectively unavoidable, but to escape the
predicament posed by declining levels of atmospheric carbon dioxide
a number of plants have found an escape route. They have evolved
a new photosynthetic pathway. Here it is necessary to explain that
the usual method is called C3 photosynthesis, because the first com-
pound (a substance known as phosphoglycerate) to be formed in the
pathway contains three carbon atoms. In contrast, the new pathway,
which arises as a response to desperately low concentrations of carbon
dioxide, first produces oxaloacetate, which has four carbon atoms,
hence C4 .48 A significant group of C4 plants are the grasses, and the
development of grassland and savannah49 had a profound effect on
mammalian evolution, notably for such herbivores as the horses, and,
many have argued, ultimately for the hominids. One should also in
passing note that declining levels of CO2 may not be the whole story
in explaining the rise of C4 photosynthesis; aridity and salinity also
appear to be significant factors.50
The C4 photosynthetic pathway is biochemically very complex,
so complex in fact that it might be thought to have evolved in only
one group of plants which fortuitously stumbled on this solution. Not
a bit of it: C4 photosynthesis is rampantly convergent and is believed
to have arisen independently at least 31 times.51 In essence the C4
process aims to protect, so far as possible, the RuBisCO activity, and
within the plant to increase the amount of carbon dioxide. For the
most part this convergence depends on three critical factors. These are
a close proximity between types of leaf tissue involved with photo-
synthesis and the initial transport of the sugars (respectively the mes-
ophyll and bundle-sheath cells, forming the so-called Kranz anatomy);
up-regulation of PEP carboxylase; and down-regulation of RuBisCO in
the mesophyll. The enzyme PEP carboxylase plays a key role because
it helps in the release of the vital carbon dioxide. Yet, as is so often the
case, this protein has been co-opted from an earlier role; in this partic-
ular instance from functions that include control of the stomata and
assimilation of nitrogen. Until recently it was also thought that an
essential aspect of C4 photosynthesis was the possession of a Kranz
anatomy, but so far as a C4 chenopodiacean plant (Borszczowia) is con-
cerned its distinctive leaf anatomy does not depend on the dual-cell
arrangement seen in classical Kranz anatomy.52
There are some interesting nuances in this story. First, although
described as C4 photosynthesis, the pathways in the respective plants
show various differences and degrees of completion in the transition
from the C3 to C4 photosynthesis.53 As is effectively inevitable in
evolutionary convergence, there are multiple pathways to the same
destination. Even so, there are a number of peculiarities in this evo-
lutionary process. One of the most obvious is that C4 photosynthesis
is not generally known among the trees, although some shrubs with
secondary woody growth are C4 photosynthesizers, and in Hawaii a
rare tree (Chamaesyce forbesii) also has this type of photosynthesis.54
Despite this restriction in land plants, what appears to be an indepen-
dent acquisition of the C4 pathway is known in some algae, specifi-
cally the diatoms.55 Second, although the story of the rise of C4 plants
focuses on events beginning about ten million years ago (see note 49),
there may be a deeper history. There are at least hints of a precipitous
drop in carbon dioxide about 90 Ma ago and a corresponding appear-
ance of C4 photosynthesis.56 More tenuous is the possibility that a
much earlier drop in atmospheric CO2 , linked to the late Palaeozoic
ice ages (late Carboniferous – early Permian, c. 280 Ma ago) also saw
an independent development of the C4 pathways.57
Once again evolutionary inevitabilities emerge. If George
Wald is correct and chlorophyll is universal (Chapter 6), so, too, is
photosynthesis. On planets where the levels of atmospheric carbon
dioxide happen to decline, and given that only on tectonically active
planets where water, plate tectonics, and mountain-building interact
will atmospheric carbon dioxide fluctuate, the C4 photosynthetic
pathway will inevitably follow. Without C4 plants, humans would
face a rather different world. Some of the most important groups of
C4 photosynthesizers are the grasses, including wheat and rice and
(originating in the New World) maize. Even within the grasses it
appears that the C4 process was arrived at independently at least four
times.58
the molecules converge 295
Now, of course, humans face the prospect of yet another world,

but this time of their own making. As carbon dioxide from the burn-
ing of fossil fuels continues to pour into the atmosphere, in principle
plants will benefit, but this is little cause for comfort because in reality
the effects are highly variable. Plants need more than carbon dioxide,
and other factors such as water stress and increasing temperatures
can work in the opposite direction and so largely negate the apparent
advantage of boosting atmospheric carbon dioxide. Not only that, but
increasing levels of carbon dioxide, not to mention the accelerating
release of the yet-more-dangerous greenhouse gas methane (note 47),
will lead to many other climatic effects that, from a human perspec-
tive, will probably be distinctly unwelcome. In parentheses one might
mention that much could be done to alleviate global warming today,
but dithering, self-interest, and, most importantly, corporate hostility
are – well – discouraging. Perhaps my children’s generation will be
entitled to ask why we did nothing?
the molecules converge

The convergence of antifreeze proteins and the multiple paths to a C4
photosynthesis can therefore be added to the other examples of molec-
ular convergence mentioned earlier, such as the ‘five-site rule’ for ver-
tebrate colour vision (Chapter 7) and probably also rhodopsin itself.
As already noted, the combinatorial vastness of protein ‘space’ would,
a priori, suggest that examples of molecular convergence would be
very rare indeed: there are, after all, so many alternatives. As it hap-
pens, however, there is a growing list of such examples. These include
the proteases59 and peptidases,60 aminoacyl-tRNA synthetases,61
cytokinases,62 proteins associated with malaria,63 NADH dehydroge-
nase 1,64 lactate dehydrogenases,65 nicotine oxidases,66 the evolution
of polysaccharide lyases,67 light-harvesting proteins,68 proteins asso-
ciated with cartilage (lamprins) and various elastic proteins,69 chitin-
binding proteins,70 HIV-protease,71 antigen receptors in sharks,72 as
well as biochemical processes such as those involved with nucleotide
binding by proteins,73 possibly DM domain factors involved with sex-
ual determination,74 and steroid signalling.75
Some of these convergences are of more than academic inter-
est, notably the structural mimicry that can arise when microbial
pathogens insinuate themselves into the cellular processes of their
hosts.76 These are particularly important, for several reasons. First, the
structural mimicry adopted by the pathogens has clearly come about

by natural selection. In addition, this mimicry requires a molecular
matching, but such a correspondence need not entail any similarity
to the amino acid sequence of the host protein. This is significant
in the context of navigating through the combinatorial vastness of
protein ‘hyperspace’: common structures will emerge notwithstand-
ing the immensity of alternative sequences. As ever, the biological
solutions are extraordinarily specific. Consider, for example, the re-
marks by Erec Stebbins and Jorge Galán with reference to one example.
This concerns a bacterial pathogen that in its invasion uses one of its
proteins to take over those signal transduction pathways that in the
host cell rely on such proteins as fibronectin.77 This procedure allows
the bacterium to bind to the host cell, which accordingly fails to rec-
ognize that it is under attack. In conclusion, Stebbins and Galán write,
‘The fact that two completely different protein structures . . . present
the same residues [two aspartic acids, one arginine] in the same posi-
tions for binding attests to the power of convergent evolution.’78 Such
examples are probably very much more widespread than is at present
realized, and they emphasize both the sophistication and the subtlety
of these molecular processes. As Stebbins and Galán conclude, in the
context of such a pathogen attack, ‘convergent evolution contributes
significantly to the dynamics of the evolutionary process.’79
That Stebbins and Galán are not wrong is now becoming clear
with other examples of molecular mimicry that go to the very heart
of the structure of life. What, for instance, are we to make of evidence
for molecular mimicry between three different proteins (respectively
known as EF-G (a translocase), eRF1 (human release factor), and RRF
(ribosome recycling factor)) and the RNA which is involved in that
part of the replication process known as translation.80 In other words,
despite having markedly different structures and compositions, these
proteins (made of amino acids) mimic structurally the RNA (made of
nucleotides). The investigator, Yoshikazu Nakamura, also points out
that the ability for such a protein to act in the process of replication,
in place of the usual RNA, might possibly have some bearing on the
origin of life, or at least on how the transfer from the hypothesized
‘RNA world’ (see note 3, Chapter 4) to the universal DNA replication-
protein system was achieved. This is a tempting and possibly believ-
able goal. We may need to remember, however, the cautionary remarks
of Andrew Ellington (note 41, Chapter 4) to the effect that the billions
the molecules converge 297
of years of molecular evolution may irrevocably sunder us from es-

tablishing the first stages in the origination of life, at least through an
examination of modern biochemistry. Moreover, if convergent evolu-
tion is an ‘eternal return’ to the ‘attractors’ of functionality, then we
cannot be surprised if history repeats itself.
One hint that this could be so comes from work on a type of RNA
known as the hammerhead ribozyme, whose simplicity and ability to
show self-cleavage makes it a tempting candidate for inclusion in the
‘RNA world’. Yet, as Kouroh Salehi-Ashtiani and Jack Szostak show,
this ribozyme has evolved several times.81 Their remarks, which
strongly echo those of Kenneth James and Andrew Ellington (note 86,
Chapter 4), are particularly pertinent: ‘Our results show that, despite
the dominance of contingency (historical accident) in some recent
discussions of evolutionary mechanisms [here they cite S. J. Gould’s
Wonderful life], purely chemical constraints (that is, the ability of
only certain sequences to carry out particular functions) can lead to
the repeated evolution of the same macromolecular structures.’82 This
book aims, if nothing else, to refute the notion of the ‘dominance of
contingency’, and these examples of molecular convergence perhaps
explain why some scientific attitudes to the reality of convergence are
perhaps ambiguous. I have already suggested (p. 128) that the associ-
ated adjectival battery of surprise may have less to do with evolution
than an uneasy sense of teleology. Self-evidently, if we cannot be sure
whether a feature is genuinely primitive as against having evolving
several times, then clearly the study of evolution, which presupposes
that organisms are related because they are similar, would be severely
compromised. This is the specific problem of homoplasy, viewed by
most cladists as both profoundly disruptive to their neat schemes and
evolutionarily of marginal interest. But, of course, evolution can be
studied: the routes are many, but the destinations few, and the land-
scapes across which all organisms must travel are adaptive.
Caution, however, is always needed.83 First, and this is a point
that presupposes the identification of any convergence, one needs
a reliable phylogeny. Even here, whether or not the end results are
judged to be convergent may depend on the sort of question asked.
Take, for example, one of the classic cases of molecular convergence,
that of the enzyme lysozyme. This is widely employed, and acts by
breaking down the walls of bacterial cells. It thereby confers protec-
tion, and is found, for example, in exposed surfaces, such as eyes and
the nose, where it serves to protect these exposed surfaces from in-
fection. Lysozyme, however, has another role. This is to assist with
the digestion of the tough plant material scooped up by cows and
other ruminants. The basic principle is to employ bacteria in a re-
action chamber, the rumen. The role of the bacteria is to attack the
plant cellulose, and so allow fermentation of the swallowed vegeta-
tion. The resultant slurry is then processed in the stomach, using
the enzyme lysozyme. This arrangement of rumen and stomach with
lysozyme activity, familiar from the cow, has evolved convergently
in one group of primates, the colobine Old World monkeys, e.g. the
langur,84 and also a rather odd bird known as the hoatzin.85 This ani-
mal, which lives in the Amazon forests and eats a diet made up largely
of leaves,86 is probably best known because of its supposed parallels
with the bird–dinosaur Archaeopteryx.87 The hoatzin has shifted its
diet from the softer fruits to tougher leaves, and here too lysozyme
is employed. Molecular convergence between the lysozymes of these
monkeys and the ruminants has significantly also been identified,88
and so, too, with the hoatzin birds.89 But is it really a convergence?
When the overall sequences of the lysozymes in primates and rumi-
nants are compared, the phylogeny is as expected, whereas if the con-
vergence were unrecognized it could be used mistakenly as a sign
of common ancestry and so would ‘artificially’ bring the two groups
together.90 This, however, is not quite the point because it is the spe-
cific similarities within the lysozyme molecule, at five or so amino
acid sites, which excite our interest,91 because of both the molecular
convergence and the implication that observations of this kind have
for molecular adaptation.92 It would be misleading, of course, to sug-
gest that these examples of molecular convergences are particularly
frequent, but examples in some key enzymes suggest that as with
DNA (p. 31) and chlorophyll (p. 110) there may be some effectively
universal constraints.
convergence and evolution

Evolutionary convergence shows that we live in a constrained world,
where all may not be possible. It does not, to be sure, rule out unique
solutions, nor does it guarantee the emergence of the identical. It
shows, however, that at many levels evolution is seeded with probabil-
ities, if not inevitabilities. As yet there is no overall metric that might
establish a scale of evolutionary likelihoods, but, as I have repeatedly
convergence and evolution 299
emphasized, the many biological parallels to those features that define

the emergence of humans suggest that something similar will emerge
elsewhere.
There are, however, several important provisos. Arguably they
should have been introduced at the very beginning of the discussion.
They have, indeed, already received passing mention. Their overall
importance is, however, clearer once the field has been reviewed. The
first, and most obvious, point is that identification of convergence
presupposes a reliable phylogeny. Indeed, assumptions about the relia-
bility of the evidence and decisions on which characters are evolution-
ary relevant introduces a constant risk of circularity in the argument.
Is a particular character the same because it evolved from a common
ancestor, or is it convergent? This point is far from trivial: those who
employ the methods of cladistics are constantly aware of what they
call homoplasy, arising from characters that carry no phylogenetic
information because they arose independently. A related question is
whether a particular character is actually primitive, but by being lost
(or transmuted beyond recognition) in some evolutionary branches
appears to be convergent because of its sporadic distribution across
the tree of life. In many of the examples I have given these alterna-
tives need to be taken seriously, and because of the fluidity of many
phylogenetic schemes the ensuing questions will not necessarily be
easy to answer. It is also worth noting that, whatever the phyloge-
netic scheme adopted, practically never can it escape convergence.
Although this is a headache for those wrestling with local phyloge-
nies – is the character in question the same because it descends from
a common ancestor or because adaptive constraints have led to con-
vergence? – for the most part the examples discussed in this book
are from distantly related groups where the evidence for separate and
independent originations is proportionally more secure.
A more significant point concerns what has been termed ‘con-
certed convergence’, which as the name implies is when a whole series
of characters change in tandem in response to an adaptive pressure.
Striking examples are available in groups as disparate as fig-breeding
fruit-flies93 and lilies.94 The likelihood that a set of characters will be
linked during evolution has attracted attention because of its phylo-
genetic implications. Clearly, if a number of characters evolve in con-
cert then they cannot be treated as heaps of nuts and bolts subject to
the atomistic scrutiny of cladists. Concerted convergence in complex
systems is probably much more widespread than is presently realized,

and it has two important implications and one qualification. First,
the interlocking inherent in concerted convergence suggests that at
least in some instances the likely outcomes of particular evolutionary
trends may be even more constrained than is at first apparent. Second,
this concept is consistent with the idea of one or more key adaptations
whereby evolution of one feature may effectively promote, if not guar-
antee, a cascade of parallel changes. The qualification is that this does
not mean that concerted convergence can be divorced from its phy-
logenetic context. Rather, as Janice Lord and her colleagues95 have
stressed, the roles of adaptation and phylogeny are complementary.
While this will tend to confer a broad predictability to the evolution-
ary process by what is termed ‘niche conservation’, it emphatically
does not debar evolutionary excursions into new adaptive territories
when the environment (however defined) happens to shift. In this
sense the much-vaunted concern about evolutionary constraints is
effectively bypassed.
The difficulties encountered in encapsulating convergence are
also exemplified by the problem of deciding how many really impor-
tant turning points there are in the history of life and, if they had been
either different or had not occurred, how distinctive would the course
of life have been? This problem may not be intractable, but the prob-
lem of facing an indefinite regress of questions (e.g. is the evolution
of mitochondria essential for intelligence?) and the unresolved diffi-
culties of evolutionary inherency, especially at the molecular levels
(e.g. if rhodopsin, then inevitably eyes?) make it very difficult ever
to identify a starting point. If DNA and proteins are universal, then
this might provide the ultimate bedrock of biological probabilities;
in other words if given molecules can operate only in given ways,
then perhaps this largely predetermines the emergence and nature of
complex structures, such as eyes, brains, and behaviours.
A final point also needs to be addressed. It might be claimed
that convergence is too elusive a concept to have any real validity:
what after all does ‘similar’ really mean in a biological context? One
response is to reject any overarching scheme, and deal with matters on
a case-by-case basis. Much of this book can, of course, be read in just
such a fashion. My overall approach, however, is to see in the recurrent
emergence of biological properties such as intelligence, memory, and
self-recognition – all, to be sure, based on prior evolutionary events but
themselves convergent – a programme that is more interested in the

definition and probability of complex states than the precise history
that led to any particular example. Not that histories are without
interest; very much the reverse, because these trajectories are a proof
of evolution as a process. They do not, however, address directly the
likelihood of the rise or maintenance of complex states in an adaptive
framework. In this sense the study of convergence is a return to a
nomothetic biology.
What then are the implications of convergence? First and fore-
most is that the various examples I have given will provide no comfort
for the ‘creation scientists’, because in their various ways they pro-
vide compelling examples of the reality of organic evolution. Thus
we see that the same ends may be arrived at along various, and some-
time wildly different, routes. Correspondingly, very seldom is the
convergence so exact as to make the organism or structure indis-
tinguishable. Discerning the nuances of difference and the paths fol-
lowed are in themselves highly informative about the evolutionary
process. As I have already emphasized, however, our fascination with
historical contingencies – the Caliph of London?, the Jesuit Centre for
Astronomy and Mathematics in Kyoto?, the coronation of an Aztec
king at Westminster in 1412? – are almost irrelevant in organic evolu-
tion. Convergence shows that we can provide first-order predictions
of the emergence of important biological properties on Earth, and by
inference elsewhere.
The realities of convergence have four implications for evo-
lution, specifically the inescapable need to consider the themes of
adaptation, trends, progress, and (a topic that has been surprisingly
neglected) whether evolution can exhaust, at least locally, its poten-
tial. All should be unremarkable, yet for what seem to be effectively
ideological reasons have been subject to sustained attack, notably
by S. J. Gould. In the case of adaptation, for example, the notion of
spandrels96 has had an entirely disproportionate influence. It would be
strange if organic ‘architecture’ lacked its equivalents to the spandrels
(or pendentives) of a building, but are they of anything more than pass-
ing interest? As has been repeatedly pointed out, a failure to devise an
adequate test for adaptation may tell us more about the shortcomings
of the investigative method than the reality or otherwise of adaptation.
This is why convergence matters, because, as a number of work-
ers have already emphasized, its very ubiquity confirms, as if there
were ever any doubt, the reality of evolutionary adaptation.97 For ex-
ample, writing on the topic of ecomorphological diversification in
freshwater fish, Kirk Winemiller98 remarked, ‘The concept of ecologi-
cal convergence and independent evolution of equivalent ecomorpho-
types deserves special attention because of its implications for general
ecological and evolutionary theory,’ and he continued, ‘the remark-
able qualities exhibited by several convergent pairings makes the phe-
nomenon difficult to dismiss as arising from stochastic processes.’99
Rob Foley100 is more trenchant: ‘Convergence is perhaps the strongest
evidence for adaptation, and indicates that despite the contingent fac-
tors that lie at the base of evolutionary episodes, the rules of survivor-
ship and reproductive success govern the final outcome.’101 Earlier I
introduced the many convergences in the subterranean mammals, and
it is no accident that Eviator Nevo’s masterly overview102 of this topic
is equally a renewed prolegomenon on behalf of the importance of the
adaptational explanation. Using multiple examples, he proclaims that
‘the adaptationist programme is as vital as ever.’103
Much as I disagree with the world picture of Richard Dawkins,
with its questionable genetic reductionism and etiolated secular
pieties, his explanations and enthusiasm for the reality of adaptation
are of great value.104 This is the way the world is, and in one way the
only interesting question is how different could it be, in terms of both
mechanism and end results? So far as the former is concerned, the
much-vaunted plurality of alternative evolutionary mechanisms ap-
pears, at best, to be of marginal relevance. Consider, for example, the
notion of ‘historical burden’, the constraints of past ‘decisions’ that
guide, restrict, and perhaps even stymie a phylogenetic ‘career’. That
such constraints exist is undeniable, but what is far more interesting
is the way in which organisms repeatedly ‘get round’ these problems,
which is why convergences are ubiquitous. This is neither to suppose
that everything is possible, nor to deny the reality of contingent hap-
penstance in biological history.105 Even to acknowledge the realities of
convergence is not to imagine that every organism is ‘trying’ to evolve
into a human. The Earth’s biosphere is patently a product of diver-
gences, but as this book emphasizes there is an underlying structure
that imposes limits and delineates probabilities of outcomes. Some
evolutionary trajectories point to increasing complexity, but simpli-
fication can be just as prevalent. Thus, there are a number of excel-
lent examples from among the animals of what appears to our eyes
as regressive evolution. These include the flatworms, whose relative

simplicity was until recently thought to represent a rather primitive
arrangement from which more advanced phyla emerged.106 More ex-
treme reduction is seen in the curious dicyemids, parasitic occupants
of the kidneys of cephalopods,107 and what is effectively a regression
to a protistan state has occurred in the myxozoans.108
The organic world is a plenitude and a marvel, but it still has
a rational structure. Simplification will arise in its own adaptive cir-
cumstances, but so, too, will complexity. And of course there are dif-
ferent solutions to the same problems, as we have seen in the case of
the eye. When we learn, however, how this optical convergence ex-
tends to the nose of the star-nosed mole, or that the olfactory systems
of insects and vertebrates are effectively the same, then we get some
inkling as to how wonderfully varied our world is, but how under-
pinned it is by deeper commonalities. It is a truism that the general
acceptance of the Darwinian explanation has not lessened the fear of
many that our animal origins erode, if not deny, our human unique-
ness. Yet, to anticipate the next chapter, there is another view that is
more optimistic. In this view of life all share the terrestrial Creation,
but we need to acknowledge that not only does our unique knowledge
reveal a transcendence in wholly remarkable ways, but it also enables
us to understand how the emergence of sentience is imprinted in the
evolutionary process. The implications of how we choose to use the
natural world, not least the other sentient species, will be obvious.
That the motor of evolution is adaptation seems incontestable,
yet I also wonder if we appreciate the full range of the adaptive frame-
work, especially with respect to molecular biology. There is, for exam-
ple, the perhaps surprising sensitivity of so many molecules to change,
so that even a single substitution may make the molecule effectively
non-functional.109 Even trivial differences can be disastrous. In a simi-
lar way, proteins with markedly different functions may differ only in
very minor ways. Conversely, there are certain types of molecule that
exist as a wide variety of variants, known as isoforms. Do they each
have more or less the same function, the differences between them
being little else than a rococo embellishment? Seemingly not. In the
case of one protein family a total of 153 isoforms has been detected,
each of which seems to have its own functional specificity.110 The
various examples of molecular convergence discussed above (p. 295)
are also relevant in providing examples of adaptation to function, yet
it is seldom that the investigator or commentator can resist express-

ing surprise. This is not because the molecules lack a function; far
from it. Rather it is the difficulty of determining how a molecule
evolved to its present precise configuration. It is partly for this rea-
son that evidence for selection at the molecular level continues to
attract particular interest.111 The importance of this is, of course, not
lost on those trying to bio-engineer molecules, and it is also of inter-
est because again it allows a rerunning of the evolutionary tape, with
predictable results.
converging trends
This ubiquity of convergence and the likelihood that the great ma-
jority of the examples are almost certainly the result of selective pro-
cesses operating in the context of adaptation has the obvious corollary
that common evolutionary destinations presuppose specific, and re-
stricted, trajectories. To mainstream evolutionary biology the exis-
tence of trends is entirely unremarkable.112 For example, Christine
Janis and John Damuth in a wide-ranging review of mammalian evo-
lution and convergence remark, ‘We regard the widespread occurrence
of detailed ecomorphic convergences to be prima facie evidence for
sustained, adaptive, phyletic trends at lower hierarchical levels,’113
and in their conclusions go on to say, ‘It should be clear at this point
that we feel that the adaptive component has been dominant in his-
torical trends observed in the Mammalia.’114 To be sure, like adap-
tation, an apparent failure to demonstrate a trend may tell us more
about the method employed than about the non-existence of the trend.
Yet trends imply directionality, and perhaps progress (note 112). It is
hardly surprising then that in his decades-long campaign against adap-
tation, S. J. Gould sought to demolish such ideas, arguing in particular
that we would do better to look at the changes in variance as lineages
evolve,115 with the corollary that the trend is no more than a diffusion
away from a restraining boundary.
So are trends with a directional component that might arise
from a persistent pressure, such as natural selection, not only real but
sufficiently common to be important? The ubiquity of convergence
would argue so, but in this context I am more interested in direc-
tionality in evolution per se. Take, for example, what is probably the
best-known evolutionary trend, that of size increase. This is known
as Cope’s rule. Yet, for the most part the reality of this rule has been
converging trends 305
Per cent change maximum size
Per cent change minimum size
figure 10.1 A graphical depiction of evolutionary changes in body size

during geological time. Upper right represents the strict case of Cope’s
Rule, with a persistent trend towards larger size. The opposite case is
shown in the lower left, while increases and decreases in the variance of
body size are shown respectively in the upper left and lower right.
(Redrawn with permission from fig. 10.5 of ‘Body size and
macroevolution’ by D. Jablonski in Evolutionary paleobiology: in honor
of James W. Valentine, edited by D. Jablonski, D. H. Erwin, and J. H.
Lipps (The University of Chicago Press, Chicago and London) c 1996 by
The University of Chicago, and with the permission of the author.)
a vague articulation rather than a specific set of hypotheses: after all,

if it occurs all over the place then what is at issue? David Jablonski,
however, injected rigour into this discussion, drawing upon his ex-
tensive knowledge of fossil marine molluscs from the Cretaceous.116
Concerning changes in size through geological time, there are four
broad possibilities (Fig. 10.1), to which can be added the character of
no change.117 Unsurprisingly, in Jablonski’s analysis all four examples
of body-size change occur, of which overall size increase, i.e. Cope’s
rule, is found in about 30% of instances, while an increase in variance
(which will include size increases) accounts for another 28%. Inter-
estingly, only a small proportion (c. 4%) shows no change, while size
decreases also occur (c. 28%). So far as these molluscs are concerned,
Cope’s rule is not universal – why should it be? – but it is highly
significant.
So, too, in a terrestrial context, when the overall history of
Tertiary mammals is examined there are striking increases in body
sizes, consistent with Cope’s rule.118 As John Alroy remarks in this

instance, ‘The basic pattern is overwhelming . . . Newly appearing
species are on average . . . 9.1% . . . larger than older congeneric species,
a highly significant difference . . . The only clear-cut hypothesis that
predicts such a pattern is the most narrow and deterministic inter-
pretation of Cope’s rule; namely, that there are directional trends
within lineages.’119 And the birds? Here an investigation120 of whether
the evolution of body size was non-random concluded that there are
‘many more instances where a large taxon putatively originated from
a smaller one than vice-versa.’121 These results, concerning the appli-
cability of Cope’s rule to mammals and birds, had already been antici-
pated in various ways. For example, Brian Maurer and his colleagues122
emphasized how the evolution of body size and its tendency to in-
crease arises as a result of the interplay between the processes of
natural selection and differences in the relative rates of speciation
as against extinction. This, in itself, is highly significant because it
suggests that microevolution and macroevolution are not decoupled;
but more importantly in this context Maurer and his colleagues ro-
bustly conclude, ‘We suggest, however, that Gould’s . . . emphasis on
randomness be replaced with an emphasis on deterministic outcomes
that result largely from the role of ecological processes in speciation
and extinction.’123
Evolutionary convergence is, therefore, consistent with what
most biologists find entirely unremarkable: the Darwinian world of
adaptation and trends. What has, in my view surprisingly, received
rather less attention is the extent to which any particular ‘design
space’ becomes saturated. Broadly, given the exploratory nature of
Darwinian mechanisms, this could be expected, and such examples
as that of ‘skeleton space’ (p. 117, Chapter 6) are consistent with a
saturation (or exhaustion) of possibilities. But is this generally true?
In an outstanding survey Peter Wagner124 demonstrates that this does
indeed appear to be a general principle. Effectively, as a cohort of evolv-
ing species (the clade) traverses geological time, so convergences come
to predominate. This obviously has important consequences for the
ability to recognize ancestral forms (unless the fossil record is taken
into account) and the reliability of phylogenies if, indeed, convergence
is rampant.
There is, however, a more general principle because the exhaus-
tion of evolutionary possibilities might be because of the particu-
lar biological constraints of a given form (metaphorically, the clade
converging trends 307
has run out of things to ‘do’), but it might equally reflect ‘persistent
selection favoring similar general forms.’125 Wagner’s study, which
encompasses a wide variety of fossil clades, was concerned with their
specific histories, which led consistently to the exhaustion of poten-
tial: in each case the tape of life was not only rerun, but rerun re-
peatedly. In a wider context, and as has already been emphasized, this
does not negate the self-evident reality of continuing diversification. It
might be objected that this undermines the principle of predictability
as inferred from the ubiquity of convergence. What evolution cannot
do is see into future diversification so far as the envelope of possibil-
ities is concerned, although it can be equally sure that a great deal of
what does one day evolve will have emerged in parallel circumstances,
in other times and places. What we can also say is that whenever the
known edge of the evolutionary envelope is reached, be it in terms
of intelligence or agriculture, then it will be explored independently
several times.
In their own ways the realities of adaptation and evolutionary
trends (and the exhaustion of possibilities) excite little comment in
the mainstream of biology; why should they? Not everything need be
adaptive, nor are random walks by any means impossible, but are they
effectively peripheral to the argument? And what about evolutionary
progress, that term that S. J. Gould126 gently refers to as ‘noxious’.127
Simply because evolution has delivered us to a point where only now
can the word ‘progress’ make any sense, need not mean that it either
has no relevance to the human condition or that it lacks an evolu-
tionary reality. That the bacteria are still with us, and that without
them the planet would soon grind to a halt in the absence of their
recycling abilities, misses the point. Neither is progress a question
of the sheer number of species, nor the supposed number of body
plans. What we do see through geological time is the emergence of
more complex worlds. Nor is this a limiting view. It might be pre-
mature to suppose that even the bacteria of today are some sort of
‘honorary fossils’, unchanged relicts from the Archaean pond-scum.
Nor need we imagine that the appearance of humans is the culmina-
tion of all evolutionary history. Yet, when within the animals we see
the emergence of larger and more complex brains, sophisticated
vocalizations, echolocation, electrical perception, advanced social
systems including eusociality, viviparity, warm-bloodedness, and
agriculture – all of which are convergent – then to me that sounds like
progress.
a possible research programme

Perhaps I am too optimistic, but I suspect that there is already a sea
change under way with respect to the realities of convergence128 in the
context of adaptational explanations as against appeals either intrin-
sic to processes (e.g. phylogenetic constraint) or to random processes.
What may be less appreciated is that the ubiquities of convergence are
clear enough evidence that life has a peculiar propensity to ‘navigate’
to rather precise solutions in response to adaptive challenges. I would
suggest that one such solution is manifested in a biological property
that we choose to call ‘mammal-ness’. So, too, within this ‘zone’ there
are more localized solutions, one of which is ‘ape-ness’. Mammals and
apes (or any other biological entity) have, of course, arisen by way of
specific historical trajectories, but in these (and many other) instances
the various convergences on mammals and apes that are documented
above indicate that although any history is necessarily unique, the re-
sultant complex end form is not simply the contingent upshot of local
and effectively random processes. On any other suitable planet there
will I suggest be animals very like mammals, and mammals much
like apes. Not identical, but similar, perhaps surprisingly similar.
This view of life is, of course, by no means the first attempt to
arrive at a general biological theory that might have potentially uni-
versal applicability.129 Nevertheless, it offers some intriguing clues
that could mark a first step towards the definition of galactic-wide
niches.130 In essence the aim will be first to define a series of hy-
perdimensional ‘boxes’, based on the combinatorial immensity of the
relevant variables that together encompass all the alternative possi-
bilities. Such ‘boxes’ range from protein ‘hyperspace’ to intelligence
and even societal ‘hyperspaces’. With respect to the last example,131
Lee Cronk reminds us not only of the universalities shared by all hu-
man societies, together a long and impressive list, but also that in
terms of combinatorial ‘hyperspace’ it is possible to envisage c.1053
alternatives, of which less than 103 are known. Cronk goes on to re-
mark that the figure of 1053 can be reduced by collapsing the number
of variables. He writes, ‘If we recalculate the total number of possible
combinations using only these variables, there are still over one hun-
dred million possible combinations. Although this is a lot less than
[1053 ] . . . it still suggests that only a tiny fraction of the points in ethno-
graphic hyperspace are occupied.’132 Such an example makes it clear
that the known zones of occupation must be an infinitesimally tiny
a possible research programme 309
fraction of the total available ‘space’. The current consensus is that

any such zone of occupation has been arrived at by chance, and there
is a very large number of alternatives, probably well in excess of the
number of habitable planets in the galaxy. Accordingly, on this basis,
it is very unlikely that the occupants of one planet will look remotely
like those of another. The phenomenon of evolutionary convergence
indicates that, to the contrary, the number of alternatives is strictly
limited, with the interesting implication that the vast bulk of any
given ‘hyperspace’ not only never will be visited during evolutionary
exploration but it never can be. These are the howling wildernesses of
the maladaptive, the 99.9% recurring of biological space where things
don’t work, the Empty Quarters of biological non-existence.
If this is correct then it suggests that an exploration of how evo-
lution ‘navigates’ to particular functional solutions may provide the
basis for a more general theory of biology.133 In essence, this approach
posits the existence of something analogous to ‘attractors’,134 by
which evolutionary trajectories are channelled towards stable nodes
of functionality.135 This, of course, begs the question as to how the
vastness of a given biological ‘hyperspace’ is actually navigated. This
is because the number of potential ‘blind alleys’ is so enormous that
in principle all the time since the beginning of the Universe would
be insufficient to find the one in a trillion-trillion solutions that ac-
tually work. This echoes the common criticism, exemplified by Fred
Hoyle’s disbelief that a tornado passing through a junkyard could con-
struct a jumbo jet, that there is anything like enough time to arrive
at biological functionality. Life, it is claimed, is simply too complex
to be assembled on any believable timescale. The stock response is to
invoke a million monkeys typing alternatives, with the invisible hand
nudging the myriad of efforts towards the correct Shakespeare sonnet
(or whatever). This really misses the point, first because it presupposes
that the correct version is known all along, and secondly because it
fails to tackle the problem of the almost illimitable size of biological
‘hyperspaces’. The metaphor of navigation, however, suggests a more
fertile research programme than the mass employment of monkeys.
This is because it will attempt to explain, not only the preferred tra-
jectories to the optimum, but also evolution’s uncanny ability to find
the short cuts across the multidimensional ‘hyperspace’ of biological
reality. It is my suspicion that such a research programme might re-
veal a deeper fabric to biology in which Darwinian evolution remains
central as the agency, but the nodes of occupation are effectively pre-
determined from the Big Bang.
One such node is, of course, that of the humanoid, and from
the present evolutionary perspective we are undeniably unique. Yet,
as I have already argued, if we had not arrived at sentience and called
ourselves human, then probably sooner rather than later some other
group would have done so, perhaps from within the primates, perhaps
from further afield, even much further afield. Yet it is our prerogative
to explore the natural world, and thereby document evolution, and
in the footsteps of Darwin marvel at its diversity. These discoveries
have not been without their cost. Not only are our animal origins plain
to see, but so, too, it is now supposed that with our origins revealed
this must banish any religious instinct: what was almost universally
believed is now to be seen as an immense delusion. Now is the time,
it is proclaimed, to adopt wholeheartedly the naturalistic view of life,
not only to expel the mysteries, but to deride the ignorant and mould
the world as we will. In the next, penultimate, chapter we shall see
that things are not necessarily so simple.
11 Towards a theology
of evolution?
Let us suppose that I am an immensely successful biochemist, and

happen to be engaged in experiments involving gene manipulation. A
couple of years ago I was attending a conference – keynote speaker,
naturally – when I fell into conversation with a curious individual,
who for some reason seemed much older than he actually looked.
As we talked, it seemed we were walking across a plain of infinite
dreariness, but his voice, his demeanour, how can I describe it? He
knew all about our work, and as conversation progressed, gave me
remarkable hints as to some avenues of research we had somehow
overlooked. I was enthralled, and as we parted he remarked, ‘I am
sure we will meet again.’ The next day, however, he had vanished, and
checking at the registration desk, I was puzzled to find no record of
my companion. It was all a little eerie, but the hints were sufficient.
Now my team and I have managed to reconfigure a gene that will
allow animals, and the poorer humans, to digest cellulose directly.
Before long, in Bangladesh and Somalia, the main meal will be recycled
newspapers. There is, unfortunately, just one small side effect, and
that is it is very likely (I can explain the details if you have time)
to induce childhood cancers in about one per cent of the population,
especially if the individual happens to live in a deprived environment.
Of course, the gene is patented, and in strictest confidence I can reveal
to you alone that the biotech company, OmegaPoint, has the product
ready for immediate marketing. Great riches beckon; surely I am to
be congratulated?
Do I detect a sense of unease,1 but pray why? Surely it is most
regrettable that a vast proportion of the world’s biomass is tied up
as indigestible cellulose,2 and at present the only simple route to the
hamburger is via the bacteria living within the rumen of the millions
of cattle contentedly grazing on thin grassland, once rain forest. It is,
of course, only these bacteria which are capable of breaking down the
refractory plant cellulose and so release compounds that elsewhere
in the cow will ultimately end up as beef mince. Nothing, of course,
will go to waste: spinal cord and the rectum have their uses, while
312 towards a theology of evolution
anything else, in the best traditions of agrobusiness, will be ground up

and returned to the cattle themselves. And if this hypothetical gene3
has side effects, what matter? The vast bulk of scientific research
undertaken by the biotechnology companies is subject neither to peer
review – the accepted norm anywhere else in science – nor available for
publication. That link to childhood cancers will take years to ‘prove’,
and in reality most probably never will be, given both the medical
inadequacies in much of the Third World and – well – facts that are
deemed ‘commercially sensitive’. And does not the race go to the best
and fastest, which in the West we equate with the accumulation of
stupendous amounts of money? So, am I not using the talent bestowed
by evolution for the best, feeding the hungry – yes, yes, on newspaper –
and making myself very, very rich?
an evolutionary embedment
It is self-evident that whatever our peculiarities as humans, we are
embedded in the natural world, and just as clearly we are one product
of an evolutionary process that began about four billion years ago. In
itself, the mechanism of evolutionary change is so unexceptional as to
be almost trivial. As Martin Carrier4 has written, ‘Darwinian theory
plays a role in evolutionary biology that is analogous to the one New-
tonian theory plays in celestial mechanics. It provides the mechanism
of change; it specifies the law-governed processes that determine how
species develop and adapt in a possibly changing environment.’5 Des-
pite this simple process most biologists will freely acknowledge that
both the routes and the products of evolution are profoundly fascinat-
ing. I have already alluded to a few of these, such as the problem of the
origin of life itself. Probably of equal moment is to discover how it is
that proteins fold so effectively and quickly. Evolution also presents
what Denis Duboule and Adam Wilkins6 have termed ‘bricolage’, that
is the surprising co-option and redeployment of biological material for
unexpected uses. An excellent example is the crystallin proteins of
the eye lens, which as discussed in Chapter 7 (p. 166) are routinely
and independently co-opted from heat-shock and other stress-related
proteins that evolved thousands of millions of years before any eye
could see.
Biologists also have, in the true Darwinian spirit, immense ad-
miration for the jury-rigging of biological design, whereby co-option
and modification lead to the functioning whole. And, if they are
an evolutionary embedment 313
honest, they may feel a sense of unease about the fluidity and grace of
adaptation. It has an almost uncanny sense of precision and balance,
which humans achieve only rarely in technology or art. Not only does
one admire the pervasiveness and ubiquity of integrated organic com-
plexity, but one cannot help but be impressed by its sensitivity, in
some circumstances, to minute changes so that sometimes even a
single amino acid substitution in a protein can lead to a major change
of functionality, if not catastrophe. And for some – probably many –
when we review the broad sweep of evolution and greet first the flick-
ering and then the full emergence of consciousness, that is all there is
to say. Science must work in a naturalistic framework, and again many
appear content to live with this arrangement. Roll on that cellulose-
digesting gene, and let us treat the world as open to unlimited
manipulation.
Yet, there are nagging doubts. Yes, it may all be due to a few
misfiring neurons, perhaps an extra dollop of neuropeptide or what-
ever, but the fact remains that humans have an overwhelming sense
of purpose. As a species we are strangely comfortable to find ourselves
embedded in a teleological matrix. So the intention of this chapter is
to begin to see whether the idea of a telos7 is redundant, to ask if some
of our predecessors who saw their religious faith either ebb or haem-
orrhage were both misinformed and over-pessimistic, and to enquire
whether some common ground can be regained.
One clue is surely our admiration for moral greatness. Rather,
however, than argue or defend any particular individual, although
there are many such men and women, let us recall the cosmic view
of G. K. Chesterton.
Reason and justice grip the remotest and the loneliest star. Look at
those stars. Don’t they look as if they were single diamonds and
sapphires? Well, you can imagine any mad botany or geology you
please. Think of forests of adamant with leaves of brilliants.
Think the moon is a blue moon, a single elephantine sapphire. But
don’t fancy that all that frantic astronomy would make the
smallest difference to the reason and justice of conduct. On plains
of opal, under cliffs cut out of pearl, you would still find a
notice-board, ‘Thou shalt not steal.’8
Let me conclude this introductory section with two other re-

marks. The first is to the effect that contrary to received wisdom I
certainly do not consider religions and metaphysics to be aberrations

of the superstitious, delusions of those still enmeshed in medieval
credulity, unwilling to face the boundlessly happy future; a future that,
strange to say, is always just around the next corner. More specifically,
although the more we explore this world the better is our understand-
ing, this is also accompanied by a growing sense of its extraordinary
strangeness and beauty. Darwin at least was no stranger to these feel-
ings, but, as we shall see, in growing older he retreated into a gloomy
agnosticism, and his world became stretched and thin. And Darwin,
of course, was (and even more so today, is) not alone in this grow-
ing sense of futility. And yet just as many are possessed of a strong
teleological instinct, and in the words of Arthur Peacocke9 , ‘Some-
how, biology has produced a being of infinite restlessness, and this
certainly raises the question of whether human beings have properly
conceived of what their true “environment” is.’10 More often these
pangs of want turn to a dissatisfaction that in turn seems to lead
only to the futile and sterile, but songs and stories tell us this need
not be so.
darwin’s priesthood
As I have already indicated, such views as those expressed by Peacocke
and others may strike a chord in places, but I suspect that they
will also be widely regarded as quaint and antiquated. Indeed
these views are under full-scale assault by a group that allows
a unique priority to Darwinian mechanisms and, in most cases,
the primacy of the gene. These ultra-Darwinists are highly promi-
nent, not least because the shoulders on which they stand in-
clude Huxley, Simpson, and Mayr, all giants in their various ways
within the Darwinian synthesis. Let me make some initial obser-
vations, which in today’s climate of accommodating pluralism and
relativism will, I suppose, seem deeply unfair. I am driven to ob-
serve of the ultra-Darwinists the following features as symptomatic.
First, to my eyes, is their almost unbelievable self-assurance, their
breezy self-confidence.11 Second, and far more serious, are partic-
ular examples of a sophistry and sleight of hand in the misuse of
metaphor, and more importantly a distortion of metaphysics in sup-
port of an evolutionary programme. Consider how ultra-Darwinists,
having erected a naturalistic system that cannot by itself possess
any ultimate purpose, still allow a sense of meaning mysteriously
darwin’s priesthood 315
to slip back in. Thus, the philosopher of science John Greene12

remarks,
Not all of the champions of the modern synthesis have been as

open as [Julian] Huxley in acknowledging the religious aspect of
their devotion to evolutionary biology, but most of them,
especially those who reject religious and philosophical approaches
to the problem of human duty and destiny, manage to smuggle in
by way of simile and metaphor the elements of meaning and value
that their formal philosophy of nature and natural science
excludes from consideration.13
Despite this, such scientists have no foundation for their reac-

tion against pointlessness other than the not unworthy and intuitive
sense that the world should be built as it is; embedded in the Uni-
verse are not only neutrons but such edicts as, to echo Chesterton,
‘Thou shalt not steal.’ Greene has other trenchant comments that are
also surely apposite. As he remarks, in the field of evolution the term
‘progress’ must be value-free and can only mean ‘survival’. He contin-
ues, ‘One would like to feel optimistic about the scientific mythology
that has grown up around the theory of evolution, but it is hard to
do so.’ And he reminds us that in the hands of some practitioners the
analysis of evolution is more like that of a myth, and one that is in-
tellectually dishonest, ‘employing teleological and vitalistic figures of
speech to describe processes that are advertised as “mechanistic” and
pretending to derive from evolutionary biology values that stem from
classical, Judaeo-Christian, and Enlightenment sources. It deifies sci-
ence, denigrates philosophy and religion, and panders to Western cul-
ture’s penchant for regarding science and technology as the guarantors
of indefinite progress toward some hazy but glorious future paradise.’
And Greene pointedly continues, ‘Worse yet, it fosters dreams of ge-
netic manipulation and control designed to reshape imperfect human
nature according to some scientistic ideal.’14
Third, as has often been noted, the pronouncements of the
ultra-Darwinists can shake with a religious fervour. Richard Dawkins
is arguably England’s most pious atheist. Their texts ring with high-
minded rhetoric and dire warnings – not least of the unmitigated evils
of religion – all to reveal the path of simplicity and straight think-
ing. More than one commentator has noted that ultra-Darwinism
has pretensions to a secular religion, but it may be noted that, how-

ever heartfelt the practitioners’ feelings, it is also without religious or
metaphysical foundations. Notwithstanding the quasi-religious en-
thusiasms of ultra-Darwinists, their own understanding of theology
is a combination of ignorance and derision, philosophically limp,
drawing on clichés, and happily fuelled by the idiocies of the so-
called scientific creationists. It seldom seems to strike the ultra-
Darwinists that theology might have its own richness and subtleties,
and might – strange thought – actually tell us things about the
world that are not only to our real advantage, but will never be re-
vealed by science. In depicting the religious instinct as a mixture
of irrational fundamentalism and wish-fulfilment they seem to be
simply unaware that theology is not the domain of pop-eyed flat-
earthers.
So does this matter? The world ticks along, and someone,
somewhere is presumably busy trying to manufacture that cellulose-
digesting gene. As a day-to-day activity science is highly pragmatic,
but it also makes much wider claims to describe the world as it is.
But to assume that science itself can produce or verify the truths upon
which it depends is, as many have pointed out, simply circular. On
this basis and as a human activity science is ultimately imperilled.
The discoveries of science, as is also widely acknowledged, are not
short of ethical and moral implications. Nowhere is this more true
than in biology, and it is pertinent to make some reference to some
of the real giants of the field as the implications of evolution sank in:
Darwin, the Huxleys (Thomas Henry and Julian), and more uneasily
Haeckel. Here I offer only a series of snapshots. Concerning Darwin
and Huxley, there is, of course, a huge literature, but as above I draw
especially on the work of John Greene.15 This is mostly because his ad-
miration for these individuals is also set in a metaphysical framework
that declines, courteously, to be browbeaten by naturalistic dogmas.
Darwin, the chief architect of evolution, whose genius is not in doubt,
had a strongly theistic background, and as a young man had an admi-
ration for Paley’s arguments for biological design reflecting the hand
of the Deity. Yet, as is equally well known, when the implications of
his theory on evolution began to sink in, he descended (and I use that
word advisedly) into a sort of pantheism, dogged by half-articulated
fears, if not terrors. For him belief seems to have slowly ebbed, almost
against his will.
darwin’s priesthood 317
Darwin was, to put it mildly, a retiring figure, and for a

more public, if not melodramatic, view of another trek away from
Damascus then it is inevitable that we turn to his staunch sup-
porter, Thomas Henry Huxley. Interestingly, for Huxley, the adop-
tion of the Darwinian world-picture also took on more than a tinge
of theology. Thus the religious imagery that accompanies Adrian
Desmond’s Huxley: From devil’s disciple to evolution’s high priest16 is
revealing,17 and while it acknowledges Huxley’s secular enthusiasm,
perhaps it also distorts the sort of man he really was. It seems that,
at first, Huxley was content that any sense of an underlying mean-
ing could be eroded from the fields of biology and evolution. These
regions of science fell into a realm of blind chance and unconscious
agony, but one that at least allowed a deistic notion that conceded,
however feebly, a cryptic arrangement of forces and laws. But in the
end this, too, was insufficient: Huxley had to become something for
which he had also to coin a name, that is, an agnostic. This was not
the decorous scepticism with which the term is usually associated:
for Huxley his position reflected genuine and unresolved intellectual
doubt. Even so, there is little evidence that he was willing to enter-
tain, let alone discuss, serious theological statements, and he had an
ill-disguised contempt – no doubt well earned in some cases – for the
clerics. Yet, Huxley was a man of transparent goodness and deeply felt
morality, and was happy to recruit such Old Testament prophets as
Micah to his cause, provided of course that they were shorn of any re-
ligious dimension. Even so, it is not clear that living in a metaphysical
vacuum brought him much peace, and towards the end of his life it
seems that he dimly discerned that the new science that he had helped
so ably and energetically to popularize was opening a Pandora’s box.
Thomas Henry Huxley was not the last of the Huxley clan to
struggle with the implications of evolution, yet arguably the vague-
ness never dissipated. Most notably his grandson, Julian Huxley, re-
turned to the fray, but he adopted a view of evolution that was akin to
what is referred to as process theology. Here in the younger Huxley’s
view the grand scheme of things was, as he wrote near the beginning
of one work,18 ‘unitary; continuous; irreversible; self-transforming;
and generating variety in novelty during its transformations’,19 sen-
timents echoed at the end of the same volume when he hoped that
‘I have given you some feeling of the unity and sweep of the pro-
cess . . . some insight into its nature as a self-transforming process,
constantly generating new patterns and novel qualities, building its

future by transcending its past.’20 And, as Greene notes, in his stimu-
lating Huxley to Huxley (note 15), this view of life ‘excludes from the
process all these elements – aim, purpose, creative ground – which run
counter to the positivistic grain of modern science and which alone
could make such a process [i.e. evolution] intelligible.’ Greene contin-
ues and concludes that ‘The result is that Julian Huxley . . . propounds
the paradox that nature, though devoid of aim and purpose, yet moves
towards ever higher levels of order and value.’21
As Julian’s grandfather had dimly perceived, the naturalistic pro-
gramme would open the door to the manipulation of a world from
which meaning had fled. Thus the march ‘towards ever higher levels
of order and value’ remained a chimaera. In reality, as Greene observes,
‘The multiplication of devices and techniques for controlling the nat-
ural environment and influencing human behavior seems only to ag-
gravate man’s lack of control over the general course of events and to
intensify the moral problems connected with human freedom . . . the
possibilities for evil multiply as rapidly as the possibilities for good,
and history affords little assurance that men or women, either indi-
vidually or collectively, will choose the general welfare of mankind
in preference to immediate personal or national advantage.’22
heresy! heresy!!
Let me continue this brief review with two other episodes that in
their different ways are almost comic, but in at least the first case
have a much darker side. Otherwise each is very different. The first
concerns Ernst Haeckel, the keeper of the Darwinian evangeliarium
in Germany, and the second the farce of the Dayton ‘Monkey Trial’ in
1925. In the mind of Haeckel, Darwin was the greatest of heroes, and
Darwinism the new beacon to lead the world from its benighted obscu-
rantism. There is an amusing, at least to the English, story of Haeckel
crossing the Channel and visiting Down House: a metaphorical, if not
literal, genuflection between Teutonic adorer and English gentleman
scientist. There was, however, no meeting of the minds. The
German was effusive, gushing, bombastic, enthusiastic, and Darwin?
well . . .
Haeckel had seized upon the Darwinian explanation and vigor-
ously and tirelessly promoted it in Germany. Not only was this done
with books, but also with lavishly illustrated lectures. Concerning
heresy! heresy!! 319
the latter, Daniel Gasman23 refers to a poster for one such lecture
in Berlin, and remarks how this example of Darwinismus-Kunst pro-
vided ‘a sinister environment for a Darwinian Passion Play.’24 It is,
moreover, Gasman who has done much to reveal the rottenness at the
heart of Haeckel’s project.25 As I remarked elsewhere26 when drawing
upon his work:
The reality is very different and . . . much darker. Haeckel’s pursuit

of Darwinism went far beyond any scientific formulation, even
further than Herbert Spencer’s uneasy rhetoric usually referred to
as Social Darwinism. Haeckel’s role as a spokesman for these
malign influences was encapsulated in the so-called Monist
league, of which he was the effective founder. This mish-mash
enshrined a set of pantheistic beliefs of supposedly cosmic
importance. Haeckel and his fellow Monists were dedicated
believers in organic progress, the end-point of which,
unsurprisingly, was the Aryan ideal. Nowhere was Haeckel’s
influence greater or more charismatic than in his book The Riddle
of the Universe. Vastly popular, endlessly reprinted and
translated, it nevertheless ‘appealed to a pseudo-educated mind . . .
without much sophistication who had sought an authoritative yet
simple account of modern science and a comprehensible view of
the world’.27 Behind the bearded sage and devotee of the little
town of Jena, was an intolerant mind wedded to racism and
antisemitism. After initial reservation, Haeckel’s admiration of
Bismarck and his autocracies grew. Predictably his support for the
German conquest of Europe was fervid. Unrepentant of German
aggression he died, still revered, in 1919. His farrago of ideas,
however, not only lived on but found a warm reception with the
Nazis. Just how much Hitler knew of Haeckel’s actual work is not
clear, but the influence of his philosophy is obvious.
Eavesdropping at Hitler’s table talk, with its hypnotic mixture of
rant, bluster and threats all set in a half-baked philosophy,
constantly echoes the monistical Haeckel.28
In the Scopes (or Monkey) Trial in Dayton, Tennessee29 the is-

sue was as much the long-standing loathing between the principals
for the defence and the prosecution, respectively William Jennings
Bryan and Clarence Darrow. Darrow was at that time the most famous
lawyer in America, already notorious for his risky and ingenious de-
fence of Nathan Leopold and Richard Loeb, two Chicago teenagers
who abducted, bludgeoned to death, and then mutilated with acid
a schoolboy and neighbour, a luckless fourteen-year-old known as
Robert Franks. This murderous pair were bunglers of a high order,
but had it seems undertaken the atrocity in the belief that they would
escape detection. There was, ostensibly, no other motive. At the trial,
Darrow undercut the prosecution by an unexpected change to a plea of
‘Guilty’, followed by a rhetorical harangue to the effect that Leopold
and Loeb retained their innocence because of the environment in
which they were nurtured, one of intellectual stimulation, but emo-
tional starvation, and in the case of Leopold, the corruption of a young
mind that inevitably results from being made to read Nietzsche. This
libertarian argument, delivered as a thrilling speech that concluded
with a quotation from Omar Khayyam, has plenty of resonances to-
day. As one writer on the Scopes Trial, Kevin Tierney, notes, at the
end of the speech, ‘many in the audience were crying. Darrow had
taken the case far beyond the bounds of reason and logic . . . It was
his most masterly oration, rousing his audience to display emotion
openly beyond what the conventions allowed . . . The spellbinder has
cast his spell once again.’30
Thus Darrow embraced a moral perspective, as exemplified by
the Leopold and Loeb trial, which led to Dayton and his battle against
Bryan. Here was a very different man. No fool: as a Democrat, he
had been quite close to winning the American presidency, and at the
time of Dayton was America’s greatest defender of fundamentalist
Christianity. Like Darrow he, too, was a brilliant orator. And it was
this engagement, between Bryan and Darrow, rather than either the
teaching of evolution or its scientific truths, which lay at the heart of
this trial. To be sure, the action began as a defence of the protection of
civil liberties and the necessary separation of Church and State, and
this was initiated as a test case by the American Civil Liberties Union
(ACLU). To their dismay, Darrow effectively imposed himself, offer-
ing to waive fees in his determination to ridicule and thereby crush
the forces of religious obscurantism and, quite incidentally, maintain
his public profile. Before long, the ACLU effectively lost control of
events as Darrow hijacked the circus for his own good purposes. Op-
posed to him, with Bryan as its senior spokesman, was a largely rural
constituency that possessed a heartfelt, if largely inarticulate, belief
heresy! heresy!! 321
in a fundamentalist Christianity that seemed imperilled on all sides

by the secular agents. Kevin Tierney spells out the real nature of this
trial:
There was a further, more sinister animosity between Bryan and

the defense. Bryan, failing though he was, was still probably the
greatest public speaker then living in the United States. The
defense team (including Darrow) wanted to pitch themselves
against him so as to seize the crown. Darrow recognized Bryan’s
power over his audience. He was a worthy competitor, to be
confronted in a final battle in which Darrow’s superiority would
be displayed. Thus the entire Scopes trial was colored by a
personal rivalry as well as a difference of opinion.31
There were (and are) indeed serious objections to fundamentalism,

rural or otherwise. But for that matter, so there were to Darrow’s
childish conception of theology, which was in its way was as dated
and credulous as Bryan’s.
The trial itself has further ironies, which while quite well
known are seldom spelt out, especially by those who regard it as a
test case between the shining uplands of science and the snarling re-
ligious reactionaries. This is not to defend the idiocies of legislation
designed (then or now) to prevent the teaching of evolution or any
other science, however uncomfortable the findings might appear to
be. Societies that ignore what we discover do so at their peril, but if
they imagine that on occasion the discoveries of evolution are neutral
in their implications, again societies delude themselves. Even so, the
technical defendant, John Thomas Scopes, was a physics teacher at the
school in Dayton. It is not even clear if he taught anything on evolu-
tion, and his own knowledge of this area was vestigial. So far as Darrow
was concerned, Scopes was a pawn, if not a stooge, and once the trial
started, he was ignored. The trial was a simple test case. Scopes had
received no warning from the School Board not to teach evolution
in defiance of the State Legislature; the Board could not care less.
The entire dynamic came first from the ACLU and thereafter from
a ruthless Darrow. Again to quote Tierney: ‘As Darrow had gained
in years and confidence, he became not only merely an unbeliever,
but a militant agnostic.’32 Ultimately his animus against religion be-
came almost obsessional, as in later years Darrow ‘continued to sift
through the religious outpourings of the day, intent upon uncovering
literalism, absurdity, and contradiction with the same energy that a

devotee of the faith might seek out heresy.’33 Underpinning the be-
liefs of what Tierney has aptly described as ‘the last village atheist’34
was a fundamentalist Darwinism. Thus ‘The theory of evolution was
close to Darrow’s philosophy about life in general, not merely to his
anti-religious bent. He believed life to be cruel, and mankind to be un-
regenerate . . . Darwinism also had appealing deterministic aspects. No
man could step aside from the march of history, which was inexorable
and inevitable.’35 Yet, in the long term, Darrow’s rhetorical brilliance
and intelligence proved a poisoned chalice. As the years slipped by he
became ‘more and more a performer, a mouther of scripts, and less a
thinker . . . as he fed his appetite for glory while travelling from city
to city for one-night stands.’36
The trial, which resulted in the jury being absent for less than
ten minutes and Scopes receiving a $100 fine, achieved little, other
than to increase the rancour and suspicion between the religious fun-
damentalists and the scientists. In America the skirmishing continues
unabated, with one side unable or unwilling to comprehend the meth-
ods of science, and the opposing party all too often exhibiting a lofty
arrogance, mingled with contemptuous disdain, which presupposes
that any religious instinct is a mental aberration. The former know
in their hearts that something is out of kilter, but their cause is hope-
less. Not only are they overly simplistic and credulous, but they are
not averse to selective quotation divorced from its context, and even
more seriously outright twisting and distortion of the evidence. So,
too, they seem unaware that theology involves rather more than scrip-
tural inerrancy, especially with respect to the Creation myth. Such a
polarization is not only regrettable; it is intellectually poisonous. For
those wedded to sociological relativism the solution is to allot each
their sphere of influence, but such apparent generosity merely con-
ceals a strategy for sidelining religion and a road to philosophical in-
coherence. Despite the antagonisms, however, there are also attempts
to find common ground between sciences and religions, most notably
in the field of cosmology. In part because of the apparent absence of
law-like properties, biology and evolution have remained largely ex-
cluded from this search for common ground, a stance that has been
reinforced by the myth of contingent forces driving the evolution of
life. Yet despite this, strange to say, biology and especially genetics
have their own fundamentalisms. These, in turn, reveal a disquieting
genetic fundamentalism 323
agenda that has curious echoes of the very systems they purport to
despise.
genetic fundamentalism
That biology can be co-opted for agendas, if not ideologies, that
promise an ever-more-perfect future, albeit across piles of corpses,
is evident from the lunacies adopted by totalitarian states. Such mad-
ness is, of course, a thing of the past – or is it? Now new distortions
beckon, not least those to be allowed by assigning a protean mal-
leability to life as engendered by genuflection to the primacy of the
gene. Now the gene is all-powerful. Susan Oyama37 describes genes as
‘molecular agencies that are immortal, omnipotent, omniscient, and
even immaterial’.38 In a related vein, Peter Koslowski’s39 critique of
this view, as it has been offered by Dawkins, defines an approach that
‘concedes a faculty for aspiration, intentionality and consciousness
to the genes. In doing so, he [Dawkins] falls into a genetic animism,
which apportions perception and decision to the genes and oversub-
scribes by far to the efficiency and the speed of Darwinian selection
mechanisms.’40
Such critiques differ greatly from the popular notion of the gene,
which does indeed seem to be able to act as a universal agency. The
spectacular examples of genetically controlled defects, not to mention
the generation of ectopic monsters with hideously sprouting organs,
reveal the potency of the gene. So, too, in examples of caste structure,
especially in eusocial animals (p. 142), the paradox of sterile workers
ceaselessly toiling for the benefit of the hive or nest is explicable by
a proportion of their genes surviving41 even though they themselves
either cannot or will not reproduce.
A closer examination, however, reveals that these particular ex-
amples fall far short of allowing the gene per se a universal application.
In her essay ‘The gene is dead – long live the gene!’ Eva Neumann-
Held42 argues that regarding the genes generally as particulate objects
of heredity is hopelessly simplistic. The genes make sense only in a
known context, but in reality to know in sufficient detail a context
that will provide the sought-after predictability may be very difficult,
perhaps impossible. That is why it is so misleading, if not dangerous,
to speak of genes for, say, schizophrenia or aggression. Given the right
set of prompts (who knows, perhaps plenty of junk food?) then the risk
could increase – or decrease. And what else are we meant to expect?
Outside its cellular milieu the DNA is biologically inert, if not use-
less. Genes may provide a switchboard for life, but the complexity of
life will depend on something else: how the same genes may be re-
cruited to make different products, how the developmental networks
change and evolve, and how apparently trivial events such as gene
duplication and protein isoforms open immense new territories for
biological exploration. Life may be impossible without genes, but to
ascribe to them powers of intentionality misses the mark.
Despite these qualifications, however, the hard-core view that
claims primacy of the gene holds sway. In some hands, perhaps most
notably those of the sociobiologist E. O. Wilson, it is a vehicle of
unbounded faith both in its power and in terms of its implications,
not least for the human prospect. This is spelt out, for example, in
Consilience,43 which is an extended belief-statement in an overarch-
ing system where all will be explained – society, art, religion – by
the gene. Wilson expresses himself with fervour and conviction, but a
more dispassionate reading of Consilience leaves me more impressed
by Wilson’s faith in the argument, accompanied by leaps in logic, un-
warranted assumptions, and over-simplification.44 And the world pic-
ture of genetic primacy has a well-known parallel, that is, the mental
equivalents of genes. They too are all-purpose, if elusive, little things
that are known as memes. Perhaps that irritating little tune that con-
tinues to bounce around your head is a meme? One cannot but no-
tice how trivial many of the examples presented are, unless it is to
portray the sheer wickedness of religious beliefs inculcated into the
brains of morally helpless humans. Happily, the alternative religions
of consumerism and shopping are arrived at by the exercise of human
dignity, unpersuaded by anything remotely like a meme. So, too, in
a way reminiscent of the notion of ‘junk’ DNA, one cannot help but
notice that the discussion of memes is often pejoratively associated
with some notion of ‘mind-parasites’. But memes are trivial, to be ban-
ished by simple mental exercises. In any wider context they are hope-
lessly, if not hilariously, simplistic. To conjure up memes not only
reveals a strange imprecision of thought, but, as Anthony O’Hear45
has remarked, if memes really existed they would ultimately deny
the reality of reflective thought.
These views on genes and memes matter very much, because
granting such molecular (or memal) hegemony puts us back firmly
on the path towards the Abolition of Man.46 And yet the rot started
genetic fundamentalism 325
at an earlier stage. As John Greene opines:47 ‘To the very end, he

[Darwin] failed to appreciate the morally ambiguous character of hu-
man progress. He failed because, like many social scientists today,
he had no adequate conception of Man.’48 As Greene also remarks,
humans are very peculiar creatures indeed; clearly a product of evo-
lution, yet a species that has, or has been allowed, to know men-
tal states that transcend (so far as we know) any other sentience on
the Earth. Again to quote Greene, ‘science becomes pointless and
even destructive unless it takes on significance and direction from
a religious affirmation concerning the meaning and value of human
existence.’49
Despite the sleights of hand, special pleading, and sanctimo-
niousness as the ultra-Darwinists attempt to smuggle back the moral
principle through the agency of the gene, only the most hardened cases
would suppose that a map of the genome will provide the blueprint of
this millennium’s equivalent to the Code of Hamurabi. And yet these
myths of genetic determinism, set in a dreary world of reductionism,
are being used to drive new agendas, most notably in eugenics. At
present it is the natural world, which according to some, should be
treated as a sort of genetic play-dough. Now vanished is the notion
that the world we have been given might have its own integrity and
values. Rather the prevailing view of scientism is that the biosphere
is infinitely malleable. Again the moral high ground is hijacked on the
assumption that all this is for our perceived good, although in reality
the benefits are far more likely to fill the coffers of the corporations
and erode the diversity of crop species, to be followed by who knows
what? There is no doubting that in at least some cases this manipu-
lation is possible. Two research, workers Temple Smith and Harold
Morowitz,50 remark ‘As a consequence of reflective thought, we have
today within our grasp the ability to assemble genetic combinations
that have nearly zero probabilities of ever being sequentially assem-
bled by nature.’ Yet, these writers seem unenthused as they remark
of this genomic programme, ‘now a gambler’s game [is being] played
by those who may not be fully aware of the stakes’.51
What follows from the genetic meddling in maize and soon pigs,
will, it may be safely assumed, be applied in due course to humans. It is
interesting to recall that T. H. Huxley, in contrast to the ever-confident
Galton, whom we met in Chapter 7 doing arithmetic by smell, drew
back in horror from this eugenic prospect, arguing simply that no man
could possibly know enough to decide. Let us recall, in the words of

C. S. Lewis,52 the prophetic voice of the head of the National Institute
of Co-ordinated Experiments, better known as N.I.C.E. Listen to the
recently ennobled, and hideous, Lord Feverstone:
‘Man has got to take charge of man. That means, remember,
that some men have got to take charge of the rest . . . we’ll get on to
biochemical conditioning in the end and direct manipulation of the
brain . . . A new type of man.’ Regrettably there are still a few dunder-
heads, ‘old women of both sexes up in arms and yapping about human-
ity’, but happily all this can be dealt with. Feverstone knows what’s
going on when he addresses the naive Mark Studdock: ‘You are what
we need; a trained sociologist with a radically realistic outlook . . . We
want you to write it down – to camouflage it . . . It’s nothing to do
with journalism. Your readers in the first instance would be commit-
tees of the House of Commons.’ Of course the dunderheads need to
be manipulated also, but fortunately there is ‘Jules . . . a distinguished
novelist and scientific populariser whose name always appeared be-
fore the public . . . He’s all right for selling [N.I.C.E.] to the great British
public in the Sunday papers and he draws a whacking salary. He’s no
use for work.’53
a path to recovery?
The corrosive view that all in this world is to be bent to our pleasure
or whim is hedged in reality with expediencies and half-truths, and
in the view of many represents the royal road to catastrophe. So how
might we begin to think about, let alone achieve, a Recovery? First,
we need to recall the limits to science. It is no bad thing to remind
ourselves of our finitude, and of those things we might never know.
Practically, we should not be afraid to acknowledge that there are ar-
eas that Roger Shattuck calls Forbidden knowledge,54 too dangerous
in our present state of understanding to explore. At its simplest it is
a precautionary principle, and more significantly a belated acknowl-
edgement that the architecture of the Universe need not be simply
physical. We should also recall, as if we needed reminding, that we
are mortal and limited, and thus should remember that the old myths
of unrestricted curiosity and the corruption of power are not neces-
sarily fables.55
Second, for all its objectivity science, by definition, is a human
construct, and offers no promise of final answers. We should, however,
a path to recovery? 327
remind ourselves that we live in a Universe that seems strangely well

suited for us. In earlier chapters I dwelt, all too briefly, on the para-
doxes of the origin of life (Chapter 4) and the many peculiarities of
the Solar System (Chapter 5) which seem to be prerequisites for our
existence. On a cosmic scale it is now widely appreciated that even
trivial differences in the starting conditions would lead to an unrec-
ognizable and uninhabitable universe. The idea of a universe suitable
for us is, of course, encapsulated in the various anthropic principles.
These come in several flavours, but they all remind us that the physi-
cal world has many properties necessary for the emergence of life. Of
these probably the best-known are those connected to the synthesis
of carbon in the interior of stars, and the many strange properties of
water (and ice).
Less widely appreciated, but of equal moment, is Howard Van
56
Till’s insistence that ‘It is not simply the numerical values of cer-
tain parameters that must be “just right” in order for life to develop.
No, it’s the entire formational economy of the universe that must
be “just right”. The full menu of the universe’s formational capabili-
ties must be sufficiently robust to make possible the actualization of
carbon-based life . . . I would argue that the formational capabilities
of the universe are more fundamental than the numerical values of
certain physical parameters.’57 Not only is the Universe strangely fit
to purpose, but so, too, as I have argued throughout this book, is life’s
ability to navigate to its solutions.
As is well known the anthropic principle, in whatever guise,
has largely attracted the interest (or scepticism) of cosmologists and
physicists. Biologists, on the other hand, have generally been content
to take such features as carbon or water as givens, with life as an emer-
gent inevitability on any suitable planet. But there are connections,
because at the heart of the study of evolution are two things. One, em-
phasized throughout this book, is the uncanny ability of evolution to
navigate to the appropriate solution through immense ‘hyperspaces’
of biological possibility. The other, equally germane and even more
mysterious, is the attempt to explain the origins of sentience, such
that the product of ultimately inanimate processes can come to un-
derstand both itself, its world, and, as I have already noted, its (and
thus our) strange sense of purpose. We need also to remember that sci-
entific explanations need not be all-embracing, and indeed it would be
surprising if they were. As Michael Polanyi, a philosopher of science
who took religion seriously, noted, other descriptions have their own
power. In Personal knowledge58 he writes:
The book of Genesis and its great pictorial illustrations, like the
frescoes of Michelangelo, remain a far more intelligent account of
the nature and origin of the universe than the representation of
the world as a chance collocation of atoms. For the biblical
cosmology continues to express – however inadequately – the
significance of the fact that the world exists and that man has
emerged from it, while the scientific picture denies any meaning
to the world, and indeed ignores all our most vital experience of
this world. The assumption that the world has some meaning
which is linked to our own calling as the only morally responsible
beings in the world, is an important example of the supernatural
aspect of experience which Christian interpretations of the
universe explore and develop.59
So, at some point and somehow, given that evolution has pro-
duced sentient species with a sense of purpose, it is reasonable to
take the claims of theology seriously. In recent years there has been a
resurgence of interest in the connections that might serve to reunify
the scientific world-view with the religious instinct. Much of the dis-
cussion is tentative, and the difficulties in finding an accommodation
remain daunting, but it is more than worth the effort. In my opinion
it will be our lifeline.
converging on convergence
The principal aim of this book has been to show that the constraints
of evolution and the ubiquity of convergence make the emergence
of something like ourselves a near-inevitability. Contrary to received
wisdom and the prevailing ethos of despair, the contingencies of bio-
logical history will make no long-term difference to the outcome. Yet
the existence of life itself on the Earth appears to be surrounded with
improbabilities. To reiterate: life may be a universal principle, but we
can still be alone. Whether or not this is literally true may never be
established, and, as many of us have argued, it is far more prudent to
assume that we are unique, and to act accordingly.60
Yet now we are faced with a special dilemma. The very scientific
method that allows us to study the natural world, be it interstellar or-
ganic molecules or memory in dolphins, also gives us tools that treat
converging on convergence 329
the world as endlessly malleable, ostensibly for the common good but
as often as not for the enrichment of the few and the impoverishment
of the many. Such attitudes fly in the face of traditional wisdoms,
and in part explain the existing antagonisms between scientific prac-
tices and religious sensibilities. Mutual misunderstandings, fuelled by
naivety and ignorance, can only lead to warfare. Although science may
emerge triumphant, it will be a pyrrhic victory; the conquered king-
dom will lie in ruins, strewn across a plain of infinite melancholy.
Constructive approaches are more difficult, and are usually viewed
with contempt, but I believe promise far more. In essence, we can
ask ourselves what salient facts of evolution are congruent with a
Creation. In my judgement, they are as follows:
(1) its underlying simplicity, relying on a handful of building

blocks;
(2) the existence of an immense universe of possibilities, but a way
of navigating to that minutest of fractions which actually work;
(3) the sensitivity of the process and the product, whereby nearly
all alternatives are disastrously maladaptive;
(4) the inherency of life whereby complexity emerges as much by
the rearrangement and co-option of pre-existing building blocks
as against relying on novelties per se;
(5) the exuberance of biological diversity, but the ubiquity of evo-
lutionary convergence;
(6) the inevitability of the emergence of sentience, and the likeli-
hood that among animals61 it is far more prevalent than we are
willing to admit.
Having already quoted G. K. Chesterton once in this chapter, let me

return to this wisdom by way of a conclusion. As he writes,62
Turning a beggar from the door may be right enough, but

pretending to know all the stories the beggar might have narrated
is pure nonsense; and this is practically the claim of the egoism
which thinks that self-assertion can obtain knowledge. A beetle
may or may not be inferior to a man – the matter awaits
demonstration; but if he were inferior by ten thousand fathoms,
the fact remains that there is probably a beetle view of things of
which a man is entirely ignorant. If he wishes to conceive that
point of view he will scarcely reach it by persistently revelling in
the fact that he is not a beetle. The most brilliant exponent of the
egoistic school, Nietzsche, with deadly and honourable logic,
admitted that the philosophy of self-satisfaction led to looking
down upon the weak, the cowardly, and the ignorant.63
So the beetle, no longer the butt of Haldane’s jibe,64 is an ex-

ample of the richness of a Creation. Whether we shall always remain
‘entirely ignorant’ of what a beetle (or a bat) thinks is open to discus-
sion. Even if we do, the complexity and beauty of ‘Life’s Solution’ can
never cease to astound. None of it presupposes, let alone proves, the
existence of God, but all is congruent. For some it will remain as the
pointless activity of the Blind Watchmaker, but others may prefer to
remove their dark glasses. The choice, of course, is yours.
12 Last word
The vehicle landed at 15.47 GMT, just over a mile from Kimmeridge,
on the south coast of England. ‘Just in time for tea?’ murmured my
companion, as we climbed through the long grass, insects rising in
the summer air. There, already sitting on the ground, were the three
extraterrestrials. As we joined them, I asked, ‘Would you like some
water?’ ‘Or perhaps something stronger?’ suggested my companion.
‘Thank you’, came the grave reply. ‘We ourselves are thirsty, on such
a warm day. And maybe something for our plants?’ The chlorophyll
of the alien species blended well with the surrounding vegetation, its
flowers a deep purple. When our visitor picked up one of the pots,
it slipped and in catching it he grazed his finger. Red blood oozed
to the surface. ‘Haemoglobin, I suppose?’ They nodded. Our hands
clasped, both warm to the touch. It seemed superfluous to ask, but
the beating of a vein hinted at the inevitable dual circulation sys-
tem and arteries with their elastic proteins. One of them sniffed the
air appreciatively; the world smelt beautiful as their and our nasal
glomeruli registered the olfactory signals. As the swallows screamed
overhead, the minute hairs in our ears and the auditory equivalents
of the extraterrestrials acted in the same way, transducing the sound
into a register of inner music. ‘Observe the pointed wings of those
flying animals – swallows, did you say? – clearly migrants, just as at
home.’ We strolled slowly back down the hill, towards the sea. Des-
pite the steepness of the slope, the aliens were confident, their bal-
ancing mechanisms the same, their walking movements controlled
by the same neural network. We stopped on the cliff edge; in the
bay three dolphins moved westwards. ‘Fast and effective swimmers,
I see; what other shape could they possibly have?’ All of us looked
at them with common delight, through camera-eyes, lenses full of
crystallins and retinas with opsin molecules making photons into
sight. By now dusk was falling, Venus already bright in the sky. It
was time to go home. ‘Before you leave, may I ask where you are
332 last word
from? Was it a very long journey?’ But we already had guessed. ‘A long
journey? Well, only in some ways. Where is our home? Why, the
planet we call Earth, of course. Surely you already knew in your
hearts that there is only one Earth?’ We were, of course, looking at
ourselves.
Notes
A note about the notes. The principal aim is to give citations, chapter and verse, to the many
facts, interpretations, and sources I give. Many more could have been given, but I hope they
form a useful introduction. Where a reference is repeated the original note is also cited. There
are also a few cross-chapter citations.
preface the cambridge sandwich

1. J. Maynard Smith’s review ‘Taking a chance on evolution’ was published in the 14 May 1992
issue on pp. 34–6.
2. Quotation is on p. 34, column 4.
5. R. C. Lewontin’s review (‘Fallen angels’) was published in the New York Review of Books,
14 June 1990, pp. 3–4, 6–7; quotation is on p. 7, column 1.
6. J. Maynard Smith (1992; citation is in note 1), p. 36, column 1.
7. R. C. Lewontin (1990; citation is in note 5), p. 7, column 1.
8. As with Maynard Smith, Lewontin’s remarks are very thoughtful, and it is only fair to put the
two quotations in their full context. Thus, in conclusion he writes: ‘So evolution may be
contingent only in a superficial and uninteresting way. The exact forms that have left
descendants visible in fossil remains may indeed be accidental variants of a historically
accidental process. But they may all be distinctions without a difference, superficial
orthographic variants of a deep structure whose rules we have yet to uncover. A description of
all the organisms that have ever been cannot decide the issue . . . We cannot know the answer
unless we have a theory of biological form that is deduced from some general principles of
biological organization, rather than inferred from the collection of objects. Or it may be that
no such principles exist, and that in this broadest sense, life has no meaning’ (citation is in
note 5, p. 7, column 1).
9. Thus the well-known false ‘thumb’, i.e. the pre-pollex (or radial sesamoid) of the giant panda
is, as pointed out by F. Wood-Jones in Proceedings of the Zoological Society of London,
vol. 109B, pp. 113–129, 1939, equivalent to the sixth digit of the mole hand. In this case the
convergence of bone structure does not extend to function; in the moles it appears that the
sixth digit serves to widen the hand and so assist in digging, whereas in pandas the ‘thumb’ is
opposable and helps to strip bamboo leaves.
10. See Peter Ward and Donald Brownlee in Rare earth: Why complex life is uncommon in the
Universe (Springer [Copernicus], New York, 2000).
11. See Stuart Ross Taylor in Destiny or chance: Our Solar System and its place in the cosmos
(Cambridge University Press, Cambridge, 1998).
1. looking for easter island

1. These are a group known as the Myxobacteria, and include the arborescent fruiting body of
Chondromyces crocatus; see H. Reichenbach in the book Myxobacteria II (M. Dworkin and
D. Kaiser, eds), pp. 13–62 Washington, DC, American Society for Microbiology (1993).
2. For the extreme accuracy of owl hearing see E. I. Knudsen and M. Konishi in Journal of
Comparative Physiology, vol. 133, pp. 13–21, 1979; see also notes 265 and 266 in Chapter 7.
While the owls have phenomenally acute hearing, so do some insects; see for example the
334 notes to pp. 2–11
paper by P. Müller and D. Robert in Journal of Experimental Biology, vol. 204, pp. 1039–1052,
2001; see also note 305 in Chapter 7.
3. For this and other remarkable examples of long-distance navigation see F. Papi and P. Luschi
in Journal of Experimental Biology, vol. 199, pp. 65–71, 1996.
4. See Simon Winchester’s book The river at the centre of the world: A journey up the Yangtze,
and back in Chinese time (Viking [Penguin], London, 1997), pp. 96–97.
5. But not necessarily; A. Eyre-Walker in Genetics (vol. 152, pp. 675–683, 1999) provides
evidence of selection in the evolution of junk DNA of junk DNA, while W. Makalowski (in
Gene, vol. 354, pp. 61–67, 2000) argues that it is more like a genomic scrapyard, to be pillaged
as and when necessary. Others, such as R. N. Mantegna et al. in Physical Review Letters, vol.
73, pp. 3169–3172, 1994 suggest that junk DNA encodes information analogous to the
structure of language (specifically an equivalent to Zipf’s law, where the total numbers of
occurrences of words in a text fall on a linear distribution when plotted on a log–log scale, and
also find evidence for redundancy whereby meaning is preserved even with deletions). These
views are questioned and replied to in a subsequent issue (vol. 76, pp. 1976–1981).
6. See L. Z. Holland and N. D. Holland in American Zoologist, vol. 38, pp. 647–658, 1998;
W. R. Jackman et al. in Developmental Biology, vol. 220, pp. 16–26, 2000; H. Toresson et al.
in Development, Genes, and Evolution, vol. 208, pp. 431–439, 1998; T. V. Venkatesh et al. in
Development, Genes, and Evolution, vol. 209, 254–259, 1999; H. Wada et al. in
Developmental Biology, vol. 213, pp. 131–141, 1999; and N. C. Williams and P. W. H. Holland
in Molecular Biology and Evolution, vol. 15, pp. 600–607, 1998.
7. See T. C. Lacalli et al. in Philosophical Transactions of the Royal Society of London B,
vol. 344, pp. 165–185, 1994, and vol. 351, pp. 243–263, 1996, as well as Acta Zoologica
(Stockholm), vol. 80, pp. 113–124, 1999 and Proceedings of the Royal Society of London B,
vol. 266, pp. 1461–1470, 1999.
8. See P. W. H. Holland in Development (Supplement 1994), pp. 125–133, 1994; American
Zoologist, vol. 38, pp. 829–842, 1998; and Seminars in Cell & Developmental Biology,
vol. 10, pp. 541–547, 1999.
9. Such examples are known from the Lower Cambrian (Chengjiang fossil assemblages): see
D. Shu et al. in Nature, vol. 384, pp. 157–158, 1996; and from the Middle Cambrian (Burgess
Shale): see papers by A. M. Simonetta et al. in Bollettino di Zoologia, vol. 62, pp. 243–252,
1995 and Italian Journal of Zoology, vol. 66, pp. 99–119, 1999.
10. This remarkable article, by Temple Smith and Harold Morowitz, entitled ‘Between physics
and history’, was published in Journal of Molecular Evolution, vol. 18, pp. 265–282, 1982.
11. See especially W. M. Elsasser’s insights into what he calls very large and immense numbers,
in Reflections on a theory of organisms: holism in biology (Johns Hopkins University Press,
Baltimore, 1998). This equally strange and stimulating book is, as its subtitle indicates,
definitely anti-reductionist. Also very relevant to this question is Information and the origin
of life by B-O Küppers (MIT Press, Cambridge, MA, 1990); see especially Chapters 6
and 7.
12. They are probably most familiar in the form of the digestive enzymes such as trypsin and
amylases, but they are essential in all parts of the cellular economy. Other enzymes are
essential for dealing with a stiff gin and tonic; here you need alcohol dehydrogenase.
13. T. Smith and H. Morowitz (1982; citation is in note 10), p. 268.
14. T. Smith and H. Morowitz (1982; citation is in note 10); both quotations are on p. 268.
15. See the short article on ‘Laws of forms revisited’, by M. Denton and C. Marshall in Nature,
vol. 410, p. 417, 2001. See also M. J. Denton et al. in Journal of Theoretical Biology, vol. 219,
pp. 325–342, 2002.
16. M. Denton and C. Marshall (2001; citation is in note 15), p. 417.
17. See, for example, the paper by Y-H. Lee et al. in Journal of Molecular Evolution, vol. 45,
pp. 278–284, 1997, where they show how pathways through RNA sequence space may be
equally parsimonious, but not equally likely.
notes to pp. 12–15 335
18. Stephen Freeland reminds me, however, that while this axiom seems sensible, ‘the only
explicit rationale for this view is Fisher’s geometric theorem (in his The genetical theory of
natural selection [Oxford, Clarendon Press, 1930; see pp. 38–41]); a simple abstract model of
evolution that predicts an inversely proportional relationship between the magnitude of
effect of a random mutation and the probability that it will represent an adaptive
improvement’; see Freeland’s article in Journal of Genetic Programming and Evolvable
Machines, vol. 3, pp. 113–127, 2002.
19. Symbiosis means ‘living together’, and, although the term is often thought to be synonymous
with mutual benefit, biologists take a symbiosis to be a neutral term that can, if sufficient
information is available, be resolved as benefit (+), loss (−), or neither (0). Parasitism usually
has a winner (the parasite) and a loser (the host), thus (+)(−), unless each parasitizes the other,
i.e. (−)(−).
20. T. Smith and H. Morowitz (1982; citation is in note 10), p. 280.
21. This conceit of ‘Different routes to similar ends’ is specifically addressed by P. H. Harvey and
L. Partridge in Nature, vol. 392, pp. 552–553, 1998, as well as by the first author writing in
Current Biology, vol. 10, p. R271, 2000.
22. That may be the natural order, but biotechnological manipulation makes it possible to
introduce novel amino acids, thus artificially expanding the genetic code. See, for example,
L. Wang et al. in Science, vol. 292, pp. 498–500, 2001, and the following paper (pp. 501–504)
by V. Döring et al., as well as the commentary by A. Böch on pp. 453–454; see also M. Ibba
et al. in Current Biology, vol. 11, pp. R563–R565, 2001.
23. See the chapter (pp. 21–27) by J. M. Diamond in Principles of animal design: the optimization
and symmorphosis debate, edited by E. R. Weibel et al. (Cambridge University Press,
Cambridge, 1998). The graphic predicament of the lift plunging to its doom is based on
remarks made by Diamond in a lecture he gave in Cambridge.
24. See Henry Petroski’s Design paradigms: case histories of error and judgement in engineering
(Cambridge University Press, Cambridge, 1994). The book is an absorbing discussion of the
trials, tribulations, and daring of bridge builders; Petroski’s remarks on the sociology of
engineering knowledge and tradition are also of great, and sometimes alarming, interest.
25. See, for example, G. M. Taylor et al. in Biological Journal of the Linnean Society, vol. 70,
pp. 37–62, 2000. They address safety factors in the claws of crabs, and also give a useful
introduction to the literature of this area.
26. Carl Gans ‘Momentarily excessive construction as the basis for protoadaptation’ is the title of
his paper in Evolution, vol. 33, pp. 227–233, 1979.
27. See Kea, bird of paradox: the evolution and behavior of a New Zealand parrot, by J. Diamond
and A. B. Bond (University of California Press, Berkeley, 1999).
28. See John Currey’s paper in Journal of Experimental Biology, vol. 202, pp. 3285–3294,
1999.
29. See J. D. Currey et al. in Proceedings of the Royal Society of London B, vol. 268, pp. 107–111,
2001, and also V. de Buffrénil and A. Casinos in Annales des Sciences naturelles, Zoologie,
Paris, vol. 16, pp. 21–32, 1995.
30. J. D. Currey (1999, citation is in note 28), p. 3289. In his paper Currey also draws attention to
the convergence between the teeth of the sea urchins (on which see R. Z. Wang et al. in
Philosophical Transactions of the Royal Society of London B, vol. 352, pp. 469–480, 1997),
which rasp the sea floor for edibles, and the gnawing incisors of the rodents.
31. The key papers are by Steve Freeland and Laurence Hurst in Journal of Molecular Evolution,
vol. 47, pp. 238–248, 1998 and Molecular Biology and Evolution, vol. 17, pp. 511–518, 2000a;
see also Trends in Biochemical Sciences, vol. 25, pp. 44–45, 2000b.
32. Concerning the optimitality of two base pairs see E. Szathmary in Proceedings of the Royal
Society of London B, vol. 245, pp. 91–99, 1991, and Proceedings of the National Academy of
Sciences, USA, vol. 89, pp. 2614–2618, 1992. See also D. A. Mac Donaill in Chemical
Communications, vol. 18, pp. 2062–2063, 2002.
33. A. L. Weber and S. L. Miller, in Journal of Molecular Evolution, vol. 17, pp. 273–284, 1981,
draw on various lines of evidence (including prebiotic synthesis, stability, and function in
proteins) to argue that the 20 terrestrial amino acids are the norm.
34. A. L. Weber and S. L. Miller (1981, citation is in note 33), p. 273.
35. A. Jiménez-Sánchez, however, suggests (in Journal of Molecular Evolution, vol. 41,
pp. 712–716, 1995) that the original arrangement in an ‘RNA world’ (see note 1, Chapter 4)
was still triplets (and one base pair (AU)), coding for seven amino acids (and a stop codon).
36. See J. T. Wong in Microbiological Science, vol. 5, pp. 174–181, 1988. R. Amirnovin in Journal
of Molecular Evolution, vol. 44, pp. 473–476, 1997, concludes that much depends on which
amino acids are assumed to have a biosynthetic linkage. M. Di Giulio and M. Medugno,
however, present a statistically based critique of Amirnovin’s idea in a subsequent issue
(vol. 50, pp. 258–263, 2000), reasserting their belief in ‘the intimate relationship between the
biosynthetic pathways of amino acids and the organization of the genetic code’ (p. 263). On
the other hand, T. A. Ronneberg et al. (in Proceedings of the National Academy of Sciences,
USA, vol. 97, pp. 13690–13695, 2000) speak strongly against the importance of biosynthetic
pathways, although they also emphasize that code expansion almost certainly occurred.
37. S. Freeland and L. Hurst (1998; citation is in note 31), p. 244.
38. S. Freeland et al. (2000b; citation is in note 31), p. 45.
39. It should be pointed out that the conclusions reached by Freeland and Hurst are not
universally accepted. See M. Di Giulio and M. Medugno in Journal of Molecular Evolution,
vol. 52, pp. 378–382, 2001, who argue that the effectiveness of the genetic code lies in a
co-evolution between the code and the biosynthetic pathways leading to the various amino
acids.
40. For example, the codons UAA and UAG, which nearly always mean stop, have been
reassigned in several unrelated groups to code for glycine. The evolvability of the genetic code
is reviewed by R. D. Knight et al. in Nature Reviews, Genetics, vol. 2, pp. 49–58, 2001, who
remark (p. 49) ‘Curiously, many of the same codons are reassigned in independent lineages,
frequently between the same two meanings . . . indicating that there may be an underlying
predisposition towards certain reassignments. At least one of these changes seems to confer a
direct selective advantage.’
41. See, for example, M. Guilio and M. Medugno in Journal of Molecular Evolution, vol. 49,
pp. 1–10, 1999.
42. See Geoffrey Irwin’s book The prehistoric exploration and colonisation of the Pacific
2. can we break the great code?

1. See, for example, the apt comments (pp. 531–532) by Christian de Duve in Proceedings of the
American Philosophical Society (vol. 142, pp. 525–532, 1998).
2. See the poignant chapter 19, ‘The self marooned in the cosmos’ in Walker Percy’s Lost in the
Cosmos: The last self-help book (Arena, London, 1994).
3. Such is argued by P. D. Ward and D. Brownlee in Rare Earth: Why complex life is uncommon
in the Universe (Copernicus, New York, 2000), which comes to a number of similar
conclusions as to the scarcity of habitable planets. These workers, however, suggest that the
origination of life is relatively commonplace but that complex life forms, such as
pen-wielding humanoids, are very rare.
4. See, for example, Christian de Duve’s Vital dust: life as a cosmic imperative (HarperCollins
[BasicBooks]), New York, 1995.
5. C. de Duve (1998; citation is in note 1), p. 527.
6. Threga IX? My imaginary planet, strangely Earth-like and home to our visitors, whom we
finally meet in Chapter 12.
7. See Blueprints: solving the mystery of evolution (Oxford University Press, Oxford, 1990) by
M. A. Edey and D. C. Johanson.
8. M. A. Edey and D. C. Johanson (1990), p. 295.

9. See Iris Fry’s The emergence of life on Earth: A historical and scientific overview (Free
Association, London, 1999)
10. George Wald, for example, provides a cogent discussion of the merits and otherwise of carbon
and silicon in his chapter (pp. 127–142) in Horizons in biochemistry. Albert Szent-Györgyi
dedicatory volume, edited by M. Kasha and B. Pullman (Academic Press, New York, 1962).
11. See, for example, Robert Shapiro and Gerald Feinberg in their chapter (pp. 248–255) in
Physical cosmology and philosophy, edited by J. Leslie (Macmillan, New York, 1990).
12. See I. Fry (1999; citation is in note 9), p. 239.
13. See I. Fry (1999; citation is in note 9), p. 241.
14. See F. H. Westheimer in Science, vol. 235, pp. 1173–1178, 1987.
15. Francis Crick and Leslie Orgel’s paper can be found in Icarus, vol. 19, pp. 341–346, 1973.
16. P. Parsons, for example, in Nature, vol. 383, pp. 221–222, 1996, reviews some current
thinking. He points out that so far as a source for panspermia is concerned, a planet orbiting a
red giant might be particularly suitable, given that such a star is relatively cool, emits less
ultraviolet radiation (which is extremely damaging, at least to terrestrial biochemistry), and
very long-lived, thus increasing the net chance of life evolving on any associated planet.
17. See, for example, F. Egami in Journal of Molecular Evolution, vol. 4, pp. 113–120, 1974. In
contrast, the related element chromium is very sparse in sea water, and correspondingly finds
no use in living organisms. The element selenium, however, is vital, but is available only in
trace amounts. See also W. R. Chappell et al. and T. H. Jukes in Icarus (vol. 21, pp. 513–517,
1974) and A. Banin and J. Navrot in Science (vol. 189, pp. 550–551, 1975).
18. For a masterly overview of this general area consult R. J. P. Williams and J. J. R. Fraústo da
Silva on The natural selection of the chemical elements (Clarendon Press, Oxford,
1996).
19. Christopher Switzer and colleagues’ article can be found in Biochemistry, vol. 32,
pp. 10489–10496, 1993.
20. C. Switzer et al. (1993; citation is in note 19), p. 10489.
23. For example, Switzer et al. (1993; citation is in note 19) also note that the type of heterocycle
ring that occurs in the oligonucleotides leads to ambiguities in the hydrogen bonding.
24. See C. Mao et al. in Nature, vol. 386, pp. 137–138, 1997.
25. See E. Winfree et al. in Nature, vol. 394, pp. 539–544, 1998.
26. See Y. Zhang and N. C. Seeman in Journal of the American Chemical Society, vol. 116,
pp. 1661–1669, 1994.
27. See A. P. Alivisatos et al. in Nature, vol. 382, pp. 609–611, 1996; and the preceding paper, on
pp. 607–609, by C. A. Mirkin et al.
28. See E. Braun et al. in Nature 391, 775–778, 1998.
30. See, for example, B. Bhat et al. in Journal of the American Chemical Society, vol. 118,
pp. 3065–3066, 1996.
31. This paper, entitled ‘Enzymatic incorporation of a new base pair into DNA and RNA extends
the genetic alphabet’, may be found in Nature, vol. 343, pp. 33–37, 1990.
32. Joseph Piccirilli et al. (1990), p. 33.
33. For another such example, see F. Seela and A. Melenewski in European Journal of Organic
Chemistry for 1999, pp. 485–496, 1999.
34. Piccirilli et al. (1990; citation is in note 32), p. 34.
35. See C. Roberts et al. in Journal of the American Chemical Society, vol. 119, pp. 4640–4649,
1997.
36. See H. Hashimoto and C. Switzer in Journal of the American Chemical Society, vol. 114,
pp. 6255–6256, 1992.
37. See J. P. Dougherty in Journal of the American Chemical Society, vol. 114, pp. 6254–6255,
1992.
38. See T. L. Sheppard and R. Breslow in Journal of the American Chemical Society, vol. 118,
pp. 9810–9811, 1996.
39. The key review papers may be found by Albert Eschenmoser and colleagues in Science
(vol. 284, pp. 2118–2124, 1999) and Origins of Life and Evolution of the Biosphere (vol. 24,
pp. 389–423, 1994 and vol. 27, pp. 535–553, 1997).
40. See K. D. James and A. D. Ellington in Origins of Life and Evolution of the Biosphere, vol. 25,
pp. 515–530, 1995; quotation is on p. 520.
41. Thus K. D. James and A. D. Ellington (1995; citation is in note 40) remark on work by the
Eschenmoser team that they ‘have concluded that very few [alternatives] can potentially form
complementary duplexes that would in theory be capable of self-replication . . . These
conclusions can be extended to nucleic acid bases as well as sugars: experiments with
alternate base pairing schemes have suggested that the current set of purines and pyrimidines
is in many ways optimal . . . the unnatural nucleic acid analogues that have been examined
experimentally have proven to be largely incapable of self-replication,’ p. 520.
42. ‘Why pentose – and not hexose – nucleic acids?’ is the main title of the paper by S. Pitsch
et al. in Helvetica Chimica Acta, vol. 76, pp. 2161–2183, 1993, and is answered with respect
to ‘a whole series of intrinsic steric handicaps’ (p. 2163). In his overview Eschenmoser (1999;
citation is in note 39) concluded that ‘The outcome of these studies has led to the conclusion
that for functional reasons, the three hexopyranosyl-(4 -6 ) oligonucleotide systems
investigated . . . could not have acted as viable competitors of RNA in the emergence of
nature’s genetic system’ (p. 2121).
43. Concerning alternatives to ribose, as five-carbon (pentose) sugars, see M. Beier et al. in
Science, vol. 283, pp. 699–703, 1999.
44. See the paper by K.-U. Schöning et al. in Science, vol. 290, pp. 1347–1351, 2000, as well as a
commentary by L. Orgel on pp. 1306–1307.
45. A. Eschemoser (1999; citation is in note 39), p. 2122.
3. universal goo: life as a cosmic principle?

1. C. de Duve (1998; citation is in Chapter 2, note 1).
2. C. de Duve (1998; citation is in Chapter 2, note 1), p. 526.
3. This quotation is given in Robert Shapiro’s Origins: A skeptic’s guide to the creation of life on
Earth (Bantam, New York, 1987), p. 187.
4. See my article in Astronomical Society of the Pacific Conference Series, vol. 213,
pp. 410–419, 2000, p. 417.
5. An overview is given by p. Ehrenfreund and S. B. Charnley in Annual Review of Astronomy
and Astrophysics, vol. 38, pp. 427–483, 2000.
6. This conceit may be found in Robert Shapiro (1987; citation is in note 3); see p. 230.
7. See S. B. Charnley et al. in Spectrochimica Acta A, vol. 57, pp. 685–704, 2001. P. Ehrenfreund
et al. (in Astrophysical Journal, vol. 550, pp. L95–L99, 2001) also point out that unless
shielded amino acids in deep space are very liable to destruction by ultraviolet radiation.
8. See, for example, L. J. Allamandola’s chapter (pp. 81–102) in The cosmic dust connection,
edited by J. M. Greenberg (Kluwer, Dordrecht, 1996), and also Astrophysical Journal, vol. 511,
L115–L119, 1999.
9. See L. Becker et al. in Geochimica et Cosmochimica Acta, vol. 61, pp. 475–81, 1997 and
Earth and Planetary Science Letters, vol. 167, pp. 71–79, 1999; and M. Zolotov and E. Shock
in Journal of Geophysical Research, vol. 104 (E6), pp. 14033–14049, 1999 and Meteoritics &
Planetary Science, vol. 35, pp. 629–638, 2000.
10. See J. P. Bradley et al. in Geochimica et Cosmochimica Acta, vol. 60, pp. 5149–5155, 1996
and Meteoritics & Planetary Science, vol. 33, pp. 765–773, 1998; K. L. Thomas-Keprta et al.
in Geochimica et Cosmochimica Acta, vol. 64, pp. 4049–4081, 2000; P. R. Buseck et al. and
D. J. Barber and E. R. D. Scott in Proceedings of the National Academy of Sciences, USA,

vol. 98, pp. 13490–13495, 2001 and vol. 99, pp. 6556–6561, 2002 respectively.
11. For example, amino acids (see J. L. Bada et al. in Science, vol. 279, pp. 362–365, 1998) and
sulphur isotopes (see J. P. Greenwood et al. in Geochimica et Cosmochimica Acta, vol. 61,
pp. 4449–4453, 1997), not to mention contaminants (e.g. A. Steele et al. in Meteoritics &
Planetary Science, vol. 35, 237–241, 2000).
12. See L. E. Borg et al. in Science, vol. 286, p. 90–94, 1999, E. R. D. Scott in Journal of
Geophysical Research, vol. 104 (E2), pp. 3803–3813, 1999; D. C. Golden et al. in Meteoritics
& Planetary Science, vol. 35, pp. 457–465, 2000; and J. M. Eiler et al. in Geochimica et
Cosmochimica Acta, vol. 66, pp. 1285–1303, 2002.
13. See, for example, H. Naraoka et al. in Earth and Planetary Science Letters, vol. 184, pp. 1–7,
2000.
14. Thus W. F. Hume, in The Cairo Scientific Journal, vol. 5, pp. 212–215, 1911, reported from a
newspaper account to the effect that one fragment ‘fell on a dog at Denshal, leaving it like
ashes in a moment’, p. 212.
15. See Kevin Yau and his colleagues’ article in Meteoritics, vol. 29, pp. 864–871, 1994.
16. K. Yau et al. (1994), Table 1 (p. 867).
17. See J. S. Lewis’s interesting Rain of iron and ice: the very real threat of comet and asteroid
bombardment (Addison-Wesley [Helix], Reading, MA, 1996), pp. 176–182.
18. See S. Veski et al. in Meteoritics & Planetary Science, vol. 36, pp. 1367–1375, 2001.
19. See P. G. Brown et al. in Science, vol. 290, pp. 320–325, 2000, with commentary by J. N.
Grossman on pp. 283–284.
20. See a special section edited by P. G. Brown et al., citation is in note 19, of Meteoritics &
Planetary Science, vol. 37(5), 2002; see also P. Ehrenfreud et al. in Proceedings of the National
Academy of Sciences, USA, vol. 98, pp. 2138–2141, 2001.
21. See G. Kminek et al. on pp. 697–701 of P. G. Brown et al., citation is in note 19. See also
D. P. Glavin in Proceedings of the National Academy of Sciences, USA, vol. 96,
pp. 8835–8838, 1999 concerning contamination in the Nakhla meteorite.
22. A leading figure in the study of extraterrestrial amino acids is J. R. Cronin, and work by him
and his colleagues can be found in Journal of Molecular Evolution, vol. 17, pp. 265–272, 1981,
two papers in Geochimica et Cosmochimica Acta (vol. 49, pp. 2259–2265, 1985 and vol. 50,
pp. 2419–2427, 1986 respectively), as well as Advances in Space Research (vol. 15 (3),
pp. 91–97, 1995) and Science (vol. 275, pp. 951–955, 1997).
23. See also J. F. Kerridge (Advances in Space Research, vol. 15, pp. 107–111, 1995), who remarks
that the style of amino acid formation is consistent with their synthesis being random and
undirected, governed by principles of thermodynamic stability and very different from any
organic activity.
24. See G. Cooper et al. in Nature, vol. 414, pp. 879–883, 2001, and commentary by M. A.
Sephton on pp. 857–859.
25. There is an interesting twist here, however, because these amino acids have been found
immediately adjacent to the ‘fireball layer’ marking the huge impact that both terminated (by
definition) the Cretaceous and its chief glory, the dinosaurs; see M. Zhao and J. L. Bada in
Nature, vol. 339, pp. 463–465, 1989). Subsequently K. Zahnle and D. Grinspoon in Nature
(vol. 348, pp. 157–160, 1990) argued that the amino acids floated to Earth in association with
comet dust that extended from the nucleus that actually collided with the Earth. But was it a
comet? The latest evidence suggests it was a meteorite: see F. K. Kyte in Nature, vol. 396,
pp. 237–239, 1998. And perhaps the overwhelming evidence for extraterrestrial mayhem has
slanted our view of the source of these strange amino acids. If organic matter was gasified by
the intense shock of impact then perhaps these amino acids are a direct result of the impact;
see E. S. Olson in Nature, vol. 357, p. 202, 1992.
26. See various articles in Nature by G. Yuen et al. in vol. 307, pp. 252–254, 1984; S. Epstein
et al. in vol. 326, pp. 477–479, 1987; and M. H. Engel et al. in vol. 348, pp. 47–49, 1990.
27. Reviews can be found in a chapter (pp. 819–857) by J. R. Cronin et al. in Meteorites and the
early Solar System, edited by J. F. Kerridge and M. S. Matthews (University of Arizona Press,
Tucson, 1988), and The molecular origins of life: assembling pieces of the puzzle, edited by
A. Brack (Cambridge University Press, Cambridge, 1998); see in particular pp. 119–146.
28. See, for example, P. G. Stoks and A. W. Schwartz in Geochimica et Cosmochimica Acta,
vol. 45, pp. 563–569, 1981.
29. See D. W. Deamer and R. M. Pashley in Origins of life and evolution of the biosphere, vol. 19,
pp. 21–38, 1989; see also J. P. Dworkin et al. in Proceedings of the National Academy of
Sciences, USA, vol. 98, pp. 815–819, 2001.
30. Right-handed, or dextral (d-), amino acids do arise in various metabolic processes; see
D-amino acids in sequences of secreted peptides of multicellular organisms, edited by
D. Jollès (Birkhauser, Basel, 1998).
31. J. L. Bada et al. (Nature, vol. 301, pp. 494–496, 1983) gave a careful critique of M. H. Engel and
B. Nagy in Nature (vol. 296, pp. 837–840, 1982) and received in return a robust response.
32. See J. R. Cronin and S. Pizzarello in Science (vol. 275, pp. 951–955, 1997) and commentary on
pp. 942–943 by J. L. Bada; see also S. Pizzarello and J. R. Cronin in Geochimica et Cosmochimica
Acta, vol. 64, pp. 329–338, 2000 and M. H. Engel & S. A. Macko in Precambrian Research,
vol. 106, pp. 35–45, 2001. Subsequently, S. Pizzarello and G. W. Cooper (in Meteoritics &
Planetary Science, vol. 36, pp. 897–909, 2001) reported that alanine was originally racemic,
but an l-enantiomeric excess in glutamic acid was probably terrestrial contamination.
33. See, for example, J. Bailey et al. in Science, vol. 281, pp. 672–674, 1998, and accompanying
commentary by R. Irion on pp. 626–627. Their observation is based on infrared radiation, but
it is assumed that the UV part of the spectrum, which is energetic enough to break chemical
bonds, is also polarized. This is not to say that all such amino acids will be non-racemic.
Laboratory studies of the analogue of ultraviolet radiation on interstellar ices produced a
variety of amino acids, but these are racemic. See M. P. Bernstein et al. and G. M. Muñoz
Caro et al. in Nature, vol. 416, pp. 401–403 and 403–406 respectively, 2002, and commentary
by E. L. Shock on pp. 380–381.
34. See the paper in Nature (vol. 405, pp. 932–935, 2000) by G. L. J. A. Rikken and E. Raupach, as
well as the commentary by L. D. Barron on pp. 895–896.
35. See K. Soai et al. in Nature, vol. 378, pp. 767–768, 1995. Again the excess is small (about 2%),
and the chemistry seems to have rather little relevance to the origin of life.
36. See R. M. Hazen et al. in Proceedings of the National Academy of Sciences, USA, vol. 98,
pp. 5487–5490, 2001. See also the paper by C. A. Orme et al. (Nature, vol. 411, pp. 775–779,
2001, with commentary by L. Addadi and S. Weiner on pp. 753, 755).
37. For an overview see Comets and the origin and evolution of life, edited by P. J. Thomas et al.
(Springer, New York, 1997), especially chapter 6 (pp. 147–173) by C. F. Chyba and C. Sagan.
38. Produced by ultraviolet irradiation in simulated interstellar conditions of PAH napthalene;
see M. P. Bernstein et al. in Meteoritics & Planetary Science, vol. 36, pp. 351–358, 2001.
Quinones are an important building block of chlorophyll, which in Chapter 6 is argued to be
the molecule of choice wherever in the Universe photosynthesis occurs.
39. See E. Pierazzo and C. F. Chyba in Meteoritics & Planetary Science, vol. 34, pp. 909–918, 1999.
4. the origin of life: straining the soup or our credulity?

1. See, for example, I. Fry’s The emergence of life on Earth: A historical and scientific overview
(Free Association Books, London, 1999).
2. The short article by A. Lazcano and S. L. Miller in Cell, vol. 85, pp. 793–798, 1996 provides a
useful and crisp overview.
3. See, for example, The RNA world, second edition: The nature of modern RNA suggests a
prebiotic RNA, edited by R. F. Gesteland et al. (Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, NY, 1998).
4. See David Bartel and Peter Unrau’s article in Trends in Biochemical Sciences, vol. 24,
pp. M9–M13, 1999 (Millennium issue).
5. Bartel and Unrau (1999; citation is in note 4), p. M9. See also notes 44 and 45.
6. See L. E. Orgel in Trends in Biochemical Sciences, vol. 23, pp. 491–495, 1998.
7. See L. E. Orgel in Trends in Biochemical Sciences, vol. 23, p. 491, 1998.
8. See, for example, A. Bar-Nun and A. Shaviv in Icarus, vol. 24, pp. 197–210, 1975; W. L.
Chameides and J. C. G. Walker in Origins of Life and Evolution of the Biosphere, vol. 11,
pp. 291–302, 1981; and J. P. Ferris and W. J. Hagan in Tetrahedron, vol. 40, pp. 1093–1120,
1984.
9. See Manfred Eigen’s influential article, entitled ‘Selforganization of matter and the evolution
of biological macromolecules’, in Die Naturwissenschaften, vol. 58, pp. 465–523, 1971; and
Steps towards life: A perspective on evolution (Oxford University Press, Oxford, 1992).
10. See John Maynard-Smith in Nature, vol. 280, pp. 445–446, 1979.
11. See D. H. Lee et al. in Current Opinion in Chemical Biology, vol. 1, pp. 491–496, 1997.
12. See D. H. Lee et al. in Nature, vol. 390, pp. 591–594, 1997.
13. See Klaus Dose’s essay ‘The origin of life: More questions than answers’ in Interdisciplinary
Science Reviews, vol. 13, pp. 348–356, 1988. Although this article is more than ten years old,
little has changed to suggest we should take a more optimistic view.
14. Dose (1988; citation is in note 13), quotations are on pp. 348 and 349.
15. See John Horgan’s stimulating article in Scientific American, vol. 264(2), pp. 100–109, 1991.
16. In terms of general scepticism, the key text is Robert Shapiro’s book Origins, with the telling
subtitle A skeptic’s guide to the creation of life on Earth (Bantam, New York, 1987); I have
frequent recourse to Shapiro’s ideas later in this chapter. A generally sceptical view can also
be found in Chapter 16 of the book Blueprints: Solving the mystery of evolution (Oxford
University Press, Oxford, 1990) by M. A. Edey and D. C. Johanson.
17. J. Horgan (1991, citation is in note 15), p. 101.
18. See Anthony Keefe and Stanley Miller’s paper on ‘Potentially prebiotic syntheses of condensed
phosphates’ in Origins of Life and Evolution of the Biosphere, vol. 26, pp. 15–25, 1996.
19. A. Keefe and S. Miller (1996; citation is in note 18), p. 15.
20. Concerning its photochemical production in a prebiotic atmosphere, see J. P. Pinto et al. in
Science, vol. 210, pp. 183–185, 1980.
21. See R. Shapiro’s ‘Prebiotic ribose synthesis: a critical analysis’ in Origins of Life and
Evolution of the Biosphere, vol. 18, pp. 71–85, 1988; see also note 16.
22. L. E. Shapiro (1988; citation is in note 21), p. 83.
23. See T. Mizuno and A. H. Weiss in Advances in Carbohydrate Chemistry and Biochemistry,
vol. 29, pp. 173–227, 1974.
24. See R. Larralde et al. in Proceedings of the National Academy of Sciences, USA, vol. 92,
pp. 8158–8160, 1995.
25. See D. Müller et al. in Helvetica Chimica Acta, vol. 73, pp. 1410–1468, 1990.
26. See G. Zubay’s paper in Origins of Life and Evolution of the Biosphere, vol. 28, pp. 13–26,
1998.
27. G. Zubay (1998; citation is in note 26), p. 19. This work is returned to by G. Zubay and T. Mui
in a subsequent issue of Origins of Life and Evolution of the Biosphere (vol. 31, pp. 87–102,
2001), as part of a discussion of nucleotide synthesis. As usual, the general tone is upbeat, but
problems remain: ‘Limited success has been achieved in the synthesis of inosine . . . The
situation for adenosine is much worse . . . For several years we have tried in vain to improve
this system’, p. 100.
28. See K-U. Schöning et al. in Science, vol. 290, pp. 1347–1351, 2000; see also note 43, Chapter 2.
29. See, for example, the papers by J. P. Ferris et al. in Nature, vol. 381, pp. 59–61, 1996 (with
commentary on pp. 20–21 by G. von Kiedrowski) and A. R. Hill et al. and R. Liu and L. Orgel
in Origins of Life and Evolution of the Biosphere, vol. 28, pp. 235–243 and 245–257, 1998
respectively.
30. See Ferris et al. (1996), and S. J. Sowerby et al. (in Proceedings of the National Academy of
Sciences, USA, vol. 98, pp. 820–822, 2001), which explored the differential adsorption of the
various nucleic acid bases on tiny particles of natural graphite, which the authors duly note
‘is not considered a dominant prebiotic material’, followed by a prompt appeal to ‘zeolites,
feldspars and silicas’, p. 821.
31. See, for example, A. W. Schwartz and L. E. Orgel on ‘Template-directed synthesis of novel,
nucleic acid-like structures’ in Science, vol. 228, pp. 585–587, 1985.
32. R. Liu and L. Orgel (1998; citation is in note 29), pp. 256–257.
33. See Graham Cairns-Smith’s Genetic takeover and the mineral origins of life (Cambridge
University Press, Cambridge, 1982). Nevertheless, so far as I am aware no experimental work
has provided a reality check on this hypothesis, and in more recent discussions of the origin
of life these ideas typically receive routine genuflection rather than any serious engagement.
34. See F. G. Mosqueira et al. in Origins of Life and Evolution of the Biosphere, vol. 26, pp. 75–94,
1996.
35. F. G. Mosqueira et al. (1996; citation is in note 34), p. 92.
36. See J. D. Bernal in The physical basis of life (Routledge & Kegan Paul, London, 1951); Bernal’s
emphasis was very much on clay minerals, as well as quartz (pp. 33–37).
37. See, for example, M. Levy and S. L. Miller in Proceedings of the National Academy of
Sciences, USA, vol. 95, 7933–7938, 1998, with specific reference to the thermal stability of
the nucleotides.
38. See Robert Shapiro in Planetary dreams: The quest to discover life beyond Earth (Wiley, New
York, 1999). This book echoes his interests and scepticisms about present attempts to
discover how life originated, but it is also both a passionate plea for the exploration of outer
space combined with a deep, almost religious, conviction that life is a universal principle.
Shapiro writes with a flair and accessibility that is difficult to match.
39. R. Shapiro (1999; citation is in note 38), p. 137.
40. G. Wächtershäuser has promoted his ideas widely and at some length; see, for example, his
papers in Progress in Biophysics and Molecular Biology, vol. 58, pp. 85–201, 1988;
Microbiological Reviews, vol. 52, pp. 452–484, 1988; and his chapter (on pp. 206–218) in The
molecular origins of life: assembling pieces of the puzzle, edited by A. Brack (Cambridge
University Press, Cambridge, 1998). An excellent overview of Wächtershäuser’s programme is
given by Fry (1999; citation is in note 1), pp. 162–172.
41. See, for example, E. Blöchl et al. in Proceedings of the National Academy of Sciences, USA,
vol. 89, pp. 8117–8120, 1992; D. Hafenbradl et al. in Tetrahedron Letters, vol. 36,
pp. 5179–5184, 1995; the two papers by C. Huber and G. Wächtershäuser in Science,
respectively vol. 276, pp. 245–247, 1997 and vol. 281, pp. 670–672, 1998 (with accompanying
commentary by G. Vogel on p. 627 and 629); W. Heinen and A. M. Lauwers in Origins of Life
and Evolution of the Biosphere, vol. 26, pp. 131–150, 1996; and G. D. Cody et al. in Science,
vol. 289, pp. 1337–1340, 2000 (with commentary by G. Wächtershäuser on pp. 1307–1308,
who remarks of these experiments ‘It remains to be established whether such conditions [200
MPa, 250 ◦ C] are geophysically possible’).
42. Experiments, promoted by those sceptical of the Wächtershäuser school, have looked at
prebiotic amino acid and nucleotide synthesis in a FeS/H2 S system and produced negative
results; see A. D. Keefe et al. in Proceedings of the National Academy of Sciences, USA,
vol. 92, pp. 11904–11906, 1995; see also the somewhat sceptical tenor of M. A. A. Schoonen
et al. in Origins of Life and Evolution of the Biosphere, vol. 29, pp. 5–32, 1999.
43. The paper by M. J. Russell and A. J. Hall in Journal of the Geological Society, London,
vol. 154, pp. 377–402, 1997 is a case in point. Despite its encouraging title ‘The emergence of
life from iron monosulphide bubbles at a submarine hydrothermal redox and pH front’, this
particular paper is strong on theory and expectations, but which pathways will ultimately
prove experimentally tractable is less obvious.
44. See Andrew Ellington’s brief overview in Biological Bulletin, vol. 196, pp. 315–319, 1999.
45. A. Ellington (1999; citation is in note 44), p. 317.
46. Literature on hydrothermal locales as the site for the shift from abiogenetic chemical
processes to life is fast expanding. Relevant material includes J. P. Amend and E. L. Shock in
Science, vol. 281, pp. 1659–1662, 1998; J. P. Ferris in Origins of life and evolution of the
biosphere, vol. 22, pp. 109–134, 1992; N. G. Holm and E. M. Andersson in Planetary and
Space Science, vol. 43, pp. 153–159, 1995; E.-i. Imai et al. in Science, vol. 283, pp. 831–833,
1999; E. L. Shock and M. D. Schulte in Journal of Geophysical Research, vol. 103E, pp. 28,
513–28, 527, 1998; and N. G. Holm and J. L. Charlou in Earth and Planetary Science Letters,
vol. 191, pp. 1–8, 2001. See also note 72.
47. See, for example, N. R. Pace in Cell, vol. 65, pp. 531–533, 1991.
48. See, for example, V. Moulton et al. in Journal of Molecular Evolution, vol. 51, pp. 416–421,
2000, who argue that the necessary folding of RNA is prejudiced at elevated temperatures,
although they are careful to note that this does not rule out other stages in the trek from
abiogenesis occuring in hot environments.
49. See M. Gogarten-Boekels et al. in Origins of Life and Evolution of the Biosphere, vol. 25,
pp. 251–264, 1995.
50. See, for example, P. Forterre et al. in Origins of Life and Evolution of the Biosphere, vol. 25,
pp. 235–249, 1995. Interestingly, in the context of Chapter 10, they argue that the key enzyme
used to stabilize DNA in hyperthermophiles, known as reverse gyrase, is actually a composite
and ‘originated from the “late” fusion of a bona fide DNA helicase and DNA topoisomerases
genes. Indeed many helicases and topoisomerases probably evolved independently from
different RNA-metabolizing enzymes of the RNA world . . . This is suggested by the existence
of several independent superfamilies of these enzymes’ (p. 241). See also note 21, Chapter 5.
51. See S. L. Miller and A. Lazcano in Journal of Molecular Evolution, vol. 41, pp. 689–692, 1995.
52. From the experimental procedures given by Huber and Wächtershäuser (1998, citation is in
note 41), note 2 on p. 672 and Heinen and Lauwers (1996, citation is in note 41), p. 132.
53. See Leslie Orgel’s critique in Proceedings of the National Academy of Sciences, USA, vol. 97,
pp. 12503–12507, 2000.
54. L. E. Orgel (2000; citation is in note 53), p. 12506.
55. A. W. Schwartz and L. E. Orgel (1985, citation is in note 31), p. 586.
56. See S. J. Sowerby and W. M. Heckl in Origins of Life and Evolution of the Biosphere, vol. 28,
pp. 283–310, 1998.
57. S. J. Sowerby and W. M. Heckl (1998; citation is in note 56), p. 291.
60. See the paper by P. A. Bachmann et al. in Nature, vol. 357, pp. 57–59, 1992.
61. See C. Böhler et al. in Origins of Life and Evolution of the Biosphere, vol. 26, pp. 1–5, 1996,
who report on the oligomerization of amino acids by micelles of CTAB, and also remark that
this process is achieved at ‘relatively concentrated aqueous solutions’ (i.e. 0.05 molar), but
with only a tenfold dilution ‘few long oligomers can be detected’, p. 1.
62. See L. E. Orgel in Nature, vol. 358, pp. 203–209, 1992.
63. L. E. Orgel (1992; citation is in note 62), p. 204.
64. L. E. Orgel (1992; citation is in note 62), p. 207
65. The original set of experiments by Stanley Miller and Harold Urey was published in Science,
vol. 117, pp. 528–529, 1953, and Journal of the American Chemical Society, vol. 77,
pp. 2351–2361, 1955.
66. Overviews by S. L. Miller et al. may be found in Journal of Molecular Evolution, vol. 9,
pp. 59–72, 1976; Cold Spring Harbor Symposia on Quantitative Biology, vol. 52, pp. 17–27,
1987; and on pp. 59–85 of The molecular origins of life: assembling pieces of the puzzle,
edited by A. Brack (Cambridge University Press, Cambridge, 1998).
67. See, for example, R. Stribling and S. L. Miller in Origins of Life and Evolution of the
Biosphere, vol. 17, pp. 261–273, 1987.
68. S. L. Miller et al. (1976; citation is in note 66), p. 64.
69. Toxic and corrosive? Within our biochemical citadels we take oxygen for granted, but rust
and raging forest fires are a more immediate reminder of the potency of oxygen.
70. S. L. Miller (1987; citation is in note 66), p. 19.

71. S. L. Miller (1987; citation is in note 66), p. 19.
72. See W. L. Marshall in Geochimica et Cosmochimica Acta, vol. 58, pp. 2099–2106,
1994.
73. See J. P. Amend and E. L. Shock in Science, vol. 281, pp. 1659–1662, 1998, who calculate the
energetics of amino acid synthesis in hydrothermal settings, but note that ‘in all the
calculations presented here [there] is the assumption that the appropriate enzymes for each
step in the amino acid synthesis pathways are present and active’, p. 1660.
74. See J. P. Dworkin et al. in Proceedings of the National Academy of Sciences, USA, vol. 98,
pp. 815–819, 2001.
75. R. Shapiro (1987; citation is in note 16).
76. See M. Levy and S. L. Miller (1998; citation is in note 37). Concerning the many other
difficulties in the prebiotic synthesis of adenine, see the article by R. Shapiro in Origins of
Life and Evolution of the Biosphere, vol. 25, pp. 83–98, 1995.
77. These quotations are from pp. 182–184 of R. Shapiro (1987; citation is in note 16).
78. In R. F. Gesteland et al. (1998; citation is in note 3).
79. See G. F. Joyce et al. in Nature, vol. 310, pp. 602–604, 1984.
80. In R. F. Gesteland et al. (1998; citation is in note 3), p. 68.
81. In R. F. Gesteland et al. (1998; citation is in note 3), p. 72.
82. See, for example, C. M. Dobson et al. in Proceedings of the National Academy of Sciences,
USA, vol. 97, pp. 11864–11868, 2000, as well as V. R. Oberbeck et al. in Journal of Molecular
Evolution, vol. 32, pp. 296–303, 1991.
83. M. Eigen (1971; citation is in note 9), p. 519 (his emphasis).
84. See Iris Fry’s paper in Biology and Philosophy, vol. 10, pp. 389–417, 1995 (see also
note 89).
85. I. Fry (1995; citation is in note 84), p. 405.
86. K. D. James and A. D. Ellington (1995; citation is in note 40, Chapter 2), p. 528.
87. See p. 188 of P. Shapiro (1987; citation is in note 16).
88. See F. Crick’s Life itself: its origin and nature (Macdonald, London, 1982).
89. F. Crick (1982; citation is in note 88), p. 38.
90. F. Crick (1982; citation is in note 88), p. 88.
91. See George Wald’s article in Scientific American, vol. 191(2), pp. 44–53, 1954.
92. G. Wald (1954; citation is in note 91), p. 46.
5. uniquely lucky? the strangeness of earth

1. This chapter is an expansion of some brief remarks I made in Chapter 9 of The Crucible of
Creation: The Burgess Shale and the rise of animals (Oxford University Press, Oxford, 1998).
It also has many obvious parallels to Rare Earth: Why complex life is uncommon in the
Universe by P. D. Ward and D. Brownlee (Springer-Verlag, [Copernicus], New York, 2000).
Concerning the thesis of a rare Earth, this chapter is in broad agreement, although as far as
possible I have emphasized the areas that Ward and Brownlee dealt with more cursorily, thus
aiming to complement rather than duplicate their account.
2. Not surprisingly, given its relative proximity and especially the Apollo lunar missions, there
is an extensive literature on the Moon and its geological history. Two of the many useful
guides are: Lunar sourcebook: a user’s guide to the Moon, edited by G. Heiken et al.
(Cambridge University Press, Cambridge, 1991) and the chapter on the Moon by S. R. Taylor
on pp. 247–275 of Encyclopedia of the Solar System, edited by P. R. Weissman et al.
(Academic Press, San Diego, 1999).
3. The original observations are available (in Latin) on p. 276 of volume 1 of The historical works
of Gervase of Canterbury, edited by W. Stubbs (Her Majesty’s Stationery Office, London,
1879). A more accessible account may be found on pp. 50–51 of J. S. Lewis’s Rain of iron and
ice: the very real threat of comet and asteroid bombardment (Addison-Wesley [Helix Books],
Reading, MA, 1996). See also J. Hartung in Meteoritics, vol. 11, pp. 187–194, 1975, and Lunar
Science, vol. VII, pp. 348–350, 1976. This interpretation was, however, queried by H. H.
Nininger and G. I. Huss in Meteoritics, vol. 12, pp. 21–25, 1977. They suggested that it might
more plausibly be interpreted as a meteorite that had entered the Earth’s atmosphere, but
happened to pass in front of the Moon’s disc, so being mistaken for an impact. There is a
further problem because such a major collision would lead to the ejection of huge quantities
of rubble, which would easily escape from the weak gravitational field of Moon and
subsequently be captured by the Earth. As P. Withers pointed out (in Meteoritics & Planetary
Sciences, vol. 36, pp. 525–529, 2001), this would have produced a spectacular week-long
meteor ‘storm’, for which there appears to be no historical evidence. Unless conditions were
very overcast in England during June 1178, the date of the inferred impact, it is unlikely that
Gervase would not have made some comment. This is because although Gervase’s chronicles
are mostly to do with matters of Church and State, he evidently had an eye for natural
phenomena. These included eclipses and, more remarkably, an earthquake in 1158 that was
felt across England. This is immediately followed by a report that in London the Thames
dried out so that people could cross it dry-shod, but this apparently is an unrelated
observation.
4 See W. K. Hartman et al. on pp. 493–512 of the book Origin of the Earth and the Moon, edited
by R. M. Canup and K. Righter (University of Arizona Press, Tucson, 2000). In their judicious
review they stress that the standard account of lunar bombardment, especially the intense
major episode at c. 3.8–4.2 Ga cannot be accepted uncritically, and various alternatives need
to be kept in mind. Evidence consistent with the late heavy bombardment episode, based on
the ages of meteorites that have come from the Moon, is given by B. A. Cohen et al. in
Science, vol. 290, pp. 1754–1756, 2000, while D. A. Kring and B. A. Cohen, in Journal of
Geophysical Research, vol. 107 (E2), pp. 4-1–4-6, 2002, review the evidence for this
cataclysmic episode throughout the inner Solar System.
5. See K. J. Zahnle and N. H. Sleep in their chapter (pp. 175–208) of Comets and the origin of
evolution of life, edited by P. J. Thomas et al. (Springer, New York, 1997).
6. A. Morbidelli et al. in Meteoritics & Planetary Science, vol. 36, pp. 371–380, 2001, suggest
that important components of the impactors were actually relatively small bolides, rather
than the monsters usually envisaged, whose high-inclination orbits relative to the planets of
the inner Solar System imparted high impact velocities.
7. See S. Chang on pp. 10–23 of Early Life on Earth, Nobel Symposium No. 84, edited by
S. Bengtson (Columbia University Press, New York, 1994).
8. These details are taken from the paper by N. H. Sleep and K. Zahnle in Journal of
Geophysical Research, vol. 103, pp. 28529–28544, 1998.
9. See, for example, A. P. Nutman et al. in Geochimica et Cosmochimica Acta, vol. 61,
pp. 2475–2484, 1997; Chemical Geology, vol. 141, pp. 271–287, 1997; and Precambrian
Research, vol. 78, pp. 1–39, 1996, as well as C. M. Fedo and M. J. Whitehouse in Science,
vol. 296, pp. 1448–1452, 2002.
10. Claims for actual fossils, such as yeast-like organisms described by H. D. Pflug in
Naturwissenschaften, vol. 65, pp. 611–615, 1978, have won almost no support.
11. Claims for an organic origin of the carbon, such as by S. J. Mojzsis et al. in Nature, vol. 384,
pp. 56–59, 1996 (with a cautious commentary by J. M. Hayes on pp. 21–22), have been
criticized; see J. M. Eiler et al. in vol. 386, p. 665, 1997. See also Fedo and Whitehouse (2002,
citation is in note 9) and M. A. van Zuilen et al. in Nature, vol. 418, pp. 627–630, 2002.
R. Schoenberg et al., in Science, vol. 418, pp. 403–405, 2002, suggest some of the carbon may
have a meteoric origin; see also A. D. Anbar et al. in Journal of Geophysical Research,
vol. 106E, pp. 3219–3236, 2001.
12. The isotopic studies of δ 13 C (i.e. the ratio of 12 C to 13 C) in the Isua Supergroup were pioneered
by M. Schidlowski (see Nature, vol. 333, pp. 313–318, 1988), and are consistent with more
recent work by M. T. Rosing in Science, vol. 283, pp. 674–676, 1999.
13. Metamorphic gneisses from north-west Canada are dated at 3960 Ma by S. A. Bowring et al.
in Geology, vol. 17, pp. 971–975, 1989; mineral grains of zircons date back to c. 4400 Ma; see
S. A. Wilde et al. in Nature, vol. 409, pp. 175–178, 2001. Although these grains have been
reworked into younger rocks they still provide some indication of what the very early Earth
was like, and notably indicate the existence of some sort of ocean; see also S. J. Mojzsis et al.
in Nature, vol. 409, pp. 178–181, 2001.
14. See K. A. Maher and D. J. Stevenson in Nature, vol. 331, pp. 612–614, 1988; see also V. R.
Overbeck and G. Fogleman in vol. 339, p. 434, 1989, and N. H. Sleep et al. in vol. 342,
pp. 139–142, 1989. A more recent overview is given by K. J. Zahnle and N. H. Sleep (1997;
citation is in note 5).
15. Respective estimates are given by B. A. Cohen et al. (2000) and D. A. Kring and B. A. Cohen
(2002; citations in note 4).
16. See, for example, M. Gogarten-Boekels et al. in Origins of Life and Evolution of the
Biosphere, vol. 25, pp. 251–264, 1995; see also N. H. Sleep et al., in Proceedings of the
National Academy of Sciences, USA, vol. 98, pp. 3666–3672, 2001.
17. See V. R. Oberbeck and G. Fogleman in Origins of Life and Evolution of the Biosphere,
vol. 19, pp. 549–560, 1989; L. E. Orgel in a later issue (vol. 28, pp. 91–96, 1998) provides a
characteristically level-headed assessment of this topic.
18. See A. Lazcano and S. L. Miller in Journal of Molecular Evolution, vol. 39, pp. 546–554, 1994.
19. See D. M. Raup and J. W. Valentine in Proceedings of the National Academy of Sciences,
USA, vol. 80, pp. 2981–2984, 1983.
20. See K. O. Stetter on pp. 1–18 (including the discussion) of a Ciba Foundation Symposium
(no. 202) Evolution of hydrothermal ecosystems on Earth (and Mars?), edited by G. R. Bock
and J. A. Goode (Wiley, Chichester, 1996).
21. See, for example, N. Galtier et al. in Science (vol. 283, pp. 220–221, 1999), as well as earlier
remarks by P. Forterre, for example, in Cell (vol. 85, pp. 789–792, 1996) and Current Opinion
in Genetics & Development (vol. 7, pp. 764–770, 1997); see also note 49, Chapter 4.
22. See N. H. Sleep and K. Zahnle (1998; citation is in note 8).
23. See J. W. Head et al. in Science (vol. 286, pp. 2134–2137, 1999). Other evidence for a much
more equable early climate on Mars is given by J. B. Pollack et al. in Icarus, vol. 71,
pp. 203–224, 1987; F. Forget and R. T. Pierrehumbert in Science, vol. 278, pp. 1273–1276,
1997; R. M. Huberle in Journal of Geophysical Research, vol. 103E, pp. 28467–28479, 1998;
and M. A. Mischna in Icarus, vol. 145, pp. 546–554, 2000.
24. One such recent announcement is of a discovery from the Oman desert, announced by E.
Gnos et al. in Meteoritics & Planetary Sciences, vol. 37, pp. 835–854, 2002.
25. Citations are in notes 9–12, Chapter 3.
26. These data and the following information are derived from B. Gladman’s paper in Icarus,
vol. 130, pp. 228–246, 1997.
27. See P. Davies on pp. 304–317 (including discussion) of G. R. Bock and J. A. Goode (1996;
28. See R. M. E. Mastrapa et al. in Earth and Planetary Science Letters, vol. 189, pp. 1–8,
2001.
29. B. Gladman (1997; citation is in note 26) comments that the thickness of the Earth’s
atmosphere and the relatively small target size of Mars make both dispatch and destination
respectively more problematic. See also C. Mileikowsky et al. in Icarus, vol. 145,
pp. 391–427, 2000.
30. The exact quotation is ‘So I tell my students: learn your biochemistry here and you will be
able to pass examinations on Arcturus’, p. 16 of Life beyond Earth and the mind of Man,
edited by R. Berendzen, NASA Scientific & Technical Information Office, 1973.
31. The title of Norman Pace’s paper in Proceedings of the National Academy of Sciences, USA,
vol. 98, pp. 805–808, 2001.
32. N. Pace (2001; citation is in note 31), p. 806.
33. This is a very fast-moving field, with new results constantly being posted on websites:
www.exoplanets.org and www.obspm.fr/planets for example. Much of the information on
these extra-solar planets given below is derived from these sources. Published reviews include
those by J. T. Lunine in Proceedings of the National Academy of Sciences, USA, vol. 96,
pp. 5353–5356, 1999 and G. W. Marcy in his chapter (pp. 1285–1311) in Protostars and
planets, edited by V. Mannings et al. (University of Arizona Press, Tucson, 2000). Relevant
books include the fairly technical (S. Clark’s Extrasolar planets: The search for new worlds
(John Wiley [Praxis], Chichester, 1998); and, more accessible, D. Goldsmith’s Worlds
unnumbered: the search for extrasolar planets (University Science Books, Sausalito, CA,
1997).
34. See A. Vidal-Madjar et al. in Planetary and Space Science, vol. 46, pp. 629–648, 1998, who
suggests that gravitational perturbations and an occultation of the star hint at the presence of
giant planets within the disc.
35. Direct detection of a planet orbiting the star HD 209458 using photometry is reported by
D. Charbonneau et al. in Astrophysical Journal, vol. 529, pp. L45–L48, 2000. This planet had
already been detected by the radial velocity technique, but transit observations of the light
flux from the star provide other data. For example, it is significantly less dense than Saturn,
the least dense of the Solar System planets.
36. See R. P. Butler et al. in Astrophysical Journal, vol. 526, pp. 916–927, 1997, and D.
Goldsmith’s report in Science, vol. 296, p. 1951, 2002.
37. So far as I am aware a coherent consideration of life in a supergravity environment is not
available. The well-known essay On being the right size by J. B. S. Haldane (published in his
Possible worlds and other essays (Chatto and Windus, London, 1927)), as are various sections
in C. J. Pennycuick’s Newton rules biology: A physical approach to biological problems
(Oxford University Press, Oxford, 1992), see especially p. 53. The role of hypergravity (and its
reverse, weightlessness) has received extensive study with respect to rocket-borne ascent and
exploration of the Solar System; see, for example, Proceedings of the Annual Meetings of the
IUPS Commission on Gravitational Physiology, published as supplements to The
Physiologist (including vol. 23 (Suppl. 6), 1980; vol. 24 (Suppl. 6), 1981; vol. 26 (Suppl. 6), 1983;
vol. 28 (Suppl. 6), 1985; vol. 35 (Suppl. 1), 1992; vol. 36 (Suppl. 1), 1993). G. A. Cavagna et al. in
Journal of Physiology, vol. 528, pp. 657–668, 2000 (see also the commentary by A. E. Minetti
in Nature, vol. 409, pp. 467–468, 2001) assess the role of gravity on human walking, both at
lower values (equivalent to Mars) and at one-and-a-half times the Earth’s gravitational field.
38. See A. C. Cameron in Astronomy & Geophysics, vol. 43, pp. 4.21–4.25, 2002.
39. See, for example, J. E. Chambers and G. W. Wetherill in Icarus, vol. 136, pp. 304–327,
1998.
40. ‘Making more terrestrial planets’ is the title of J. E. Chambers’s paper in Icarus, vol. 152,
pp. 205–224, 2001.
41. J. E. Chambers (2001; citation is in note 40), p. 212.
42. J. E. Chambers (2001; citation is in note 40), p. 223.
43. See D. N. Lin et al. in Nature, vol. 380, pp. 606–607, 1996; see also J. C. B. Paploizou and
J. D. Larwood in Monthly Notices of the Royal Astronomical Society, vol. 315, pp. 823–833,
2000. In at least one case there is also compelling evidence that this process was both early
and relatively rapid; see J. I. Lunine in Proceedings of the National Academy of Sciences,
USA, vol. 98, pp. 809–814, 2001.
44. See G. Israelian et al. in Nature, vol. 411, pp. 163–166, 2001. The evidence is based on the
presence of a lithium isotope (6 Li), which is expected to occur in the early atmospheres of
giant planets, but is later destroyed as the star evolves.
45. See F. A. Rasio and E. B. Ford in Science, vol. 274, 954–956, 1996; S. J. Weidenschilling and
F. Marzari in Nature, vol. 384, pp. 619–621, 1996; N. Murray et al. in Science, vol. 279,
pp. 69–72, 1998; and F. Marzari and S. J. Weidenschilling in Icarus, vol. 156, pp. 570–579,
2002. In this last paper a ‘jumping Jupiter’ model is proposed to explain the ‘hot Jupiter’ close
to its sun and to infer the existence of a more distant planet forming a stable system arising
from a chaotic beginning.
46. S. J. Weidenschilling and F. Marzari (1996; citation is in note 45), p. 620.
47. For an interesting discussion of the possibility of habitable moons orbiting extra-solar giant
planets see D. M. Williams et al. in Nature, vol. 385, pp. 234–236, 1997 (and the
accompanying commentary by C. F. Chyba on p. 201, which in part is also a paean for Carl
Sagan).
48. G. Gonzalez et al. in Icarus, vol. 152, pp. 185–200, 2001.
49. See, for example, papers by K. K. Khurana et al. in Nature, vol. 395, pp. 777–780, 1998 (and
accompanying commentary on pp. 749, 751 by F. Neubauer); R. T. Pappalardo et al. in Journal
of Geophysical Research, vol. 104E, pp. 24015–24055, 1999; J. S. Kargel et al. in Icarus,
vol. 148, pp. 226–265, 2000; and M. G. Kivelson et al. in Science, vol. 289, pp. 1340–1343,
2000; and X-P. Wu et al. in Geophysical Research Letters, vol. 28, pp. 2245–2248, 2001. There
is also a possibility that some sort of sub-surface ocean exists on another Galilean satellite,
Callisto; see J. Ruiz in Nature, vol. 412, pp. 409–411, 2001, and the commentary by K. A.
Bennett on pp. 395–396.
50. See C. K. Gessmann and D. C. Rubie in Earth and Planetary Science Letters, vol. 184,
pp. 95–107, 2000, and A. G. W. Cameron in Meteoritics & Planetary Science, vol. 36,
pp. 9–22, 2001; the latter account provides a helpful introduction to the earlier literature.
51. See C. Alibert et al. in Geochimica et Cosmochimica Acta, vol. 58, pp. 2921–2926, 1994.
52. See the papers by M. T. McCulloch in Earth and Planetary Science Letters, vol. 126, pp. 1–13,
1994; C. J. Allègre and et al. in Geochimica et Cosmochimica Acta, vol. 59, pp. 1445–1456,
1995; M. Ozima and F. A. Podosek in Meteoritics & Planetary Sciences, vol. 33 (4, Suppl),
p. A120, 1998; and F. A. Podosek in Science, vol. 283, pp. 1863–1864, 1999.
53. See, for example, D-C. Lee et al. in Earth and Planetary Science Letters, vol. 198,
pp. 267–274, 2002.
54. An argument based on tungsten isotopes; see A. N. Halliday and D-C. Lee in Geochimica et
Cosmochimica Acta, vol. 63, pp. 4157–4179, 1997.
55. See R. M. Canup and E. Asphaug in Nature, vol. 412, pp. 708–712, 2001, as well as the
accompanying commentary by J. Melosh on pp. 694–695.
56. See A. G. W. Cameron and W. Benz in Icarus, vol. 92, pp. 204–216, 1991.
57. This paper, by A. G. W. Cameron, the fifth in a series exploring the so-called Giant Impact
Hypothesis, is in Icarus, vol. 126, pp. 126–137, 1997. For an update and overview see A. G. W.
Cameron’s paper in Meteoritics & Planetary Science, vol. 36, pp. 9–22, 2001.
58. See R. M. Canup and L. W. Esposito in Icarus, vol. 119, pp. 427–426, 1996.
59. R. M. Canup and L. W. Esposito (1996; citation is in note 58), p. 429.
60. See S. Ida et al. in Nature, vol. 389, pp. 353–357, 1997.
61. This is the title of a commentary on the paper by S. Ida et al. (1997; citation is in note 60),
published in the same issue of Nature, on pp. 327–328.
62. A. G. W. Cameron and W. Benz (1991; citation is in note 56).
63. A. G. W. Cameron and W. Benz (1991; citation is in note 56), p. 215.
64. See P. D. Ward and D. Brownlee (2000; citation is in note 1), pp. 222–234.
65. See J. Laskar and R. Robutel in Nature, vol. 361, pp. 615–617, 1993.
66. The principal exponent of these ideas is G. Williams, who has published extensively on this
topic, including a recent paper in Sedimentary Geology, vol. 120, pp. 55–74, 1998; see also
Nature, vol. 396, pp. 453–455, 1998.
67. See J. Laskar and P. Robutel, in Nature, vol. 361, pp. 608–612, 1993.
68. See Neil Comins’s intriguing book What if the Moon didn’t exist: voyages to Earths that
might have been (HarperCollins, New York, 1993), which in addition to the title’s
consideration looks at a series of other questions, such as what the Earth would be like if its
obliquity were much greater (as for Uranus), what might happen if a black hole met the Earth,
and if the Earth were smaller; this last topic is returned to briefly below.
69. See J. P. Vanyo and S. M. Awramik in Precambrian Research, vol. 29, pp. 121–142, 1985.
70. See W. Benz et al. in Icarus, vol. 74, pp. 516–528, 1988.
71. See, in particular, D. J. Stevenson in Annual Review of Earth and Planetary Sciences, vol. 15,
pp. 271–315, 1987.
72. Mars has two moons (Deimos and Phobos), but these are very small, respectively with
maximum lengths of 16 and 26 km. See P. C. Thomas on pp. 309–314 of Encyclopedia of the
Solar System, edited by P. R. Weissman (Academic Press, San Diego, 1999). These moons are
usually interpreted as captured asteroids, but, as Thomas notes, this is very difficult to
explain given their equatorial orbits. In addition, the long-term future of Phobos is uncertain,
and calculations suggest that its orbit will decay by tidal friction, with the moon crashing
onto the Martian surface within 100 Ma.
73. His stimulating article How common are habitable planets? may be found on pp. C11–C14 in
a special issue of Nature (vol. 402, 1999). N. F. Comins (see note 68) also explores in some
detail the consequences of living on a planet smaller than the Earth.
74. G. Gonzalez in Astronomy and Geophysics, vol. 40, pp. 318–320, 1999 explores some of these
points, remarking on both the roundness of the Moon (and the Sun) and the exactness of the
apparent sizes of each body (a tiny 3.7 arc minutes) allowing the Moon to block the bright
photosphere but not the dramatic chromosphere. Gonzalez also reminds us of the various
happy coincidences – the nature of the Sun, the Earth’s axial stability, the origin of the Moon
by impact, and the size of the Earth – that are probably prerequisites for astronomers to
evolve.
75. See pp. 210–213 of M. E. Bakich’s gazetteer to the Solar System, The Cambridge Planetary
Handbook (Cambridge University Press, Cambridge, 2000). See also D. H. Levy’s Impact
Jupiter: The crash of comet Shoemaker–Levy 9 (Plenum, New York, 1995).
76. George Wetherill’s paper ‘Possible consequences of absence of “Jupiters” in planetary
systems’ was published in Astrophysics and Space Science, vol. 212, pp. 23–32, 1994; see also
his commentary in Nature, vol. 373, p. 470, 1995. The role of Jupiter as ‘goal-keeper’ is also
stressed by Ward and Brownlee (2000; citation in note 1), see pp. 235–242.
77. See P. R. Weissman in Nature, vol. 344, pp. 825–830, 1990, as well as his earlier paper
(pp. 272–282) in Dynamics of the Solar System, edited by R. L. Duncombe (Reidel, Dordrecht,
1979).
78. G. Wetherill (1994; citation is in note 76), p. 23.
79. Two accessible introductions are W. Alvarez’s T. rex and the crater of doom (Princeton
University Press, Princeton, 1997) and C. Frankel’s The end of the dinosaurs: Chicxulub
crater and mass extinctions (Cambridge University Press, Cambridge, 1999).
80. See J. W. Kirchner and A. Weil in Nature, vol. 404, pp. 177–180, 2000, and commentary by D.
Erwin on pp. 129–130.
81. See the paper by J. K. Schubert and D. J. Bottjer in Geology, vol. 20, pp. 883–886, 1992, and the
summary by D. J. Bottjer et al. in Geological Society of London Special Publication, vol. 83,
pp. 7–26, 1995.
82. See P. B. Wignall and R. J. Twitchett in Sedimentology, vol. 46, pp. 303–316, 1999. Other
parallels to sedimentary features typical of the earlier Precambrian environment might
include large-scale precipitation of sea-floor carbonate cements; see A. D. Woods et al. in
Geology, vol. 27, pp. 645–648, 1999.
83. See S. D’Hondt et al. in Science, vol. 282, pp. 276–279, 1998.
84. See the reports by G. A. Logan et al. in Nature, vol. 376, pp. 53–56, 1995, with an
accompanying commentary on pp. 16–17 by M. Walter; and in Geochimica et Cosmochimica
Acta, vol. 61, pp. 5391–5409, 1997.
85. See J. S. Lewis (1996; citation is in note 3), Chapter 14.
86. Concerning displacement of bodies from the asteroid belt see, for example, the papers by
P. Farinella et al. in Nature, vol. 371, pp. 314–317, 1994 and B. J. Gladman et al. in Science,
vol. 277, pp. 197–201, 1997.
350 notes to pp. 96–100
87. See, for example, the interesting studies by J. J. Sepkoski on pp. 211–255 of Evolutionary
paleobiology: in honor of James W. Valentine, edited by D. Jablonski et al. (University of
Chicago Press, Chicago; 1996), and with others in Paleobiology, vol. 26, pp. 7–18, 2000.
88. Concerning the final destiny of the Earth and the Solar System see the discussion by K. P.
Rybicki and C. Denis in Icarus, vol. 151, pp. 130–137, 2001. See also notes 105–107.
89. Evidence that the bulk of the Earth’s water derives from the outer asteroid belt is addressed by
A. Morbidelli et al. in Meteoritics & Planetary Science, vol. 35, pp. 1309–1320, 2000.
90. See J. I. Lunine (2001; citation is in note 43).
91. See C. F. Chyba and C. Sagan on pp. 147–173 of Comets and the origin and evolution of life,
edited by P. J. Thomas et al. (Springer, New York; 1997), and A. H. Delsemme in Advances in
Space Research, vol. 15(3), pp. 49–57, 1995.
92. C. F. Chyba in Advances in Space Research, 15 (3), pp. 45–48, 1995.
93. See papers by Z. Sekanina in Icarus, vol. 27, pp. 123–133, 1976, and more especially T. A.
McGlynn and R. D. Chapman in The Astrophysical Journal, vol. 346, pp. L105–L108,
1989.
94. Evidence for this is discussed by A. Vidal-Madjar et al. in one of their series of papers on β
Pictoris, specifically in Astronomy and Astrophysics (vol. 290, pp. 245–258, 1994) (see also in
the same journal A. M. Lagrange-Henri et al., vol. 190, pp. 275–282, 1988, and H. Beust et al.,
vol. 236, pp. 202–216, 1989).
95. See S. A. Stern et al. in Nature, vol. 345, pp. 305–308, 1990.
96. See S. A. Stern et al. in Icarus, vol. 91, pp. 65–75, 1991; this search entailed 17 nearby stars
and the authors emphasize the problems of detectability with present technology.
97. See G. J. Melnick et al. in Nature, vol. 412, pp. 160–163, 2001.
98. Although Chyba (1995; citation is in note 92) reminds us that it is possible to envisage a solar
system whose inner planets have received volatile delivery, but lacks an equivalent of the
Oort Cloud.
99. Three such episodes have been identified in the Precambrian, one about 2200 Ma ago (see, for
example, D. A. Evans et al. in Nature, vol. 386, pp. 262–266, 1997; D. McB. Martin in
Geological Society of America Bulletin, vol. 111, pp. 189–203, 1999; and J. L. Kirschvink
et al. in Proceedings of the National Academy of Sciences, USA, vol. 97, pp. 1400–1405,
2000), and two at c. 700 Myr and 600 Myr ago (see P. Hoffman et al. in Science, vol. 281,
pp. 1342–1346, 1998, and supporting blasts in Nature, vol. 397, p. 384, 1999 and vol. 400,
p. 708, 1999; but see also G. S. Jenkins and C. R. Scotese in Science, vol. 282, pp. 1644–1645,
1998 (response by Hoffman et al. follows in pp. 1645–1646), G. S. Jenkins and L. A. Frakes in
Geophysical Research Letters, vol. 25, pp. 3525–3528, 1998; W. T. Hyde et al. in Nature,
vol. 405, pp. 425–429, 2000; and R. A. Kerr in Science, vol. 287, pp. 1734–1736, 2000; and M. J.
Kennedy et al. in Geology, vol. 29, pp. 443–446, 2001).
100. See H. C. Jenkyns and P. A. Wilson in American Journal of Science, vol. 299, pp. 341–392,
1999.
101. See, for example, James Lovelock’s Gaia: A new look at life on Earth (Oxford University
Press, Oxford, 1979) and The ages of Gaia: A biography of our living Earth (Oxford University
Press, Oxford, 1988).
102. See, for example, Michael Hart’s paper in Icarus, vol. 37, pp. 351–357, 1979.
103. A topic explored by J. Laskar in Nature, vol. 338, pp. 237–238, 1989.
104. See J. F. Kasting et al. in Icarus, vol. 101, pp. 108–138, 1993 and J. F. Kasting in Origins of Life
and Evolution of the Biosphere, vol. 27, pp. 291–307, 1997.
105. See S. Franck et al. in Naturwissenschaften, vol. 88, pp. 416–426, 2001; see also the papers by
S. Franck et al. in Journal of Geophysical Research, vol. 105E, pp. 1651–1658, 2000 and
Planetary and Space Science, vol. 48, pp. 1099–1105, 2000 (see also note 88).
106. See the papers by S. Franck et al. in Chemical Geology, vol. 159, pp. 305–317, 1999, and
Tellus, vol. 52B, pp. 94–107, 2000.
107. Franck et al. (2001; citation is in note 105), p. 420. One should note that they qualify this by
remarking that chemolithotrophic hyperthermophiles might persist for longer but ‘all higher
notes to p. 100–107 351
forms of life would certainly be eliminated’. Not all authors are quite so pessimistic;
T. M. Lenton and W. von Bloh (in Geophysical Research Letters, vol. 28, pp. 1715–1718, 2001)
give an extra 300 Ma (and perhaps a little more).
108. See C. H. Lineweaver in Icarus, vol. 151, pp. 307–313, 2001.
109. See for example, G. Gonzalez et al. in the Astrophysical Journal, vol. 511, pp. L111–L114,
1999. Although this correlation of metallicity and the presence of planets is strong, it is not
inevitable. In Astronomy & Geophysics (vol. 40, pp. 5.25–5.29, 1999) G. Gonzalez has also
suggested that in a number of respects, including its metallicity, our Sun may be anomalous.
However, I. N. Reid (in Publications of the Astronomical Society of the Pacific, vol. 114,
pp. 306–329, 2002) suggests that while there is a strong correlation between stellar metallicity
and the presence of planets, our Sun falls in the mid-range of such metallicities.
110. However, S. J. Vogt et al. report (in Astrophysical Journal, vol. 536, pp. 902–914, 2000) six
new planetary systems, of which two are metal poor.
111. See, for example, R. Jimenez et al. in Astrophysical Journal, vol. 561, pp. 171–177, 2001.
112. See the article ‘How old is ET?’ by R. P. Norris on pp. 103–105 of When SETI succeeds: The
impact of high-information contact, edited by A. Tough (Foundation for the Future, Bellevue
WA, 2000).
113. R. P. Norris (2000; citation is in note 112), p. 105.
114. See James Annis’s article in Journal of the British Interplanetary Society, vol. 52, pp. 19–22,
1999.
115. J. Annis (1999; citation is in note 114), p. 22.
116. See the paper by G. Gonzalez et al. in Icarus, vol. 152, pp. 185–200; 2001a, as well as a more
popular overview in Scientific American, vol. 285 (4), pp. 52–59, 2001b. In the context of a
special position of the Solar System with respect to our galaxy, see also the earlier paper by
L. S. Marochnik in Astrophysics and Space Science, vol. 89, pp. 61–75, 1983.
117. G. Gonzalez et al. (2001b; citation is in note 116), p. 59, left-hand column.
118. G. Gonzalez et al. (2001b; citation is in note 116), p. 59, right-hand column.
6. converging on the extreme

1. This conceit is taken from my article in Proceedings of the Royal Institution of Great
Britain, vol. 70, pp. 21–62, 1999, see p. 24.
2. See The crucible of creation: The Burgess Shale and the rise of animals (Oxford University
Press, Oxford, 1998).
3. This has not prevented, of course, speculation of possible alternatives. In terms of biology
there is D. Dixon’s entertaining After Man: A zoology of the future (London, Granada; 1981).
Not surprisingly most emphasis, however, has been on historical counterfactuals, such as If
it had happened otherwise: lapses into imaginary history (Longmans, Green, London, 1931),
Virtual history: alternatives and counterfactuals (Macmillan, London, 1998) and What if,
military historians imagine what might have been (London, Pan, 2001) edited respectively by
J. C. Squire, N. Ferguson, and R. Cowley. The realm of alternative histories has also attracted
novelists, such as Kingsley Amis in his witty The alteration (Penguin, London, 1988), and the
chilling The sound of his horn (Tartarus, Horam, 1999) by Sarban (John William Wall). Others
have considered how past events might be manipulated; of these, Orson Scott Card’s poignant
Pastwatch: the redemption of Christopher Columbus (Tom Doherty, New York, 1996) is a
haunting example of a world that might have been.
4. After a rather dull period, matters have become quite polarized between enthusiasts for the
Search for Extra-Terrestrial Intelligence (SETI) engaged in debate with the sceptics, which
include not only myself but such individuals as I. Crawford; see part 4, pp. 24–26, and part 6,
p. 19 in Astronomy & Geophysics, vol. 38, 1997, as well as Scientific American, vol. 283 (1),
pp. 28–33, 2000.
5. A. Y. Mulkijanian and W. Junge, in Photosynthesis Research, vol. 51, pp. 27–42, 1997, suggest
that the origins of photosynthesis lie in protective measures of pigmented proteins against
352 notes to pp. 107–109
ultraviolet radiation. Given that this UV flux will be the norm on primitive planets, and the
fact that such building blocks of chlorophyll as quinones are synthesized in interstellar space
(see Chapter 3), so chlorophyll may be a universal molecule (note 13).
6. The general consensus seems to be that this endosymbiotic capture of the
cyanobacteria/chloroplast (and the comparable case of α-proteobacteria captured to yield in
due course the mitochondria) were unique events (see, for example, A. J. Roger and C. F.
Delwiche respectively in American Naturalist, vol. 154, pp. S146–S163 and S164–S177, 1999;
but see J. W. Stiller et al. in Journal of Phycology, vol. 39, pp. 95–105, 2003). Even so, although
rare, there are a number of other cases of what is called secondary endosymbiosis. This occurs
when two separate eukaryotes (or at least the plastid component) combine to form a single
organism. See, for example, papers by S. P. Gibbs in Annals of the New York Academy of
Sciences, vol. 361, pp. 193–207, 1981, S. E. Douglas et al. in Nature, vol. 350, pp. 148–151,
1991; S. Köhler et al. in Science, vol. 275, pp. 1485–1489, 1997, Z-d. Zhang et al. in Nature,
vol. 400, pp. 155–159, 1999; K. Takishita and A. Uchida in Phycological Research, vol. 47,
pp. 207–216, 1999, and Journal of Molecular Evolution, vol. 51, pp. 26–40, 2000 respectively;
and J. F. Saldarriaga et al. in Journal of Molecular Evolution, vol. 53, pp. 204–213, 2001 (who
also present evidence for multiple loss of plastids). Helpful overviews of this area are provided
by G. McFadden and P. Gilson in Trends in Ecology & Evolution, vol. 10, pp. 12–17, 1995 and
T. Cavalier-Smith in Trends in Plant Sciences, vol. 5, pp. 174–182, 2000. Finally, and to
reinforce the likelihood that symbiotic unions between species are, in the grand order of
things, overwhelmingly probable, see the case in the insects where the symbiotic bacteria
have themselves symbiotic bacteria; see C. D. von Dohlen et al. in Nature, vol. 412,
pp. 433–436, 2001 (see also note 48).
7. Interestingly, the first major glaciations appear to date from the same interval (see, for
example, J. L. Kirschvink et al. in Proceedings of the National Academy of Sciences, USA,
vol. 97, pp. 1400–1405, 2000). One possibility is that as levels of oxygen increased so the
amount of reactive methane, a powerful greenhouse gas, declined. This in turn led to
planetary cooling and ultimately severe, perhaps catastrophic, glaciation (see A. A. Pavlov
et al. in Journal of Geophysical Research, vol. 105E, pp. 11981–11990, 2000).
8. See, however, Chapter 5 (pp. 72–73) for a more extended discussion of the significance of the
oldest sedimentary carbon.
9. The third carbon isotope, carbon 14 (14 C), is radioactive and is produced principally by cosmic
rays striking atoms of nitrogen. The relatively short half-life of 14 C means that it plays no
effective part in the biochemistry of photosynthesis and the geological burial of carbon.
10. As ever, this statement requires scientific qualification. Fractionation of carbon isotopes,
denoted by δ 13 C (the ratio of 12 C to 13 C against an agreed standard that by definition has a
δ 13 C of zero), occurs in other biochemical systems, such as methane generation by some
bacteria (methanogens). To the first approximation different metabolic processes imprint
separable isotopic signatures, with photosynthetic fractionation averaging c. 27‰ δ 13 C.
11. For helpful reviews of RuBisCO see the articles by F. C. Hartman and M. R. Harpel in Annual
Review of Biochemistry, vol. 63, pp.197–234, 1994 and G. Schneider et al. in Annual Review
of Biophysics and Biomolecular Structure, vol. 21, pp. 119–143, 1992. Some interesting
aspects of its molecular structure and possibly severe constraints on its function are addressed
by E. A. Kellogg and N. D. Juliano in American Journal of Botany, vol. 84, pp. 413–428,
1997.
12. The prokaryotic alga Acaryochloris lives within the tissues of tunicates and contains
significant proportions of chlorophyll d; see H. Miyashita et al. in Nature, vol. 383, p. 402,
1996, H. Schiller et al. in FEBS Letters, vol. 410, pp. 433–436, 1997. Concerning a similar
feature in the green alga Ostreobium, see P. Halldal in Biological Bulletin, vol. 134,
pp. 411–424, 1968, and B. Koehne et al. in Biochimica et Biophysica Acta, vol. 1412,
pp. 94–107, 1999. In part, at least, the red-shift is due to the shaded habitats of these
photosynthesizers.
notes to pp. 110–112 353
13. This quotation (pp. 13–14) is taken from George Wald’s stimulating and thoughtful paper
‘Fitness in the Universe: Choices and necessities’ in Origins of Life and Evolution of the
Biosphere, vol. 5, pp. 7–27, 1974.
14. R. D. Wolstencroft and J. A. Raven, in Icarus, vol. 157, pp. 535–548, 2002, have an interesting
discussion of how likely it is that photosynthesis has evolved on remote planets, and the
degree of similarity there might be to Earth. They conclude that the development of an
oxygenated atmosphere is very likely, although the energetics of photon capture and
specifically the number required to reduce a molecule of CO2 might differ.
15. D. J. DeRosier gives a helpful introduction to the flagellar motors of bacteria in Cell, vol. 93,
pp. 17–20, 1998, as do R. M. Berry and J. P. Armitage in Advances in Microbial Physiology,
vol. 41, pp. 291–337, 1999. See also in a thematic series ‘The molecular physics of biological
movement’, in Philosophical Transactions of the Royal Society of London B, vol. 355, 2000
the papers by H. C. Berg (pp. 491–501) and R. M. Berry (pp. 503–509). Rotary activity is also
reported in eukaryotic cells, albeit involving the microtubules within the flagellum; see
C. K. Omoto and G. B. Witman in Nature, vol. 290, pp. 708–710, 1981.
16. More typical speeds are of the order of 200 rev/second, but Y. Magariyama et al. report speeds
of up to 1700 rev/second in the bacterium Vibrio alginolyticus; see Nature, vol. 371, p. 752,
1994.
17. See, for example, the account of how the filament is constructed given by K. Yonekura et al. in
Science, vol. 290, pp. 2148–2152, 2000, with an accompanying commentary by R. M. Macnab
on pp. 2086–2087, where he remarks that the assembly of this structure by the bacterium ‘is a
much more sophisticated process than any of us could have envisaged’, p. 2087.
18. See, in particular, M. J. Behe’s Darwin’s black box: The biochemical challenge to evolution
(The Free Press, New York, 1996). Unsurprisingly, the views of Behe have come under fierce
and justified criticism. As might be expected, orthodox Darwinists are willing to wade in, but
so are those who are keen to reconcile science and religion: see, in particular, K. R. Miller’s
thoughtful Finding Darwin’s God: A scientist’s search for common ground between God and
evolution (HarperCollins [Cliff Street Books], New York, 1999), especially Chapter 5.
19. This is F1 -ATPase, a component of the enzyme ATP synthase, which is located in the
mitochondria. See K. Kinosita et al. in Cell, vol. 93, pp. 21–24, 1998, two papers in
Philosophical Transactions of the Royal Society of London B, vol. 355, 2000, by A. G. W.
Leslie and J. E. Walker (pp. 465–472) and K. Kinosita et al. (pp. 473–489) respectively; and
M. L. Hutcheon et al. in Proceedings of the National Academy of Sciences, USA, vol. 98,
pp. 8519–8524, 2001.
20. See the paper Locomotion like a wheel? by R. Full et al. in Nature, vol. 365, p. 495, 1993. An
analogy of a kind can also be found in the aquatic larvae of mosquitoes, which execute a
series of somersaults and still travel in a straight line through the water; see J. Brackenbury in
Journal of Experimental Biology, vol. 202, pp. 2521–2529, 1999 for a revealing description of
this machine-like behaviour.
21. One should also note the convergences between the pangolin (an Old World group) and the
New World xenarthrans, which include the anteaters and armadillos; see F. Delsuc et al. in
Proceedings of the Royal Society of London B, vol. 268, pp. 1605–1615, 2001, where their
phylogenetic analysis ‘confirms that morphological similarities between xenarthrans and
pangolins are a spectacular example of convergence in relation to myrmecophagy [ant-eating]’
(p. 1612). The extent of the similarities between pangolins and xenarthrans is emphasized by
K. Z. Reiss in American Zoologist, vol. 41, pp. 507–525, 2001, where she has an interesting
discussion of the nature of convergences in both a phylogenetic and adaptational framework.
Importantly she comments that ‘there is no evidence that phylogenetic constraints have
facilitated or prevented the evolution of cranial muscle specializations for myrmecophagy’
(p. 520). These conclusions on convergence echo earlier work by K. D. Rose and R. J. Emry on
pp. 81–102 of Mammal phylogeny: Placentals, edited by F. S. Szalay et al. (Springer,
New York, 1993). Such critical assessments are of considerable importance in phylogenetics,
354 notes to p. 112
because prior analysis executed by hard-line cladists had strongly supported a link between
the pangolins and xenarthrans. As Rose and Emry comment, a credulous oversimplification
of ‘character states to facilitate algorithm-driven analyses often leads to unwarranted
confidence in some phylogenetic conclusions – for instance, the conclusion of synapomorphy
where homoplasy is more probable’ (p. 82). In the context of convergence among the anteaters
it is also worth remarking that a number of fossil forms, variously assigned to ant-eating
groups, may well be mistakenly allied because of unappreciated convergences in form and
function. Thus, Ernanodon from the Paleocene of China is probably convergent on the
xenarthrans (Rose and Emry 1993, p. 96; see also T. J. Gaudin in Fieldiana: Geology, vol. 41
(new series), pp. 1–38, 1999), while the much-discussed Eurotamandua from the famous
Messel oil shale (and also the Geiseltal lignite) of Eocene age in Germany poses serious
problems concerning the biogeographic dispersal of the xenarthran anteaters, problems that
as Delsuc et al. (2001) remark can be resolved if ‘the striking morphological resemblances
between this taxon [Eurotamandua] and Myrmecophagidae [anteaters] might be the result of
adaptive convergence towards fossoriality and ant feeding. This example once again
underlines how morphological adaptation to similar ecological niches could be positive
misleading in terms of phylogenetic signal’ (p. 1613).
22. A useful escape reaction, not least because the aboriginal tribe regards the pangolin as a great
delicacy. Details (of the escape tactics, not the recipe) are given by R. R. Tenaza in Journal of
Mammalogy, vol. 56, p. 257, 1975.
23. These examples are mentioned by R. L. Caldwell in Nature, vol. 282, pp. 71–73, 1979, a paper
whose principal purpose is to draw attention to the backward somersaulting abilities of the
stomatopods (see also note 20).
24. G. A. Gill and A. G. Coates, in Lethaia, vol. 10, pp. 119–134, 1977 remark, it is less clear
whether rolling of these more or less spherical coral colonies is achieved only when nudged
by another animal, say a fish, or even whether the colony remains in situ and grows by
periodic ‘press-ups’; see also J. B. Lewis’s discussion of spherical Siderastrea (S. radians) in
Coral Reefs 7, pp. 161–167, 1989.
25. LaBarbera’s paper is entitled ‘Why the wheels won’t go’, published in American Naturalist,
vol. 121, pp. 395–408, 1983. Also of interest are the remarks in ‘Why legs and not wheels’ by I.
Walker in Acta Biotheoretica, vol. 39, pp. 151–155, 1991, although her emphasis is more on
the contrast between biological and man-made systems.
26. Philip Pullman, in the last book of his neo-secular fantasy, The amber spyglass (Scholastic,
London, 2000), provides an ingenious portrayal of the wheeled malefa. Their wheels, however,
are derived from seed-pods, and the axle is effectively formed by a powerful claw with an
extending spur that locks into the pod. The reliance on the trees and their pods leads to an
ecological interdependence that provides Pullman with some graphic and imaginative writing.
27. See R. A. Fortey’s book Life: An unauthorized biography (HarperCollins, London, 1997).
28. R. A. Fortey (1997; citation is in note 27), p. 325.
29. R. A. Fortey was not the first to speculate on the existence of giant floating organic bladders.
Carl Sagan, for example, in Cosmos (Macdonald, London, 1981, see pp. 40–43) conjures up
what he dubs ‘floaters’ living high in the turbulent atmosphere of Jupiter-like planets. They
would, he proposed, either consume organic material synthesized abiotically in the reducing
atmosphere (in a Miller–Urey like fashion, see Chapter 4, pp. 60–62) or make their own food
in a way analogous to photosynthetic plants. These organisms, unlike the terrestrial Fortean
bladders, had, of course, a ready supply of hydrogen. Jupiter, too, is conceivably an abode of
life, but as R. D. MacElroy stresses in his chapter (pp. 69–93) in Chemical evolution of the
giant planets, edited by C. Ponnamperuma (Academic Press, New York, 1976) there are very
considerable difficulties. So far as Fortean bladders are concerned, the extreme atmospheric
turbulence would seem to make their survival problematic. As MacElroy writes, ‘We surmise
that the origin of life on Earth required surfaces and some degree of stability. Stability appears
not to be a general characteristic of Jupiter, except for the Great Red Spot. Postulating
notes to pp. 112–113 355
surfaces with stability in a turbulent atmosphere requires imagination and faith, but there is
historical precedence for the use and overuse of these qualities. Fanciful description of life in
places not visited has been one of the most wonderful and revealing aspects of man’s history.
Wonderful, because the descriptions show a basic belief that anything imaginable is possible;
revealing, because of the fundamental acceptance that the natural order of the Universe is to
be populated with live creatures. Inevitably, it is harshly revealed that the historical
descriptions of such magnificent beasts as unicorns and mermaids have been the result of
naı̈veté and the absence of facts. A biologist considering the possibilities of life on other
planets and their satellites is the most naı̈ve of scientists’ (p. 83).
30. See also my article ‘Palaeodiversifications: mass extinctions, ‘clocks’, and other worlds’ in
GeoBios, vol. 32, pp. 165–174, 1999.
31. Useful accounts of this organ are given by R. McN. Alexander in Biological Reviews, vol. 41,
pp. 141–176, 1966, and J. B. Steen on pp. 413–443 in vol. 4 of Fish Physiology, edited by W. S.
Hoar and D. J. Randall (Academic Press, New York, 1970).
32. G. N. Lapennas and K. Schmidt-Nielsen provide a helpful review in Journal of Experimental
Biology, vol. 67, pp. 175–196, 1977. The crystals of guanine are very thin, and when the
swim-bladder contracts they can strongly fold.
33. See his review on pp. 70–72 of Marine biology: its accomplishment and future prospect,
edited by J. Mauchline and T. Nemoto (Hokusen-Sha, Tokyo, 1991). Guanine, which is one of
the four bases of DNA, also acts as a reflective agent on the surface of fish, as well as in the
eyes of both sharks and scallops; see E. J. Denton in Philosophical Transactions of the Royal
Society of London B, vol. 258, pp. 286–313, 1970.
34. Counter-current systems are widespread and convergent in animals, especially those
involving heat exchange and thermoregulation. A well-known example is in long-legged
birds, whereby the feet, perhaps perched on the ice of a frozen lake, are markedly colder than
the rest of the body because the heat carried in the blood descending the legs is transferred by
the counter-current system to the ascending stream.
35. This interesting octopus is described by A. Packard and M. Wurtz in Philosophical
Transactions of the Royal Society of London B, vol. 344, pp. 261–275; 1994. More
speculatively A. Seilacher and M. LaBarbera have proposed (in Palaios, vol. 10, pp. 493–506,
1995) that the buoyancy of the ammonites, whose closest living relatives are squids,
depended upon a bladder-like arrangement with an associated gas gland. Seilacher and
LaBarbera suggest that the last-formed chamber retained a flexible wall, only later to calcify
as a subsequent chamber was delimited. In association with a putative gas gland this would
have allowed it to act as an analogue of the swim-bladder.
36. Concerning the siphonophores see the review by G. O. Mackie et al. in Advances in Marine
Biology, vol. 24, pp. 97–262, 1987.
37. Details of the gas gland are provided by D. E. Copeland in Biological Bulletin, vol. 135,
pp. 486–500, 1968. He suggests that carbon monoxide is produced because it is less markedly
soluble than carbon dioxide and so ‘more readily retained by the highly hydrated float tissue’
(p. 497). Other siphonophores also secrete carbon monoxide (see, for example, G. V. Pickwell
et al. in Science, vol. 144, pp. 860-862, 1964), as does the bladder kelp, an enormous brown
alga; see G. B. Rigg and B. S. Henry in American Journal of Botany, vol. 22, pp. 362–365, 1935.
In this seaweed, Nereocystis, the distal end bears a float with a capacity of four litres; see N.
L. Nicholson in Journal of Phycology, vol. 6, pp. 177–182, 1970. There is another interesting
convergence because as Rigg and Henry (1935) also remark ‘The inner surface of this float is
covered with a cobweb-like layer of tubular cells which bear such a striking resemblance to
the sieve tubes of higher plants in both structure and function that they [too] are commonly
referred to as sieve tubes’ (p. 363). Various other convergences between these giant brown
algae and other plants are addressed on pp. 19–20 of K. J. Niklas’s Plant biomechanics: An
engineering approach to plant form and function (University of Chicago Press, Chicago,
1992).
356 notes to pp. 113–114
38. A detailed description of this siphonophore is given by P. R. Pugh in Transactions of the

Royal Society of London B, vol. 301, pp. 165–300, 1983. Thermopalia is very delicate, but
spends most of its time attached to the basaltic sea floor like a tethered balloon, anchored by
its float. Pugh provides a helpful review of siphonophore floats, and notes that the
mechanisms whereby the respiratory processes in the adjacent tissues are not poisoned by the
carbon monoxide are not really understood, although he suggests that the chitinous lining
provides a barrier to diffusion.
39. See the review by A. E. Walsby in Microbiological Reviews, vol. 58, pp. 91–144, 1994.
40. The gas helium, which is much preferable to hydrogen for reasons of safety, is much less
likely to be employed as it is difficult to imagine a way of concentrating the trace amounts
leaking out of the interior of Earth; see K. A. Farley and E. Neroda in Annual Review of Earth
and Planetary Sciences, vol. 26, pp. 189–218, 1998.
41. See the paper by R. Conrad in Advances in Microbial Ecology, vol. 10, pp. 231–283, 1988.
42. The unexpectedly high production of hydrogen by microbial mats, and the possible
consequences for changes in the composition of the atmosphere of the early Earth (see also
note 7), is described by T. M. Hoehler et al. in Nature, vol. 412, pp. 324–327, 2001, with the
accompanying commentary by B. B. Jorgensen on pp. 286–287, 289.
43. Organelles are tiny structures in the eukaryotic cell, and are probably most familiar in the
form of chloroplasts (containing chlorophyll), which are derived from the endosymbiosis of
cyanobacteria-like cells (see note 6) and mitochondria, also of endosymbiotic origin.
44. Hydrogenosomes are probably derived from mitochondria, and are employed in anaerobic
environments by such single-celled organisms as ciliates, various flagellate protistans, and
chytridiomycete fungi; see M. Müller in Journal of General Microbiology, vol. 139,
pp. 2879–2889, 1993; T. M. Embley et al. in Proceedings of the Royal Society of London B,
vol. 262, pp. 87–93, 1995 and Trends in Ecology & Evolution, vol. 12, pp. 437–441, 1997; J. H.
P. Hackstein et al. in Trends in Microbiology, vol. 7, pp. 441–447, 1999; and M. Benchimol in
Tissue & Cell, vol. 32, pp. 518–526, 2000. Because hydrogenosomes generally lack a genome,
establishing their origins has not been straightforward, but in one ciliate the evidence for a
mitochondrial origin is now very strong; see A. H. A. M. van Hoek et al. in Molecular Biology
and Evolution, vol. 17, pp. 202–206, 2000. Hydrogenosomes must have evolved on numerous
occasions; J. D. Palmer in Science, vol. 275, pp. 790–791, 1997, writes, ‘Hydrogenosomes are a
spectacular example of the repeated evolution of biochemically similar organelles,’ p. 790.
45. Concerning hydrogenase, see the review by L. Casalot and M. Rousset in Trends in
Microbiology, vol. 9, pp. 228–237, 2001. So, too, in an article on the origin and evolution of
hydrogenase, D. S. Horner et al. (in Molecular Biology and Evolution, vol. 17, pp. 1695–1709,
2000) remark, ‘The extraordinary capacity of eukaryotes to repeatedly evolve organelles
capable of producing hydrogen (hydrogenosomes) is a fascinating biological puzzle,’ p. 1706.
46. Thus hydrogenase has been found in the green algae Chlamydomonas; see the papers in
European Journal of Biochemistry by T. Happe and J. D. Naber, and T. Happe et al. in
vol. 214, pp. 475–481, 1993 and vol. 222, pp. 769–774, 1994, respectively.
47. A helpful summary on the vestimentiferans and their remarkable metabolism is given by C.
L. van Dover in Chapter 7 of The ecology of deep-sea hydrothermal vents (Princeton
University Press, Princeton, 2000); see also C. Arndt et al. in Journal of Experimental
Biology, vol. 204, pp. 741–750, 2001.
48. Concerning the associations between insects and symbiotic bacteria see, for example, E. J.
Houk and G. W. Griffiths in Annual Review of Entomology, vol. 25, pp.161–187, 1980, and N.
A. Moran and A. Telang in BioScience, vol. 48, pp. 295–304, 1998. Moran and Telang
emphasize that such insect–bacterial associations have evolved a number of times
independently.
49. Sure enough, the collected issue of the Fortean Times 42–46, edited by P. Sieveking, is
entitled If pigs could fly (J. Brown, London, 1994). The title is based on a brief report in
volume 45, p. 36, and involves the unexpected discovery of a giant turd on the first-floor
notes to pp. 114–116 357
balcony of a dwelling in the Huntington Beach area, California. The report continues, ‘a vet
analysed a sample and concluded it was the work of a huge pig’, which presumably had
detached itself from the passing flotilla in order to relieve itself.
50. The problem of envisaging a biological equivalent to an airship is addressed in his Cat’s paws
and catapults: Mechanical worlds of nature and people (Norton, New York, 1998); see
especially pp. 151–152. S. Vogel’s entertaining and informative books on biomechanics are
very well worth reading. So, too, is his succinct chapter ‘Convergence as an analytical tool in
evolutionary design’ (pp. 13–20) in Principles of animal design: The optimization and
symmorphosis debate, edited by E. R. Weibel et al. (Cambridge University Press, Cambridge,
1998).
51. See, for example, the paper by R. A. J. Taylor and D. Reling in Environmental Entomology,
vol. 15, pp. 431–435, 1986. To be sure, there are local concentrations of an ‘aerial plankton’,
especially in connection with atmospheric convergences, but as V. A. Drake and R. A. Farrow
(in Trends in Ecology & Evolution, vol, 4, pp. 381–385, 1989) emphasize, these are temporary,
and are usually associated with migration. Locusts are perhaps the most familiar example.
52. An excellent introduction to the spiders, with a helpful description of the complexity of the
silk glands and associated apparatus, is given in R. F. Foelix’s Biology of spiders (Oxford
University Press, Oxford, 1996 [Second edition]).
53. See the article by J. W. Schultz in Biological Reviews, vol. 62, pp. 89–113, 1987.
54. For a helpful overview of the mechanical design of spider silks, see J. M. Gosline et al. in
Journal of Experimental Biology, vol. 202, pp. 3295–3303, 1999.
55. To a first approximation silk proteins are arranged either as α-helices or parallel β-pleated
sheets, the latter more familiar in the form of spider silk and commercial silk. A third variety
is the cross β-pleated sheet; see Craig (1997, note 61) for a helpful overview.
56. The gene family that encodes the fibroins is reported by P. A. Guerette et al. in Science,
vol. 272, pp. 112–115, 1996.
57. Conservation and convergence of spider silk fibroin sequences is reported by J. Gatesy et al.
in Science, vol. 291, pp. 2603–2605, 2001.
58. See C. L. Craig et al. in Evolution, vol. 48, pp. 287–296, 1994; quotation is from p. 293.
59. So far as I am aware the extrusion of liquid proteins to form silks is found only in the
arthropods. But there are other parallels, notably in the formation of the dogfish egg case.
This is composed of the protein collagen, which is extruded in a liquid crystalline state, and
like silk hardens with the loss of water. Helpful accounts of the egg-case production are given
by D. Feng et al. and D. P. Knight et al. in Philosophical Transactions of the Royal Society of
London B, vol. 343, pp. 285–302, 1994 and vol. 351, pp. 1205–1222, 1996 respectively, as well
as by D. P. Knight et al. in Biological Reviews, vol. 71, pp. 81–111, 1996. In addition, it is
worth noting that while silk proteins in arthropods are typically thought of in the context of
fibroins, ‘a collagen-like silk is produced by sawflies in the family Perigidae (Hymenoptera)’
(Craig, 1997; citation is in note 61, p. 244), while this family and the related Argidae also
employ polyglycine, an apparent precursor of collagen (see also the earlier review of arthropod
silks by K. M. Rudall and W. Kenchington in Annual Review of Entomology, vol. 16, p. 73–96,
1971, especially pp. 81–84).
60. This is Bombyx mori, which as E. B. Ford explains in Moths (Collins, London, 1955) produces
a silken cocoon of three layers of which the middle can be unwound to form ‘a single
unbroken thread about 1500 feet in length’ (p. 101).
61. The evolution of silks in arthropods is reviewed by C. L. Craig in Annual Review of
Entomology, vol. 42, pp. 231–267, 1997.
62. See J. Brackenbury in Physiological Entomology, vol. 21, pp. 7–14, 1996, where he describes
how certain caterpillars can descend from danger on a silk lifeline.
63. For an account of the New Zealand glow-worm, see J. E. Lloyd’s chapter on insect
bioluminesence (pp. 241–272) in Bioluminescence in action, edited by P. J. Herring (Academic
Press, London, 1978).
358 notes to pp. 116–119
64. See J. F. Jackson’s interesting account of web-building in other dipterans, and their connection
to Arachnocampa, published in American Midland Naturalist, vol. 92, pp. 240–245, 1974.
This paper is of additional interest because it addresses a macro-evolutionary fantasy put
forward earlier by Richard Goldschmidt, which he subsequently abandoned, to explain how
web-building, bioluminescence, and carnivory could have all evolved simultaneously.
65. See B. M. Walshe’s account of chironomid larval feeding in Proceedings of the Zoological
Society of London, vol. 121, pp. 63–79, 1951. The web is spun at frequent intervals, and then
consumed with its trapped contents.
66. A review of filter-feeding in aquatic insects is given by J. B. Wallace and R. W. Merritt in
Annual Review of Entomology, vol. 25, pp. 103–132, 1980.
67. J. M. Edington, in Journal of Animal Ecology, vol. 37, pp. 675–692, 1968, reports nets of
Plectrocnemia conspersa as being built of a series of irregular strands, up to c. 7.5 cm across.
68. See A. G. Hildrew and C. R. Townsend in Journal of Animal Ecology, vol. 45, pp. 41–57, 1976.
69. This information on weaver ants is from The ants by B. Hölldobler and E. O. Wilson
(Springer, Berlin, 1990), see pp. 620–622. Silk production in ants is typically associated with
the larvae, but as B. L. Fisher and H. G. Robertson demonstrate (in Insectes Sociaux, vol. 46,
pp. 78–83, 1999), adults of Melissotarsus also secrete silk threads.
70. Ants, incidentally, are by no means the only arthropods to make tents by the expedient of
folding together leaves. Various caterpillars of butterflies construct tents and other shelters by
bringing leaves, including those of grass, together with silk threads (see E. B. Ford’s Butterflies
(Collins, London, 1945, pp. 89–90); so, too, does the orthopteran Camptonotus (see Craig
1997, note 61, p. 251).
71. Concerning this behaviour see, for example, the papers by A. P. Brooke in Journal of Zoology,
London, vol. 221, pp. 11–19, 1990 and T. H. Kunz and G. F. McCracken in Journal of Tropical
Ecology, vol. 12, pp. 121–137, 1996. E. Cholewa et al. explain, in Biological Journal of the
Linnean Society, vol. 72, pp. 179–191, 2001, how damage to the leaves by the bats, necessary
for folding and achieved by chewing, is contrived so that water supply is maintained and the
leaf remains fresh.
72. Not that this prevents other animals using silk for their own purposes, such as birds
employing spider silk in their nest construction; a trait evidently arrived at independently
several times. See Bird nests and construction behaviour by M. Hansell (Cambridge
University Press, Cambridge, 2000).
73. See George Ledyard Stebbins’s paper Natural selection and the differentiation of angiosperm
families in Evolution, vol. 5, pp. 299–324, 1951.
74. Their work is published on pp. 283–294 in Constructional morphology and evolution, edited
by N. Schmidt-Kittler and K. Vogel (Springer, Berlin, 1991), and as a parallel paper in
Evolution, vol. 47, pp. 341–360, 1993. Thomas et al. (in Science, vol. 288, pp. 1239–1242, 2000)
extended this work to an analysis of how quickly the ‘skeleton space’ was occupied in the
Cambrian ‘explosion’. By reference to the Burgess Shale fauna, which shows an exceptionally
complete cross-section of Cambrian life, they conclude that occupation was very rapid.
75. Such explorations can be demonstrated, for example, in the case of the aperture of the
protistan foraminiferans (see V. Mikhalevich and J-P. Debenay in Journal of
Micropalaeontology, vol. 20, pp. 13–28, 2001). This seemingly minor example is important
because there have been relatively few investigations of such evolutionary ‘explorations’; it
also illustrates the general principle.
76. See the paper ‘Stingray jaws strut their stuff’ by A.P. Summers et al. in Nature, vol. 395,
pp. 450–451, 1998.
77. The bone-like structure in Ibla is described by H. A. Lowenstam and S. Weiner in Proceedings
of the National Academy of Sciences, USA, vol. 89, pp. 10573–10577, 1992. These authors
emphasize the convergence with lamellar bone, but are also careful to note various
differences. Barnacles themselves also show interesting convergences, one of which is a
striking trend in the reduction in the number of parietal plates in the acorn (or balanomorph)
notes to pp. 119–122 359
barnacles, a point noted by Darwin in his famous monograph on the cirripedes. A. R. Palmer,
in Paleobiology, vol. 8, pp. 31–44, 1982, argues that this reduction is a direct result of
predation pressure exerted by drilling gastropods, whose preferred mode of attack is at plate
margins: fewer plates, fewer margins and greater safety.
78. See D. A. Kelly in Integrative and Comparative Biology, vol. 42, pp. 216–221, 2002. Kelly has
some interesting comments on the nature of convergence, and also remarks ‘If, as
hypothesized, mammals and turtles independently evolved inflatable penises from ancestors
who lacked intromittent organs, the similarity of the anatomical designs . . . to produce penile
stiffness is astonishing’, p. 219. Interestingly, many mammals also possess a penis bone,
known as the baculum. They have their uses: a zoological club, to which I was invited to give
a lecture, keeps order by having to hand the baculum of a large sea-mammal, a walrus as I
recall.
79. The basic cycle of burrowing shown by many groups of invertebrates, entailing alternate
formation of the penetration and terminal anchors in the sediment, is strongly convergent;
see E. R. Trueman’s The locomotion of soft-bodied animals (Edward Arnold, London, 1975).
80. This group is relatively diverse today, but has a very poor fossil record; see S. E. Evans and D.
Sigogneau-Russell in Palaeontology, vol. 44, pp. 259–273, 2001.
81. The paper by J. C. O’Reilly et al. on hydrostatic locomotion in the caecilians is in Nature,
vol. 386, pp. 269–272, 1997.
82. That, in essence, is the thesis of Pavane by Keith Roberts (reprinted by VGSF, London, 1995) a
counterfactual science fiction novel: twentieth-century England is still firmly Catholic, and
the much-delayed scientific revolution is emerging in a Europe where there is still a chance of
avoiding the disasters that have arisen in the real world, where it is assumed that rather than
our being the custodians of God’s great creation, nature is infinitely malleable to our whims.
83. See A. Donovan’s Antoine Lavoisier: Science, administration and revolution (Cambridge
University Press, Cambridge, 1993). Lavoisier’s laboratory in Paris was one of the most
advanced of its type, equipped with superbly constructed instruments and a team of
enthusiastic and dedicated acolytes. Asked to advise on the atrocious state of French
gunpowder, he transformed both its quality and methods of production. A footnote to
history? Possibly not. It can be argued that the American War of Independence, which
depended critically on French supplies to the rebels, would have failed. An interesting
thought, and not for the usual reasons of injured British pride. It has been argued that an
important aim of the British government was to restrict the westward expansion of the
colonists (see R. Harvey’s A few bloody noses: The American War of Independence (John
Murray, London, 2001)), one result of which was the genocide of indigenous tribes and
nations, with the destruction of a cultural richness that is our common loss.
84. See Ferguson (1998; citation is in note 3) for interesting discussions of how the British might
have avoided the disaster of the trenches. John Adamson addresses Charles the First’s crucial
mistake in not attacking the Scots’ army on the Tweed in 1639. Andrew Roberts and Niall
Ferguson explore the consequences of England falling in 1940 by Nazi invasion.
85. Key papers are those by the groups led by R. E. Lenski and M. Travisano. The initial paper in
this series, by R. E. Lenski et al., was published in American Naturalist, vol. 138,
pp. 1315–1341, 1991. An overview by R. E. Lenski and M. Travisano is provided in
Proceedings of the National Academy of Sciences, USA, vol. 91, pp. 6808–6814, 1994, and
more recently by M. Travisano in Current Biology, vol. 11, pp. R440–R442, 2001.
86. The paper, entitled Experimental tests of the roles of adaptation, chance, and history in
evolution, is published in Science, vol. 267, pp. 87–90, 1995.
87. See the paper by F. Vasi et al. in American Naturalist, vol. 144, pp. 432–456, 1994, where they
looked at the effects of fluctuating resources: feast and famine. Taking twelve independently
evolving lines they asked whether the ‘derived genotypes exhibit the same suite of adaptive
changes in their life histories, or did they find alternate solutions to the challenges imposed
by the seasonal environment?’ They remark that superficially the evidence points to
360 notes to pp. 122–124
‘substantial divergence of the derived genotypes in the life-history bases of their improved
fitness’, but they continue, ‘it is equally important to emphasize that direction of
evolutionary change was highly parallel for most of the fitness components and life-history
traits’ (their emphasis). They concluded ‘there was near-perfect uniformity in the directional
responses exhibited by the replicate populations’ (quotations on p. 449).
88. A. Joshi (in Current Science, vol. 72, pp. 944–949, 1997) looked at how E. coli responded to
changes in temperature, again weighing up the relative roles of chance, history, and adaptive
convergence. He concluded, ‘In the case of adaptation to novel temperature, too, the
contribution of adaptation to fitness at 20 ◦ C [the bacteria having been ‘trained’ to ‘enjoy’
substantially higher temperatures of between 32 ◦ C and 42 ◦ C, either constant or fluctuating]
was significantly greater than the combined contribution of history and chance, despite a
significant increase over 1000 generations in the latter’, p. 946.
89. See, for example, the discussion by M. Travisano et al. in Evolution, vol. 49, pp. 189–200,
1996. This concerns the response of the bacteria to sugar substrates, where there are a
number of important qualifications when the bacteria encounter novel, i.e. non-glucose,
environments. Here the variation in evolved fitness was much greater, possibly owing to
heterogeneous pleiotropy.
90. The general topic of evolutionary ‘dead ends’ and how they are avoided is perhaps somewhat
neglected. Especially interesting is the problem of specialization, seemingly reducing the
range of evolutionary possibilities, and thereby increasing the risk of extinction. A good
example concerns the chrysomeline leaf beetles, as described by A. Termonia et al. in
Proceedings of the National Academy of Sciences, USA, vol. 98, pp. 3909–3914, 2001. As
larvae these animals synthesize chemical compounds to assist in defence. Convergently,
however, at least two groups of these beetles have learnt how to derive chemicals from a
specific group of host plants, known as the salicaceans, that confer defence. The advantage is
that metabolic energy does not need to be wasted by the insect in making defensive
chemicals, and as a bonus these larvae also derive a source of sugar. The problem, however, is
that they seemingly irretrievably commit themselves to the salicacean plants, so entering an
evolutionary cul-de-sac. This, however, did not happen because one group was able to shift its
metabolic pathways in a novel direction and so open up a whole new set of host–plant
associations. Another example of how specialization does not necessarily lead to an
evolutionary dead end is given by C. D’Haese in Cladistics, vol. 16, pp. 255–273, 2000. It
concerns another group of insects, the primitive and wingless collembolans. Here, too,
convergence is rife, and involves multiple occupation of habitats by generalized descendants
of specialized ancestors.
91. Just how dynamic the genome is during evolution is addressed by D. Papadopoulos et al. in
Proceedings of the National Academy of Sciences, USA, vol. 96, pp. 3807–3812, 1999. They
studied genome evolution in E. coli for 10 000 generations, which occurs over 1500 days. A
number of interesting results emerged, including the decoupling of genomic and phenotypic
(e.g. cell size, fitness) evolution. Thus adaptive changes slowed down markedly after the
2000th generation, but genomic evolution continued unabated.
92. Moreover, in experiments concerned with adaptation to temperature and a wider variety of
metabolic substrates (see M. Travisano and R. E. Lenski in Genetics, vol. 143, pp. 15–26,
1996) the results are more complex and reflect various sorts of divergence.
93. See number 6 in the series, published by M. Travisano in Genetics, vol. 146, pp. 471–479,
1997.
94. Specifically those selected for glucose and now transferred to maltose showed substantial
genetic variation but no improvement in fitness, whereas those adapted to maltose and
transferred to glucose showed effectively the opposite, i.e. equal fitness improvement
irrespective of whether it was glucose or maltose, but little genetic variation.
95. M. Travisano (1997; citation is in note 93), p. 476.
96. A. Joshi (1997; citation is in note 88), p. 947.
notes to pp. 124–126 361
97. The experiments are described on pp. 947–948 of Joshi (1997). He notes ‘the fact that for
fitness (or a trait strongly correlated with fitness), the effect of current ongoing selection
tends to obliterate the effects of historical and random divergence among lineages’, p. 948.
98. Joshi (1997), p. 949. The quote continues, ‘With a change in selection pressure, the genetic
structure of a population is rapidly shuffled and reorganized, much as what happens to a
particular sequence of sand dunes upon a sudden change in the direction of the wind’. The
wind bloweth where it will, but sand dunes still have a structure and so does the evolution of
life.
99. See H. Teotónio and M. R. Rose in Nature, vol. 408, pp. 463–466, 2000 (with commentary by
J. J. Bull on pp. 416–417).
100. H. Teotónio and M. R. Rose (2000; citation is in note 99), p. 465.
101. D. Schluter’s paper on ‘Tests for similarity and convergence of finch communities’ is
published in Ecology, vol. 67, pp. 1073–1085, 1986.
102. The question as to whether entire communities of organisms can converge in any exact sense
is generally greeted with negatives (see, for example, J. Blondel et al. in Evolutionary Biology,
vol. 18, pp. 141–213, 1984). Even so, the United States/International Biological Program
(US/IBP) Synthesis Series 3 and 5, respectively Convergent evolution in warm deserts: An
examination of strategies and patterns in deserts of Argentina and the United States, edited
by G. H. Orians and O. T. Solbrig, and Convergent evolution in Chile and California:
Mediterranan climate ecosystems, edited by H. A. Mooney (Dowden, Hutchinson & Ross,
Stroudsburg, PA, 1977) have not only many examples of pair-wise comparisons within
particular groups, but also some interesting remarks on the overall degree of convergence
between the North and South American communities. An overview of community
convergence is given by G. H. Orians and R. T. Paine in their chapter (pp. 431–458) in
Coevolution, edited by D. J. Futuyma and M. Slatkin (Sinauer, Sunderland, MA; 1983). For an
interesting example of community parallelisms in the fossil record, specifically between
Devonian brachiopod communities, see P. Wallace in Lethaia, vol. 11, pp. 259–272, 1978.
103. See his interesting discussion in Systematic Biology, vol. 41, pp. 403–420, 1992. Losos also
notes, perhaps surprisingly, that in two of the islands (Puerto Rico and Jamaica) ‘the assembly
of anole faunas . . . followed a highly similar evolutionary sequence’ (p. 415), suggesting in
turn that the evolutionary trajectories are channelled and thereby directive. Losos is careful
to point out the origins of such a ‘decision’ might lie in the existence of a particular
ecomorph, such as an arboreal generalist, and that another starting point, e.g. a grass-bush
ecomorph, might ‘exhibit a quite different trajectory of community assembly’ (p. 416). He
also has an interesting discussion of certain ecomorphs that ‘ought’ to exist on some islands
but do not, such as the absence of a grass-bush ecomorph in Jamaica. Subsequently Losos
et al. (in Science, vol. 279, pp. 2115–2118, 1998, with a commentary by G. Vogel on p. 2043)
extended this analysis of anolid lizards and concluded that: ‘adaptive radiation in similar
environments can overcome historical contingencies to produce strikingly similar
evolutionary outcomes’ (p. 2115). On the other hand, among the aquatic anolid lizards, a
habit that has evolved several times, there is no evidence for convergence; see M. Leal et al.
in Evolution, vol. 56, pp. 785–791, 2002. These workers suggest that the lack of convergence
may be because there are several distinct habitats, or alternatively ‘more than one way may
exist to adapt to a single habitat’ (p. 787).
104. A. Ben-Moshe et al., in American Naturalist, vol. 158, pp. 484–495, 2001, have compared the
community organization (and morphology in terms of teeth and skull patterns) in Old and
New World rodents, and identified patterns of ecomorphological convergence, that ‘strongly
indicates general rules governing ecological communities or guilds’, p. 484.
105. See, for example, the paper by R. Korona in Genetics, vol. 143, pp. 637–644, 1996, looking at
fitness convergence in a soil bacterium. Here, too, genetic variability was maintained even
though the replicates converged towards similar adaptive end points. See also R. C. McLean
and G. Bell in American Naturalist, vol. 160, pp. 569–581, 2002.
362 notes to pp. 126–130
106. Amelie Karlstrom and colleagues’ paper is published in Proceedings of the National Academy
of Sciences, USA, vol. 97, pp. 3878–3883, 2000.
107. Both crustaceans and trilobites are arthropods, but trilobites disappeared 250 Ma ago, rather
insignificant victims of the Permian mass extinctions. The Jurassic trilobite-mimic is
described by A. Radwanski in Acta Geologica Polonica, vol. 45, pp. 9–25, 1995. Trilobites
show their own convergences. N. C. Hughes et al. (in Evolution & Development, vol. 1,
pp. 24–35, 1999) give a particularly interesting example where a proetid trilobite
(Aulacopleura konincki) converges in form on the older olenimorph trilobites. It appears that
the proetid is reoccupying a specific habitat, low in oxygen. This paper is also important
because Hughes et al. provide cogent arguments in support of adaptational explanations in
the context of a trend to a more efficient mechanism, specifically protective enrolment.
108. The appropriately named Archimedes has a characteristic spiral column from which lace-like
fans arise. It is a widespread form, best known in Carboniferous strata. P. D. Taylor and F. K.
McKinney describe a mimic (in Journal of Paleontology, vol. 70, pp. 218–229, 1996), from the
Eocene of North Carolina. The second author also describes in the same journal (vol. 72,
pp. 819–826, 1998) another interesting case of convergence in the bryozoans, specifically in
the independent invention of structures that functionally mimic the specialized, non-feeding
zooids known as the avicularia.
109. Palaeontologists usually refer to taxa that bow out of the fossil record and then at some later
point in the stratigraphic column reappear as either Lazarus or Elvis taxa, terms coined
respectively by D. Jablonski in his chapter (pp. 183–229) in Dynamics of extinction, edited by
D. K. Elliott (Wiley [Wiley-Interscience], New York, 1986) and D. H. Erwin and M. L. Droser
in Palaios, vol. 8, pp. 623–624, 1993. The former are taxa that survive an extinction event, but
remain undetected, that is ‘dead’ to the fossil record (see the article by E. Fara in Geological
Journal, vol. 36, pp. 291–303, 2001 for an overview), only to re-emerge at some later date.
Elvis taxa, on the other hand, are those that which look like the pre-catastrophe forms, but
are mere imitations.
110. See the paper by D. H. Erwin and H-Z. Pan on pp. 223–229 of Biotic recovery from mass
extinction events (edited by M. B. Hart), Geological Society of London Special Publication
102, 1996.
111. Despite the similarities that define any evolutionary convergence, there are also subtle and
interesting differences, a point of very considerable importance for those interested in
explaining the realities of evolution: destinations are similar, seldom identical. Even so, on
occasion the degree of similarity is quite astonishing. K. C. Emberton (see Evolution, vol. 49,
pp. 469–475, 1995), for example, notes a convergence between two land snails so precise that
only a dissection allows them to be distinguished
112. This comparison is discussed by H. Ulrich in Natur und Museum, vol. 95, pp. 499–508, 1965.
113. A helpful review of mantispid biology is given by K. E. Redborg in Annual Review of
Entomology, vol. 43, pp. 175–194, 1998.
114. See the paper by U. Aspöck and M. W. Mansell in Systematic Entomology, vol. 19,
pp. 181–206, 1994. Writing of the Rhachiberothidae they comment, ‘We assume that the
raptorial forelegs have evolved independently of those of the Mantispidae. Parallel evolution
in related taxa is certainly more frequent than generally assumed, and quite natural due to a
similar gene pool’ (p. 204).
115. The mechanics of the predatory strike are described by P. T. A. Gray and P. J. Mill in Journal
of Experimental Biology, vol. 107, pp. 245–275, 1983.
116. See S. N. Gorb and R. G. Beutel in Naturwissenschaften, vol. 88, pp. 530–534, 2001, as well as
a more extended treatment by R. G. Beutel and S. N. Gorb in Journal of Zoological
Systematics and Evolutionary Research, vol. 39, pp. 177–207, 2001. In addition, similar
attachment structures have evolved independently in other arthropods; and these authors
comment on the interesting analogies with attachment devices in the vertebrates. It is also
worth mentioning the evidence for ‘remarkable convergence’ (p. 1509) in digital adhesive pad
notes to p. 130 363
design in some lizards; see E. E. Williams and J. A. Peterson in Science, vol. 215,
pp. 1509–1511, 1982, who draw attention to similarities with the beetles: as ever, different
starting point, same destination.
117. See the paper by D. Tshudy and U. Sorhannus in Journal of Paleontology, vol. 74, pp. 474–486,
2000, who note the independent evolution of pectinate claws in at least four lineages of
crustaceans. This example is quite telling because the claws are often found isolated as fossil
specimens and ‘Convergence in this claw form developed to the extent that isolated fossil
claws . . . commonly have been misidentified at high taxonomic levels’, p. 474.
118. See C. L. Morrison et al. in Proceedings of the Royal Society of London B, vol. 269,
pp. 345–350, 2002. These authors suggest that the trend towards ‘crab-ness’ is irreversible,
and they have an interesting discussion as to whether this form represents some sort of key
innovation or whether it is facilitated by developmental pathways.
119. Even so, C. L. Morrison et al. (2002; citation is in note 118) remark that in at least some cases
the key step in the transition to a crab morphology appears to have occurred in shallow
water.
120. There is an extensive literature on the sabre-tooth cats. Papers by W. A. Akersten in Los
Angeles County Museum Contributions to Science, vol. 356, pp. 1–22, 1985; M. Antón et al.
in Zoological Journal of the Linnean Society, vol. 124, pp. 369–386, 1998; and M. Antón and
A. Galobart in Journal of Vertebrate Paleontology, vol. 19, pp. 771–784, 1999, provide helpful
introductions.
121. Concerning the thylacosmilids, see the papers by W. D. Turnbull on pp. 399–414 in
Development, function and evolution of teeth, edited by P. M. Butler and K. A. Joysey
(Academic Press, London, 1978), and C. S. Churcher in Australian Mammalogy, vol. 8,
pp. 201–220, 1985.
122. A very useful overview of the sabre-tooth cats is given by A. Turner and M. Antón in their
book The big cats and their fossil relatives: An illustrated guide to their evolution and
natural history (Columbia University Press, New York, 1997). They point out that sabres
have evolved at least three times in the placentals, twice in the more advanced nimravids
(e.g. Barbourofelis) and felids (e.g. Homotherium and Smilodon), as well as in the more
primitive creodont oxyaenids (e.g. Apataelurus).
123. See Bob Bakker’s paper published in Gaia (Lisbon), vol. 15, pp. 145–158, 1998. One obvious
difference between the mammalian sabre-tooth cats and the allosaurids is that the latter do
not develop the giant canines, but Bakker suggests ‘that the entire upper jaw of allosaurs
functioned as one huge, saw-edged Samoan war-club, with each small, individual tooth acting
as a mega-serration. Polynesian warriors glued shark teeth to clubs, and the individual shark
tooth functioned as a mega-serration that was, on a much smaller scale, also serrated. Such a
club inflicts a long-jagged wound with concomitant trauma and blood loss, especially if the
mega-serrated blade is pulled backwards as it strikes its target’, pp. 152–153.
124. See, for example, J. A. W. Kirsch’s thoughtful paper on the adaptedness of marsupials in
American Scientist, vol. 65, pp. 276–288, 1977. Interestingly he suggests that the diagnostic
marsupium, that is, the pouch in which the young develop, actually arose several times
independently within the marsupials, an argument that presupposes that primitively this
group of mammals lacked the pouch.
125. This judgement is based in part on the near extirpation of the South American marsupial
faunas, including the thylacosmilids, following the so-called Great American Interchange
when island bridges (at c. 6 Myr) and then land bridges (at c. 2 Ma) were established that
linked this continent to North America. It now seems, however, that this is too simplistic,
and human activities may have also played a significant part in these extinctions. See, for
example, the paper assessing extinction patterns in the late Pleistocene mammals of South
America by E. P. Lessa and R. A. Farina in Palaeontology, vol. 39, pp. 651–662, 1996.
126. See, for example, W. V. Koenigswald and F. Goin’s remarks on enamel structure in
Palaeontographica, Abteilung A, vol. 255, pp. 129–168, 2000. In passing it is also worth
364 notes to pp. 130–135
drawing attention to a striking convergence between the tooth enameloid of two groups of
actinopterygian fish (characidids and sphyraenidids) and modern sharks, involving the
independent development of highly ordered, bending resistant structure; see the paper by
W-E. Reif in Scanning Electron Microscopy, 1979, pp. 547–554, 1979.
127. Specifically H. N. Bryant and C. S. Churcher write ‘Diverse dental anatomies among
sabre-tooths suggest that, despite necessary functional similarities, precise adaptations
varied. The hunting adaptations of these animals were surely diverse’; see Nature, vol. 325,
p. 488, 1987.
128. See his chapter (pp. 157–278) in The biology of marsupials, edited by D. Hunsaker (Academic
Press, New York, 1977). The examples given concern (a) the pronounced lamination of cells
and fibres found in the lateral geniculate nucleus of various marsupials and also primates and
carnivores, which as Johnson remarks, ‘raise many hypotheses about arboreality and visual
evolution’ (p. 262); (b) the occurrence of small cells in distinct subgroups in the phalanger
Trichosaurus that ‘resemble in many ways the “barrels” first described in sensory cortex of
mice and related to projections of vibrissae [whiskers] and other regions of high receptor
density in rats’ (pp. 221–222); and (c) ‘The formation of gyri and sulci at predictable loci in the
sensory representation in neocortex’ (p. 262) of both marsupials and placentals. Johnson
concludes by remarking, ‘Marsupials represent the great alternative case of mammalian
adaptive radiation, and when the same result happens in two such separate phylogenetic
lines, we can begin to identify the determining factors in brain evolution’ (p. 262).
L. Krubitzer, in Trends in Neurosciences, vol. 18, pp. 408–417, 1995, also has an interesting
discussion of some convergences in brain structure between placental and marsupial
mammals. It is also worth mentioning convergences of brain structure within the placentals,
of which one of the best known is the independent evolution of cytochrome oxidase-rich
‘blobs’ in the primary visual cortex of primates and cats; see T. M. Preuss in Brain, Behavior
and Evolution, vol. 55, pp. 287–299, 2000.
129. See, for example, W. R. Scott in Journal of Morphology, vol. 5, pp. 301–406, 1891.
130. See the series of comparisons drawn by G. Dubost in Revue écologie (Terre Vie), vol. 22,
pp. 3–28, 1968.
131. These parallels among placental mammals are addressed by O. Madsen et al. in Nature,
vol. 409, pp. 610–614, 2001. The following paper (pp. 614–618) by W. J. Murphy et al. is also
directly relevant.
132. See R. T. Bakker’s chapter (pp. 350–382) in Coevolution, edited by D. J. Futuyma and M.
Slatkin (Sinauer, Sunderland, MA, 1983).
133. R. T. Bakker (1983; citation is in note 132) p. 354.
134. See Kirk Winemiller’s paper in Ecological Monographs, vol. 61, pp. 343–365, 1991; see also
N. Lamouroux et al. in Ecology, vol. 83, pp. 1792–1807, 2002, who document community
convergence amongst freshwater fish on the basis of stream hydraulics, especially in terms of
the dimensionless Froude number.
135. The paper by R. Hanel and C. Sturmbauer is in Journal of Molecular Evolution, vol. 50,
pp. 276–283, 2000; see also J. J. Day in Biological Journal of the Linnean Society, vol. 76,
pp. 269–301, 2002.
136. The description of replicate evolution in Lake Tanganyika cichlids is given by L. Rüber et al.
in Proceedings of the National Academy of Sciences, USA, vol. 96, pp. 10230–10235, 1999;
see also L. R. Rüber and D. C. Adams in Journal of Evolutionary Biology, vol. 14, pp. 325–332,
2001. This paper makes a specific attempt to remove the ‘burden of history’, whereby a
particular group has a limited evolutionary choice because of phylogenetic ‘decisions’ by its
ancestors. See also the general overview of the Rift Lake cichlids by M. L. J. Stiassny and A.
Meyer in Scientific American, vol. 280 (2), pp. 44–49, 1999, and the figure on p. 48 with the
remark that ‘distantly related cichlids. . .evolved to become uncannily alike.’
137. See R. Felger and J. Henrickson in Haseltonia, 1997 (part 5), pp. 77–85, 1997.
138. For a helpful review of the topic of plant adaptation to desert environments, see the chapter
(pp. 67–106) in Orians and Solbrig (1977; citation is in note 102). R. K. Peet, in American
notes to pp. 135–136 365
Naturalist, vol. 112, pp. 441–444, 1978, offers a short but incisive discussion on various
aspects of convergence, including the adaptations shown by various desert plants. His point
that there may be more than one adaptive solution to a problem (such as extreme aridity) is
important, and a useful counterbalance to over-simplistic thinking. Thus he points out that
in Australia native succulents are rare, but have flourished when introduced, e.g. the prickly
pear. In the end, what matters is how many adaptive solutions exist, and the extent to which
prior conditions govern final outcomes.
139. See the papers by R. M. Cowling et al., for example in Vegetatio, vol. 43, pp. 191–197, 1980,
Journal of Biogeography, vol. 21, pp. 651–664, 1994, and Australian Journal of Ecology,
vol. 19, pp. 220–232, 1994, as well as H. A. Mooney and E. L. Dunn in Evolution, vol. 24,
pp. 292–303, 1970.
140. This is known as myrmecochory. An interesting example of convergence between Australia
and South Africa species is discussed by A. V. Milewski and W. J. Bond on pp. 89–98 of
Ant-plant interactions in Australia, edited by R. C. Buckley (Junk, The Hague, 1982).
141. This overview of convergence across the globe is provided by P. B. Reich et al. in Proceedings
of the National Academy of Sciences, USA, vol. 94, pp. 13730–13734, 1997. As they remark,
‘What selection pressures would lead to such common “solutions” among phylogenetically
different groups? We hypothesize that there are interrelated constraints set by biophysics and
natural selection’ (p. 13733). Thus, in this case certain types of leaf simply cannot exist, while
others are effectively inevitable.
142. Many other examples of convergence in plants are given in K. J. Niklas’s authoritative The
evolutionary biology of plants (University of Chicago Press, Chicago, 1997); see in particular
his Chapter 7. See also Vogel (1998, citation is in note 50) for a series of examples including
twist-to-bend ratios, vascular fluid flow, and leaf shape. More specifically, T. J. Givnish et al.
(in Molecular Phylogenetics and Evolution, vol. 12, pp. 360–385, 1999) give an extensive list
of convergences in the commelinoid flowering plants, ranging from type of pollination, flower
structure, fruits, silica bodies, and xeric adaptations.
143. See the paper by K. J. Niklas in Evolution, vol. 45, pp. 734–750, 1991, which looks at the
biomechanical convergence in the petioles of flowering plants (angiosperms) and ferns.
144. The tree habit has evolved multiple times; for a specific example involving tree-ferns see
J. Galtier and F. M. Hueber in Proceedings of the Royal Society of London B, vol. 268,
pp. 1955–1957, 2001. Such comparisons also span geological time; H. W. Pfefferkorn et al. (in
Historical Biology, vol. 15, pp. 235–250, 2001), for example, draw attention to convergences
between trees growing in areas subject to flooding and high sedimentation rates in South
America today and the Carboniferous deltas of Laurasia. Many other types of plant habit have
also converged. Consider the Mediterranean-like evergreen sclerophyll shrubs, where
Mooney and Dunn (1970; citation is in note 139) explore the convergences imposed by
drought, fire, and mineral shortages.
145. Pollen and ovule each have half the complement of chromosomes, and hence are haploid, in
the same manner as egg and sperm. Successful fertilization combines the chromosomes, to
give the diploid seed that germinates and grows into the plant in the garden (known as the
sporophyte) where ultimately the sex cells are produced, female ovule to await male pollen,
and so the cycle continues . . .
146. Together with his famous metaphor of life’s diversity as a ‘tangled bank’, this phrase must be
one of the most quoted of Darwin’s remarks; first published in More letters of Charles
Darwin. A record of his work in a series of hitherto unpublished letters, edited by F. Darwin
and A. C. Seward (John Murray, London, 1903), see vol. II, pp. 20–21 for an extract from a
letter to J. D. Hooker written on 22 July 1879.
147. Represented by three genera: Ephedra, Gnetum, and Welwitschia.
148. Helpful overviews of the male and female reproductive structures in Gnetales are given
respectively by L. Hufford and P. K. Endress in International Journal of Plant Sciences,
vol. 157 (Supplement 6), pp. S95–S112 and S113–S125, 1996. To a first approximation the
gnetalacean flowers bear separate sexes, but various degrees of bisexuality are known.
366 notes to pp. 136–138
149. The identification and significance of the gnetalacean double fertilization is addressed by J. S.
Carmichael and W. E. Friedman in American Journal of Botany, vol. 83, pp. 767–780, 1996.
150. See, for example, the review paper by P. R. Crane et al. in Nature, vol. 374, pp. 27–33, 1995.
151. In the context of the water vascular system it is also worth noting that it, too, appears to be
convergent, having evolved independently in the bryophytes; see R. Ligrane et al. in New
Phytologist, vol. 156, pp. 491–508, 2002. Significantly, this convergence is not simple, but
entails considerable biochemical complexity.
152. See, for example, A. Hansen et al. in Molecular Biology and Evolution, vol. 16, pp. 1006–1009,
1999; T. Kh. Samigullin in Journal of Molecular Evolution, vol. 49, pp. 310–315, 1999; and the
three papers in Proceedings of the National Academy of Sciences, USA by K.-U. Winter et al.
(vol. 96, pp. 7342–7347, 1999); S-M. Chaw et al. (vol. 97, pp. 4086–4091, 2000); L. M. Bowe
et al. (vol. 97, pp. 4092–4097, 2000); Y-L. Qiu et al. in Nature, vol. 402, pp. 404–407, 1999 (as
well as the preceding paper by P. S. Soltis on pp. 402–404; Y.-L. Qiu et al. in Molecular Biology
and Evolution, vol. 18, pp. 1745–1753, 2001; S. Shindo et al. in Evolution & Development,
vol. 1, pp. 180–190, 1999, and International Journal of Plant Sciences, vol. 162, pp. 1199–1209,
2001; E. Gugerli et al. in Molecular Phylogenetics and Evolution, vol. 21, pp. 167–175, 2001;
S. Magallon and M. J. Sanderson in American Journal of Botany, vol. 89, pp. 1991–2006, 2002;
and D. E. Soltis et al. in American Journal of Botany, vol. 89, pp. 1670–1681, 2002.
153. Interestingly, it is now known that at least some basal angiosperms have a diploid, rather
than triploid, endosperm; see J. H. Williams and W. E. Friedman in Nature, vol. 415,
pp. 522–526, 2002. It is also possible that the triploid condition evolved twice.
154. In support of the angiosperm–Gnetales connection the following characters in common are
often listed: lignin biochemistry (marked by the distinctive Mäule reaction (see R. D. Gibbs’
chapter (pp. 269–312) in The physiology of forest trees, edited by K. V. Thimann (Ronald, New
York, 1958), although he emphasizes that the corresponding lignin biochemistry has probably
evolved several times (p. 307)); a double integument; tunica formation in the apical meristem;
pollen wall (exine) structure; and lack of archegonia; see, for example, J. A. Doyle and M. J.
Donoghue in Botanical Review, vol. 52, pp. 321–431, 1986, p. 356, and P. Crane et al. (1995;
citation is in note 150), Box 1 on p. 29.
155. Thus S. Carlquist in International Journal of Plant Sciences, vol. 157 (Supplement 6),
pp. S58–S76, 1996 provides evidence to support the idea that xylem vessels in the Gnetales
are evolutionarily independent of those developed in the angiosperms.
156. The idea of a common ancestry to double fertilization is effectively the argument presented
by W. E. Friedman and S. K. Floyd in Evolution, vol. 55, pp. 217–231, 2001, who reasonably
remind us that key information in a number of plant groups is still wanting, especially with
respect to the origin of the endosperm.
157. There are, however, some dissenting voices. For example, M. Schmidt et al. (in Journal of
Molecular Evolution, vol. 54, pp. 715–724, 2002) argue on the bases of phytochrome genes
that Gnetales are basal to gymnosperms. This does not make Gnetales and angiosperms a
sister-group, but in principle it could mean that their common ancestor shared reproductive
organs preadapted for evolution towards flowers. A close relationship between Gnetales and
conifers was also questioned by C. Rydin et al. (in International Journal of Plant Sciences,
vol. 163, pp. 197–214, 2002).
158. See P. K. Endress (1996; citation is in note 148).
159. The question, of course, arises that if the triploid endosperm of angiosperms has a clear
nutritive value for the adjacent developing embryo, what then is the function of the diploid
supernumerary zygote in the gnetalaceans? The standard explanation (see, for example,
W. E. Friedman in Proceedings of the National Academy of Sciences, USA, vol. 92,
pp. 3913–3917, 1995) looks to the evolutionary phenomenon known as inclusive fitness,
whereby the embryo (and thereby its genes) receive the benefit of a nutritional supplement by
the ‘sacrifice’ of a genetically similar (or identical) partner.
160. Thus W. E. Friedman and S. K. Floyd (2001; citation is in note 156) remark, ‘Thus, it would
not be surprising if double fertilization, per se, represents an “apomorphic tendency”
notes to p. 138 367
[effectively the repeated likelihood that a particular character will appear; for further
discussion see P. D. Cantino in Systematic Botany, vol. 10, pp. 119–122; 1985] . . . of seed
plants that evolved several times, predicated upon the fact that all seed plants form pairs of
both male and female nuclei that may participate in separate fertilization events . . . If the
evolution of double fertilization is homoplasious among seed plant lineages, this
phenomenon would stand out as a premier example of an apomorphic tendency predicated
upon what has been referred to as an “underlying synapomorphy”. . . in essence, a case of
parallelism that results from an underlying developmental constraint or bias’ (pp. 225–226).
This, of course, opens an indefinite regression: are seeds, land plants, and sex all evolutionary
inevitable? If so, then sooner or later flowers.
161. S. C. Tucker, in International Journal of Plant Sciences, vol. 158 (Supplement 6),
pp. S143–S161, 1997, has an interesting discussion of convergences in flower structure. See
also L. Hufford’s analysis of floral homoplasies in the same issue, pp. S65–S80. More specific
examples of convergence in flowering plants are discussed by various authors, including P. K.
Endress et al. in Nordic Journal of Botany, vol. 3, pp. 293–300, 1983 (addressing the
appearances of apocarpy) and R. Dahlgren in Botaniska Notiser, vol. 123, pp. 551–568, 1970
(concerning South African leguminoseans).
162. See, for example, M. Proctor et al.: The natural history of pollination (HarperCollins, London,
1996), as well as more recent literature such as P. Bernhardt’s discussion of convergence in
beetle-pollinated flowers in Plant Systematics and Evolution, vol. 222, pp. 239–320, 2000,
and E. L. Borba and J. Semir’s discussion of convergence in fly pollination of orchids in Annals
of Botany, vol. 88, pp. 75–88, 2001.
163. See, for example, papers by V. Bretagnolle in The American Naturalist in vol. 142,
pp. 141–173, 1993, and P.-A. Crochet et al. in Journal of Evolutionary Biology (vol. 13,
pp. 47–57, 2000).
164. This topic is discussed by M. Mönkkönnen in Evolutionary Ecology (vol. 9, pp. 520–528,
1995), who points out that not only has long-distance bird migration evolved repeatedly, but
there is a strong tendency in the main feathers (the primaries) for the distal set to elongate
and the proximal ones to shorten, thereby bringing the wing-tips closer to the leading edge.
This topic was returned to in considerable detail by R. Lockwood et al. in Journal of Avian
Biology, vol. 29, pp. 273–292, 1998.
165. The osteology of migrant birds is addressed by R. G. Calmaestra and E. Moreno in Journal of
Zoology, London, vol. 252, pp. 495–501, 2000. The convergences are not in themselves
particularly surprising because all are related to the muscle insertions of the major flight
muscles.
166. The paper, specifically exploring convergences between carduelid finches and turdid chats, is
by A. Landmann and N. Winding in Oikos, vol. 73, pp. 237–250, 1995.
167. See the thoughtful paper by A. Barbosa and E. Moreno in Netherlands Journal of Zoology,
vol. 45, pp. 291–304, 1995, who analyse the bills, wings, and legs of four families of aerially
foraging birds. Their analysis is important because they weigh the relative influences of
adaptation for this demanding way of life against the constraints to change imposed by the
phylogenetic ‘burden’ inherited from the ancestral form that did not pursue this way of life.
Also relevant in the context of aerial insectivory is the paper by A. Keast et al. in The Auk,
vol. 112, pp. 310–325, 1995 on convergence, of a sort, between the flycatchers (tyrannids) and
the American Redstart.
168. This concerns the proposed convergence between the neotropical Honey Creepers,
specifically the nectar-adapted warblers (Parulidae) and tanagers (Thraupidae), a topic
addressed by W. J. Beecher in The Wilson Bulletin, vol. 63, pp. 247–287, 1951.
169. This unexpected convergence emerged as part of a wider study of bird phylogeny by M. Van
Tuinen et al. in Proceedings of the Royal Society B, vol. 268, pp. 1345–1350, 2001. The
considerable morphological similarities of grebes and loons had led them, by the irrefutable
certainties of cladistic analysis, to be regarded as closely allied. The molecular data, on the
other hand, strongly indicates that the similarities are the result of convergence. As these
368 notes to pp. 138–139
authors trenchantly remark, ‘We propose that this unusual alliance of birds has been
overlooked because of the exceptional adaptations to their respective aquatic niches have
obscured evolutionary history,’ p. 1349.
170. See the chapter (pp. 149–188) on adaptive radiations in birds by R. W. Storer in Avian biology,
vol 1, edited by D. S. Farner et al. (Academic Press, New York, 1971). Among the many
examples he gives are those of graviportal locomotion, climbing, webbing, diving (see also
citation in note 169), carnivory including multiple evolution of shrike-like forms, the carrion
eaters (e.g. vultures), and nectar feeders. Evidence for repeated convergence of feeding types is
also stressed by A. H. Bledsoe in The Auk, vol. 105, pp. 504–515, 1988 on the basis of
molecular data that cut radically across various morphological assumptions.
171. See, for example, the remarks by G. A. Bartholomew in his chapter (pp. 11–37) in New
directions in ecological physiology, edited by M. E. Feder et al. (Cambridge University Press,
Cambridge, 1987). In addition to noting that ‘Sphinx moths and hummingbirds are
aerodynamically similar and resemble each other when in flight despite totally different
morphological organization,’ he also shows that ‘The energy costs of hovering flight in these
two structurally and physiologically different types of organisms are remarkably similar’ (p.
21). He continues, ‘One can also translate this similarity into a general statement of
evolutionary interest. The aerodynamic constraints imposed by the combination of forward
flight, while searching for flowers, and sustained hovering, while feeding on nectar, are
associated with the evolution of similar body size, body shape, and wing shape in sphingids
and hummingbirds despite fundamentally different patterns of morphological organization.
The energy expenditures during hovering flight in these convergent insects and birds are
virtually identical and extremely high . . . sphingids and hummingbirds are approaching the
biological limit of aerodynamic performance for animals of their size’ (p. 22). In another
context R. K. Colwell, in American Naturalist, vol. 156, pp. 495–510, 2001, illustrates an
example of convergence (specifically Rensch’s rule > in larger species, > in smaller
species) between hummingbirds and flower mites, both of which compete for the same
resource and are subject to similar constraints in terms of sexual selection and
courtship.
172. The discovery of the toxin homobatrachotoxin in the skin and feathers of the bird Pitohui is
reported by J. P. Dumbacher et al. in Science, vol. 258, pp. 799–801, 1992; see also the
informative commentary by J. M. Diamond in Nature, vol. 360, pp. 19–20, 1992, where he
remarks that this ‘discovery provides an astonishing example of convergent evolution’, p. 19.
173. The bird in question, Ifrita, is probably not closely related to Pitohui, but its systematic
position is still enigmatic. It is unknown, therefore, whether its ability to make this alkaloid
toxin is an example of convergence; see J. P. Dumbacher et al. in Proceedings of the National
174. The convergence in chemical ecology between heteropterans and the bird Aethia cristatellais
is reported by H. D. Douglas et al. in Naturwissenschaften, vol. 88, pp. 330–332, 2001. They
point out that the ultimate source of the toxin is not known, but it may act to repel
ectoparasites. Nor is this the only example of such a chemical convergence. An ability to
sequester pyrrolizidine alkaloids (PA) and cardiac glycosides (CG) from plants, to be employed
for defence or sometimes as pheromones, is well developed in arctiidid and ctenuchid moths
but has also evolved independently in nymphalid butterflies; see M. Wink and E. von
Nickirsch-Rosenegk in Journal of Chemical Ecology, vol. 23, pp. 1549–1568, 1997. In
addition, these authors remark on an extension of this example of convergence with the
independent evolution of PA sequestration in such beetles as Oreina. This example is
important because ‘The biochemical mechanisms that enable insects to sequester PAs or CGs
are certainly intricate’ (p. 1552), thus emphasizing that convergences are far from simplistic.
175. See the summary (pp. 213–214) of D. J. Futuyama’s contribution (pp. 207–231) in Coevolution,
edited by D. J. Futuyama and M. Slatkin (Sinauer, Sunderland, MA, 1983).
176. See W. Cooper in Journal of Zoology, London, vol. 257, pp. 53–66, 2002.
notes to pp. 139–143 369
177. See the remarks on p. 60 of G. G. Simpson’s paper in Bulletin of the Museum of Comparative
Zoology, vol. 139 (1), pp. 1–86, 1970, as well as an earlier paper, specifically on Necrolestos by
B. Patterson, in Breviora, vol. 94, pp. 1–14, 1958.
178. A key worker in this area is Eviatar Nevo. Particularly helpful is his Mosaic evolution of
subterranean mammals: Regression, progression and global convergence (Oxford University
Press, Oxford, 1999). This is an outstanding source book, not only for a detailed discussion of
convergences in the burrowing mammals, but also because of Nevo’s emphatic and repeated
emphases on the reality of adaptation and its central role in evolutionary biology.
179. Concerning powerful claws see also the paper by M. A. Coombs in Transactions of the
American Philosophical Society, vol. 73 (7), pp. 1–96, 1983, where she reviews the
occurrences of large mammalian herbivores with prominent claws. Three principal categories
of activity are identified (digger/tearer, bipedal browser, and climber), and each is identified by
a complex of associated characters.
180. ‘How to eat a carrot?’ is the intriguing main title of the paper by H. Burda et al. in
Naturwissenschaften, vol. 86, pp. 325–327, 1999. Concerning the use of smell by subterranean
rodents to detect and discriminate against plant food, see G. Heth et al. in Behavioral Ecology
and Sociobiology, vol. 52, pp. 53–58, 2002, who emphasize its convergent nature.
181. For overviews, see M. J. Mason and P. M. Narins in American Zoologist, vol. 41,
pp. 1171–1184, 2001, and Journal of Comparative Psychology, vol. 116, pp. 158–163, 2002. In
Chapter 8, we shall see how the burrowing termites and ants also employ seismic
communication (Chapter 8, notes 78 and 79).
182. There are, however, gradients of fierceness, and in desert environments behaviour in the mole
rat Spalax is unusually pacific, probably as an adaptation ‘to minimize overheating, water and
energy expenditure’; see E. Nevo et al. in Behaviour, vol. 123, pp. 70–76, 1992. The
physiological basis underlying this tendency towards pacifism is rather interesting, and may
involve the steroid corticosterone. In threatening situations, which may lead to overt
aggression, this chemical is generated to release metabolic energy. In doing so, however, it
also leads to elevated rates of filtration by the kidneys and thereby increases water
consumption, a problem for mole-rats living in arid environments who also need to avoid
overheating; see G. Ganem and E. Nevo in Behavioural Ecology and Sociobiology, vol. 38,
pp. 245–252, 1996. Ganem also suggests, in a subsequent issue (vol. 42, pp. 365–367, 1998),
that another group of mole rats (the bathyergids) may have followed a similar pathway to
pacifism (see also ensuing discussion by H. Burda on pp. 369–370). This is very much the
exception, and as the writers also note the ‘animals were easily handled by us and caressed
freely like pets, an inconceivable act with other known Spalax species’ (p. 72).
183. See the paper by K. D. Rose and R. J. Emry in Journal of Morphology, vol. 175, pp. 33–56, 1983,
who note the phylogenetic confusion that these animals have sown whereby convergence is
mistaken for phyletic affinities to such groups as the xenarthrans (citation is in note 21).
184. See The biology of the naked mole-rat, edited by P. W. Sherman et al. (Princeton University
Press, Princeton, 1991), p. viii.
185. The independent evolution of eusociality in bathyergid mole-rats is described by J. U. M.
Jarvis and N. C. Bennett in Behavioural Ecology and Sociobiology, vol. 33, pp. 253–260, 1993,
with specific reference to the Damaraland mole-rat (see also J. U. M. Jarvis et al. in Trends in
Ecology & Evolution, vol. 9, pp. 47–51, 1994, and C. G. Faulkes et al. in Proceedings of the
Royal Society of London B, vol. 264, pp. 1619–1627, 1997). The review by H. Burda et al., in
Behavioural Ecology and Sociobiology, vol. 47, pp. 293–303, 2000, has an interesting
discussion as to why eusociality evolves in these mammals.
186. See the brief report by N. G. Solomon in Trends in Ecology & Evolution, vol. 9, p. 264, 1994.
187. See B. J. Crespi (in Nature, vol. 359, pp. 724–726, 1992); B. D. Kranz et al. (in Insectes Sociaux,
vol. 48, pp. 315–323, 2001); and T. W. Wills et al. (in Naturwissenschaften, vol. 88, pp. 526–
529, 2001), as well as a general review by Crespi et al. in Annual Review of Entomology,
vol. 42, pp. 51–71, 1997.
370 notes to pp. 143–145
188. Eusociality in an ambrosia beetle (Australoplatypus incompertus) is documented by D. S.

Kent and J. A. Simpson in Naturwissenschaften, vol. 79, pp. 86–87, 1992. This group is
important in another context, that of the independent evolution of fungal agriculture (see
Chapter 8).
189. The physical dimorphism which distinguishes the reproductives and helpers in a mole-rat
colony is described by M. J. O’Riain et al. in Proceedings of the National Academy of
Sciences, USA, vol. 97, pp. 13194–13197, 2000, and is marked by a lengthening of the
vertebrae and corresponding enlargement of the abdominal region, which in turns allows
more space for both reproductive and digestive tissue in the breeding female. These authors
conclude by remarking that ‘Termite, ant, and naked mole-rat “queens” with their enlarged
abdomens and long life-spans provide a striking example of convergent evolution in response
to such selective pressure’ (p. 13197), which they argue will be ecological and extrinsic.
190. See J. E. Duffy in Nature, vol. 381, pp. 512–514, 1996, with a commentary by J. Seger and N.
A. Moran on pp. 473–474.
191. This was demonstrated by J. E. Duffy et al. in Evolution, vol. 54, pp. 503–516, 2000, who
emphasized the possible link between eusociality in these shrimps and the role of
competition and enemy pressure; see also J. E. Duffy et al. in Behavioral Ecology and
Sociobiology, vol. 51, pp. 488–495, 2002.
192. See A. Perle et al. in Nature, vol. 362, pp. 623–626, 1993, and associated commentary by A. C.
Milner on p. 589. A more complete description of Mononykus is given by A. Perle et al. in
American Museum Novitates, no. 3105, pp. 1–29, 1994.
193. See, for example, the overview by L. M. Chiappe in Nature, vol. 378, pp. 349–355, 1995.
Concerning Mononykus, he is upbeat about its avian affinities.
194. Thus Perle et al. (1993) write, ‘The highly modified forelimb of Mononychus [a name
subsequently changed to Mononykus; see Perle et al. 1994, p. 3; citation is in note 192] is
similar to that of digging animals. [whereas] . . . The short forelimb and long, gracile hindlimb
are, paradoxically, incongruous with a burrowing habitus’ (p. 625), a view reiterated by Perle
et al. (1994, p. 26; citation is in note 192).
195. See his article in The Auk, vol. 112, pp. 958–963, 1995. A reply to Zhou is offered by L.
Chiappe et al. in a subsequent issue, vol. 114, pp. 300–302, 1997. Even if Mononykus is a bird,
a number of its features must be convergent with the modern birds; see Perle et al. (1994,
pp. 20–24; citation is in note 192).
196. See G. M. Taylor et al. in Brain, Behavior and Evolution, vol. 45, pp. 96–109, 1995, who
remark of the fossorial frog brain, ‘A reduced optic tectum and enlarged main olfactory bulb
and torus semicircularis would appear to be in response to the unique selection pressures of
living underground’ (p. 105).
197. See the paper by M. S. Y. Lee in Biological Journal of the Linnean Society, vol. 65,
pp. 369–453, 1998, as well as the references therein. The author offers several interesting
remarks on the role of convergence, emphasizing the functional constraints that lead to
adaptive complexes associated with reptilian burrowing (‘small body size, a robust skull, a
long slender body, and reduced appendages’, p. 419), as well as the dangers of treating such
characters independently in a phylogenetic analysis.
198. The expert is E. N. Arnold, and useful introductions to his careful and imaginative work may
be found in Philosophical Transactions of the Royal Society of London B, vol. 344,
pp. 277–290, 1994; Journal of Zoology, London, vol. 235, pp. 351–388, 1995; and Bulletin of
the Natural History Museum of London (Zoology), vol. 65, pp. 165–171, 1999. This work is
also of particular importance because Arnold casts his analysis of evolutionary trajectories
and convergence in a statistical context that makes it possible to assess the order in which
character traits are acquired in different lineages: evolution is real (hence the order varies),
but the destination is often the same.
199. Arnold (1999), p. 165. With specific respect to iguanian lizards he notes extensive ear
reduction has occurred independently five times, and reiterates the important point ‘that the
sequence of ear modification has been different in some groups even though the end results
notes to pp. 145–147 371
have substantial similarity’ (p. 170). While ear reduction in these lizards is understandable as
an adaptation to a burrowing mode of life (protection, and perhaps also better transmission of
sound), Arnold is also careful to point out that such reduction occurs in some tree-dwelling
forms.
200. This is the famous lizard, Basiliscus, which lives in areas of central and South America, and
in many regions is referred to by the local inhabitants as Largarto Jesus Cristo, the Jesus
Christ lizard. For an account of its aquatic bipedalism see J. Laerm in American Midland
Naturalist, vol. 89, pp. 314–333, 1973, and Ecology, vol. 55, pp. 404–411, 1974.
201. C. Luke, in Biological Journal of the Linnean Society, vol. 27, pp. 1–16, 1986, provides
evidence that lizard toe fringes have evolved independently at least 26 times. While
emphasizing convergence in an adaptive context, such as the 12 (or 13) independent
originations of triangular toe-fringes among lizard inhabiting wind-blown sand, Luke is
careful to explore what may be non-adaptive aspects. Thus, two other fringe types are also
found in the habitat of wind-blown sand.
202. See J. J. Wiens and J. L. Slingluff in Evolution, vol. 55, pp. 2303–2318, 2001, with specific
respect to anguid lizards. Interestingly, they suggest that this trend to a snake-like form is
gradual and there is no need to invoke any macroevolutionary jumps.
203. A helpful overview of convergences among desert mammals is given by M. A. Mares et al. on
pp. 107–163 in Orians and Solbrig (1977; citation is in note 102), in a chapter that also has
interesting comparisons of various desert dwelling groups, including the amphibians, lizards
and ants. A parallel investigation with emphasis on macroecological similarities (or
otherwise) is given by M. L. Cody et al. on pp. 144–192 in Mooney et al. (1977; citation is in
note 102).
204. Key papers by M. A. Mares on convergence in the desert rodents may be found in Proceedings
of the National Academy of Sciences, USA, vol. 72, pp. 1702–1706, 1975; Paleobiology,
vol. 2, pp. 39–63, 1976; Bulletin of the Carnegie Museum of Natural History, vol. 16,
pp. 1–51, 1980; and BioScience, vol. 43, pp. 372–379, 1993.
205. Specifically, Mares (1980) writes, ‘it is likely that future evolution within these deserts will
be that of refinement within a guild, rather than the development of a new, exploited adaptive
zone. Only in the Monte and in Australia do there seem to be vacant zones awaiting
exploitation by rodents. Given the evenness with which such zones have been filled in other
desert areas, I feel that the evolution of such species in these deserts is highly probable,’ p. 42.
206. The relationships of these somewhat enigmatic marsupials are discussed by M. R.
Sanchez-Villagra in Zoological Journal of the Linnean Society, vol. 131, pp. 481–496, 2000,
and Palaeontology, vol. 43, pp. 287–301, 2001.
207. G. G. Simpson (1970; citation is in note 177), p. 49. Other aspects of this convergence include
the common possession in the argyrolagids and kangaroo rats of spaces within the palate, for
which no functional explanation appears to exist (it may be adaptive or possibly coincidental);
Simpson remarks that it is ‘baffling’ (see p. 50). Another shared feature is a very globular
bulla, which may be linked to hearing in arid environments.
208. See M. A. Mares (1975; citation is in note 204).
7. seeing convergence
1. So far as humans are concerned probably not, and in Chapter 9 I review how each step leading
towards humans has its parallels elsewhere. Concerning the woodpecker, J. M. Diamond has
stressed (e.g. in Chapter 12 of The rise and fall of the third chimpanzee (Radius, London,
1991)) how only this one group of birds has developed the complex of adaptations, such as
specialized beak, sawdust filters, skull structure, and musculature that together make
percussive attack on trees possible. He uses this example to argue that, in an analogous way,
humans (at least in a radio-building capacity) are also unique.
2. Vangids are a group of birds endemic to the island of Madagascar. They show a spectacular
adaptive radiation, one that rivals the Galapagos finches or Hawaiian honeycreepers.
372 notes to p. 147
S. Yamagishi and K. Eguchi note in Ibis, vol. 138, pp. 283–290, 1996 (see also S. Yamagishi
et al. in Journal of Molecular Evolution, vol. 53, pp. 39–56, 2001) that ‘The bill shapes in this
family vary so widely that it is difficult to believe these vangids belong to a single family’
(p. 284). Their paper is an exploration of the wide range of foraging ecologies. In particular,
analysis of the Sickle-billed Vanga (Falculea palliata) ‘suggests that this species is a true
substitute for woodpeckers’. Yamagishi and Eguchi are careful to add that ‘Even with this
species, however, the niche of woodpeckers, with special adaptations for boring through the
concealing wood and for probing the exposed tunnel to catch the retreating insect, is not
completely occupied. Thus we may conclude that several vangid species each partially fill the
role of woodpeckers on this island’ (p. 288). M. Cartmill, in his chapter (pp. 655–670) in
Prosimian biology, edited by R. D. Martin et al. (Duckworth, London, 1974), also presents a
list of other islands, uncolonized by true woodpeckers, that in one way or another possess an
avifauna that has ‘developed adaptations analogous to the picid [woodpecker] specialisations
for tunnelling and probing in wood’ (p. 657).
3. Specifically these are the strongly convergent aye-aye (Daubentonia), certain marsupial
possums (Dactylopsila) (see Cartmill 1974), and an extinct group known as the apatemyids
(see W. v. Koenigswald and H.-P. Schierning in Nature, vol. 326, pp. 595–597, 1987). Of
Daubentonia Cartmill wrote that it ‘forages for wood-boring grubs in the rain forest . . . of
Madagascar. In feeding, it begins by tapping the surface of a branch . . . turning its large ears
towards the spot being investigated. Detecting an insect, it probes the crevices of the infested
plant with a grotesquely attentuated third finger, exposing deep-burrowing larvae by biting
and tearing at the wood with its rodent-like incisor teeth’ (p. 657). Cartmill continues by
addressing the similarities to Dactylopsila, remarking, ‘it must be assumed that all the
peculiar cranial traits [of Daubentonia and Dactylopsila] are . . . functionally related to their
other adaptations for preying on wood-boring insects. This assumption is further supported
by the fact that comparable peculiarities . . . are characteristic of woodpeckers’ (p. 662). In the
opinion of T. Iwano and C. Iwakawa (Folia Primatologica, vol. 50, pp. 136–142, 1988) the
aye-aye is better regarded as being convergent on squirrels, a group that is also absent from
Madagascar. So woodpecker or squirrel mimic? In one sense both, because N. Garbutt points
out in Mammals of Madagascar (Pica Press, Sussex, 1999) that the aye-aye is adept at opening
nuts and scooping out the contents with its extraordinary middle finger, and is also known to
raid coconut plantations. This animal also spends considerable time searching for insect
grubs, as described above. In any event, J. M. Diamond is less than impressed with these
convergences on woodpeckers (see his chapter (pp. 157–164) in Extraterrestrials: Where are
they?, edited by B. Zuckerman and M. H. Hart (Cambridge University Press, Cambridge, 1995
[Second edition]) He has a point, but to infer an absence of extraterrestrial intelligences on the
basis of woodpecker monophyly, and against the mass of other evidence for convergences at
all levels, suggests that it is perhaps a belief system that is being called into question.
4. This is the celebrated Bombardier Beetle. A helpful introduction is given by W. Agosta in his
book Bombardier beetles and fever trees: A close-up look at chemical warfare and signals in
animals and plants (Addison- Wesley, Reading, MA, 1996), see pp. 38–41. The mechanism of
bombardment depends on combining a mixture of hydroquinones plus hydrogen peroxide,
and driving the chemical reaction with enzymes. It produces a scaldingly hot (100 ◦ C) and
toxic defensive spray; see D. J. Aneshansley et al. in Science (vol. 165, pp. 61–63, 1969). In one
group of Bombardiers, the paussine beetles, the accuracy of the discharge is dependent on the
so-called Coanda Effect. This is familiar from the irritating habit of certain brands of teapot to
dribble the tea down the spout rather than into the cup. In the case of the Bombardier,
however, small flanges, that hitherto seemed to be without an adaptive function, ‘serve as
launching guides’ for the corrosive fluid when the animal chooses to shoot forwards; see
T. Eisner and D. J. Aneshansley in Science, vol. 215, pp. 83–85, 1982, as well as a subsequent
article by T. Eisner et al. in Psyche, vol. 96, pp. 153–160, 1989. Whether or not the bombardier
habit evolved independently in a variety of beetles is still debated; see for example T. Eisner
notes to pp. 147–149 373
et al. in Chemoecology, vol. 2, pp. 29–34, 1991. Squirting hot and toxic irritants in the faces of
your foes is not the only way to deal with trouble. Chapter 15 of The biology of the Coleoptera
(Academic Press, London, 1981) by R. A. Crowson gives many other fascinating examples of
how beetles defend themselves. Explosive discharges of protective chemicals are not confined
to the insects; some millipedes spray quinone-containing secretions, while the polydesmid
millipedes produce a protective cloud of hydrogen cyanide; see the description by T. Eisner
et al. of the defensive secretions of millipedes on pp. 41–72 of Arthropod venoms, edited by S.
Bettini (Springer, Berlin, 1978). Another fascinating example of defence is the release of fine
hairs that entangle a would-be predator, such as occurs in the larvae of the dermestid beetles
(more familiar as the ‘woolly bear’ and carpet beetle; my brother, unfamiliar with these
beasts, recently sent me a specimen for identification drowned in gin) (see M. Ma et al. in
Annals of the Entomological Society of America, vol. 71, pp. 718–723, 1978). An interesting
convergent arrangement has been arrived at by the polyxenid millipedes; see T. Eisner et al.
in Proceedings of the National Academy of Sciences, USA, vol. 93, pp. 10848–10051,
1996.
5. A group known as the tunicates (or sea-squirts), which are quite closely related to the
vertebrates, have a thick outer tunic which contains a proportion of the biopolymer cellulose;
see D. D. P. Delmer in Annual Review of Plant Physiology and Plant Molecular Biology,
vol. 50, pp. 245–276, 1999. Cellulose is otherwise unknown in animals but is typical of plants
(as well as algae and some bacteria; see Y. van Daele et al. in Bulletin de Société Royale
Sciences de Liège, vol. 59, pp. 329–417, 1990, and Biology of the Cell, vol. 76, pp. 87–96, 1992;
and S. Kimura et al. in Protoplasma, vol. 194, pp. 151–163, 1996, vol. 204, pp. 94–102, 1998,
and vol. 216, pp. 71–74, 2001). It is not clear whether cellulose has evolved more than once.
Given, however, that cellulose is just one example of a carbohydrate polymer (others being
chitin, which is typical of insect exoskeletons, and hyaluronan, which is important in skin
and cartilage), and that there is some evidence that hyaluronan synthase has probably evolved
twice (see A. P. Spicer and J. A. McDonald in Journal of Biological Chemistry, vol. 273,
pp. 1923–1932, 1998), then it would not be particularly surprising if the same held true for
cellulose (see p. 1930).
6. As discussed below the acuity of the compound eye is substantially less than that of the
camera-eye, and it is perhaps unlikely that the eyes of a mosquito could specifically resolve
my image. However, wasps have a visual acuity that permits identification of other
individuals; see E. A. Tibbetts in Proceedings of the Royal Society of London B, vol. 269,
pp. 1423–1428, 2002.
7. The details of convergence of vision, hearing, and olfaction are returned to below. The
similarities of motor control of locomotion in vertebrates and insects are reviewed by K. G.
Pearson in Annual Review of Neuroscience, vol. 16, pp. 265–297, 1993, and Current Opinion
in Neurobiology, vol. 5, pp. 786–791, 1995. In the former article, for example, he notes that
‘Many of the observations on the afferent control of the stance to swing transition in the cat
are closely paralleled by findings in the walking systems of crustacea and insects’ (p. 279).
(Citation is in note 28.)
8. See D. Sandeman’s absorbing paper ‘Homology and convergence in vertebrate and
invertebrate nervous systems’, published in Naturwissenschaften, vol. 86, pp. 378–387, 1999.
This article is a rich
source of insights on convergences, some of which are returned to below. As Sandeman
remarks, the topic remains of perennial interest because ‘Convergences provide us with a
glimpse of what may be a “best way” to achieve an end’, p. 386.
9. D. Sandeman (1999; citation is in note 8), p. 383.
10. The phylogenetic position of the strepsipterans has long been problematic, but molecular data
published by A. Rokas et al. in Insect Molecular Biology, vol. 8, pp. 527–530, 1999, suggests
this group is not close to dipterans, leaving little doubt that the evolution of halteres, as they
remark, is ‘a remarkable case of convergent evolution’ (p. 529). (See also note 11.)
374 notes to pp. 149–151
11. More details of these rather remarkable insects may be found in the review by J.
Kathirithamby in Systematic Entomology, vol. 14, pp. 41–92, 1989. Additional papers by
Kathirithamby on the strepsipterans include those on the elusive females (in Zoological
Journal of the Linnean Society, vol. 128, pp. 269–287, 2000), and a special issue (number 1) of
vol. 27 of International Journal of Insect Morphology and Embryology, 1998. The last paper
(pp. 53–60) in this issue, by M. F. Whiting, considers the phylogenetic position of the
strepsipterans, and contrary to Rokas et al. (1999) concludes that they are close to the
dipterans (but see the discussion by M. F. Whiting and J. P. Huelsenbeck in Systematic
Biology, vol. 47, pp. 134–138 and 519–537 respectively, 1998; see also W. C. Wheeler et al. in
Cladistics, vol. 17, pp. 113–169, 2001).
12. J. Kathirithamby (1989; citation is in note 11) notes, ‘Structurally and functionally the
mesothoracic wings of Strepsiptera are analogous to the dipteran halteres’ (p. 52), and this
comparison is pursued in greater detail by W. Pix et al. in Naturwissenschaften, vol. 80,
pp. 371–374, 1993.
13. Elytra are modified hind-wings characteristic of beetles, and although sometimes involved in
flight their principal function seems to be to provide a protective wing-case. However, in the
Uganda lymexylid beetle (Atractocercus brevicornis), which unusually for this group is a fast
flyer, the elytra have been modified into haltere-like organs. If they are removed, stable flight
is no longer possible. Interestingly, despite the use of these natural gyroscopes, the sensory
equipment of this African beetle is simpler than that seen in the dipterans. Details of this
remarkable beetle may be found in P. L. Miller’s paper in The Entomologist, vol. 104,
pp. 105–110, 1971.
14. See A. D. Imms’s A general textbook of entomology including the anatomy, physiology,
development and classification of insects (9th edition, revised by O. W. Richards and R. G.
Davies) (Methuen, London, 1957).
15. A. D. Imms (1957; citation is in note 14), p. 452.
16. An overview of this type of gravity perception, and especially the minerals employed, may be
found on pp. 190–196 in H. A. Lowenstam and S. Weiner’s On biomineralization (Oxford
University Press, New York, 1989); see also M. D. Ross in Advances in Space Research, vol. 4
(12), pp. 305–314, 1984.
17. These statocysts occur in the so-called Müller vesicles, found in a group of ciliates known as
the loxodids; see T. Fenchel and B. J. Finlay in Journal of Protozoology, vol. 33, pp. 69–76,
1986.
18. For a brief resumé see D. Sandeman (1999; citation is in note 8). More detailed accounts are
available in two consecutive papers by D. Sandeman and C. Janse, published in Journal of
Comparative Physiology, vol. 130, pp. 95–111, 1979, and D. Sandeman in Fortschritte der
Zoologie, vol. 28, pp. 213–229, 1983.
19. D. Sandeman (1983; citation is in note 8), p. 228.
20. Reference to the literature on eyes is given below, including a number of general reviews. In
this latter category should also be added vol. 2 (Evolution of the eye and visual system),
edited by J. R. Cronly-Dillon and R. L. Gregory, of Vision and visual dysfunction (Macmillan,
London, 1991), Animal eyes by M. F. Land and D-E. Nilsson (Oxford University Press, Oxford,
2002), the paper in Annual Review of Neuroscience, vol. 15, pp. 1–29, 1992 by M. F. Land and
R. D. Fernald simply entitled The evolution of eyes, and the review with the same title by M.
F. Land in Berlin-Brandenburgische Akademie der Wissenschaften Berichte und
Abhandlungen, vol. 8, pp. 311–334, 2000.
21. I should emphasize that there are many other types of eye (see references and citations note
20). These demonstrate a whole series of remarkable adaptations, not least the amazing
triple-lens system of the copepod Pontella. Although I touch on a few such examples my
emphasis here is on convergences and the limits of possibility.
22. A helpful overview of cephalopod eyes (and the other sense-organs) is given by B. U.
Budelmann in Marine and Freshwater Behaviour and Physiology, vol. 25, pp. 13–33,
1994.
notes to pp. 153–154 375
23. Not all cephalopods have lenses; for an interesting discussion of the primitive pinhole eye of
the cephalopod Nautilus see J. B. Messenger on p. 374 of Cronly-Dillon and Gregory (1991;
citation is in note 20). Land (2000; citation is in note 20) expresses surprise that Nautilus has
failed to evolve a lens: ‘The problem, from an evolutionary point of view is that almost any
blob of jelly, placed in the region behind the pupil, would improve both image quality and
light gathering power . . . Why did this not happen in Nautilus? I still find this a bigger
mystery than the origin of really good eyes, which so concerned Darwin’ (p. 321). This is a fair
point, but may in part be answered by reference to the olfactory capabilities of Nautilus; see J.
A. Basil et al. in Journal of Experimental Biology, vol. 203, pp. 1409–1414, 2000, as well as its
crepuscular habit and low metabolic rate (concerning the latter see R. G. Boutilier et al. in
Nature, vol. 382, pp. 534–536, 1996). When Nautilus ‘needs’ a lens, so it will be ‘provided’.
24. See the paper by B. Willekens et al. in Tissue & Cell, vol. 16, pp. 941–950, 1984, where they
document the lens structure of Sepiola.
25. For more information on the cephalopod lens see the review by J. M. Arnold on pp. 265–311
of Experimental embryology of marine and fresh-water invertebrates, edited by G. Reverberi
(North-Holland, Amsterdam, 1971). Other relevant papers by J. M. Arnold are in Journal of
Ultrastructure Research, vol. 14, pp. 534–539, 1966 and vol. 17, pp. 527–543, 1967
respectively, as well as T. J. C. Jacob and G. Duncan in Nature, vol. 290, pp. 704–706, 1981.
Other helpful reviews are in chapters 17 (by H. R. Saibil, pp. 371–397) and 18 (by I. A.
Meinertshagen, pp. 399–419) in Squid as experimental animals, edited by D. L. Gilbert et al.
(Plenum, New York, 1990).
26. The idea that focusing depends on the movement of the cephalopod lens is raised by J. G.
Sivak in Journal of Comparative Physiology, vol. 147A, pp. 323–327, 1982, although this idea
goes back many years; see J. S. Alexandrowicz in Archives de zoologie expérimentale et
générale, vol. 66, pp. 71–134, 1927.
27. The focal properties of the cephalopod lens are addressed by J. G. Sivak in Canadian Journal
of Zoology, vol. 69, pp. 2501–2506, 1991.
28. So, too, flying insects have a sophisticated and integrated control system that not only links
specific sensory inputs to motor control, but has fail-safe mechanisms whereby input of
faulty data does not lead to catastrophic error. See, for example, H. Reichert and C. H. F.
Rowell in Trends in Neurosciences, vol. 9, pp. 281–283, 1986, who note, ‘It would not be
surprising if similar principles of neuronal organization were found in all animals that need to
integrate a rhythmic motor output with non-phase-locked sensory input’, p. 283.
29. See the papers by B. U. Budelmann and J. Z. Young in Philosophical Transactions of the Royal
Society of London, B, vol. 306, pp. 159–189, 1984, the overview by B. U. Budelmann 1994
(citation is in note 22) as well as the detailed tabular comparison between cephalopod and
vertebrate vestibulo-oculomotor reflexes by B. U. Budelmann and Y. Tu in Vie et Milieu,
vol. 47, pp. 95–99, 1997. Details of the statocyst organization in octopus are addressed by B. U.
Budelmann et al. in Philosophical Transactions of the Royal Society of London B, vol. 315,
pp. 305–343, 1987. These authors emphasize the complexity of the system, and while noting
the differences from the vertebrate system they also stress ‘striking analogies’, p. 340.
30. See H. Collewijn’s paper comparing eye movement of cuttlefish and rabbit, in Journal of
Experimental Biology, vol. 52, pp. 369–384, 1970. In addition A. McVean in Comparative
Biochemistry and Physiology, vol. 78A, pp. 711–718, 1984, emphasizes that in comparison
with the mammals, the extra-ocular muscles in the octopus are inferior in such matters as
speed of contraction and muscle viscosity. Interestingly, the arrangement of the musculature
that moves the head of the fly is analogous to the optic muscles so that the entire head
capsule, with its paired compound eyes, is equivalent to the vertebrate eye; see
N. J. Strausfeld et al. and J. J. Milde et al. in Journal of Comparative Physiology, vol. 160A,
pp. 205–224 and 225–238 respectively, 1987.
31. Earthworms are oligochaetes, and although once thought to be the most primitive annelids,
the almost entirely marine polychaetes are usually cast in this role. The leeches, famous for
their blood-sucking abilities, are thought to have arisen from the oligochaetes.
376 notes to p. 154
32. Key papers are those by C. O. Hermans and R. M. Eakin in Zeitschrift für Morphologie der
Tiere, vol. 79, pp. 245–267, 1974, and G. Wald and S. Raypart in Science, vol. 196,
pp. 1434–1439, 1977. Wald and Raypart are enthusiastic in their identification of the
convergences (with words such as ‘remarkable’ and ‘extraordinary’), whereas although
Hermans and Eakin (1974) accept convergence, at the same time they struggle to find an
underlying homology between the alciopid camera-eye and that of the
cephalopods.
33. Annelids and molluscs, as well as various other groups such as the brachiopods, nemerteans,
and platyhelminthes, are placed in a group known as the lophotrochozoans. Such similarities
as there are in terms of the camera-eyes are due to similar modes of embryological
development, and not because those of cephalopod and alciopid polychaete evolved from a
common ancestor that possessed such an eye.
34. G. Wald and S. Raypart (1977; citation is in note 32) remark, ‘The presence of accessory
retinas in alciopid eyes offers a prime instance of the phenomenon of evolutionary
convergence’, p. 1437. They suggest that they may function to detect light of long wavelength,
which penetrates more deeply into the sea, and could thus be used to judge depth.
35. A general overview of the camera-like eyes in both the heteropods (such as Pterotrachea) and
also Littorina and Strombus (see notes 37 and 38 respectively), is given by G. H. Charles on
pp. 455–521 of vol. II of Physiology of Mollusca, edited by K. M. Wilbur and C. M. Yonge
(Academic Press, London, 1966); see also the superb review by M. F. Land on pp. 471–592 of
Handbook of sensory physiology, vol. VII/6B: Invertebrate visual centers and behavior I,
edited by H. Autrum (Springer, Berlin, 1981), and also the chapter (pp. 364–397) by Messenger
in Cronly-Dillon and Gregory (1991; citation is in note 20).
36. Although reviewed in more recent literature, the key reference on heteropods was published
more than a century ago by R. Hesse in Zeitschrift für Wissenschaftliche Zoologie, vol. 68,
pp. 379–477, 1900. The complex structure of the photoreceptors in Pterotrachea, which have
some resemblances to those of vertebrates, is addressed by P. N. Dilly in Zeitschrift für
Zellforschung und mikroskopische Anatomie, vol. 99, pp. 420–429, 1969. A curiosity of the
heteropods is that although the eye is camera-like, the retina consists of a narrow ribbon. In
life there are rapid scanning movements in which the eye is rapidly flicked downwards and
then slowly returns to the horizontal position. M. F. Land, in Journal of Experimental
Biology, vol. 96, pp. 427–430, 1982, suggests that this scanning makes possible the detection
of stationary objects. Interestingly, in the heteropod Atlanta the movement of the eyes is
synchronous and their overlap suggests the possession of binocular vision (see notes 86 and
89).
37. Concerning the well-known winkle Littorina littorea, see the paper by G. F. Newell in
Proceedings of the Zoological Society of London, vol. 144, pp. 75–86, 1965. The eye structure
of L. irrorata is discussed by P. V. Hamilton et al. in Journal of Comparative Physiology,
vol. 152A, pp. 435–445, 1983.
38. Aspects of the sophistication of the eye of Strombus are addressed by H. L. Gillary in Journal
of Experimental Biology, vol. 60, pp. 383–396, 1974; vol. 66, pp. 159–171, 1977; and vol. 107,
pp. 243–310, 1983. Further details of the ultrastructure of these eyes are given by H. L. Gillary
and E. W. Gillary in Journal of Morphology, vol. 159, pp. 89–116, 1979; the light-sensitive
pigments (rhodopsin and retinochrome) are discussed by K. Ozaki et al. in Vision Research,
vol. 26, pp. 691–705, 1986.
39. For accounts of the structure of the cubozoan eye, see V. J. Martin in Canadian Journal of
Zoology, vol. 80, pp. 1703–1722, 2002, G. Laska and M. Hündgen in Zoologische Jahrbücher
Abteilung für Anatomie und Ontogenie der Tiere, vol. 108, pp. 107–123, 1982, and J. S. Pearse
and V. B. Pearse in Science, vol. 199, p. 458, 1978. See also J. Piatigorsky et al. in Journal of
Comparative Physiology, vol. 164A, pp. 577–587, 1989.
40. The suggestion that the lens might be adjustable was made by F. W. Berger in Memoirs from
the Biological Laboratory of the Johns Hopkins University, vol. 4(4), pp. vi + 84, 1900; see
pp. 44 and 58–60, see also Piatigorsky et al. (1989).
notes to pp. 154–156 377
41. See, for example, the work by M. J. F. Blumer et al. in Zoomorphology, vol. 115, pp. 221–227,
1995. Most probably the light-sensitive protein employed will be rhodopsin, the importance
of which is returned to below. C. Musio et al. report (in Journal of Comparative Physiology,
vol. 187A, pp. 79–81, 2001) a rhodopsin-like molecule in Hydra, but this animal lacks
eye-spots, and, curiously, expression of the protein is widely distributed across the
animal.
42. Evidence for the visual abilities of cubozoans is given by W. M. Hamner et al. in Marine
Freshwater Research, vol. 46, pp. 985–990, 1995, and they cite some intriguing observations
by an earlier investigator, J. H. Barnes. R. F. Hartwick (in Hydrobiologia, vol. 216/217,
pp. 171–179, 1991) also has some interesting comments on the possible role of vision in the
cubozoans, especially with respect to search for mates. G. I. Matsumoto (in Marine and
Freshwater Behavior and Physiology, vol. 26, pp. 139–148, 1995) also reviews the optical
acuity of the cubozoans, and leaves little room for doubt that these animals genuinely see
objects. See also S. E. Stewart in Marine and Freshwater Behavior and Physiology, vol. 27,
pp. 175–188, 1996.
43. This rather remarkable behaviour, for a cnidarian, is discussed by B. Werner in Marine
Biology, vol. 18, pp. 212–217, 1973, and entails transfer of the spermatophores to the female
in a series of specific steps. See also Hartwick (1991).
44. Comparable remarks apply also to the other cnidarians, such as Hydra, with their nerve nets.
For example, G. O. Mackie (in American Zoologist, vol. 30, pp. 907–920, 1990) points to
various under-appreciated complexities, including pacemakers and giant axons, as well as
quite sophisticated behaviours, remarking, ‘most modern workers would have to agree with
Parker [a distinguished biologist] in dismissing as inadequate both von Uexkull’s description
of sea anemones as “a bundle of reflexes” and, at the other extreme, Gosse’s [“hero” of
Edmund Gosse’s Father and son] picture of them as creatures endowed by consciousness and
will’, p. 917.
45. Unless one labels a pacemaker as a brain, and this is still remote from the bilaterian brain.
Nick Strausfeld has suggested to me that the earliest function of the brain was ‘to assess
asymmetries in the sensory surround and to compensate for these by appropriate motor
efferent reply. From this basal condition brains might have evolved specialized connections to
exploit asymmetries of the sensory world by mediating appropriate downstream asymmetries
of motion, which would provide for goal-directed behavior’ (personal communication;
10/02/2002).
46. See Matsumoto (1995, citation is in note 42), p. 146. He also hints that the unique lens
proteins (the crystallins; see Piatigorsky et al. (1989, citation is in note 39; see also note 114)
developed by the cubozoans might also be a significant factor. While the origin of these
cubozoan crystallins is, unsurprisingly, separate from other animals (see J. Piatigorsky et al.
in Journal of Biological Chemistry, vol. 268, pp. 11894–11901, 1993; see also J. Piatigorsky (in
Proceedings of the National Academy of Sciences, USA, vol. 98, pp. 12362–12367, 2001),
these novel lens proteins actually provide an excellent example of molecular convergence
(citation is in note 114 and Chapter 10).
47. The ‘unexpected complexity’ refers specifically to the enteric nervous system of echinoderms
as documented by J. E. Garcı́a-Arrarás et al. in Journal of Experimental Biology, vol. 204,
pp. 865–873, 2001.
48. Richard Satterlie and Thomas Nolan’s paper is in Journal of Experimental Biology, vol. 204,
pp. 1413–1419, 2001.
49. R. Satterlie and T. Nolen (2001), p. 1418.
50. However speculative intelligence based on a nerve net might be, so far as brains are
concerned we shall see in the following chapters that there are indeed several but convergent
routes to ‘orthodox’ intelligence.
51. An account of the predatory behaviour of a Panamanian species of Dinopis is given by M. H.
Robinson and B. Robinson in The American Midland Naturalist, vol. 85, pp. 85–96, 1971; the
behaviour of an Australian species is addressed by A. D. Austin and A. D. Blest in Journal of
378 notes to pp. 156–158
Zoology, London, vol. 189, pp.145–156, 1979. The sensitivity of the eyes of Dinopis to its
night-time prey is discussed below, but this spider may have a further trick up its ‘sleeve’. As
Austin and Blest note, ‘There is a suggestion that Dinopis may enhance prey/background
contrast by an ingenious device . . . Dinopis always defecates on completing
snare-construction . . . the white splash of faeces makes a conspicuous spot on the ground
roughly in the middle of the field which the spider surveys, and over which the prey must
run’, p. 156.
52. This account of the Dinopis eye is taken from the paper by A. D. Blest and M. F. Land in the
Proceedings of the Royal Society of London B, vol. 196, pp. 197–222, 1977. Further insights
into the rather remarkable sensitivity of this eye are given by S. Laughlin et al. in Journal of
53. See the two consecutive papers by M. F. Land in Journal of Experimental Biology, vol. 51,
pp. 443–470 and 471–493, 1969, as well as the review on pp. 53–78 of Neurology of arachnids,
edited by F. G. Barth (Springer, Berlin, 1985).
54. See D. L. Clark and G. W. Uetz in Animal Behaviour, vol. 40, pp. 884–890, 1990.
55. See D. S. Williams and P. McIntyre in Nature, vol. 288, pp. 578–580, 1980. In addition to
describing the retinal structure that makes a telephoto component possible, they also
note a similarity to certain birds, remarking, ‘Both a group of vertebrates and invertebrates
have therefore adopted the same strategy to improve the visual acuity despite a restricted
cephalic space’, p. 580. Further, these workers demonstrate that the lenses of the jumping
spiders are able to correct for spherical aberration in the same way as the net-casting
spiders.
56. Concerning the spectral sensitivities of these jumping spiders, see A. D. Blest et al. in Journal
of Comparative Physiology, vol. 145A, pp. 227–239, 1981; S. Yamashita on pp. 103–117 of
Barth (1985; citation is in note 53); and A. G. Peaslee and G. Wilson in Journal of
57. M. F. Land (1981; citation is in note 35), p. 515. Ten years later, Messenger, in Cronly-Dillon
and Gregory (1991; citation is in note 20), could write regarding Strombus, ‘Indeed we have no
clues to the function of these remarkable eyes’, p. 376.
58. Evidence of biotic interaction and displacement of the strombids against the aporrhaids is
reviewed by K. Roy in Paleobiology, vol. 22, pp. 436–452, 1996.
59. K. Roy (1996; citation is in note 58), p. 441. Further information on strombid behaviour and
ecology, especially its remarkable activity, is given by C. J. Berg in Behaviour, vol. 51,
pp. 274–322, 1974, and Bulletin of Marine Science, vol. 25, pp. 307–317, 1975.
60. See F. Evans, in Proceedings of the Zoological Society of London, vol. 137, pp. 393–402, 1961,
who describes the various behaviours of some Ghanaian winkles.
61. See P. V. Hamilton in Marine Behaviour and Physiology, vol. 5, pp. 255–271, 1978.
62. One should note there are other ways of dealing with the problem of seeing on featureless
tidal flats. In some shore-dwelling crabs, such as the fiddler crabs, the compound eyes are
mounted on long stalks like periscopes and have a high resolving power conferred by a
specific arrangement of the components (ommatidia) that provides a so-called acute zone; see
J. Zeil et al. in Journal of Comparative Physiology, vol. 159A, pp. 801–811, 1986. These
workers also remark that ‘visual systems adapted to the geometry of vision in a flat world’
(p. 810) have evolved independently in the insects, among the hemipterans and dipterans, and
an analogy can also be found between the acute zone of arthropods and the visual streak
found in the eyes of various vertebrates that live in open habitats.
63. Further discussion in the optical acuity of winkles can be found in the paper by P. V.
Hamilton and M. A. Winter in Animal Behaviour, vol. 30, pp. 752–760, 1982. Subsequently
these authors extended their study in vol. 32, pp. 51–57, 1984 to a variety of other snails,
emphasizing a range of visual abilities. Littorina irrorata was significantly better at shape
recognition than even a related snail, living in similar habitats, known as Tectarius
muricatus.
notes to pp. 158–159 379
64. A. de Queiroz has an interesting discussion as to whether image-forming eyes promote

evolutionary diversifications; see his article in Evolution, vol. 53, pp. 1654–1664, 1999. He
concluded that no such relationships could be established in terms of net speciation, but did
acknowledge that a connection existed between image-forming capability and activity.
65. See the article by R. G. Northcutt and J. H. Kaas in Trends in Neurosciences, vol. 18,
pp. 373–379, 1995, where they emphasize the independent evolution of visual areas in cats
and monkeys.
66. In the context of convergence in fibre optics, it is worth mentioning that in the light organs of
some squid the diffuse area of luminosity ‘is the result of a hemispherical cushion of light
guide fibres overlying the photogenic tissue’ (p. 213 of Herring 1978; citation is in note 117).
67. See the paper by E. M. Kampa in Vision Research, vol. 5, pp. 475–481, 1965, where he notes
that the eyes of the deep-sea crustaceans are ‘strikingly similar to each other’ (p. 476), not
only in terms of ommatidial structure but also in their developmental pattern.
68. T. H. Oakley and C. W. Cunningham in Proceedings of the National Academy of Sciences,
USA, vol. 99, pp. 1426–1430, 2002 present a phylogenetic analysis based on molecular data
that indicates how the compound eyes of the mydocopids, the only group of ostracods to
possess such eyes, must have arisen independently of the other arthropods. The likelihood of
this convergence was also posited by G. Fryer in Biological Journal of the Linnean Society,
vol. 58, pp. 1–55, 1996; see pp. 37–40.
69. F. B. Krasne and P. A. Lawrence describe the compound eye of the sabellid Branchiomma in
Journal of Cell Science, vol. 1, pp. 239–248, 1966. Further information on their microstructure
is given by A. Kerneis in Journal of Ultrastructure Research, vol. 53, pp. 164–179, 1975.
70. A detailed description of the compound eye in the arc-shell (Arca) is given by W. Patten in
Mitteilungen aus der Zoologischen Station zu Neapel, vol. 6, pp. 542–756, 1886, a topic that
is returned to by T. R. Waller in Smithsonian Contributions to Zoology, vol. 313, pp. iii + 58,
1980. An overview of bivalve eyes and their evolution is given by B. Morton in Oceanography
and Marine Biology: an Annual Review, vol. 39, pp. 165–205, 2001. Interestingly, he remarks,
‘It thus seems that there is a general picture emerging of increasing sophistication with regard
to the bivalve pallial eye . . . There also seems to be a general picture of increasing
sophistication with time’ (p. 193), although he also stresses that not all simple eyes will
inevitably evolve towards greater complexity.
71. Their occurrence in the genera Bispira and Megalomma is noted by K. Fitzhugh (in
Zoological Journal of the Linnean Society, vol. 102, pp. 305–322, 1991), as part of a cladistic
analysis of the sabellids. In an earlier paper (Bulletin of the American Museum of Natural
History, no. 192, 1–104, 1989) Fitzhugh also noted that the evolution of the compound eyes in
these sabellid worms is convergent; see p. 54.
72. See D-E. Nilsson’s description of the sabellid and bivalve eyes and their functional
significance in Philosophical Transactions of the Royal Society of London B, vol. 346,
pp. 195–212, 1994.
73. In passing one should note that not all arthropod eyes are compound, and some are simpler
eye-spots, including the curious example of those butterflies that have eyes located on their
genitalia (where no doubt they have their uses). Don’t believe me? See K. Arikawa and K.
Aoki in Nature, vol. 288, pp. 700–702, 1980, and Journal of Comparative Physiology,
vol. 148A, pp. 483–489, 1982. See also Y. Miyako et al. in Journal of Comparative Neurology,
vol. 327, pp. 458–468, 1993.
74. Discussion of the nature and degree of convergence between the eyes of the strepsipterans
(and to a lesser extent some other living arthropods) and trilobites is given by G. Horvath
et al. in Historical Biology, vol. 12, pp. 229–263, 1997, as well as by E. Buschbeck et al. in
Science, vol. 286, pp. 1178–1180, 1999. The similarities are certainly striking, but they do not
encompass the unique calcitic mineralogy of the trilobite eye. Concerning the optics of
strepsipteran eyes and the likely differences between trilobite and strepsipteran eyes, see W.
Pix et al. in Journal of Experimental Biology, vol. 203, pp. 3397–3409, 2000.
380 notes to pp. 159–164
75. See the paper by J. Aizenberg et al. in Nature, vol. 412, pp. 819–822, 2001 (and commentary
by R. Sambles on p. 783). Nor is this the first report of the possible role of the calcitic
skeleton in brittle-stars for optical sensitivity. For example, S. Johnsen in Journal of
Experimental Biology, vol. 195, pp. 281–291, 1994 suggests that parts of the skeleton are
sensitive to polarized light, and he subsequently reported the presence of probable rhodopsin
at the arm-tips (see Biological Bulletin, vol. 193, pp. 97–105, 1997).
76. A standard laboratory demonstration is to place a large crystal of clear calcite, known as
iceland spar, above a dot. Viewed through the crystal the dot becomes double, and as the
crystal rotates so one dot remains stationary, while the other dot rotates around it. If the
crystal is viewed along the optic axis, then the two dots coalesce into one.
77. For further information on the sensitivity of some brittle-stars, especially spectacular colour
changes, see G. Hendler in Marine Ecology, vol. 5, pp. 379–401, 1984.
78. Schizochroal eyes are particularly characteristic of the phacopid trilobites, and consist of
large biconvex lenses set in cup-like sclera and surrounded by calcitic skeleton.
79. Concerning the investigation of the doublet structure of schizochroal eyes in trilobites the
key figure is E. N. K. Clarkson, whose papers in Palaeontology, vol. 22, pp. 1–22, 1979, and
with J. Miller in Philosophical Transactions of the Royal Society of London B, vol. 288,
pp. 461–480, 1980, provide the necessary details. A further development in this area is the
suggestion, by J. Gál et al. in Vision Research, vol. 40, pp. 843–853, 2000, that this type of
trilobite lens could also act as a bifocal.
80. K. M. Towe, in Geological Society of America Abstracts with Program, vol. 11, p. 529, 1979,
questions whether the supposed doublet structure is not an artefact of diagenesis, specifically
recrystallization and migration of ions.
81. This famous experiment was undertaken by K. M. Towe and reported in Science, vol. 179,
pp. 1007–1009, 1973, and involved capturing images of ‘happy faces’; see his Fig. 1j.
82. See the paper by D-E. Nilsson and R. F. Modlin in Journal of Experimental Biology, vol. 189,
pp. 213–236, 1994.
83. D-E. Nilsson and R. F. Modlin (1994; citation is in note 82), p. 216.
84. D-E. Nilsson and R. F. Modlin (1994; citation is in note 82), p. 227.
85. See K. Kirschfeld’s chapter on pp. 354–370 of Neural principles in vision, edited by F. Zettler
and R. Weiler (Springer, Berlin, 1976), as well as the additional remarks in Land (1981; see
note 35 for citation) on pp. 551–553. More details on what arthropod eyes are, and are not,
capable of is given by E. J. Warrant and P. D. McIntyre in Progress in Neurobiology, vol. 40,
pp. 413–461, 1993. The relative ineffectiveness of the compound eye has, moreover, been long
appreciated; see the paper by A. Mallock in Proceedings of the Royal Society of London,
vol. 55, pp. 85–90, 1894.
86. See, for example, the remarks by J. D. Pettigrew on the evolution of binocular vision
(pp. 271–283) in Cronly-Dillon and Gregory (1991; citation is in note 20). He also remarks
that such vision may extend at least to fish and amphibians. For an interesting exception to
avian binocular vision see note 231.
87. See M. F. Land’s paper in Journal of Insect Physiology, vol. 38, pp. 939–951, 1992.
88. M. F. Land (1992; citation is in note 87), p. 947.
89. See S. Rossel in Nature, vol. 302, pp. 821–822, 1983, and Y. Toh and J.-Y. Okamura in Journal
of Experimental Biology, vol. 204, pp. 615–625, 2001.
90. See, for example, the remarks by N. J. Strausfeld and J.-K. Lee in Visual Neuroscience, vol. 7,
pp. 13–33, 1991, which not only document some of the extraordinary complexity of the visual
centres in the insect brain, but also touch on the canonical properties of such nervous
systems as are shared by insect and primate.
91. See the fascinating paper by J. D. Pettigrew et al. in Current Biology, vol. 9, pp. 421–424,
1999, as well as the commentary by M. V. Srinavasan in Nature, vol. 399, pp. 305–306, 1999.
Concerning ‘This remarkable convergence’ (p. 424) Pettigrew et al. present a list (their Table
1) of the convergent features between the optics and behaviours of the chameleons and
notes to pp. 164–165 381
sand-lances. Thirteen are listed, of which five are not otherwise known in any other lizard or
fish. Srinivasan remarks ‘in these two animals, eye design cannot be predicted by evolutionary
origin – rather, it has been crafted almost exclusively by environmental constraints’, p. 305.
92. Rapid strikes at prey by using a tongue are also familiar in the frogs, where as K. C.
Nishikawa (in BioScience, vol. 47, pp. 341–354, 1997) notes that the more derived and
sophisticated method known as ‘inertial elongation of the tongue [has] evolved as many as
eight times independently’, and that ‘Surprisingly, novel sensory fibers (i.e. afferents) in the
tongue have evolved independently four to five times in frogs that use inertial elongation to
protract their tongues’, p. 344.
93. Luitfried v. Salvini-Plawen and Ernst Mayr’s much-cited paper ‘On the evolution of
photoreceptors and eyes’ was published in Evolutionary Biology, vol. 10, pp. 207–263, 1977.
94. L. v. Salvini-Plawen and E. Mayr (1977; citation is in note 93), p. 249.
95. See the helpful overview by M. F. Land in Scientific American, vol. 239 (part 6), pp. 88–99,
1978, as well as M. F. Land (1981), pp. 536–538, and Messenger (1991; both citations are in
note 35).
96. See his review on pp. 118–135 of J. R. Cronly-Dillon and R. L. Gregory (1991; citation is in
note 20).
97. M. F. Land (1991; citation is in note 20), p. 124 (his emphasis).
98. Guanine, however, is not a universal reflector, and in at least some cephalopods the material
employed appears to be chitin; see the paper by E. J. Denton and M. F. Land in Proceedings of
the Royal Society of London B, vol. 178, pp. 43–61, 1971.
99. In his 1981 paper M. F. Land (citation is in note 35) remarks on the surprising fact that the
deep-sea squids lack an equivalent tapetum, despite the manifest advantages of some such
light-gathering mechanism in the inky depths of this environment. Land, however, goes on to
note that one group of invertebrates, the lycosid spiders, has managed this feat.
100. This is particularly pronounced in the hatchet fish; see the paper by E. J. Denton on pp. 59–86
of Marine biology: its accomplishment and future prospect, edited by J. Mauchline and T.
Nemoto (Hokusen-Sha, Tokyo, 1991), as well as the paper by E. J. Denton et al. in
Proceedings of the Royal Society of London B, vol. 225, pp. 63–97, 1985.
101. There is a wide literature on protistan eye-spots. Useful overviews are given by K. W. Foster
and R. D. Smyth in Microbiological Reviews, vol. 44, pp. 572–630, 1980 and J. D. Dodge on
pp. 323–340 of Cronly-Dillon and Gregory (1991; citation is in note 20). More specific papers
include those by K. W. Foster et al. in Nature, vol. 311, pp. 756–759, 1984; G. Kreimer et al. in
FEBS Letters, vol. 293, pp. 49–52, 1991; W. Deininger et al. in EMBO Journal, vol. 14,
pp. 5849–5858, 1995; and E. Ebnet et al. in The Plant Cell, vol. 11, pp. 1473–1484, 1999.
Information on the genes connected to eye-spot formation, which has to occur every time the
cell divides, is given by M. R. Lamb et al. and D. G. W. Roberts et al. in Genetics, respectively
in vol. 153, pp. 721–729, 1999, and vol. 158, pp. 1037–1049, 2001.
102. What P. A. Kivic and P. L. Walne describe in the title of their paper as ‘multiple parallel
evolutions’ of protistan eye-spots is discussed in BioSystems, vol. 16, pp. 31–38, 1983. They
also emphasize that nearly always the photosensory apparatus arises as a modification of a
pre-existing system. Even bacteria have eye-spots; see P. Albertano et al. in Micron, vol. 31,
pp. 27–34, 2000.
103. See K. Schaller and R. Uhl in Biophysical Journal, vol. 73, pp. 1573–1578, 1997, where they
note that the reflectance is concentrated in the yellow part of the spectrum. The preceding
and companion paper by Schaller et al. (pp. 1562–1572) shows the remarkable phototactic
response of Chlamydomonas.
104. These are more familiar, and more unwelcome than is sometimes realized because
dinoflagellate ‘blooms’ are responsible for so-called red tides and in some instances for the
release of toxins whose transfer to shellfish can lead to dire consequences for the gourmet.
Information on these toxins is given by Y. Shimizu on pp. 282–315 of The biology of
dinoflagellates, edited by F. J. R. Taylor (Blackwell, Oxford, 1987).
382 notes to pp. 165–166
105. For a succinct review of dinoflagellate eye-spots see Dodge (1991; citation is in note 20).
106. Dinoflagellate ‘eyes’ are typical of the warnowiids, and interesting descriptions are given by
C. Greuet in Protistologica, vol. 4, pp. 209–230, 1968, and vol. 13, pp. 127–143, 1977; and
Cytobiologie, vol. 17, pp. 114–136, 1978. An overview is provided by the same author on
pp. 119–142 of Taylor (1987; citation is in note 104).
107. The dioptric properties of the crystalline lens are discussed by D. Francis in Journal of
Experimental Biology, vol. 47, pp. 495–501, 1967.
108. In his review of dinoflagellate evolution F. J. R. Taylor in BioSystems, vol. 13, pp. 65–108,
1980, remarks, ‘The ocelli exhibit an uncanny parallelism to the structure of metazoan eyes,
with all the major functional components except for conduction of signals, constructed
entirely at the subcellular level’, p. 76 (his emphasis).
109. See the chapter (pp. 207–227) by J. D. Dodge in The chromophyte algae: Problems and
perspectives, edited by J. C. Greene et al. (Oxford University Press, Oxford, 1989). It also
appears that in the dinoflagellates the chloroplasts were acquired by secondary
endosymbiosis, which is again a convergent feature (see Chapter 6, note 6).
110. Ester Piccinni and Pietro Omodeo’s paper is in Bollettino di Zoologia, vol. 42, pp. 57–79, 1975.
111. E. Piccinni and P. Omodeo (1975; citation is in note 110), p. 72. This remark is echoed by
Taylor (1980; citation is in note 108) when he writes of the dinoflagellate eye-spot ‘What can
they do with a focussed image without a nervous system?’, p. 76.
112. Entanglement of prey is achieved by thread-like cnidocysts that are stored in specialized
organelles; see the papers by L. Mornin and D. Francis in Journal de Microscopie (Paris),
vol. 6, pp. 759–772, 1967 and C. Greuet in Protistologica, vol. 7, pp. 345–355, 1971; see also
the paper by C. Greuet and R. Hovasse in Protistologica, vol. 13, pp. 145–149, 1977, and
Greuet (1987; citation is in note 106).
113. The cnidocysts are convergent on the nematocysts of the cnidarians, which are painfully
familiar from jellyfish stings. What is probably another independent invention of the
cnidocyst is found in the myxozoans. Long thought to be protistans, it is now recognized they
represent parasitic metazoans. The cnidocyst-like structures, known as polar capsules,
seemed consistent with their being degenerate cnidarians (see, for example, M. E. Siddall
et al. in Journal of Parasitology, vol. 81, pp. 961–967, 1995 and M. Schlegel et al. in Archiv für
Protistenkunde, vol. 147, pp. 1–9, 1996), but it is now clear they are some sort of bilaterian, as
argued by J. F. Smothers et al. in Science, vol. 265, pp. 1719–1721, 1994, and by B. Okamura
et al. in Parasitology, vol. 124, pp. 215–223, 2002 (see also A. S. Monteiro et al. in Molecular
Biology and Evolution, vol. 19, pp. 968–971, 2002). Okamura et al. remark, ‘The resemblance
in morphogenesis and final structure of cnidarian nematocysts and myxozoan polar capsules
remains to be explained. We suggest the possibility that this may have arisen by independent
incorporation of eukaryotic symbionts into Cnidaria and Myxozoa which then evolved as
nematoblasts and capsulogenic cells respectively’, p. 222. These cnidocysts are only one of a
variety of so-called extrasomes, dischargeable thread-like structures (e.g. trichocysts,
toxicysts, etc.), found in various protistans; see K. Hausmann in International Review of
Cytology, vol. 52, pp. 197–276, 1978. His comments on the parallels are interesting:
‘Surprisingly, some enigmatic similarity exists between extrasomes and organelles of
systematically widely differing organisms such as . . . toxicysts of ciliates and the nematocysts
of cnidarians’, pp 267–268.
114. There is a rich and absorbing literature on the crystallins. Useful introductions can be found
in N. J. Clout et al. (Nature Structural Biology, vol. 4, p. 685, 1997); H. Janssens and W. J.
Gehring (Developmental Biology, vol. 207, pp. 204–214, 1999); J. Piatigorsky et al. (Journal of
Comparative Physiology, vol. 164A, pp. 577–587, 1989; Journal of Biological Chemistry,
vol. 267, pp. 4277–4280, 1992; Science, vol. 252, pp. 1078–1079, 1991; and especially their
wide-ranging review in European Journal of Biochemistry, vol. 235, pp. 449–465, 1996); P. J. L.
Werten et al. (Proceedings of the National Academy of Sciences, USA, vol. 97,
pp. 3282–3287, 2000); G. Wistow et al. (Proceedings of the National Academy of Sciences,
notes to pp. 166–168 383
USA, vol. 87, pp. 6277–6280, 1990; and Trends in Biochemical Sciences, vol. 18, pp. 301–306,
1993); and H. Chang et al. (Biophysical Journal, vol. 78, pp. 2070–2080, 2000).
115. The explanation for lens transparency was arrived at only relatively recently: see the paper in
Nature (vol. 302, pp. 415–417, 1983) by M. Delaye and A. Tardieu, as well as the
accompanying commentary by G. Benedek on pp. 383–384.
116. See the paper by J. Piatigorsky et al. in Journal of Biological Chemistry, vol. 275,
pp. 41064–41073, 2000. There appears to be only one crystallin in this bivalve mollusc, and it
is homologous to one of those found in the cephalopod eye.
117. A helpful review of these photophores is given by P. J. Herring in his chapter (pp. 199–240) in
Bioluminescence in action, edited by P. J. Herring (Academic Press, London, 1978).
A remarkable range of animals, not only in the sea but on land (e.g. fireflies) can produce
light, and Herring emphasizes how ‘luminescence has been developed on numerous separate
occasions’ (p. 239). Light production is also well known from dinoflagellates, sometimes
giving a beautiful oceanic phosphorescence, while glowing woods are illuminated by
bioluminescent fungi.
118. See M. K. Montgomery and M. J. McFall-Ngai in Journal of Biological Chemistry, vol. 267,
pp. 20999–21003, 1992. Nor is this the only such example of transformation of musculature
into a bioluminescent organ because I. A. Johnston and P. J. Herring describe such an example
(in Proceedings of the Royal Society of London B, vol. 225, pp. 213–218, 1985) in the
scopelarchid fish Benthalbella. It does not appear to be known whether this latter
transformation involves production of crystallins. The evolutionary versatility of the
modification of musculature extends yet further. This includes the development of electrical
organs, which will be considered below. In addition, rather remarkably an eye muscle of some
fish is modified as a heater organ to keep the eye and adjacent parts of the brain warm. See B.
A. Block and G. de Metrio et al. in Journal of Morphology, vol. 190, pp. 169–189, 1986 and
vol. 234, pp. 89–96, 1997 respectively.
119. For example, Piatigorsky et al. (1989; see note 114) remarks ‘The heterogeneity of proteins
that have been used as lens crystallins in different vertebrate and invertebrate species is
astounding. It is not known whether this is a consequence of evolutionary pragmatism with
many proteins being able to fulfil the requirements of lens transparency or whether each
species has differences which need lens proteins with special properties. In any case, the
diversity of lens crystallins throughout the animal kingdom is consistent with the occurrence
of convergent evolution’ (p. 584). A more specific example of convergence comes from the
crystallins found in the eye-lenses of frogs and lizards (geckos). These have been
independently recruited from the same protein superfamily, specifically a group of
stress-related enzymes known as the aldo-keto reductases; see M. A. M. van Boekel et al. in
Journal of Molecular Evolution, vol. 52, pp. 239–248, 2001. The geckos also show their own
example of ocular convergence. This concerns the shift from nocturnal vision to a diurnal
existence. This happened independently at least three times, and is established on the basis of
a study of the crystallins; see B. Röll in Naturwissenschaften, vol. 88, pp. 293–296, 2001.
120. See, for example, J. Nathans in Biochemistry, vol. 31, pp. 4923–4931, 1992.
121. Not to mention the story (or is it a myth?) of the pilots of the Luftwaffe being encouraged to
eat carrots (rich in vitamin A) before embarking on another night of flying in an attempt to
destroy the Royal Air Force.
122. See the papers by G. H. Jacobs et al. in Nature, vol. 382, pp. 156–158, 1996, and P. M. Kainz
et al. in Vision Research, vol. 38, pp. 3315–3320, 1998.
123. Evidence for trichromatic vision in Coquerel’s sifaka and the red ruffed lemur is presented by
Y. Tan and W.-H. Li in Nature, vol. 402, p. 36, 1999. These authors suggest trichromacy is
more primitive than has been thought, but the likelihood is that at least some features of
trichromatic vision in the primates are convergent; see G. K. Jacobs in Vision Research,
vol. 38, pp. 3307–3313, 1998, who emphasizes the case of the howler monkey (See also note
122).
384 notes to p. 168
124. Thus C. A. Arrese et al. (in Current Biology, vol. 12, pp. 657–660, 2002) suggest ‘the potential
for trichomacy in marsupials may have a different evolutionary origin from that in primates’,
p. 659.
125. A popular suggestion for the origin of trichromatic vision in these primates was to facilitate
the recognition of coloured fruits against the green background of the ancestral jungle habitat;
see the papers by B. C. Regan et al. in Vision Research, vol. 38, pp. 3321–3327, 1998, and
N. G. Caine and N. I. Mundy in Proceedings of the Royal Society of London B, vol. 267,
pp. 439–444, 2000.
126. See N. J. Dominy and P. W. Lucas in Nature, vol. 410, pp. 363–366, 2001, who emphasize the
maintenance of trichromatic vision (and its independent evolution in the New World howler
monkeys) as being linked to selection for young leaves, which in the tropics are coloured red,
rich in protein, and tender, and can provide an alternative in times of scarcity of fruit.
127. Vision in many of the New World monkeys, including the squirrel monkey and marmoset, is
therefore polymorphic (see J. D. Mollon et al. in Proceedings of the Royal Society of London
B, vol. 222, pp. 373–399, 1984) making them suitable for colour discrimination experiments,
such as those conducted by Caine and Mundy (2000; citation is in note 125). Despite their
classic status as dichromatic/trichromatic a number of details concerning the genetic
regulation are only now being worked out; see S. Kawamura et al. in Gene, vol. 269,
pp. 45–51, 2001. For an overview of the evolutionary genetics of primate colour vision, see
W-H. Li et al. in Evolutionary Biology, vol. 32, pp. 151–178, 2000. A similar polymorphism
leading to both dichromatic and trichromatic vision is also inferred in some prosimians; see
G. H. Jacobs et al. in Vision Research, vol. 42, pp. 11–18, 2002. It is uncertain whether this
condition is convergent with that found in the platyrrhine monkeys; see also note 123.
128. See the paper by S.-K. Shyue et al. in Science, vol. 269, pp. 1265–1267, 1995, where they
discuss the evidence for polymorphic trichromacy evolving independently in the marmosets
and squirrel monkeys.
129. See the discussion by M. J. Tovée et al. in Vision Research, vol. 32, pp. 867–878, 1992, see
especially p. 877.
130. See the anonymous remarks in a wartime issue of Nature, vol. 146, p. 266, 1940. The short
article is more to do with the implications of camouflage for the colour-blind, but this issue
of Nature also has a lively correspondence on the competence (or otherwise) of the military in
this area.
131. See E. B. Ford’s Moths (Collins, London, 1955) where he remarks that on visiting moorland in
search of the larvae of the Emperor moth their cryptic coloration makes their discovery quite
difficult even when fully exposed, yet the colour-blind can spot them ‘when 20 or 30 yards
distant’ (p. 95). Ford goes on to point out the potentially adaptive advantage of this sort of
colour blindness.
132. A useful overview of colour vision in mammals is given by G. H. Jacobs in Biological
Reviews, vol. 68, pp. 413–471, 1993.
133. Among the vertebrates ultraviolet (UV) vision is widespread, and has been documented in
fish, amphibians, reptiles, birds, and mammals; see G. H. Jacobs in American Zoologist,
vol. 32, pp. 544–554, 1992. See also note 144.
134. Male red–green colour blindness is because of a linkage to the X sex chromosome; see the
classic paper by J. Nathans et al. in Science, vol. 232, pp. 193–202, 1986, as well as a recent
update by S. Yokoyama and F. B. Radlwimmer in Genetics, vol. 153, pp. 919–932, 1999.
135. The spectral tuning of the protein necessary for red–green colour vision is discussed by
various authors, including J. Neitz et al. in Science, vol. 252, pp. 971–974, 1991.
136. Evidence for spectral tuning at the key sites being sometimes dependent on significant
changes in the molecular size of the amino acid is given by S-K. Shyue et al. in Journal of
Molecular Evolution, vol. 46, pp. 697–702, 1998.
137. See G. G. Kochendoerfer et al. in Trends in Biochemical Sciences, vol. 24, pp. 300–305, 1999.
In effect the absorption depends on very localized reconfigurations within the protein
molecule that alter either the electrostatic or dipolar electric charges.
notes to p. 169 385
138. See R. Yokoyama and S. Yokoyama in Proceedings of the National Academy of Sciences,
USA, vol. 87, pp. 9315–9318, 1990. Curiously the fish in question is a blind cave-dweller, but
it has evidently migrated to this habitat recently and the rhodopsin is still functional. It is
also worth noting that red–green vision is retained in the blind mole-rat (see p. 142, Chapter
6). Here it is suggested that it has been co-opted for the maintenance of circadian rhythms.
Details of this changed functionality are given by Z. K. David-Gray et al. in Nature
Neuroscience, vol. 1, pp. 655–656, 1998, who also show that the sensitivity is strongly
red-shifted, perhaps because the reduced eye is covered with skin and red blood vessels.
139. See Yokoyama and Radlwimmer (1999; citation is in note 134), where they review the
‘five-site rule’ for the mammals, reptiles (chameleon), birds (pigeon), and goldfish, and a
further update by these authors in Genetics, vol. 158, pp. 1697–1710, 2001.
140. For the various subtleties of substitution leading to red–green vision of mammals see the
paper by S. Yokoyama and F. B. Radlwimmer in Molecular Biology and Evolution, vol. 15,
p. 560–567, 1998.
141. Sensitivity to blue light in the Conger eel is discussed by S. Archer and J. Hirano in
Proceedings of the Royal Society of London B, vol. 263, pp. 761–767, 1996, who also review
parallel site substitutions in humans, birds, and other fish. See also B. S. W. Chang et al. in
Molecular Phylogenetics and Evolution, vol. 4, pp. 31–43, 1995, who emphasize convergences
at sites in the rhodopsin associated with a blue shift.
142. ‘For whales and seals the ocean is not blue’ is the main title of the article by L. Peichl et al. in
European Journal of Neurosciences, vol. 13, pp. 1520–1528, 2001. Not only is the loss of blue
sensitive s-cones convergent (and may have also occurred twice within the whales), but it
probably confers an adaptive advantage.
143. Six sites are identified by A. J. Hope et al. (in Proceedings of the Royal Society of London B,
vol. 264, pp. 155–163, 1997), of which two are of key importance.
144. In the case of birds the sensitivity to UV radiation can be achieved by a single amino acid
substitution, but the mechanisms adopted by other groups evidently differ, although they
depend on a limited number of site changes; see Y. Shi in FEBS Letters, vol. 486, pp. 167–172,
2000; and S. Yokoyama et al. and Y. Shi et al. in Proceedings of the National Academy of
Sciences, USA, vol. 97, pp. 7366–7371, 2000, and vol. 98, pp. 11731–11736, 2001 respectively.
145. An overview of visual pigments in the invertebrates is, however, given by W. Gärtner and P.
Towner in Photochemistry and Photobiology, vol. 62, pp. 1–16, 1995.
146. Details of squid rhodopsin are addressed by A. Davies et al. in Journal of Molecular Biology,
vol. 314, pp. 455–463, 2001. There are some interesting differences from the mammals, not
least the high degree of ordering of the rhodopsin in the photoreceptor membranes. This
arrangement is linked to the ability both to detect polarized light and to optimize light
capture.
147. This example of convergence between squid and mammal rhodopsin is documented by A.
Morris et al. in Proceedings of the Royal Society of London B, vol. 254, pp. 233–240, 1993.
148. Concerning spectral sensitivities in the hymenopterans see D. Peitsch et al. in Journal of
Comparative Physiology 170A, pp. 23–40, 1992. See also B. S. W. Chang et al. in Gene,
vol. 173, pp. 215–219, 1996, where they report the detailed structure of honeybee
rhodopsin.
149. See the paper by A. D. Briscoe in Molecular Biology and Evolution, vol. 18, pp. 2270–2279,
2001. Briscoe suggests that in both these cases the molecular substitution may well be under
positive Darwinian selection, a topic returned to in its molecular context in Chapter 10. In
addition, Briscoe draws attention to parallels with crayfish (see also note 150), and
subsequently (Molecular Biology and Evolution, vol. 19, pp. 983–986, 2002) addressed the
parallel substitutions for red–green sensitivity in butterflies and bees.
150. The ability to absorb specific wavelengths of light in various species of freshwater crayfish is
discussed by K. A. Crandall and T. W. Cronin in Journal of Molecular Evolution, vol. 45,
pp. 524–534, 1997. One should also note the adaptations for vision in deep-water shrimps,
where an otherwise unexpected near-ultraviolet sensitivity is linked to bioluminesence; see
386 notes to pp. 169–172
T. W. Cronin and T. M. Frank in Proceedings of the Royal Society of London B, vol. 263,
pp. 861–865, 1996.
151. See the paper by J. P. Carulli et al. in Journal of Molecular Evolution, vol. 38, pp. 250–262,
1994.
152. See, for example, the use of oil droplets to confer ultraviolet vision in hummingbirds,
described by T. H. Goldsmith in Science, vol. 207, pp. 786–788, 1980.
153. See J. Marshall et al. in Current Biology, vol. 9, pp. 755–758, 1999.
154. For example, T. W. Cronin and N. J. Marshall, in Nature, vol. 339, pp. 137–140, 1989, report
ten or more types of photoreceptor sensitive to different wavelengths of light.
155. See the paper by D.-E. Nilsson and E. J. Warrant, in Current Biology, vol. 9, pp. R535–R537,
1999, which is a commentary on Marshall et al. (1999, note 153).
156. See R. D. Fernald in Current Opinion in Neurobiology, vol. 10, pp. 444–450, 2000.
157. R. D. Fernald (2000; citation is in note 156), p. 446.
158. There is a large literature on bacteriorhodopsin; for a helpful introduction see the article by
U. Haupts et al. in Annual Review of Biophysics and Biomolecular Structure, vol. 28,
pp. 367–399, 1999.
159. This is a slight simplification, inasmuch as there are also varieties that pump chloride (Cl)
ions (halorhodopsin) and are phototactic (sensory rhodopsin); the latter is returned to below
(citation is in note 171) in the context of colour discrimination.
160. See, for example, the paper by H. Luecke et al. in Journal of Molecular Biology, vol. 291,
pp. 899–911, 1999; and K. Palczewski et al. in Science, vol. 289, pp. 739–745, 2000 (and
commentary by H. R. Bourne and E. C. Meng on pp. 733–734).
161. See the papers by O. Béjà et al. in Science, vol. 289, pp. 1902–1906, 2000 (and accompanying
commentary by E. Pennisi on p. 1869), and Nature, vol. 411, pp. 786–789, 2001.
162. Recent papers on rhodopsin in the green algae include a review by J. L. Spudich et al. (Israel
Journal of Chemistry, vol. 35, pp. 495–513, 1995), as well as by M. Beckmann and P.
Hegemann (Biochemistry, vol. 30, pp. 3692–3697, 1991), W. Deininger et al. (1995), E. Ebnet
et al. (1999; both citations are in note 101), and L. Barsanti et al. in FEBS Letters, vol. 482,
pp. 247–251, 2000.
163. See W. Deininger et al. 1995 (citation is in note 101). The non-involvement of
chlamyrhodopsin in photoreception is addressed by M. Fuhrmann et al. in Journal of Cell
Science, vol. 114, pp. 3857–3863, 2001.
164. See G. Nagel et al. in Science, vol. 296, pp. 2395–2398, 2002 and O. A. Sineshchekov et al. in
Proceedings of the National Academy of Sciences, USA, vol. 99, pp. 8689–8694, 2002, as well
as the overview by K. D. Ridge in Current Biology, vol. 12, pp. R588–R590, 2002.
165. The paper describing fungal phototaxis is by J. Saranak and K. W. Foster in Nature, vol. 387,
pp. 465–466, 1997. Rhodopsin has also been identified in other fungi, notably Neurospora, but
here its function in terms of light regulation is less obvious, see J. Bieszke et al. in Proceedings
of the National Academy of Sciences, USA, vol. 96, pp. 8034–8039, 1999a. See also the same
group’s report of rhodopsin from a yeast in Biochemistry, vol. 38, pp. 14138–14145, 1999b.
166. See the papers by R. Henderson and G. F. X. Schertler in Philosophical Transactions of the
Royal Society of London B, vol. 326, pp. 379–389, 1990, and J. Soppa in FEBS Letters, vol. 342,
pp. 7–11, 1994. Soppa does not rule out descent from a common ancestor, but emphasizes the
lack of sequence similarities.
167. Most notably the pocket within the bacteriorhodopsin that binds the retinal has no
convincing counterpart in animal rhodopsin; see, for example, R. Henderson et al. in Journal
of Molecular Biology, vol. 213, 899–929, 1990.
168. See J. L. Spudich et al. (in Annual Review of Cell and Developmental Biology, vol. 16,
pp. 365–392, 2000, see pp. 379–380).
169. J. L. Spudich et al. (2000; citation is in note 168), p. 385. Despite the likelihood of
convergence between the rhodopsins it is only fair to point out that others regard them as
derived from a common protein; see, for example, Bieszke et al. (1999a), who tentatively
notes to pp. 172–174 387
suggest that their discovery of rhodopsin in Neurospora might provide a link, although
in their 1999b paper (both citations in note 165) they prefer to emphasize the differences.
170. Reviewed by H. G. Kohorana in Proceedings of the National Academy of Sciences, USA,
vol. 90, pp. 1166–1171, 1993; see also W. Zhang et al. in a subsequent issue, vol. 93,
pp. 8230–8235, 1996.
171. Evidence for this colour discrimination is given by J. L. Spudich and R. A. Bogomolni in
Nature, vol. 312, pp. 509–513, 1984, with a commentary by L. Stryer on pp. 498–499. A
mutation analysis of this sensory rhodopsin is given by K-H. Jung and J. L. Spudich in Journal
of Bacteriology, vol. 180, pp. 2033–2042, 1998. Spectral sensitivity is also reported in the
proteorhodopsin; see Béjà et al. (2001; citation is in note 161).
172. The molecular basis of circadian rhythms is addressed by J. C. Dunlap in Cell, vol. 96,
pp. 271–290, 1999.
173. For an overview, albeit emphasizing their occurrence in mice, see A. Sancar in Annual
Review of Biochemistry, vol. 69, pp. 31–67, 2000.
174. See, for example, papers by Y. Miyamoto and A. Sancar in Proceedings of the National
Academy of Sciences, USA, vol. 95, pp. 6097–6102, 1998, and G. T. J. van der Horst et al. in
Nature, vol. 398, pp. 627–630, 1999. Similar functions in Drosophila are addressed by P.
Emery et al. in Cell, vol. 95, pp. 669–679, 1998.
175. See, for example, S. Folkard in Philosophical Transactions of the Royal Society of London B,
vol. 327, pp. 543–553, 1990. As Folkard notes (using the 24-hour clock), night is the realm of
accidents. Three Mile Island? 04.00; Chernobyl? 01:23; Bhopal? ‘just after midnight’.
Single-vehicle accidents are three times more likely to occur between 21.00 and 09.00 than
they are during the rest of the day, notwithstanding lower traffic density. If this is taken into
account, then the increase in vehicle accidents rises to twelve times. So far as the avoidance
of major catastrophes is concerned, Folkard suggests that ‘more extreme measures may be
required. The best solution here would appear to be create a nocturnal sub-society that not
only works at night, but also remains on a nocturnal routine on rest days’, p. 103. I assume
this suggestion is tongue-in-cheek, although I must say I am so glad I am an alpha . . .
176. The possibility is reviewed by A. Sancar (2000; citation is in note 173), see p. 62.
177. D. B. Small et al. in Plant Molecular Biology, vol. 28, pp. 443–454, 1995 discuss
cryptochromes in the alga Chlamydomonas.
178. See, for example, H-Q. Yang et al. in Cell, vol. 103, pp. 815–827, 2000.
179. This evidence for molecular convergence is addressed by A. R. Cashmore et al. in Science,
vol. 284, pp. 760–765, 1999.
180. See A. Sancar in Biochemistry, vol. 33, pp. 2–9, 1994.
181. The evidence against photolyases in humans is addressed by Y. F. Li et al. in Proceedings of
the National Academy of Sciences, USA, vol. 90, pp. 4389–4393, 1993.
182. So how much time do we need? The famous article by D-E. Nilsson and S. Pelger, entitled ‘A
pessimistic estimate of the time required for an eye to evolve’ (in Proceedings of the Royal
Society of London B, vol. 256, pp. 53–58, 1994) suggests that the transformation from simple
eye-spot to fully functioning camera-eye can be achieved in substantially less than a million
years.
183. See his Hereditary Genius: An enquiry into its laws and consequences: first published in
1869, and its second edition in 1892. The latter is available as a reprint, with an introduction
by C. D. Darlington, and published by Fontana in 1962.
184. Chesterton’s book is entitled Eugenics and other evils (Cassell, London, 1922).
185. Francis Galton’s book entitled Memories of my life (Methuen, London, 1908); unfortunately it
has been long out of print.
186. Evident also from his panache and chutzpah when travelling through remote regions,
including his journey in Southern Africa, described in Narrative of an explorer in tropical
South Africa, being an account of a visit to Damaraland in 1851 (Ward Lock, London, 1889).
388 notes to pp. 174–178
Even better is his The art of travel; or shifts and contrivances available in wild countries
(John Murray, London, 1860 [3rd edition]).
187. Colin Beavan’s book is Fingerprints: Murder and the race to uncover the science of identity
(Fourth Estate, London, 2002).
188. C. Beaven (2002; citation is in note 187), p. 98.
189. This article was published in The Psychological Review, vol. 1, pp. 61–62, 1894.
190. F. Galton (1908; citation is in note 185), p. 284.
191. For an interesting discussion of some similarities between the ways olfactory information is
processed and language perceived, see T. S. Lorig in Neuroscience and Biobehavioral
Reviews, vol. 23, pp. 391–398, 1999, who suggests that there is at least some cortical overlap
in the brain.
192. Details of how the burrows are modified to allow habitation of such unpromising substrates
is given by G. C. Hickman in Canadian Journal of Zoology, vol. 61, pp. 1688–1692, 1983,
adaptations that he regarded as admirable. The exceptional diving abilities of the star-nosed
mole, and its respiratory implications, are addressed by I. W. McIntyre et al. in Journal of
Experimental Biology, vol. 205, vol. 45–54, 2002.
193. For an engaging account, with excellent illustrations, see the article by T. L. Yates in Natural
History, vol. 92 (11), pp. 54–61, 1983. Yates suggests that the divergence of the star-nosed
moles from other talpids is quite ancient, perhaps 30 million years ago, but that migration to
America may have occurred no earlier than the later Pliocene. Elsewhere, however, the
star-nosed mole is extinct.
194. The principal publications by K. C. Catania can be found in Nature, vol. 375, pp. 453–454,
1995 and Journal of Comparative Physiology, vol. 185A, pp. 367–372, 1999; see also the paper
by K. C. Catania and J. H. Kaas in BioScience, vol. 46, pp. 578–586, 1996.
195. For a review of Eimer’s organ in the mole group see K. C. Catania in Brain, Behavior and
Evolution, vol. 56, pp. 146–174, 2000. He emphasizes that although these remarkable
structures are effectively unique to the talpids, not only can their epidermal origin be inferred
with some confidence, but convergently derived structures do occur in other mammals,
notably the push-rod mechanoreceptors of the duck-billed platypus and echidna. Concerning
these latter structures see, for example, P. R. Manger and J. D. Pettigrew in Brain, Behavior
and Evolution, vol. 48, pp. 27–54, 1996, who also remark on the similarity with the Eimer’s
organ.
196. The way in which this fovea-like region gets a head-start in capturing a significant portion of
the nervous supply is described by K. C. Catania in Nature Neuroscience, vol. 4, pp. 353–354,
2001. He also notes that similar circumstances are found in the development of the fovea of
primate eyes; see P. Azzopardi and A. Cowey in Nature, vol. 361, pp. 719–721, 1993. See also
R. N. S. Sachdev and K. C. Catania in Journal of Neurophysiology, vol. 87, pp. 2602–2611,
2002.
197. K. C. Catania (1999; citation is in note 194), p. 367. The sensory representations of the
star-nosed mole in the overall context of mammalian cortical maps is addressed by J. H. Kaas
and K. C. Catania in BioEssays, vol. 24, pp. 334–343, 2002.
198. See the paper by K. C. Catania and M. S. Remple in Proceedings of the National Academy of
199. P. W. Lucas and P. F. Roche, in Monthly Notices of the Royal Astronomical Society, vol. 314,
pp. 858–864, 2000, present such evidence for free-floating planets (and also brown dwarfs) in
Orion; see also M. R. Z. Osorio et al. in Science, vol. 290, pp. 103–107, 2000, with
commentary by R. Irion on p. 26.
200. Concerning the navigational abilities of the blind mole-rats and the possible role of a
magnetic compass, see the papers by T. Kimchi and J. Terkel in Animal Behaviour, vol. 61,
pp. 171–180, 2001 and The Journal of Experimental Biology, vol. 204, pp. 751–758, 2001.
201. Just how widespread the construction of such cognitive maps might be in the animal
kingdom may yet yield some surprises. Desert ants (Cataglyphis), for example, are famous for
notes to pp. 178–180 389
returning to their nests across hundreds of metres in a straight line. Evidently the ants know
how to sum up their outward journey and perform a path integration that allows them, in
ways not fully understood, to expedite their return; see S. Wohlgemuth et al. in Nature,
vol. 411, pp. 795–798, 2001 (and commentary by M. V. Srinivasan on pp. 752–755); see also
S. Åkesson and R. Wehner in Journal of Experimental Biology, vol. 205, pp. 1971–1978, 2002.
202. ‘Why do snakes have eyes?’ ask X. Bonnet et al. in Behavioral Ecology and Sociobiology,
vol. 46, pp. 267–272, 1999, where they document the ecological flexibility of these blinded
snakes. And how did the tiger snakes get to Carnac Island? Evidently they were released there
by a travelling showman, after his wife was bitten by one and subsequently died.
203. Among the most famous of these is the infrared sensitivity, such as in the pit vipers. Rather
remarkably some insects have independently evolved a heat-sensitive organ, which has some
striking similarities to that of the snakes; see H. Schmitz et al. in Naturwissenschaften,
vol. 89, pp. 226–229, 2002.
204. See J. S. Edwards et al. in Journal of Neurobiology, vol. 20, pp. 101–114, 1989.
205. See K. C. Catania (2000, citation is in note 195).
206. See the paper by K. C. Catania et al. in Journal of Experimental Biology, vol. 202,
pp. 2719–2726, 1999 (and also Catania, 2000; citation is in note 195).
207. Changes in the masticatory apparatus of the star-nosed mole are discussed by T. E. Grand
et al. in Journal of Mammalogy, vol. 79, pp. 492–501, 1998. Concerning convergence amongst
the xenarthran and pangolin anteaters, see note 21, Chapter 6.
208. The glomeruli are dense, more-or-less spherical, aggregations of nervous tissue that consist of
first-order synaptic neuropil. It is clear that specific parts of the glomerulus are dedicated to
specific types of olfaction, such as the legendary sensitivity of some insects to sex
pheromones. There is an extensive literature on this topic. An introduction can be found in
such papers as those by B. S. Hansson et al. in Science, vol. 256, pp. 1313–1315, 1992; L. A.
Oland and L. P. Tolbert in Journal of Neurobiology, vol. 30, pp. 92–109, 1996; N. J. Vickers
et al. and P. P. Laissue et al. in Journal of Comparative Neurology, vol. 400, pp. 35–56, 1998,
and vol. 405, pp. 543–552, 1999 respectively; and R. Ignell et al. in Brain, Behavior and
Evolution, vol. 57, pp. 1–17, 2001.
209. For an overview see L. Dryer in Bioessays, vol. 22, pp. 803–810, 2000. In the fruit-fly some 59
genes have been identified as coding for the seven trans-membrane domain proteins used in
olfaction. See, for example, the paper by A. Gao and A. Chess in Genomics, vol. 60, pp. 31–39,
1999, where they note that the total number of olfactory genes involved could climb
substantially. While emphasizing the differences with vertebrates, Gao and Chess also note
some similarities, including ‘zoning’ of receptors. Vertebrates have about ten times as many
known olfactory genes, but they also code for transmembrane domain proteins, albeit with
little structural similarity to those of the insects; for an overview see P. Mombaerts in Annual
Review of Neuroscience, vol. 22, pp. 487–509, 1999 and Science, vol. 286, pp. 707–711, 1999.
210. See, for example, the paper entitled ‘Mechanisms of olfactory discrimination: converging
evidence for common principles across phyla’ in Annual Review of Neuroscience, vol. 20,
pp. 595–631, 1997 by J. G. Hildebrand and G. M. Shepherd, as well as the paper entitled
‘Olfactory systems: common design, uncommon origins?’ by N. J. Strausfeld and J. G.
Hildebrand in Current Opinion in Neurobiology, vol. 9, pp. 634–639, 1999.
211. In contrast to the insects, and most famously the moths, the action of sex pheromones in
vertebrates is less explored, but P. W. Sorenson et al. (in Current Opinion in Neurobiology,
vol. 8, pp. 458–467, 1998) emphasize the likely analogies in the use of these chemical
languages.
212. N. Strausfeld and J. G. Hildebrand (1999; citation is in note 210), p. 634.
213. N. Strausfeld and J. G. Hildebrand (1999; citation is in note 210), p. 635 (all
quotations).
214. The realization that insects probably emerged from the aquatic crustaceans has only become
apparent in recent years, largely thanks to molecular biology; see, for example, the papers by
390 notes to pp. 180–181
J. L. Boore et al. in Nature, vol. 392, pp. 667–668, 1998; E. Garcı́a-Machado et al. in Journal of
Molecular Evolution, vol. 49, pp. 142–149, 1999; and K. Wilson et al. in Molecular Biology
and Evolution, vol. 17, pp. 863–874, 2000. The ramifications of this proposal, not least in
terms of the origin of the insect wings from some sort of aquatic appendage (see M. Averof
and S. M. Cohen in Nature, vol. 285, pp. 627–630, 1997), and the hitherto unappreciated
similarities between the brains of crustaceans and insects (see, for example, M. Utting et al.
in Journal of Comparative Neurology, vol. 416, pp. 245–261, 2000) has led to some fertile
research.
215. This possibility is raised by G. Laurent and H. Davidowitz in their paper on insect olfaction
and oscillatory neural assemblies in Science, vol. 265, pp. 1872–1875, 1994 (see also the
accompanying commentary by D. W. Tank et al. on pp. 1819–1820). An update and overview,
with continued emphasis on the similarities between not only the vertebrates and insects,
but also the molluscs (in the form of the slug), are given by A. Gelperin in Journal of
Experimental Biology, vol. 202, pp. 1855–1864, 1999.
216. See, for example, the paper by V. V. Gurevich in Journal of Biological Chemistry, vol. 273,
pp. 15501–15506, 1998.
217. The shared presence of seven trans-membrane domain proteins is important because these
olfactory proteins belong to the same class (G protein-coupled receptors; for an overview see
H. G. Dohlman et al. in Annual Review of Biochemistry, vol. 60, pp. 653–688, 1991) as
rhodopsin (used in vision), and also function in taste (see M. A. Hoon et al. in Cell, vol. 96,
pp. 541–551, 1999). If proteins themselves are universal, so perhaps there is a strong
preference to use specific protein architectures to see, smell, and taste in much the same way.
218. See C. E. Merrill et al. in Proceedings of the National Academy of Sciences, USA, vol. 99,
pp. 1633–1638, 2002, and the commentary by A. Nighorn and J. G. Hildebrand on
pp. 1113–1114.
219. See J. Riesgo-Escovar et al. in Proceedings of the National Academy of Sciences, USA,
vol. 92, pp. 2864–2868, 1995, who document a shared function in Drosophila vision and
olfaction for a phospholipase C.
220. Helpful introductions to this fascinating topic can be found in papers by M. B. Fenton in
Quarterly Review of Biology, vol. 59, pp. 33–53, 1984; U. M. Norberg and J. M. V. Rayner in
Philosophical Transactions of the Royal Society of London B, vol. 316, pp. 335–427, 1987; and
E. C. Teeling et al. in Nature, vol. 403, pp. 188–192, 2000. Concerning the phylogenetic
context of bat echolocation, see M. S. Springer et al. in Proceedings of the National Academy
of Sciences, USA, vol. 98, pp. 6241–6246, 2001. Bat echolocation has its examples of
convergence, notably the independent evolution of nasal-emitting sounds in rhinolophids and
phyllostomids; see S. C. Pedersen’s chapter (pp. 174–213) in Ontogeny, functional ecology
and evolution of bats, edited by R. A. Adams and S. C. Pedersen (Cambridge University Press,
Cambridge, 2000); see also E. C. Teeling et al. in Proceedings of the National Academy of
Sciences, USA, vol. 99, pp. 1431–1436, 2002. So, too, two genera in these groups (Rhinolophus
and Pteronotus parnelli) have convergently arrived at Doppler-sensitive sonar; see M. Vater
on p. 145 of the same volume.
221. M. Ruedi and F. Mayer, in Molecular Phylogenetics and Evolution, vol. 21, pp. 436–448, 2001,
discuss evidence for such convergence, and emphasize that ‘Independent adaptive radiations
among species of the genus Myotis therefore produced strikingly similar evolutionary
solutions in different parts of the world’ (p. 436). Various other examples of convergence are
presented, with the important corollary that ‘morphological similarity amongst these bats is
a poor predictor of their genetic similarities’ (p. 444). Ruedi and Mayer are unlikely to please
those biologists who appeal to contingent factors in evolution, given that the title of their
paper includes reference to ‘deterministic ecomorphological convergences’.
222. This ability of the blind to navigate by reflected sound is very variable, but the exceptional
few who surge through city streets, past crowds, obstacles, and traffic, are – at least to the
sighted – a matter of some astonishment. In Chapter 12 his book Listening in the dark: the
notes to pp. 181–182 391
acoustic orientation of bats and men (Yale University Press, New Haven, 1958) D. R. Griffin
has a very interesting discussion of this topic. In the 1974 reprint (Dover, New York) Griffin
draws attention to more recent work, including two papers in Science on the use of sonar by
blind humans, respectively by W. N. Kellogg, vol. 137, pp. 399–404, 1962 and C. E. Rice in
vol. 155, pp. 656–664, 1967.
223. J. P. Rauschecker and U. Kniepert (in European Journal of Neuroscience, vol. 6, pp. 149–160;
1994) discuss how in ‘visually deprived cats’ the auditory skills are markedly
enhanced.
224. For further details see the paper by J. Simmons et al. in Biological Bulletin, vol. 191,
pp. 109–121, 1996.
225. Concerning the perception in moths of the ultrasonic sound frequencies that the bats
generate see the papers by K. D. Roeder in Journal of Experimental Zoology, vol. 134,
pp. 127–157, 1957 (with A. E. Treat) and Animal Behaviour, vol. 10, pp. 300–304, 1962. On
the topic of evasion by the hunted moths see M. B. Fenton and J. H. Fullard in American
Scientist, vol. 69, pp. 266–275, 1981 and T. K. Werner in Canadian Journal of Zoology,
vol. 59, pp. 525–529, 1981. What may well be a convergent example of such sensitivity is the
evidence for ultrasound detection by a clupeid fish (the American shad), which would allow
evasive action on hearing the ultrasonic noises projected by the echolocatory and predatory
cetaceans; see D. A. Mann et al. in Nature, vol. 389, p. 341, 1997 and Journal of the
Acoustical Society of America, vol. 104, pp. 562–568, 1998.
226. This topic is reviewed by W. E. Connor in Journal of Experimental Biology, vol. 202,
pp. 1711–1723, 1999. He points out that while jamming of the bat’s echolocation system is a
widely accepted explanation, alternatives include startling the bat or advertising the moth’s
distastefulness (see, for example, D. C. Dunning and M. Krüger in Biotropica, vol. 27,
pp. 227–231, 1995). A helpful overview of the tactics and counter-tactics of insects and bats,
prey and predator, associated with sound and hearing is given by L. A. Miller and
A-M. Surlykke in BioScience, vol. 51, pp. 570–581, 2001. (See also note 256).
227. For a succinct account of this topic see chapters 5–7 of H. C. Hughes’ enjoyable book Sensory
exotica: A world beyond human experience (MIT Press, Cambridge; 1999).
228. D. R. Griffin (1958; citation is in note 222).
229. D. R. Griffin (1958; citation is in note 222), p. 287.
230. M. Konishi and E. I. Knudsen discuss how oilbirds hear and echolocate in Science, vol. 204,
pp. 425–427, 1979.
231. Concerning the visual system of oilbirds see the article by J. D. Pettigrew and M. Konishi in
National Geographic Society Research Report, vol. 16, pp. 439–450, 1984, as well as
comments by Pettigrew (1991; citation is in note 86). In the former paper they remark that
vision is more important than echolocation, and also draw parallels to the optics of other
nocturnal birds.
232. Details of these experiments are given by D. R. Griffin and R. A. Suthers in Biological
Bulletin, vol. 139, pp. 495–501, 1970. Somewhat similar conclusions were arrived at by M. B.
Fenton in Biotropica, vol. 7, pp. 1–7, 1975, where he noted that even obstacles only 1.5 mm
thick could be avoided. More recent work on echolocation in the swiftlets includes papers by
D. R. Griffin and D. Thompson in Behavioral Ecology and Sociobiology, vol. 10, pp. 119–123,
1982, and D. M. Smyth and J. R. Roberts in Ibis, vol. 125, pp. 339–345, 1983. The actual
mechanism of production of the echolocating clicks is discussed by R. A. Suthers and D. H.
Hector in Journal of Comparative Physiology, vol. 148A, pp. 457–470, 1982.
233. Thus Smyth and Roberts (1983) remark, ‘It is obvious that sensitivity of echolocation [in the
swiftlets] does not match the spectacular feats demonstrated by insectivorous bats in
detecting, ranging and tracking minute objects and in distinguishing between a variety of
shapes and textures’ (p. 345), although more precise comparisons can be made with some of
the fruit-bats, such as Rousettus. This animal belongs to the so-called mega-bats that as a
group evidently lost their ability to echolocate, so suggesting that its presence in the fruit-bat
392 notes to pp. 182–183
Rousettus is a secondary acquisition; see Springer et al. (2001; citation in note 220). One
might also note that among the fruit-bats nectarivory has evolved at least twice, and probably
several times; see J. A. W. Kirsch et al. in Australian Journal of Zoology, vol. 43, pp. 395–428,
1995 and L. J. Hollar and M. S. Springer in Proceedings of the National Academy of Sciences,
USA, vol. 94, pp. 5716–5721, 1997.
234. Electrosensitivity in the star-nosed mole has been posited by Grand et al. (1998; see note 207
for citation; see also Journal of Mammalogy vol. 74, pp. 108–116, 1993), but this proposal
receives short shrift from Catania (2000; see note 195 for citation).
235. Thus the duck-billed platypus is electroreceptive, using glands in its bill; see the two papers
in Nature by respectively H. Scheich et al. (vol. 319, pp. 401–402, 1986) and J. E. Gregory
et al. (vol. 326, pp. 386–387, 1987), as well as the article by U. Proske et al. in Journal of
Comparative Physiology, vol. 173A, pp. 708–710, 1993. So also is the related echidna; see, for
example, J. E. Gregory et al. in Journal of Physiology, vol. 414, pp. 521–538, 1989; K. H.
Andres et al. in Anatomy and Embryology, vol. 184, pp. 371–393, 1991; and P. R. Manger
et al. in Proceedings of the Royal Society of London B, vol. 264, pp. 165–172, 1997. This
electrical sensitivity arose independently of other vertebrates, notably the mormyrid and
gymnotid fish that are discussed below. Details of the electroreceptor structure are given by
Manger and Pettigrew (1996; citation is in note 195), and they also have a helpful discussion
of the similarities and differences in electroreception in the fish and monotremes. They list
three significant differences, but emphasize the more numerous similarities and stress ‘that
particular constraints are acting upon the evolution of electroreception’ (p. 49), where the
convergence emerges by the combination of physical constraints and adaptation to the
environment.
236. Concerning the basic similarities of the strongly convergent electric organs of the fish see the
review by H. H. Zakon and G. A. Unquez in Journal of Experimental Biology, vol. 202,
pp. 1427–1434, 1999, as well as Electric fishes: history and behavior (Chapman and Hall,
London, 1995), a book written largely by P. Moller but with various contributions by a
number of his colleagues.
237. In the context of the early recognition of electric fish, especially with respect to medicine, see
P. Kellaway’s essay in Bulletin of the History of Medicine, vol. 20, pp. 112–137, 1946, and also
D’Arcy Thompson’s discussion of the torpedo ray in A glossary of Greek fishes (Oxford
University Press, London, 1947), pp. 169–171. A more recent overview of the historical
importance of electric fish, including the debate between Galvani and Volta (hence galvanic
and volts), is given in the first chapter of Moller (1995).
238. The only surviving work by Scribonius Largus is his Compositiones medicae; see the edition
prepared by S. Sconocchia (BSB B. G. Teubner, Leipzig, 1983). Reference to the electrical uses
of the torpedo fish may be found in sections 11 and 162, translations of which are given by
Kellaway (1946, p. 130).
239. See, for example, J. M. Riddle’s Dioscorides on pharmacy and medicine (University of Texas
Press, Austin, 1985). Rather little is known of Dioscorides, other than that he was born in
Asia Minor, in Anazarbus, and he studied botany and pharmacology in Tarsus, ‘no mean city’,
as St Paul observed (Acts 22.39). Riddle casts doubt on the popular legend that Dioscorides
marched with the Roman legions.
240. See the edition of The Greek herbal of Dioscorides. Illustrated by a Byzantine A.D. 512.
Englished by J. Goodyer (Oxford University Press, Oxford, 1933), p. 97; ‘The sea torpedo being
applyed in griefs in long continuance about ye head, doth assuage the fiercenesse of the grief:
the same too applyed doth stay up the seate, being either overturned, or else fallen downe’,
p. 97.
241. Not that the treatment fell out of fashion for many centuries. Thus, in his
Panzoologicomineralogia. Or a compleat history of animals and minerals (Godwin, Oxford,
1661) Robert Lovell writes of the aptly named Cramp-fish, i.e. torpedo, with reference to
Dioscorides that ‘Applied to the head, they help old paines thereof, and restraine the falling
out of the fundament’, p. 191.
notes to pp. 183–185 393
242. See Alison Winter’s Mesmerized: Powers of mind in Victorian Britain (University of Chicago
Press, Chicago, 1998).
243. For an account of these experiments see vol. III (May 26, 1836 – Nov. 9, 1839) of Faraday’s
diary, being the various philosophical notes of experimental investigation made by Michael
Faraday, etc., edited by T. Martin (G. Bell and Sons, London, 1933).
244. P. Kellaway (1946; citation is in note 237), also gives some graphic descriptions of treatment
of various ailments in South America using the electric eel, which he refers to as ‘heroic
remedies’ (p. 136). Moreover, a report in B. Kramer (1996; citation is in note 246) says that
‘according to natives, the puraqué (local name for electric eel) is said to harvest palm fruit
which it has been observed to eat, by electroshocking the base of a tree’, p. 18.
245. The phylogenetic relationships of the gymnotids and mormyrids are addressed by J. A. Alves-
Gomes in Journal of Experimental Biology, vol. 202, pp. 1167–1183, 1999, and a detailed
assessment of the gymnotids is provided by J. S. Albert in Miscellaneous Publications
Museum of Zoology, University of Michigan, no. 190, vi + 127 pp., 2001. As Alves-Gomes
points out ‘these two lineages of fish separated more than 140 × 106 years ago, and current
evidence suggests that their common ancestor was neither electroreceptive nor electrogenic.
It is quite remarkable, therefore, that after this prolonged period of independent evolution,
these two fish clades have evolved a number of very elaborate similarities associated with
their sensory and motor biology, completely independently of one another. The electrogenic
and electrosensory systems . . . are used for electrolocation and communication, and the
design, physiology and modus operandi of the two systems are extraordinarily similar, if not
identical, in several ways’, p. 1180.
246. The topic of electric fish and mormyrids specifically has attracted many outstanding
scientists, but specific mention should be made of Hans Lissman, a refugee from Nazi
Germany and a distinguished Cambridge physiologist, who pioneered this field; see, in
particular, his papers in Nature, vol. 167, pp. 201–201, 1951, and Journal of Experimental
Biology, vol. 35, pp. 156–191 (in which he emphasizes the various convergences) and
pp. 457–486, 1958, the latter with K. F. Machin. Otherwise, the key reference to this topic is
the book Electroreception, edited by T. H. Bullock and W. Heiligenberg (Wiley, New York,
1986). Also helpful are special issues of Journal of Comparative Physiology, vol. 173A,
pp. 657–763, 1993, and Journal of Experimental Biology, vol. 202, pp. 1167–1458, 1999; see
also B. Kramer’s Electroreception and communication in fishes, as vol. 42 of Progress in
Zoology (Fischer, Stuttgart, 1996). The topic of electrical generation in fish is an area of
considerable research activity; key groups include those led by C. D. Hopkins (see, for
example, papers in American Zoologist, vol. 21, pp. 211–222, 1981; Current Opinion in
Neurobiology, vol. 5, pp. 769–777, 1995); M. Kawasaki (see Journal of Neuroscience, vol. 17,
pp. 1761–1768, 1997; American Zoologist, vol. 33, pp. 86–93, 1993; Journal of Comparative
Physiology, vols. 173A, pp. 9–22, 1993 and 174A, pp. 133–144, 1994 respectively; Journal of
Neuroscience, vols. 16, pp. 380–391, 1996 and 18, pp. 7599–7611, 1998 respectively); and
B. Kramer (see, for example, Ethology, vol. 103, pp. 404–420, 1997, and Naturwissenschaften,
vol. 84, pp. 119–121, 1997).
247. This is with reference to the active detection of objects near the fish; see G. von der Emde in
Journal of Experimental Biology, vol. 202, pp. 1205–1215, 1999.
248. This ability to deceive the fish was established by Hans Lissman in 1951 (citation is in note
246), where he fed the fish’s own electrical pulses back into the tank, eliciting an attack on
the electrodes. Concerning the employment of an artificial dipole, see Chapter 10 by
P. Moller and J. Serrier, in P. Moller (1995, citation is in note 236).
249. In one group of gymnotids, known as the apteronotids, the electric organ derives from the
nervous system; see S. G. Waxman et al. in Journal of Cell Biology, vol. 53, pp. 210–224, 1972.
250. A general introduction to the process of electrogeneration in fish can be found in M. V. L.
Bennett’s chapter (pp. 347–491) in vol. 5 of Fish physiology, edited by W. S. Hoar and D. J.
Randall (Academic Press, New York, 1971). Also helpful is Zakon and Unquez (1999; citation
is in note 236).
394 notes to pp. 185–186
251. Comments on the four principal types of electric organ in the mormyrid fish and their
evolution are provided by J. Alves-Gomes and C. D. Hopkins in Brain, Behavior and
Evolution, vol. 49, pp. 324–351, 1997. The question of the phylogenetic relationships within
the mormyrids and the evolution of the various electrocytes is pursued further by S. Lavoué
et al. in Molecular Phylogenetics and Evolution, vol. 14, pp. 1–10, 2000, and J. P. Sullivan
et al. in Journal of Experimental Biology, vol. 203, pp. 665–683, 2000.
252. Much of the literature cited in this section of the book is directly relevant to the type of
electrical signals generated by these fish, but a thorough overview is given by Moller (1995;
253. This possibility is explored by P. A. Aguilera et al. in Journal of Experimental Biology,
vol. 204, pp. 185–198, 2000.
254. The role of predation in gymnotid, and probably mormyrid, diversification is addressed by P.
K. Stoddard in Nature, vol. 400, pp. 254–256, 1999.
255. The detection by catfish of a mormyrid, the bulldog fish, is discussed by S. Hanika and B.
Kramer in Naturwissenschaften, vol. 86, pp. 286–288, 1999, and Behavioral Ecology and
Sociology, vol. 48, pp. 218–228, 2000. In the latter paper, they discuss (pp. 226–227) the
inferences drawn by Stoddard (1999) on electric fish diversification but are evidently more
sceptical, at least so far as the mormyrids are concerned.
256. P. K. Stoddard (1999; citation is in note 254) points out that in this example of signalling there
is a parallel in the ctenuchine moths. These insects use an ultrasonic acoustic signal to
advertise themselves to bats, who evidently associate the sound with the disgusting taste of
the moths (they are wasp mimics and contain toxic chemicals), but in addition the signalling
has been co-opted for sexual attraction. This ultrasonic signalling has in turn evolved
independently several times in this group of moths; see papers by W. E. Conner et al. in
Journal of Insect Behavior, vol. 8, pp. 19–31, 1995 and vol. 9, pp. 909–919, 1996 respectively,
as well as a helpful overview (W. E. Connor, 1999; citation is in note 226) on acoustic
communication in moths.
257. See, for example, the paper by P. Belbenoit in Journal de Physiologie (Paris), vol. 59,
pp. 344–345, 1967. The ‘electric images’ that the fish senses are both depicted and discussed
in Journal of Experimental Biology by C. Assad et al. (in vol. 202, pp. 1185–1193, 1999) and
R. Budelli and A. A. Caputi (in vol. 203, pp. 481–492, 2000) respectively, as well as by G. von
der Emde and S. Schwarz in Philosophical Transactions of the Royal Society of London B,
vol. 355, pp. 1143–1146, 2000.
258. In a series of experiments on mormyrid fish P. Cain et al., in Ethology, vol. 96, pp. 33–45,
1994, found that once accustomed to a novel environment they no longer depended on active
electrolocation, with the interesting implication that familiar environments are stored within
the brain as a sort of cognitive map (see also Kawasaki (1997; citation is in note 263) for
evidence of ‘an internal representation’ (p. 474) with respect to the processing of electrical
signals in the context of jamming avoidance. Subsequently P. Cain (in Ethology, vol. 99,
pp. 332–349, 1995) extended this work, and suggested that the relative hydrostatic pressure,
and thereby water depth was an important cue in mormyrid navigation.
259. See the chapter (pp. 313–393) by M. V. L. Bennett in the book Lateral line detectors, edited by
P. H. Cahn (Indiana University Press, Bloomington, 1967). In describing these receptors he
writes ‘The similarities of tonic [for d.c. reception] and phasic [a.c. reception] electroreceptors
in gymnotids and mormyrids constitute an instance of convergent evolution as remarkable as
that of the electric organs themselves’ (p. 386). His earlier contribution in Cold Spring Harbor
Symposia on Quantitative Biology, vol. 30, pp. 245–262, 1965 is also helpful.
260. An apt phrase used by C. D. Hopkins in his review of electrical signals in a mormyrid fish
community, published in American Zoologist, vol. 21, pp. 211–222, 1981, see p. 216.
261. See the paper by P. K. McGregor and G. W. M. Westby in Animal Behaviour, vol. 43, 977–986,
1992. There is also evidence, in at least one gymnotid, for plasticity in the electrical signal;
produced, most probably as a sexual signal; see C. R. Franchina et al. in Journal of
notes to pp. 186–187 395
262. C. D. Hopkins (1981; citation is in note 260), p. 216.

263. The jamming avoidance response (JAR) is discussed by W. Heiligenberg et al. in Journal of
Comparative Physiology, vol. 127, pp. 267–286, 1978, and more recently by M. Kawasaki (and
colleagues) in papers published in Biological Bulletin, vol. 191, pp. 103–108, 1996; Current
Opinion in Neurobiology, vol. 7, pp. 473–479, 1997; and Journal of Experimental Biology,
vol. 202, pp. 1377–1386, 1999; and in the same journal by W. Metzer (vol. 202, pp. 1365–1375)
and B. Kramer (vol. 202, pp. 1387–1398); see also his earlier paper in Journal of Experimental
Biology, vol. 130, pp. 39–62, 1987, where the JAR is discussed in the context of sexual
dimorphism.
264. Concerning the sensitivity and timing of these signals see, for example, C. D. Hopkins and A.
H. Bass in Science, vol. 212, pp. 85–87, 1981, and G. Rose and W. Heiligenberg in Nature,
vol. 318, pp. 178–180, 1985, as well as the overview and update given in M. Kawasaki (1997;
265. See M. Konishi (in Journal of Comparative Physiology, vol. 173A, pp. 700–702, 1993), who
draws attention to the similar neural algorithms possessed by the electric fish, specifically
the gymnotid Eigenmannia, and owls. In the fish the problem is to resolve different signals
that might lead to jamming; in owls to distinguish sounds received by the asymmetrical ears
(which have evolved independently at least five times; see R. Å. Norberg in Philosophical
Transactions of the Royal Society of London B, vol. 280, pp. 375–408, 1977) to enable
localization of source, say, a mouse. Konishi concludes that his ‘findings show that there are
computational rules that transcend different sensory systems and animals’, p. 702.
266. It is worth noting that ‘sound-processing circuits associated with ear asymmetries in owls
[see also note 265] . . . may have evolved as many as five to seven times independently’, see
K. C. Nishikawa in BioScience, vol. 47, pp. 341–354, 1997, p. 341, reporting work by S. F.
Volman in his chapter (pp. 292–314) on directional hearing in owls in Perception and motor
control in birds: an ecological approach, edited by M. N. O. Davies and P. R. Green (Springer,
Berlin, 1994).
267. Once again this convergence has elicited a sense of surprise. For example, C. C. Bell and
T. Szabo remark in their chapter (pp. 375–421) in Electroreception (see note 246 for citation)
that ‘The similarities [between gymnotiform and mormyrid fish] are so striking that it is
difficult to believe in their independent evolution. But the taxonomic distance . . . is so
large . . . electroreception must have evolved independently’ (p. 407).
268. So, too, does the electrosensitive duck-billed platypus. This animal is a highly effective
predator in the streams of Australia, emerging from its burrow to hunt at night, with its eyes
closed; see U. Proske et al. (1993; citation is in note 235).
269. Details of the retinal structure are discussed by M. R. McEwan in Acta Zoologica, Stockholm
vol. 19, pp. 427–465, 1938.
270. M. R. McEwan (1938; citation is in note 269), p. 442.
271. A more recent discussion of the visual abilities of mormyrid fish is given by C. Teyssedre and
P. Moller in Zeitschrift f ür Tierpsychologie, vol. 60, pp. 306–312, 1982, and an overview of
mormyrid vision may be found on pp. 368–370 of Moller (1995, note 236).
272. ‘just another example of convergent development’ as F. Kirschbaum puts it, as part of the title
of his article in Environmental Biology of Fishes, vol. 10, pp. 3–14, 1984.
273. Concerning this convergent evolution of ecology, and especially feeding, in the gymnotids
and mormyrids, see T. R. Roberts in Bulletin of the Museum of Comparative Zoology,
vol. 143, pp. 117–147, 1972, and in Environmental Biology of Fishes the papers by C. Marrero
and K. O. Winemiller (vol. 38, pp. 299–309, 1993) and K. O. Winemiller and A. Adite (vol. 49,
pp. 175–186, 1997) respectively.
274. The role of the electric field of the gymnotid Apteronotus in the capture of its small prey is
discussed by M. E. Nelson and M. A. MacIver in Journal of Experimental Biology, vol. 202,
pp. 1195–1203, 1999. See also the subsequent paper by M. A. MacIver et al. in vol. 204,
pp. 543–557, 2001, which dwells on the various behavioural characteristics that enhance
successful prey interception.
396 notes to pp. 187–190
275. T. R. Roberts (1972; citation is in note 273), p. 141.

276. See C. C. Bell’s paper, entitled ‘Memory-based expectations in electrosensory systems’, in
Current Opinion in Neurobiology, vol. 11, pp. 481–487, 2001.
277. See, in particular, the discussion of distance perception in Gnathonemus by G. von der Emde
et al. in Nature, vol. 395, pp. 890–894, 1998, as well as the commentary by W. Metzner on
pp. 838–839.
278. See B. R. Brown in Journal of Experimental Biology, vol. 205, pp. 999–1007, 2002.
279. Integration of visual and electroreceptive sensory information in the gymnotid Apteronotus
is addressed by J. Bastian in Journal of Comparative Physiology, vol. 147, pp. 287–297, 1982.
An overview of this topic is given on pp. 376–379 of Moller (1995, note 233); see also R. Rojas
and P. Moller in Brain, Behavior and Evolution, vol. 59, pp. 211–221, 2002.
280. See the paper by R. Budelli and A. A. Caputi in Journal of Experimental Biology, vol. 203,
pp. 481–492, 2000, where the authors model the generation of electrical images according to
the impedance of the perceived objects. They draw on analogies of sight and light, and
thereby propose that the system of electroreception allows the discrimination of classes of
objects, i.e. recognition is distant invariant, which they term ‘electrical colors’ (p. 490).
281. See, for example, Synaesthesia: Classic and contemporary readings, edited by S. Baron-Cohen
and J. E. Harrison (Blackwell, Oxford, 1997). That synaesthesia, despite its relative rarity,
might be more than a curiosity is made clear in this book, with implications (perhaps) in the
areas of consciousness and artistic appreciation, which can be taken to include even the
ephemeral but strangely satisfying experience of fireworks (see pp. 29–30). As our
understanding of animal sentience grows, as well as the various sensory modalities that are
accommodated in their different brains, so we may find some deeper clues as to the presently
intangible nature of qualia and other sense perceptions.
282. See, for example, R. Yaka et al. in European Journal of Neuroscience, vol. 11, pp. 1301–1312,
1999.
283. See M. J. Weissburg et al. in Journal of Comparative Neurology, vol. 440, pp. 311–320, 2001.
284. Concerning these possible parallels see E. Mugaini and L. Males et al. in Journal of
Comparative Physiology, vol. 173A, pp. 683–685, 1993. The preceding paper (pp. 682–683) by
S. Coombs et al. on possible neurological similarities of the lateral line, electrosensory
systems, and audition is also directly relevant. While emphasizing both possible differences
(and lack of information) Coombs et al. also comment that ‘Regardless of how these parallels
[between the three sensory modalities] may have arisen, they suggest that there may be some
fundamental principles of organization and cellular interactions that apply to all three
systems’ (p. 682). In a related review J. C. Montgomery et al. (in Auditory Neuroscience,
vol. 1, pp. 207–231, 1995) and C. Bell et al. (in Brain, Behavior and Evolution, vol. 50
(Supplement 1), pp. 17–31, 1997) draw attention to the remarkable similarities in the
operation of cerebellum-like structures in four groups of fish (including the gymnotids and
mormyrids), as well as the dorsal cochlear nucleus of the mammalian auditory system and
the optic tectum of teleost fish.
285. See the papers by G. Diesselhorst and E. Stipetic in Zeitschrift f ür Vergleichende Physiologie,
286. See H. Y. Yan and W. S. Curtsinger in Journal of Comparative Physiology, vol. 186A,
pp. 595–602, 2000 and L. B. Fletcher and J. D. Crawford in Journal of Experimental Biology,
vol. 204, pp. 175–183, 2001.
287. Sound production in this mormyrid is discussed by J. D. Crawford and X-f. Huang in Journal
of Experimental Biology, vol. 202, pp. 1417–1426, 1999. It is also interesting to note that the
sound frequencies generated by this mormyrid imply exceptionally fast contraction of the
drumming muscles, perhaps without rival among the vertebrates.
288. This is one of the triumphs of palaeontological investigation, revolving around key work on
the Upper Devonian Acanthostega (and related forms) by J. A. Clack; see her Gaining
ground: The origin and evolution of tetrapods (Indiana University Press, Bloomington, 2002).
notes to pp. 190–192 397
289. Concerning aspects of the potentially radical transformations between aquatic sound pressure
perception and subaerial hearing from a neurological perspective, see B. Fritsch in Brain,
Behavior and Evolution, vol. 50, pp. 38–49, 1997.
290. See the overview by J. A. Clack in Brain, Behavior and Evolution, vol. 50, pp. 198–212, 1997,
as well as J. A. Clack (2002; citation is in note 288).
291. See, for example, the article by R. Nobili et al. entitled ‘How well do we understand the
cochlea?’ in Trends in Neurosciences, vol. 21, pp. 159–167, 1998.
292. Concerning these sensory cells, G. A. Manley et al., in Journal of Comparative Physiology,
vol. 164A, pp. 289–296, 1989, wrote ‘Considering the independent evolutionary origin of
these two vertebrate classes from different groups of reptile ancestors, it is remarkable to find
indications of a clear functional parallel between avian and mammalian hair-cell populations.’
They then continue with the important observation that ‘The parallel evolution of such
hair-cell populations suggests that there are fundamental properties of hair cells which lend
themselves to a particular kind of division of labour.’ An update and overview is given by G.
A. Manley and C. Köppl in Current Opinion in Neurobiology, vol. 8, pp. 468–474, 1998.
293. The question of the adaptive optimal sound for vocal communication in the tunnels of the
blind mole-rat is addressed by G. Heth et al. in Experientia, vol. 42, pp. 1287–1289, 1986.
294. Seismic communication is reviewed by E. Nevo in Mosaic evolution of subterranean
mammals: Regression, progression, and global convergence (Oxford University Press, Oxford,
1999), see especially pp. 105–107; see also M. J. Mason and P. M. Narins, in American
Zoologist, vol. 41, pp. 1171–1184, 2001.
295. The similarities in the method of sound conduction is addressed by H. Burda et al. in Journal
of Morphology, vol. 214, pp. 49–61, 1992. A notable exception to this example of convergence
is the stapes ear-bone, which shows remarkable shape variation, even within a genus.
296. See the paper by T. Lindenlaub et al. in Journal of Morphology, vol. 224, pp. 303–311, 1995.
See also note 200.
297. The astonishing acoustic sensitivity of the mosquito, especially in the male to whom the
wing-beats of the female spell a special magic, is reported by M. C. Göpfert and D. Robert in
Proceedings of the Royal Society of London B, vol. 267, pp. 453–457, 2000. Göpfert and
Robert also point out that while the female’s hearing is less acute, it still far outstrips that of
most other insects.
298. See their paper, ‘Active auditory mechanics in mosquitoes’, in Proceedings of the Royal
Society of London B, vol. 268, pp. 333–339, 2000.
299. M. C. Göpfert and D. Robert (2000; citation is in note 298), pp. 337–338.
300. See the paper by J. de Wilde in Archives Neerlandaises de Physiologie de l’homme et des
animaux, vol. 25, pp. 381–400, 1941.
301. For an overview see R. R. Hoy and D. Robert in Annual Review of Entomology, vol. 41,
pp. 433–450, 1996. Hoy and Robert also touch on the question of convergence, a conclusion
well supported by the remarkably diverse positions of the tympanal organ across the body
surface of the insect.
302. Interestingly, the chordotonal organ (see L. H. Field and T. Matheson in Advances in Insect
Physiology, vol. 27, pp. 1–228, 1998) appears to have been co-opted from its primary function,
which is to act as a stretch receptor for movement of various parts of the body, e.g. segments.
Concerning this derivation see G. S. Boyan in Journal of Insect Physiology, vol. 39,
pp. 187–200, 1993, and M. J. van Staaden and H. Römer in Nature, vol. 394, pp. 773–776, 1998.
303. See, for example, the discussion of tympanal (and atympanal) hearing in hawkmoths, by M.
C. Göpfert et al. in Proceedings of the Royal Society of London B, vol. 269, pp. 89–95, 2002.
This example is relevant to the discussion of convergence because these ears ‘evolved
independently on the basis of homologous structures’, p. 93.
304. Although this is not to say that crickets cannot defend themselves from digger wasps that are
attempting to parasitize them by one of several mechanisms, including kicking; see W.
Gnatzy and R. Heusslein in Naturewissenschaften, vol. 73, pp. 212–215, 1986.
398 notes to pp. 192–193
305. See the paper by D. Robert et al. in Science, vol. 258, pp. 1135–1137; 1992. As they aptly
remark ‘This example of convergent evolution in a hearing organ demonstrates the
constraints on morphological design that are imposed by behavioral function as well as by
principles of physical acoustics’ (p. 1135). An analysis of this ear’s morphological origins from
pre-existing structures is provided by R. S. Edgecomb et al. in Cell & Tissue Research,
vol. 282, pp. 251–268, 1995, and in a subsequent issue (vol. 301, pp. 447–457) D. Robert and U.
Willi discuss the histological architecture of the ormiinid ear, and also echo the earlier
remarks when they comment ‘It is quite remarkable that in insects the sense of hearing has
probably evolved many times independently in at least seven orders, and that, in each
instance, however diverse the auditory organs are, they are of the chordotonal type endowed
with scolopidial mechanoreceptors’, pp. 454–455.
306. D. Robert et al. (1992; citation is in note 305), p. 1137.
307. A suggestion made by D. Robert et al. in Cell & Tissue Research, vol. 284, pp. 435–448, 1996.
308. See the paper by A. C. Mason et al. in Nature, vol. 410, pp. 686–690, 2001, and the
accompanying commentary by P. M. Narins on pp. 644–645.
309. Thus D. Robert et al. (1996; citation is in note 307) note ‘Remarkably, morphological and
behavioral evidence points to the presence of prosternal hearing organs in two calyptrate fly
species from another big family of parasitoids, the Sarcophagidae . . . Hence, it is likely that
acoustic parasitism evolved twice independently within calyptrate flies’, p. 446.
310. See the paper by R. G. Walker et al. in Science, vol. 287, pp. 2229–2234, 2000. Their specific
example touches on mechanosensory transduction in the fruit-fly and a homologue in worms
(i.e. C. elegans), but the link to vertebrate hair cells is also addressed.
311. The role of Math 1 in the generation of the inner hair cells is addressed by N. A. Bermingham
et al. in Science, vol. 284, pp. 1837–1841, 1999.
312. For the activity of atonal as a proneural gene, see A. P. Jarman et al. in Development, vol. 121,
pp. 2019–2030, 1995.
313. The role of atonal in the glomeruli of the olfactory lobe is addressed by D. Jhaveri and V.
Rodriques in Development, vol. 129, pp. 1251–1260, 2002.
314. The literature on the Pax-6 gene is very extensive; it includes the following papers: R. L.
Chow et al. in Development, vol. 126, pp. 4213–4222, 1999; A. Cvekl and J. Piatigorsky in
BioEssays, vol. 18, pp. 621–630, 1996; C. Desplan in Cell, vol. 19, pp. 861–864, 1997;
G. Halder et al. in Science, vol. 267, pp. 1788–1792, 1995; R. Quiring et al. in Science,
vol. 265, pp. 785–789, 1994; S. I. Tomarev et al. in Proceedings of the National Academy of
Sciences, USA, vol. 94, pp. 2421–2426, 1997; and C. Punzo et al. in Genes & Development,
vol. 15, pp. 1716–1723, 2001.
315. For this suggestion, and a consideration of the early evolution of the Pax genes see D. J. Miller
et al. in Proceedings of the National Academy of Sciences, USA, vol. 97, pp. 4475–4480, 2000.
316. The role of Pax-6 in nasal development is addressed by J. C. Grindley et al. in Development,
vol. 121, pp. 1433–1442, 1995; see also M-D. Franco et al. in Journal of Experimental Biology,
vol. 204, pp. 2049–2061, 2001, who among other matters point out that Pax-6 appears to play
different roles in development of olfactory epithelia in frog and mouse (see p. 2059).
317. Pax-6 expression in the brain is addressed by M. Takahashi and N. Osumi in Development,
vol. 129, pp. 1327–1338, 2002; P. Callaerts et al. in Journal of Neurobiology, vol. 46,
pp. 73–88, 2001; and several earlier papers in Development, see vol. 113, pp. 1435–1450, 1991
(C. Walther and P. Gruss), vol. 126, pp. 5569–5579, 1999 (P. Chapouton et al.), vol. 127,
pp. 2357–2365 (E. Matsunaga et al.), and vol. 127, pp. 5267–5178, 2000 (R. Pratt et al.),
as well as M. Schwarz et al. in Mechanisms of Development, vol. 82, pp. 29–39,
1999.
318. See C. Kioussi et al. in Proceedings of National Academy of Sciences, USA, vol. 96,
pp. 14378–14382, 1999.
319. The role of Pax-6 in the glucagon producing α cells of the pancreas is addressed by L. St-Onge
et al. in Nature, vol. 387, pp. 406–409, 1997.
notes to pp. 193–199 399
320. D. Arendt et al. in Development, vol. 129, pp. 1143–1154, 2002, present some interesting data
on polychaete eyes, noting Pax-6 activity only in the larval eyes, and not those of the adult.
321. See, for example, D. A. Nelson and P. Marler’s discussion of categorical perception of birdsong
in wild swamp sparrows in Science, vol. 244, pp. 976–978, 1989.
322. See R. A. Wyttenbach et al. in Science, vol. 273, pp. 1542–1544, 1996.
323. R. A. Wyttenbach et al. (1996; citation is in note 322), p. 1543.
324. See R. Nieuwenhuys and C. Nicholson’s chapter (pp. 107–134) in Neurobiology of cerebellar
evolution and development, edited by R. Llinás (American Medical Association, Chicago,
1969). The following chapter (pp. 135–169), by the same authors, provides details of the
histology of the mormyrid brain. More recent discussions are in the chapter (pp. 375–421) by
C. C. Bell and T. Szabo in Bullock and Heiligenberg (1986; see note 246 for citation), and an
article by Szabo in Acta Morphologica Hungarica, vol. 31, pp. 219–234, 1983.
325. R. Nieuwenhuys and C. Nicholson (1969; citation is in note 324), p. 107.
326. See the paper by G. E. Nilsson in Journal of Experimental Biology, vol. 199, pp. 603–607,
1996. He stresses that the extraordinary energetic cost in the mormyrids is not only because
of a large brain, but also because the fish are ectothermic. Moreover, given that these fish may
find themselves in poorly aerated water, it is not surprising to learn of their ability to
scavenge oxygen effectively, although L. J. Chapman and K. G. Hulen (in Journal of Zoology,
London, vol. 254, pp. 461–472, 2001) question whether even increase in gill size is really
sufficient.
327. See J. S. Albert et al. on pp. 647–656 of Proceedings of the 5th Indo-Pacific Fish Conference,
Nouméa, 1997, edited by B. Séret and J-Y. Sire (Société Française d’Ichtyologie, Paris, 1999).
Interestingly, among the fish it is the chondrichthyans, notably the galeomorph sharks and
batoid rays, which have proportionally the largest of brains.
328. See the chapter (pp. 319–373) by C. E. Carr and L. Maler in T. H. Bullock and W. Heiligenberg
(1986; citation is in note 246).
329. See C. Bell et al. (1997; citation is in note 284).
8. alien convergences?
1. These two genera belong to a larger group of ants, known as the attines (Tribe Attini), all of
which are fungus-growers. This section very largely concentrates on the advanced leaf-cutters.
2. Useful introductions are available in K. Dumpert’s book The social biology of ants (Pitman,
Boston, 1981); the masterly work by B. Hölldobler and E. O. Wilson entitled simply The ants
(Springer, Berlin, 1990); R. C. Buckley’s overview on pp. 111–114 (excluding bibliography) of
Ant–plant interactions in Australia, edited by R. C. Buckley (Junk, The Hague, 1982); N. A.
Weber’s article forming vol. 92 of Memoirs of the American Philosophical Society,
Philadelphia (1972); and J. K. Wetterer’s chapter (pp. 309–328) in Nourishment and evolution
in insect societies, edited by J. H. Hunt and C. A. Nalepa (Westview Press, Boulder, CO,
1994).
3. More primitive attine species are more opportunistic and collect a wide variety of other
material, including flowers, seeds, faeces, and insect corpses; see I. R. Leal and P. S. Oliveira
in Insectes Sociaux, vol. 47, pp. 376–382, 2000.
4. See the article on Acromyrmex in the Arizona desert by J. K. Wetterer et al. in Sociobiology,
vol. 37, pp. 633–649, 2001.
5. The species in question is found in Paraguay, and details are given by H. G. Fowler and S. W.
Robinson in Ecological Entomology, vol. 4, pp. 239–247, 1979. Further information is
provided by S. P. Hubbell et al. in Biotropica, vol. 12, pp. 210–213, 1980. These researchers
noted that almost never did the tree-top harvesters carry any material back down to the
ground, and that the ants lugging the plant material back to the nest were significantly more
robust than their companions dropping the pieces from above. F. Roces, in Biological
Bulletin, vol. 202, pp. 306–313, 2002, has an interesting discussion of the balance between
colony-wide activity and individual autonomy.
400 notes to pp. 199–200
6. See Hubbell et al. (1980). The adaptive benefits of this leaf transfer, in the context of speeds of
locomotion, are addressed by C. Anderson and J. L. V. Jadin in Insectes Sociaux, vol. 48,
pp. 404–405, 2001. An overview of insect ‘bucket brigades’, which have evidently evolved
independently several times, is given by C. Anderson et al. in Insectes Sociaux, vol. 49,
pp. 171–180, 2002.
7. The benefits and drawbacks of leaf caching by the ant Atta are addressed by A. G. Hart and F.
L. W. Ratnieks in Animal Behaviour, vol. 62, pp. 227–234, 2001.
8. Concerning the energetic cost to the colony of clearing and maintaining these trails, see J. J.
Howard’s analysis in Behavioral Ecology and Sociobiology, vol. 49, pp. 348–356, 2001.
9. The highway patrol undertaken by the attine minors is documented by W. O. H. Hughes and
D. Goulson in Behavioral Ecology and Sociobiology, vol. 49, pp. 503–508, 2001.
10. See T. A. Linksvayer et al. in Biotropica, vol. 34, pp. 93–100, 2002.
11. Although the cut surfaces do provide a source of nutriment via the plant sap; see, for example,
M. Littledyke and J. M. Cherrett in Bulletin of Entomological Research, vol. 66, pp. 205–217,
1976.
12. See R. J. Quinlan and J. M. Cherrett in Ecological Entomology, vol. 2, pp. 161–170, 1977.
13. Interestingly, but perhaps not surprisingly, the enzymatic activity of the fungi and that in the
ants’ digestive system are distinct, each tending to attack different substrates. Such a lack of
overlap is consistent with this mutualism system being ancient. It is also possible that one of
the enzymatic properties the larvae possess has been transferred from the fungus. See P.
D’Ettorre et al. in Journal of Comparative Physiology, vol. 172B, pp. 169–176, 2002.
14. Details of this process are given by C. R. Currie and A. E. Stuart in Proceedings of the Royal
Society of London B, vol. 268, pp. 1033–1039, 2001. Weeding makes horizontal (or lateral)
transmission between nests substantially more difficult, but helps to ensure vertical
transmission (from generation to generation) within the nest; see A. N. M. Bot et al. in
Evolution, vol. 55, pp. 1980–1991, 2001.
15. As F. Roces and C. Kleineidam observe in Insectes Sociaux (vol. 47, pp. 348–350, 2000)
‘Relocation of fungus gardens to promote [fungal] growth, like the repotting of flowering
plants by humans, clearly illustrates the skill of leaf-cutting ants as true “gardeners”,’
p. 350.
16. At least on the basis of experimental work; see M. Bollazzi and F. Roces in Insectes Sociaux,
vol. 49, pp. 153–157, 2002.
17. See M. Bass and J. M. Cherrett in Functional Ecology, vol. 10, pp. 55–61, 1996.
18. Manuring is addressed by M. M. Martin on pp. 163–166 of his chapter reviewing the use of
ingested enzymes in insects in Invertebrate–microbial interactions, edited by J. M. Anderson
et al. (Cambridge, Cambridge University Press, 1984).
19. The last of these compounds includes sterols that are essential for hormone synthesis; see
P. Maurer et al. in Archives of Insect Biochemistry and Physiology, vol. 20, pp. 13–21, 1992.
20. The threats posed to the attine ant colonies by this parasitic fungus are graphically brought
out by C. R. Currie in Oecologia, vol. 128, pp. 99–106, 2001. See also C. R. Currie et al. in
Science, vol. 299, p. 386–388, 2003. Smaller colonies are at particular risk, whereas the larger
ones remain clear of infection.
21. In fact the use of antibiotics is widespread in the social insects and termites; see R. B.
Rosengaus et al. in Journal of Chemical Ecology, vol. 26, pp. 21–39, 2000.
22. For more details see the paper by C. R. Currie et al. in Nature, vol. 398, pp. 701–704, 1999, as
well as the overview by C. R. Currie in Annual Review of Microbiology, vol. 55, pp. 357–380,
2001.
23. These include secretions from the metapleural glands, which are especially well developed in
the attines. See, for example, the papers by D. Ortius-Lechner et al. in Journal of Chemical
Ecology, vol. 26, pp. 1667–1683, 2000; A. N. M. Bot et al. in Insectes Sociaux, vol. 48,
pp. 63–66, 2001; and M. Poulsen et al. in Behavioural Ecology and Sociobiology, vol. 52,
pp. 151–157, 2002. Experiments in which a fungicide is introduced quickly lead to a change in
notes to pp. 200–201 401
foraging behaviour, which may in addition lead to this behaviour being transmitted to
another colony; see R.D. North et al. in Physiological Entomology, vol. 24, pp. 127–133,
1999.
24. See M. Poulsen et al. in Insectes Sociaux, vol. 49, pp. 15–19, 2002, who document these
alternative defence mechanisms in two species of Acromyrmex.
25. This case of agro-predation by the ant Megalomyrmex is described by R. M. M. Adams et al.
in Naturwissenschaften, vol. 87, pp. 549–554, 2000. In this situation these ants have, of
course, to manipulate the fungi, and perhaps not surprisingly their ‘behavior closely
resembles that of attine ants when they add substrate to or reconstruct gardens . . . suggesting
that it has been convergently derived in Megalomyrmex sp. ants’ (p. 551). These authors also
point out that the continued health of the occupied gardens suggests that ‘Megalomyrmex sp.
may have evolved a similar repertoire of maintenance behaviors and chemical secretions’
(p. 550) to those of the attine ants.
26. For a graphic account of the defence and eventual overwhelming of an attine colony by the
army ant Nomamyrex see the article by M. B. Swartz in Biotropica, vol. 30, pp. 682–684, 1998.
27. See, in particular, W. H. Gotwald’s Army ants: the biology of social predation (Cornell
University Press [Comstock], Ithaca; 1995). The best-documented convergence of army ant
adaptive syndrome is between the New World Eciton and Old World taxa such as Dorylus and
Aenictus. Gotwald also draws attention to at least two more instances of army ant-like
behaviour, these being in two species of Pheidologeton where his comments include the
observation that ‘their foraging behavior is uncanny in its resemblance to that of army ants’
(p. 38), and also in the minuscule leptanillinine ants. A helpful summary of the behaviour of
leptanillinine ants and Pheidologeton is given on pp. 590–592 and pp. 593–594 of B.
Hölldobler and E. O. Wilson (1990; citation is in note 2).
28. The reality versus the folklore of army ants is addressed by Gotwald (1995), see pp. 237–238,
249, 252. In agriculture army ants may be helpful in pest-control, but they can also be
destructive of crops. One should note that the mandibles of ants have been used to suture
both natural and surgical wounds, pp. 249–250.
29. See N. R. Franks’ article ‘Army ants: A collective intelligence’ in American Scientist, vol. 77,
pp. 138–145, 1989. So, too, S. C. Pratt et al. (in Behavioral Ecology and Sociobiology, vol. 52,
pp. 117–127, 2002) describe how in the context of colony emigration another ant
(Leptothorax) society behaves ‘as a single information-processing unit’, (p. 117). These
authors draw attention to similar quorum sensing, not only in other social insects, notably
bee (see note 36), but even bacteria and immune response T-cells.
30. Well, not all of them, because the army ants have their own predators, including birds and
such anteaters as the pangolins.
31. As N. R. Franks et al. (in Proceedings of the Royal Society of London B, vol. 266,
pp. 1697–1701, 1999) show, the net result in New World Eciton and Old World Dorylus is
achieved by rather different means. In essence Dorylus uses more and smaller workers to
carry collectively larger loads, but at a slower speed. Further insights into the convergent
teamwork in Eciton and Dorylus are provided by N. R. Franks et al. in Animal Behaviour,
vol. 62, pp. 635–642, 2001. Here they show that the division of labour typically involves a
team with an unusually large and unusually small ant. As the authors point out, an analogy
for this exists in the form of the penny-farthing bicycle, where each wheel has a separate
sub-task, the ‘penny’ providing load-bearing and drive, the ‘farthing’ stability. The net result
is a synergistic contraption whose effectiveness is more than the sum of the parts.
32. This is only one example of the recurrent property of self-assembly in social insects, a topic
reviewed by C. Anderson et al. in Insectes Sociaux, vol. 49, pp. 99–110, 2002. These workers
identify at least 18 different ways in which insect individuals link themselves together to
form structures as disparate as bivouacs, bridges, ladders, and rafts. Not only are many of
these behaviours convergent, but, as importantly, they reflect new levels of complexity
unavailable to simpler societies.
402 notes to pp. 201–203
33. This topic is addressed by N. R. Franks in Physiological Entomology, vol. 14, pp. 397–404,
1989.
34. See G. P. Boswell et al. in Proceedings of the Royal Society of London B, vol. 268,
pp. 1723–1730, 2001.
35. See, for example G. J. Vermeij’s article in American Naturalist, vol. 120, pp. 701–720, 1982,
especially pp. 710–713.
36. The complexity and sophistication of the bee colony is probably still underestimated, and
their degree of integration with selection operating at the group level of the colony (see, for
example, the article by T. D. Seeley in American Naturalist, vol. 150, pp. 522–541, 1997)
makes them, at least potentially, an interesting evolutionary alternative to the emergence of
advanced cognizance. Consider, for example, how a bee swarm dispatches scout bees to
numerous sites that are then ‘assessed’ before a collective decision is arrived at; see
S. Camazine et al. in Insectes Sociaux, vol. 46, pp. 348–360, 1999 and T. D. Seeley and
S. C. Buhrmann in Behavioral Ecology and Sociobiology, vol. 45, pp. 19–31, 1999.
37. For example, T. S. Collett et al., in Journal of Comparative Physiology, vol. 181A,
pp. 343–353, 1997, present evidence for contextural learning in honey-bees, and M. Giurfa
et al. (in Nature, vol. 410, pp. 930–933, 2001) show that not only can bees place the outside
world in a context and memorize it, but engage in higher-order functions, that is they can
‘master abstract interrelationships, such as sameness and difference’, p. 930.
38. See the papers by W. Kaiser in Nature, vol. 301, pp. 707–709, 1983 (with J. Steiner-Kaiser) and
Journal of Comparative Physiology, vol. 163A, pp. 565–584, 1988. In the latter paper Kaiser
remarks how four characteristics – decreased motility, lowered body temperature, reduced
muscle tone, and raised reaction threshold – ‘strongly resemble [those] characteristic features
of sleep in humans, mammals and birds’ (p. 565). So do bees dream?
39. See, for example, G. Buzsaki in Journal of Sleep Research, vol. 7 (Supplement 1), pp. 17–23,
1998, as well as more recent contributions in Nature Neuroscience, vol. 3, pp. 1237–1238 (R.
Stickgold et al.) and pp. 1335–1339 (S. Gais et al.), 2000. A useful overview of the adaptive
contexts of sleep is given by J. L. Kavanau in Animal Behaviour, vol. 62, pp. 1219–1224, 2001,
and in his earlier article in Brain Research Bulletin, vol. 42, pp. 245–264, 1997, where the
potential importance of endothermy (see p. 223 of this Chapter) is stressed.
40. The honey bees themselves show an interesting convergence with another eusocial species,
the common wasp. This entails so-called ‘worker policing’ in which any eggs laid by the
workers, which will be parthenogenetic and male, are quickly destroyed by other members of
the squad – egg production of course being the prerogative of the queen. As K. R. Foster and F.
L. W. Ratnieks, in Proceedings of the Royal Society of London B, vol. 268, pp. 169–174, 2001,
remark ‘The worker policing by mutual egg-eating in V. vulgaris (the common wasp) is
strikingly similar to that found in the honeybee A. mellifera . . . This is not due to common
ancestry since the vespine wasps and honeybees belong to lineages that have evolved
eusociality independently’ (p. 172). There are, however, differences between these insects in
terms of their genetic relatedness of the workers: effectively honeybee queens are
polyandrous, that is they receive multiple matings on their nuptial flights, so workers are not
closely related and there is a premium in destroying any workers’ eggs (‘not your genes,
destroy! destroy!’). In the wasps, however, relatedness of the workers is higher, and so another
explanation for the policing has to be found. Here Foster and Ratnieks suggest that the benefit
may arise from reducing intra-colony reproductive aggression. They also remark that such
policing may well be a feature of leaf-cutter and other ant colonies; see also remarks by
P. Villesen et al. (in Proceedings of the Royal Society of London B, vol. 269, pp. 1541–1548,
2002) on the convergent emergence of multiple matings in leaf-cutter ants and other very
advanced social insects.
41. Stingless they may be, but they are highly effective in defending the colony, ‘locking their
mandibles in catatonic spasms so that before the grip is broken, their heads tear loose from
the body. The Trigona flaveola group of species in tropical America also eject a burning liquid
notes to pp. 203–205 403
from their mandibles which in Brazil has earned them the name cagafogos, meaning ‘fire
defecators’, p. 89 of E. O. Wilson’s The insect societies (Belknap, Cambridge, MA; 1971). For
an interesting account of the highly sophisticated methods of communication in the stingless
bees, see the article by J. C. Nieh in American Scientist, vol. 87, pp. 428–435, 1999, which he
notes has probably evolved independently of the honey-bee; see also note 43.
42. On this topic see the paper by M. S. Engel in Proceedings of the National Academy of
43. See the paper by S. A. Cameron and P. Mardulyn in Systematic Biology, vol. 50, pp. 194–214,
2001. Their conclusions are highly resonant with the thrust of this book: ‘If indeed the highly
eusocial behavior of Apini and Meliponini evolved twice independently – the hypothesis we
support as being more likely – then the comparative behavior of these bees takes on an added
interest from an evolutionary perspective. The apparent convergent similarities in behavior
between these two tribes suggest that highly eusocial organization has limited permutations.
Each tribe has been channeled along a similar behavioral track, responding in similar fashion
to similar contingencies’, p. 209. The notion of convergence in advanced eusocial behaviour
actually goes back many years, see, for example, M. L. Winston and C. D. Michener in
Proceedings of the National Academy of Sciences, USA, vol. 74, pp. 1135–1137, 1977. These
authors are also struck by the implications of convergence; they write, ‘The development of
such [highly social] behavior, probably the most elaborate of any invertebrate, is remarkable
enough; independent parallel development of two highly eusocial systems would be even
more noteworthy’ (p. 1135). Engel (2001), however, regards advanced eusociality in bees as
monophyletic (as does F. B. Noll in Cladistics, vol. 18, pp. 137–153, 2002). He bases his
analysis on the fossil record, especially amber preservation. S. A. Cameron and P. Mardulyn
(2001) come to the same conclusion on the basis of morphological data, but this is in conflict
with molecular data.
44. The general consensus is that advanced eusocial species, with such features as division of
labour and colony homeostasis, are effectively ‘locked in’ to this arrangement. It should be
pointed out that there is evidence that in the case of less organized social behaviours this does
not preclude a return to a solitary mode of life and that this has occurred independently a
number of times. See, for example, the papers by R. Gadagkar (in Current Science, vol. 72,
pp. 950–956, 1997) and W. T. Wcislo and B. N. Danforth (in Trends in Ecology & Evolution,
vol. 12, pp. 468–474, 1997), as well as the useful overview by B. J. Crespi, on pp. 253–287 of
Phylogenies and the comparative method in animal behavior, edited by E. P. Martins (Oxford
University Press, New York, 1996).
45. M. S. Engel (2001; citation is in note 42); see p. 1663. This writer is careful to emphasize that
competition need not be the sole cause of extinction in the eusocial bees, and that climate
cooling might also play some role.
46. Part of the title of G. P. Boswell et al. (2001; citation is in note 34).
47. N. R. Franks (1989; citation is in note 29), p. 139.
48. Ulrich Mueller reminds me that pheromones are not the only currency of exchange in ant
communication, and that tactile contact and food exchange can also be important.
49. N. R. Franks (1989; citation is in note 29), p. 144.
50. See his chapter (pp. 281–298) in Insect movement: mechanisms and consequences, edited by
I. P. Woiwod et al. (CABI, Wallingford, 2001, p. 286, his emphases). See also the papers by E.
Bonabeau entitled ‘Social insect colonies as complex adaptive systems’ in Ecosystems, vol. 1,
pp. 437–443, 1998, and B. J. Cole in Biological Bulletin, vol. 202, pp. 256–261, 2002.
51. See the stimulating article by A. D. M. Rayner and N. R. Franks on ‘Evolutionary and
ecological parallels between ants and fungi’ in Trends in Ecology & Evolution, vol. 2,
pp. 127–133, 1987.
52. C. Kleinidam et al. (in Naturwissenschaften, vol. 88, pp. 301–305, 2001) have confirmed that
the ventilation is wind-induced, a not surprising necessity given the large quantities of
oxygen required by the inhabitants and their fungi.
404 notes to pp. 205–208
53. See, for example, the remarkable photograph (Fig. 20c) in the paper on nest structure
by J. C. M. Jonkman in Zeitschrift f ür Angewandte Entomologie, vol. 89, pp. 217–246, 1980.
54. See A. G. Hart and F. L. W. Ratnieks in Behavioral Ecology and Sociobiology, vol. 49,
pp. 387–392, 2001, and also Behavioral Ecology, vol. 13, pp. 224–231, 2002.
55. Waste dumps are frequently infected with the parasitic fungus Escovopsis (see note 20) and
mortality of the workers can be significantly increased if the waste material is not quickly
cleared; see A. N. M. Bot et al. in Ethology, Ecology and Evolution, vol. 13, pp. 225–237, 2001.
Bot et al. also remark that ‘There are a number of interesting parallels between waste
management in agricultural societies of ant and human. Both isolate waste in specific places:
ants use refuse chambers or piles where we use landfill sites . . . natural selection has probably
led to a series of general attitudes of disgust towards different types of waste products . . .
Apparently, natural selection and evolution have produced converged methods by which
to address these problems in both human and insect societies,’ p. 235.
56. See the paper by U. G. Mueller et al. in Science, vol. 281, pp. 2034–2038, 1998, as well as the
stimulating commentary by J. Diamond on pp. 1974–1975, where he explores the similarities
to and differences from human agriculture. The likely origins of the attine ant-fungus
association are addressed by U. G. Mueller et al. in Quarterly Review of Biology, vol. 76,
pp. 169–197, 2001.
57. See The origins of agriculture: An international perspective, edited by C. W. Cowan and
P. J. Watson (Smithsonian, Washington, DC, 1992). In particular, the chapter (pp. 143–171) by
E. M. de Tapia specifically rejects the generally popular notion of population increase in the
instance of Mesoamerica.
58. See U. G. Mueller et al. (2001; citation is in note 56).
59. N. A. Weber (1972; citation is in note 2), p. 1.
60. J. C. M. Jonkman in Zeitschrift f ür Angewandte Entomologie, vol. 89, pp. 135–143, 1980
provides some interesting figures on both population sizes and biomass accumulated by the
colony.
61. This topic is addressed by R. M. M. Adams et al. in Naturwissenschaften, vol. 87, pp. 491–493,
2000; see also A. M. Green et al. in Molecular Ecology, vol. 11, pp. 191–195, 2002.
62. Quoted (p. 763) in the article by L. Ariniello in Bioscience, vol. 49, pp. 760–763, 1999.
63. See B. D. Farrell et al. in Evolution, vol. 55, pp. 2011–2027, 2001, and the review by R. A.
Beaver on pp. 121–143 in Insect–fungus interactions, edited by N. Wilding et al. (Academic
Press, London, 1989). See also C. C. Labandeira et al. in American Journal of Botany, vol. 88,
pp. 2026–2039, 2001.
64. See A. R. Berryman’s chapter (pp. 145–159) in Wilding et al. (1989).
65. A. R. Berryman (1989), p. 149.
66. See the overview by T. D. Paine et al. in Annual Review of Entomology, vol. 42, pp. 176–206,
1997.
67. Evolution of defensive compounds, such as resin and latex in plants, is rampantly convergent.
B. D. Farrell et al., in American Naturalist, vol. 138, pp. 881–900, 1991, estimate that ‘Canal
systems containing latex or resin appear to have originated at least 40 times,’ p. 889.
68. B. D. Farrell et al. (2001; citation is in note 63), p. 2017.
69. See the chapter (pp. 105–130) by J. P. E. C. Darlington in J. H. Hunt and C. A. Nalepa (1994;
citation is in note 2). See also D. K. Aanen et al. (in Proceedings of the National Academy of
Sciences, USA, vol. 99, pp. 14887–14892, 2002), who document the evolution of this
mutualism, as well as the differences and similarities between termites and attine ants, and
J. Korb and D. K. Aanen in Behavioural Ecology and Sociobiology, vol. 53, pp. 65–71, 2003).
70. See the description by J. P. E. C. Darlington in Journal of Zoology, London, vol. 198,
pp. 237–247, 1982, which together with Darlington (1994) provides the information given
here; see also G. Schuurman and J. M. Dangerfield in Sociobiology, vol. 27, pp. 29–38, 1996.
71. The example of convergent evolution between the Echidnophora and Termitoxeniiae is given
by R. H. L. Disney in Systematic Entomology, vol. 20, pp. 195–206, 1995.
notes to pp. 208–213 405
72. The molecular phylogenetics of Termitomyces are addressed by C. Rouland-Lefèvre et al. in

Molecular Phylogenetics and Evolution, vol. 22, pp. 423–429, 2002.
73. R. Heim’s book Termites et champignons: Les champignons termitophiles d’Afrique Noire et
d’Asie méridionale (Boubé, Paris, 1977) is principally devoted to Termitomyces.
74. R. Heim (1977; citation is in note 73), p. 43.
75. J. P. E. C. Darlington (1994; citation is in note 69 and in note 2), p. 124.
76. In fact such evidence is now available: see, for example, P. Mora and C. Lattaud in Insect
science and its application, vol. 19, pp. 51–55, 1999, who report the activity of the enzyme
laccase (employed in lignin degradation) in both fungus-cultivating termites and
Termitomyces, as well as the overview by C. Rouland-Lefèvre on pp. 289–306 of Termites:
Evolution, sociality, symbioses, ecology, edited by T. Abe et al. (Kluwer, Dordrecht,
2000).
77. J. P. E. C. Darlington (1994; citation is in note 2), pp. 125–126.
78. W. H. Kirchner et al. discuss vibrational alarm communication in the New World termite
Zootermopsis, in Physiological Entomology, vol. 19, pp. 187–190, 1994. This is a relatively
primitive termite, inhabiting rotten wood, and is not a fungus cultivator.
79. See H. Markl in Science, vol. 149, pp. 1392–1393, 1965 and Zeitschrift f ür vergleichende
Physiologie, vol. 57, pp. 299–330, 1967. More recently F. Roces and B. Hölldobler in
Behavioral Ecology and Sociobiology, vol. 39, pp. 293–299, 1996, have also demonstrated that
this stridulation serves to attract workers to the leaf-cutting site.
80. See A. Röhrig et al. in Insectes Sociaux, vol. 46, pp. 71–77, 1999, who note that such
long-distance communication is unique among the insects. Röhrig et al. also remark that the
intensity of drumming determines the type of reaction, at its most intense leading to a mass
retreat, with the intriguing possibility ‘that alarm signals help to avoid costly direct
interactions between [adjacent termite] colonies, battles that can kill thousands of soldiers
from both nests’, p. 76.
81. See the chapter by S. E. Kleinfeldt on pp. 283–294 of Insects and the plant surface, edited by
B. Juniper and R. Southwood (Edward Arnold, London, 1986).
82. This area is reviewed by U. G. Mueller et al. (2001; citation is in note 56), pp. 176–177.
83. See D. W. Davidson et al. in Ecology, vol. 69, pp. 801–808, 1988.
84. Importing vertebrate faeces (p. 1146) and the possibility of pathogen control (p. 1151) are
mentioned in a general review of ant-garden ecology by D. W. Davidson in Ecology, vol. 69,
pp. 1138–1152, 1988.
85. A helpful summary is given by W. H. Gotwald (1995; citation is in note 27); an earlier paper
on dorylomimine mimics is by D. H. Kistner in Annals of the Entomological Society of
America, vol. 59, pp. 320–340, 1966.
86. C. H. Seevers, in Fieldiana: Zoology, vol. 47, pp. 137–351, 1965, suggests a total of at least
seven transformations of staphylinids to army-ant mimics.
87. See the chapter (pp. 339–413) by D. H. Kistner in Social Insects (volume I), edited by H. R.
Hermann (Academic Press, New York, 1979), where he discusses the colour mimicry
between the beetle Ecitomorpha nevermanni and army ant Eciton burchelli. Subsequently,
Kistner and H. R. Jacobson, in Sociobiology, vol. 17, pp. 333–465, 1990, point out that colour
variation in a given colony is somewhat wider than previously realized.
88. D. H. Kistner (1979; citation is in note 87), p. 357.
89. D. H. Kistner (1979; citation is in note 87), p. 357, see also his Fig. 8.
90. A. P. Mathew in Journal of the Bombay Natural History Society, vol. 52, pp. 249–263, 1958.
91. A. P. Mathew (1958; citation is in note 90), p. 250.
92. The observation of the spider mimicking an ant in distress is based on laboratory
investigations. Field observations by A. P. Mathew (1958; citation is in note 90) also indicate
this spider stalking and pouncing on ants that stray too far.
93. For a review of this insect–spider mimicry, see J. D. McIver and G. Stonedahl in Annual
Review of Entomology, vol. 38, pp. 351–379, 1993.
406 notes to pp. 213–214
94. This is known as transformational mimicry. For example, A. P. Mathew, in Journal of the
Bombay Natural History Society, vol. 37, pp. 803–813, 1935, explains how the hemipteran
bug Riptortus pedestris, in a series of moults, successively resembles the ant Prenolepis
longicornis (‘almost indistinguishable’, p. 804), then Plagiolepis (‘mimicry at this stage is
almost perfect’, p. 804), and then the already encountered Oecophylla smaragdina ‘which
presented a very perfect resemblance to the Red Ant . . . even more perfect than the mimicry
which I was observing and studying in the Attid ants’ (p. 803). J. D. McIver and G. Stonedahl
(1993; citation is in note 93) provide an overview on this topic, with further citations.
95. See J. D. McIver and G. Stonedahl (1993), pp. 364–365.
96. A helpful overview is given by L. E. Gilbert in his chapter (pp. 263–281) in Coevolution,
edited by D. J. Futuyama and M. Slatkin (Sinauer, Sunderland, MA, 1983), who concludes by
remarking ‘Mimicry is possibly the most compelling evidence that community patterns are
more than random noise,’ p. 281.
97. But see, for example, the paper by J. Berger and J. Kaster (in Evolution, vol. 33, pp. 511–513,
1979). Berger and Kaster argue that a striking resemblance between a caddisfly larva
(Helicopsyche borealis) and an aquatic snail (Physa integra) arises from a common selection
pressure imposed by the hydrodynamic regime rather than a mimicry imposed by
predation.
98. See the interesting overview by C. Anderson and D. W. McShea in Insectes Sociaux, vol. 48,
pp. 291–301, 2001. One might draw attention to the ant equivalent of ‘cow-sheds’ to house
honeydew-yielding aphids, and ‘pens’ to look after caterpillars. In their conclusion the
authors remark, ‘What is certainly needed is more detail about construction mechanisms.
Where such information was available, we observe a good example of convergent evolution:
trenches and arcades appear to be similar in both structure and construction behavior in new
world army ants, such as Labidus, old world African driver ants, such as Dorylus, and Asian
marauder ants, Pheidologeton. Arcade construction [in the last genus] has certainly arisen
independently,’ p. 299.
99. See K. Wada in Ethology, vol. 96, pp. 270–280, 1994 and Crustaceana, vol. 68, pp. 524–526,
1995 (with J. K. Park); and J. Kitaura et al. in Molecular Biology and Evolution, vol. 15,
pp. 626–637, 1998.
100. See the paper by G. A. Croll and J. B. McClintock in Journal of Experimental Marine Biology
and Ecology, vol. 254, pp. 109–121, 2000.
101. As an aside on the topic of convergence, I cannot resist the example of the discharge of ink by
various cephalopods to baffle their attackers. And the parallel? The discharge of a reddish
cloud by dwarf sperm-whales, albeit from the anus; see M. D. Scott and J. G. Cordaro in
Marine Mammal Science, vol. 3, pp. 353–354, 1987.
102. A useful review of the brains of octopuses and other advanced cephalopods is given by B. U.
Budelmann on pp. 115–138 of The nervous system of invertebrates: An evolutionary and
comparative approach edited by O. Breidbach and W. Kutsch (Birkhäuser, Basel, 1995); see
also the chapter (pp. 399–413) by B. U. Budelmann et al. in Cephalopod neurobiology:
Neuroscience studies in squid, octopus, and cuttlefish, edited by N. J. Abbott et al. (Oxford
University Press, Oxford, 1995) and J. Z. Young’s monograph The anatomy of the nervous
system of Octopus vulgaris (Oxford, Clarendon Press, 1971). In describing the cerebellum-like
structure in the squid brain, J. Z. Young (in Nature, vol. 264, pp. 572–574, 1976) remarked
how the pattern of nervous control associated with movement of the eye as well as the whole
animal ‘shows remarkable similarities between cephalopods and vertebrates, in spite of the
fact that the gross anatomical entities are totally different’, p. 574.
103. See the stimulating paper by A. Packard entitled Cephalopods and fish: the limits of
convergence in Biological Reviews 47, 241–307, 1972. The question of relative brain sizes is
addressed on pp. 265–267. Packard’s overall conclusion is that the cephalopods are good, but
not as good as the fish. Even so, if the fish eventually win out it is going to take time. J. J.
Madan and M. J. Wells also point out, in Nature, vol. 380, p. 590, 1996, that the deep-sea
notes to pp. 214–215 407
squid have a refuge of sorts by inhabiting sea water with low levels of oxygen, a mode of life
made possible by having very thin – and delicate – gills.
104. See chapter 27 (pp. 446–457) by M. Bundgaard et al. in N. J. Abbott et al. (1995; citation is in
note 102), as well as two companion papers by N. J. Abbott et al. in Journal of Physiology,
vol. 368, pp. 197–212 and 213–226, 1985.
105. M. Bundgaard et al. (1995; citation is in note 104), p. 446.
106. See H. Jaaro et al. in Trends in Neurosciences, vol. 24, pp. 79–85, 2001. Interestingly, the most
advanced sort of brains in the arthropods, found in the insects, apparently lack neurotrophins,
at least to judge from a survey of the fruit-fly genome. This is not to say that insect brains are
simple; the reverse is very much the case. Jaaro et al. suggest, however, that neurotrophins
predetermine the ‘differences between a “plastic” nervous system with a potential for
evolving complexity, versus a more constrained nervous system that cannot evolve beyond
certain limitations’ (p. 84). It may be unwise, however, to underestimate the plasticity of
insect brains; see, for example, the overview by I. A. Meinertzhagen in Advances in Insect
Physiology, vol. 28, pp. 84–167, 2001.
107. Chromatic expression is controlled by a specific area of the brain and is achieved by
pigmented areas known as chromatophores; see, for example, the papers by E. Florey in
American Zoologist, vol. 9, pp. 429–442, 1969; R. T. Hanlon and J. B. Messenger in
Philosophical Transactions of the Royal Society of London B, vol. 320, pp. 437–487, 1988, and
Cephalopod behaviour (Cambridge University Press, Cambridge, 1996); and A. Packard in
chapter 21 (pp. 331–367) of Abbott et al. (1995; citation is in note 102). J. B. Messenger (in
Biological Reviews, vol. 76, pp. 473–528, 2001) remarks ‘Cephalopods are such complex and
behaviourly advanced animals . . . that they must be able to transmit messages of a far subtler
kind [than straightforward messages] and it seems likely that future work will reveal this’
(p. 520). See also A. C. Crook et al. (in Philosophical Transactions of the Royal Society of
London B, vol. 357, pp. 1617–1624, 2002), who note that in contrast to other animal
communication systems (see note 140, Chapter 9)) the visual signals of cuttlefish do not
follow Zipf’s law.
108. Relevant literature includes papers by G. Fiorito and P. Scotto in Science, vol. 256,
pp. 545–547, 1992; J. Z. Young in Biological Bulletin, vol. 180, pp. 200–208, 1991a, Visual
Neuroscience, vol. 7, pp. 1–12, 1991b; and J. G. Boal et al. in Journal of Comparative
Psychology, vol. 114, pp. 246–252, 2000. Fiorito and Scotto show how one octopus can learn
by observing the activities of another octopus. J. Z. Young in his 1991a paper demonstrates
how there are effectively two memory systems, one chemo-tactile and one visual. He remarks
on the complexity of the brain structure, noting that the learning ability ‘is distributed with
high redundancy in a series of matrices networks, with recurrent circuitry, up to a late stage
where funneling to a few cells occurs’ (p. 206). Elsewhere (in chapter 26 (pp. 432–443) of
N. J. Abbott et al. (1995; citation is in note 102)), Young remarks that ‘The information
already available shows that the octopus memory involves selection during learning among
large number of parallel pathways. The capacity to set up and hold the memory is distributed
among many distinct matrices. This seems to be the case also in the complex memories of
mammals, birds, and insects’, p. 442.
109. T. Moriyama and Y-P. Gunji (in Ethology, vol. 103, pp. 499–513, 1997), show how the octopus
can choose a seemingly inefficient action (e.g. reduced swimming speed) the more easily to
achieve its final objective, say a tasty piece of shrimp.
110. On octopus living on coral atolls in French Polynesia, see J. W. Forsythe and R. T. Hanlon in
Journal of Experimental Marine Biology and Ecology, vol. 209, pp. 15–31, 1997.
111. J. W. Forsythe and R. T. Hanlon (1997; citation is in note 110), p. 26.
112. See the paper by J. A. Mather and R. C. Anderson in Journal of Comparative Psychology,
vol. 107, pp. 336–340, 1993. Their analysis depends on the difference between octopus
individuals, their patent ability to learn in the face of novel environments, and
temperamental behaviour, which also define in part human personality. This work is
408 notes to pp. 215–216
extended by D. L. Sinn et al. in the same journal (vol. 115, pp. 351–364, 2001). They suspect
that the range of octopus temperaments may not be fully explored, while Mather and
Anderson comment that there may be commonalities across phyla; addressing an octopus by
name need not be completely daft.
113. The rubber-like proteins in the octopus aorta are documented by R. E. Shadwick and
J. M. Gosline in Journal of Experimental Biology, vol. 114, pp. 239–257, 1985. Such proteins,
of which the vertebrate elastin, insect resilin, and bivalve mollusc abductin are another three
examples, are possibly convergent. These authors note that ‘The tremendous variation in
amino acid composition among the four known protein rubbers . . . suggests to us that each
protein arose independently during evolution in response to selection for similar mechanical
design’, p. 256.
114. R. E. Shadwick notes in Journal of Experimental Biology, vol. 202, pp. 3305–3313, 1999, that
‘In animals as different as Nautilus [a cephalopod] and the rat, there is a striking similarity in
the microscopic appearance of the aortic wall’, p. 3309. See also his earlier articles in Marine
Behavior and Physiology, vol. 25, pp. 69–85, 1994, and Journal of Experimental Biology,
vol. 114, pp. 259–284, 1985 (with J. M. Gosline); and R. Schipp et al. in Zoomorphology,
vol. 118, pp. 79–85, 1998. It is also worth mentioning that the convergence in arterial wall
structure also extends to the crustaceans; see R. E. Shadwick et al. in Physiological Zoology,
vol. 63, pp. 90–101, 1990.
115. Helpful reviews of the circulatory system of cephalopods are given by M. J. Wells in
pp. 239–290, volume 5 of The Mollusca, edited by A. S. M. Saleuddin and K. M. Wilbur
(Academic Press, New York, 1983), and in Experientia vol. 43, pp. 487–499, 1987 (with
P. J. S. Smith), where they remark, ‘The basic design of the molluscan circulatory system
was . . . well suited to a slow-moving animal . . . but a disaster for large, high-metabolic rate
invertebrates, competing with fish and other vertebrates. Natural selection works on the
material available, and the cephalopod circulatory system is a miracle of making the best of a
bad lot’ (pp. 497–498). They continue, ‘Jet propulsion is inescapably expensive . . . It costs a
squid . . . something like twice as much energy, to travel half as fast, as a salmon. The
difference would be even more extreme if comparisons were made with mackeral or tuna . . .
Add to these problems the appallingly low carrying capacity [of oxygen] of cephalopod blood,
and it is evident that the circulatory pumps must be achieving something rather remarkable
in terms of cardiac output’ (p. 498). A similar conclusion is reached by R. E. Shadwick et al.
(in Journal of Experimental Biology, vol. 130, pp. 87–106, 1987) where they also note that
amongst the large squids the arterial haemodynamics is likely to be much more similar to
that of the active vertebrates. R. Schipp also has a short, but helpful, update in Vie et Milieu,
vol. 47, pp. 111–116, 1997.
116. The topic of these branchial hearts is addressed in the literature cited in note 115, and also by
A. Fielder and R. Schipp in Experientia, vol. 43, pp. 544–553, 1987.
117. These are not the only differences. Despite the systemic heart in the cephalopods, many of
the other veins are also muscular and contractile. They also lack true capillaries, but this
might be because the blood protein consists of aggregates of haemocyanin, whereas in the
vertebrates the squeezing of the red blood corpuscles through the narrow capillaries
facilitates oxygen release.
118. This perichondrial tissue is described by A. Bairati et al. in Tissue & Cell, vol. 27,
pp. 515–523, 1995. In a general review of invertebrate cartilages, P. Person and D. E. Philpott,
in Biological Reviews, vol. 44, pp. 1–16, 1969, remark of cephalopod cartilage ‘The cell
processes penetrate the matrix, and often give rise to extensive systems of intercellular
canaliculae, very similar to those seen in vertebrate bone and cartilage’, p. 8.
119. A helpful overview is given in The mechanosensory lateral line: neurobiology and evolution
edited by S. Coombs et al. (Springer, New York, 1989).
120. The lateral line analogue in the cephalopods is known as the epidermal lines; see
G. Sundermann in Cell and Tissue Research, vol. 232, pp. 669–677; 1983. Confirmation of a
sensory role was given by B. U. Budelmann and H. Bleckmann in Journal of Comparative
notes to pp. 216–218 409
Physiology, vol. 164A, pp. 1–5, 1988, where they noted it is ‘another fascinating example,
beside the eyes [see notes 22–27, Chapter 7] and equilibrium receptor systems [see note 28,
Chapter 7], of convergent evolution between a sophisticated cephalopod and vertebrate
sensory system’ (p. 4). More details were given by H. Bleckmann et al. in vol. 168A,
pp. 247–257, 1991 of the same journal. More recent contributions are an overview by B. U.
Budelmann in Marine and Freshwater Behaviour and Physiology, vol. 25, pp. 13–33, 1994;
S. Lenz et al. in Zoologisches Anzeiger, vol. 234, pp. 145–157, 1995; and S. Lenz in Vie et
Milieu, vol. 47, pp. 143–147, 1997.
121. See S. Coombs et al. in Philosophical Transactions of the Royal Society of London B,
vol. 355, pp. 1111–1114, 2000.
122. See R. L. Reep et al. in Brain, Behavior and Evolution, vol. 59, pp. 141–152, 2002.
123. See E. J. Denton and J. Gray in Proceedings of the Royal Society of London B, vol. 226,
pp. 249–261, 1985.
124. See Jared Diamond’s chapter (pp. 3–8) in Ecological restoration of New Zealand islands,
edited by D. R. Towns et al. (Conservation Sciences Publication, No. 2, 1990).
125. J. Diamond (1990; citation is in note 124), p. 3.
126. See the review by R. A. Cooper and P. R. Millener in Trends in Ecology & Evolution, vol. 8,
pp. 429–433, 1993.
127. The quotation, which paraphrases a sentence in the article ‘Invertebrate mice’ by G. W.
Ramsay in New Zealand Entomologist, vol. 6, p. 400, 1978, is from a stimulating review
paper by C. H. Daugherty et al. in Trends in Ecology & Evolution, vol. 8, pp. 437–442, 1993,
see p. 439.
128. Further information on the amazing wetas is available in a paper by A. M. Richards in Journal
of Zoology, vol. 169, pp. 195–236, 1973. One weta, Deinacrida heteracantha, can grow up to
82 mm and thus is one of the largest insects in the world. See also The biology of wetas, king
crickets and their allies, edited by L. H. Field (CAB International, Wallingford, 2001), where
the mouse-like characters of wetas are remarked upon by M. McIntyre, on p. 226 of her
chapter (pp. 225–242). I should emphasize, however, that the editor, Larry Field, is dismissive
of wetas being regarded as ecological analogues of mice (personal communication).
129. J. Diamond (1990, citation is in note 124), p. 4. This wren is extinct, eaten by a cat.
J. Diamond also remarks on the attempts of short-tailed bats and wetas to ‘become’ mice.
130. See C. H. Daugherty et al. (1993; citation is in note 127). C. H. Daugherty et al. also remark
on the other parallels, with large flightless birds occupying the niches elsewhere taken by
browsing and grazing mammals.
131. A brief account of this flightless parrot, which can however glide, is available in A field guide
to the birds of New Zealand and outlying islands (Collins, London, 1966) by R. A. Falla et al.
(see pp. 170–171).
132. An overview of moa species and their ecological impact is given by A. Cooper et al. in Trends
in Ecology & Evolution, vol. 8, pp. 433–437, 1993, and a general account is given by A.
Anderson in Prodigious birds: Moas and moa-hunting in prehistoric New Zealand
133. See A. Cooper et al. in Proceedings of the National Academy of Sciences, USA, vol. 89,
pp. 8741–8744, 1992, and also A. Cooper et al. in Nature, vol. 409, pp. 704–707, 2001. Cooper
et al. suggest that the kiwis and moas represent separate colonizations, with kiwis arriving
earlier. Giant, flightless birds have evolved repeatedly. Other examples in addition to those in
New Zealand include the Madagascan elephant birds and the fearsome South American
phorusrhacids. Another independent development, from the Eocene of Egypt, is reported by
D. T. Rasmussen et al. in Palaeontology, vol. 44, pp. 325–337, 2001. This occurrence is
important because Africa appeared to be the only significant landmass without an endemic
large, ground-dwelling bird (the ostriches migrated to Africa from Asia in the Miocene).
134. Despite earlier scepticism there is now little doubt that the Maoris were an important
element in the extinction of the moa, by both direct and profligate hunting, and also burning
of forest habitats. See the chapters by M. Trotter and B. McCullock (pp. 708–727) and
410 notes to pp. 218–220
A. Anderson (pp. 728–740) in Quaternary extinctions: a prehistoric revolution, edited by P. S.

Martin and R. G. Klein (University of Arizona Press, Tucson, 1984).
135. See the papers by W. A. Calder in Scientific American, vol. 239 (1), pp. 102–110, 1978, and
BioScience, vol. 29, pp. 461–467, 1979.
136. See R. A. Falla et al. (1996; citation is in note 131), pp. 18–20.
137. See the note by S. M. Monro in Avicultural Magazine, vol. 80, p. 202, 1974. The budgie’s
owner noticed that it looked a bit under the weather, and she was even more surprised when
the bird apparently gave birth to ‘two perfectly formed chicks’. In particular, no sign of any
shells was noted. Whatever the budgie thought of it is obscure, not least because shortly after
this event the bird died.
138. See the paper entitled ‘Why are there no viviparous birds?’ by D. G. Blackburn and H. E.
Evans in American Naturalist, vol. 128, pp. 165–190, 1986. See also R. L. Dunbrack and M. A.
Ramsay, in vol. 133, pp. 138–148, 1989 of the same journal, who argue that an absence of
effective lactation has been an important factor against ovoviviparity.
139. See D. J. Anderson et al. in American Naturalist, vol. 130, pp. 941–947, 1987.
140. W. A. Calder (1978, citation is in note 135), p. 110.
141. See Richard Shine’s chapter (pp. 361–369) in the book Egg incubation: its effects on
embryonic development in birds and reptiles, edited by D. C. Deeming and M. W. J. Ferguson
(Cambridge University Press, Cambridge, 1991). He has an interesting discussion of the
factors that predispose the squamates to this mode of reproduction, and why it has never
evolved in the crocodiles, turtles, or birds.
142. Daniel Blackburn’s overview of both viviparity and matrotrophy (that is supply of nutrients
by the mother) is given in American Zoologist, vol. 32, pp. 313–321, 1992; see also M. H.
Wake’s article in the same issue, on pp. 256–263.
143. See, for example, the extraordinary case of a Namibian toad (Nectophrynoides occidentalis)
reported by F. Xavier on pp. 545–552 in Major patterns in vertebrate evolution, edited by M.
K. Hecht et al. (Plenum, New York, 1977). Xavier remarks that it is ‘quite remarkable to find
the same structures and the same hormones in the Nimba toad and in mammals, along with
the same functions, i.e. to contain fetuses in the maternal uterus, to feed them and then to
expel them’ (p. 550). He also emphasizes, however, there are fundamental differences, of
which the most obvious is the absence of a placenta. In addition, the hormones are used in a
different way from the mammalian system.
144. For useful reviews see J. P. Wourms in American Zoologist, vol. 21, pp. 473–515, 1981 and
Israel Journal of Zoology, vol. 40, pp. 551–568, 1994, and J. P. Wourms and J. Lombardi in
American Zoologist, vol. 32, pp. 276–293, 1992. Wourms (1994) estimates that fish viviparity
has evolved independently at least 42 times, and in this review comments both on the major
changes and on the many advantages of viviparity (see p. 554). It is especially frequent among
the sharks and rays (see also Dulvy and Reynolds 1997; citation is in note 154), but it also
known in the teleost fish, as well as the famous ‘living fossil’, the coelacanth Latimeria. The
paper by Wourms and Lombardi (1992) also remarks on various other sorts of convergence
seen in the viviparous fish. These include the so-called trophotaenia, projections of the
embryonic hind-gut that facilitate nutrient transfer, and seemingly have evolved ‘in four
distantly related orders of teleosts’ (p. 287). Of even greater interest is ‘The pericardial
amniochorion which occurs in the embryos of poeciliid fishes [see also B. D. Grove and J. P.
Wourms in Journal of Morphology, vol. 209, pp. 265–284, 1991 and vol. 220, pp. 167–184,
1994 respectively], some goodeids [see also I. Kokkala and J. P. Wourms, and F. Hollenberg and
J. P. Wourms in Journal of Morphology, vol. 219, pp. 35–46 and pp. 105–129 respectively,
1994] and in a modified form . . . in Anableps is of particular interest because of its striking
similarity to the anterior amnio-chorionic fold of amniotes. This is an apparent anomaly in
that the amnion and chorion first appear in the tetrapod lineage as one of the major steps in
the evolution of the cleidoic egg . . . Besides being a remarkable example of evolutionary
convergence in an extraembryonic membrane, it helps one to understand the developmental
basis for the evolution of the amnion and chorion in amniotes,’ p. 286.
notes to pp. 220–224 411
145. An extensive review of reptile placentation is given by Z. Yaron on pp. 527–603 in vol. 15 of
Biology of the Reptilia, edited by C. Gans and G. Billett (Wiley, New York, 1985). See also
D. G. Blackburn in Journal of Experimental Zoology, vol. 266, pp. 414–430, 1993, where he
remarks that ‘The evolution of placentomes in three separate lineages [of squamate]
represents an extraordinary manifestation of convergence’ (p. 427). It has long been
recognized that the development of the reptile placenta is quite independent of the mammals
and might also indicate how it arose in the latter group. See also D. N. Reznick et al. in
Science, vol. 298, pp. 1018–1020, 2002 (with commentary by V. Morell on p. 945) who
document the independent evolution of a placenta in the fish Poeciliopsis, and also make
some interesting remarks on the rapid emergence of complex biological structures.
146. Daniel Blackburn and his colleagues’ paper is published in Proceedings of the National
147. D. G. Blackburn et al. (1984; citation is in note 146), p. 4862.
148. Concerning the placenta and viviparity of Mabuya bistriata, see L. J. Vitt and
D. G. Blackburn in Copeia, for 1991, pp. 916–927, 1991. A helpful overview of reproduction in
Mabuya is given by D. G. Blackburn and L. J. Vitt on pp. 150–164 of Reproductive biology of
South American vertebrates, edited by W. C. Hamlett (Springer, New York, 1992), where they
write, ‘The independent evolution of these features among mammals and lizards represents
one of the most striking cases of evolutionary convergence in reproductive specializations to
be documented among terrestrial vertebrates’, p. 161.
149. See D. G. Blackburn in Amphibia–Reptilia, vol. 3, pp. 185–205, 1982.
150. See D. G. Blackburn (1992, citation is in note 142). Further details of this remarkable placenta
are given by D. G. Blackburn in Journal of Morphology, vol. 216, pp. 179–195, 1993.
151. J. P. Wourms and J. Lombardi (1992; citation is in note 144), p. 290, as well as earlier remarks
by J. P. Wourms and colleagues in the first chapter (pp. 1–134) of Fish physiology, vol. XIB,
edited by W. S. Hoar and D. J. Randall (Academic Press, San Diego, 1988). In the latter work
they remark ‘The selachian yolk-sac placenta has probably obtained the pinnacle of its
evolutionary development in the spadenose shark . . . Here it appears to function with the
same degree of efficiency as a mammalian placenta.’ (p. 69). See also note 154.
152. D. G. Blackburn (1992; citation is in note 142), p. 320.
153. D. G. Blackburn (1992; citation is in note 142), p. 320.
154. See the paper by N. K. Dulvy and J. D. Reynolds in Proceedings of the Royal Society of
London B, vol. 264, pp. 1309–1315, 1997. See also M. S. Y. Lee and R. Shine, in Evolution,
vol. 52, pp. 1441–1450, 1998, who present evidence for reversals in the tendency towards
viviparity in the reptiles. While such reversals are not ruled out absolutely, in reality there are
only a handful of possible examples. Such entrenchments impose a strong directionality on
the history of life.
155. See D. G. Blackburn (1982; citation is in note 149). See also R. M. Andrews (in Physiological
and Biochemical Zoology, vol. 75, pp. 145–154, 2002), who explores the link between
coolness and oxygen availability in determining the likelihood of viviparity, and especially
R. Shine (in American Naturalist, vol. 160, pp. 582–593, 2002), who presents sound adaptive
reasons why cooler thermal regimes favour in utero development of reptile eggs.
156. N. K. Dulvy and J. D. Reynolds (1997; citation is in note 154), p. 1314.
157. See B. Heinrich’s The hot-blooded insects. Strategies and mechanisms of thermoregulation
(Springer, Berlin, 1993), and The thermal warriors: Strategies of insect survival (Harvard
University Press, Cambridge, MA, 1996).
158. See the papers by B. A. Block et al. in Science, vol. 260, pp. 210–214, 1993 (and commentary
by T. Watson on pp. 160–161), and B. A. Block and J. R. Finnerty in Environmental Biology of
Fishes, vol. 40, pp. 283–302, 1994.
159. See the paper by F. G. Carey et al. in Memoirs, Southern California Academy of Sciences,
vol. 9, pp. 92–108, 1985, where they also note similarities with tuna.
160. Reviewed in B. A. Block and J. R. Finnerty (1994; citation is in note 158).
412 notes to pp. 224–227
161. See the paper by B. A. Block and F. G. Carey in Journal of Comparative Physiology, 156B,
pp. 229–236, 1985.
162. B. A. Block and F. G. Carey (1985; citation is in note 161), p. 234.
163. B. A. Block et al. (1993; citation is in note 158), p. 213.
164. The grim topic of malignant hyperthermia in humans and its equivalent in pigs (porcine
stress syndrome) is addressed by D. H. MacLennan and M. S. Phillips in Science, vol. 256,
pp. 789–794, 1992.
165. The similarities between malignant hyperthermia and ‘tuna burn’ are reviewed by B. A. Block
in Annual Review of Physiology, vol. 56, pp. 535–577, 1994; see also B. A. Block and
J. R. Finnerty (1994; citation is in note 158).
166. C. G. Farmer’s paper, entitled ‘Parental care: The key to understanding endothermy and other
convergent features in birds and mammals’ was published in American Naturalist, vol. 155,
pp. 326–334, 2000.
167. C. G. Farmer (2000; citation is in note 166), p. 330; as given here the quotation has been
rewritten by combining parts of a paragraph.
168. Convergences of maternal care also occur within the mammals, such as between the
placental ungulates and macropod marsupials, e.g. kangaroo. See D. O. Fisher et al., in
Evolution, vol. 56, pp. 167–176, 2002.
169. C. G. Farmer (2000; citation is in note 166), p. 330.
170. The descent of Man, and selection in relation to sex (Murray, London, 1889, second edition),
p.181. Or is this an ‘urban myth’, like the last Cornish monoglot, Dolly Pentreath (d. 1777)
and her apocryphal posthumous parrot?
171. See P. M. Gray et al. in Science, vol. 291, pp. 52–54, 2001.
172. This concerns the male palm cockatoo that selects a fresh stick and shapes it (by removing
foliage, chewing the ends off and stripping bark) into a drumstick suitable for drumming on a
hollow log; see G. A. Wood in Corella, vol. 8, pp. 94–95, 1984.
173. P. M. Gray et al. (2001; citation is in note 171), p. 52.
174. See I. M. Pepperberg in Animal Learning and Behavior, vol. 27, pp. 15–17, 1999 (citations are
in notes 137 and 139 of Chapter 9).
175. Allison Doupe and Patricia Kuhl’s paper, entitled ‘Birdsong and human speech: Common
themes and mechanisms’ is in Annual Review of Neuroscience, vol. 22, pp. 567–631; 1999.
The realization that there are important parallels between birdsong and human speech
include, for example, second language acquisition (or bilingualism) (see I. M. Pepperberg and
D. M. Neapolitan in Ethology, vol. 77, pp. 150–168, 1988), and evidence for cultural
transmission (see R. B. Payne et al. in Behaviour, vol. 77, pp. 199–221, 1981).
176. A. Doupe and P. Kuhl (1999; citation is in note 175), p. 567.
180. See G. F. Striedter in Journal of Comparative Neurology, vol. 343, pp. 35–56, 1994;
subsequently, E. D. Jarvis et al. (in Nature, vol. 406, pp. 628–632, 2000) reported striking
similarities in the forebrain region of hummingbirds, songbirds, and parrots. This idea is
further developed by M. A. Farrier in Brain, Behavior and Evolution, vol. 58, pp. 80–100,
2001, who writes, ‘nature has conducted what are in effect three independent experiments in
the evolution of vocal learning in birds: parrots, hummingbirds, and songbirds. By studying
the evolution of similar behavioral capacities from similar starting points, it might be
possible to address some important questions in evolutionary biology . . . For example, how
much freedom does evolution have to find different solutions to the same problem?’ p. 96.
181. There are plenty of examples of convergence in such grinding teeth, such as those between
the molars of the rhinoceros Elasmotherium and horses (see W. B. Scott in Journal of
Morphology, vol. 5, pp. 301–406, 1891), and within the elephants (see M. O. Thomas et al. in
Proceedings of the Royal Society of London B, vol. 267, pp. 2493–2500, 2000).
notes to pp. 227–229 413
182. In passing, one should note an emphasis on a mammal-like occlusion in the teeth of a much
more primitive group of chordates, the extinct conodonts; see P. C. J. Donoghue and
M. A. Purnell in Paleobiology, vol. 25, pp. 58–74, 1999. This example is interesting because
the marine conodonts, whose precise taxonomic position is controversial, lack anything like
a jaw. See also P. C. J. Donoghue in Proceedings of the Royal Society of London B, vol. 268,
pp. 1691–1698, 2001.
183. The tribosphenic molar is typically of the primitive marsupial and placental mammals, and
their immediate forebears. Z-x. Luo et al. (in Nature, vol. 409, pp. 53–57) argue that the
tribosphenic condition evolved twice, both in the aforementioned groups and independently
in the ancestors of the monotremes (whose extant representatives are toothless); see also
T. H. Rich et al. (in Journal of Vertebrate Paleontology, vol. 22, pp. 466–469, 2002) for possible
qualification of this conclusion. The arrangement of interlocking cusps permits precise
occlusion and shearing, and not surprisingly there are further examples of convergence,
J. P. Hunter and J. Jernvall, in Proceedings of the National Academy of Sciences, USA,
vol. 92, pp. 10718–10722, 1995, estimate that the evolution of an additional cusp, the
hypocone, happened more than 20 times, and regard it as a key pre-adaptation to evolutionary
success. See also J. Jernvall’s follow-up paper in the same journal (vol. 97, pp. 2641–2645,
2000).
184. See the paper by R. L. Nydam et al. in Journal of Vertebrate Paleontology, vol. 20,
pp. 628–631, 2000, although the authors are careful not to over-emphasize the rather striking
similarity. See also R. L. Nydam and R. L. Cifelli in a subsequent issue, vol. 22, pp. 276–285,
2002. A somewhat similar occurrence is also known in the living lizard Uromastyx, which
shows a dramatic change in dentition in its life as the animal moves from a diet of small
animals to plants when an adult. More unusually, the enamel of these teeth has convergently
evolved a prismatic structure, also with tubules, that is reminiscent of the mammals, but is
otherwise unknown in the reptiles; see J. S. Cooper and D. F. G. Poole in Journal of Zoology,
London, vol. 169, pp. 85–100, 1973.
185. See J. M. Clark et al. in Science, vol. 244, pp. 1064–1066, 1989, and X-C. Wu et al. in Nature,
vol. 376, pp. 678–680, 1995.
186. See O. Kuhn’s description of remarkably preserved material from the famous site at Geiseltal
in Germany, in Nova Acta Leopoldina, vol. 6 (N.F.), pp. 311–329, 1938, p. 323.
187. See P. C. Sereno et al. in Science, vol. 282, pp. 1298–1302, 1998.
188. ‘Spinosaurs as crocodile mimics’ is the title of the commentary by T. R. Holtz that
accompanies Sereno et al. (1998); see pp. 1276–1277. This ‘intriguing case of convergence’
(p. 1277) is evidently a result of a shift by this dinosaur to a piscivorous diet.
9. the non-prevalence of humanoids?

1. See L. F. Laporte’s George Gaylord Simpson: Paleontologist and evolutionist (Columbia
University Press, New York; 2000), as well as H. B. Whittington’s encomium in Biographical
Memoirs of Fellows of the Royal Society, vol. 32, pp. 527–539, 1986.
2. The article was published in Science, vol. 143, pp. 769–775, 1964a, and promptly provoked a
vigorous discussion, see vol. 144, pp. 613–615, 1964. Effectively the same paper, but shorter,
was published by Simpson as Chapter 13 in This view of life: the world of an evolutionist
(Harcourt, Brace and World, New York, 1964b).
3. Together with the other giants of evolutionary study, notably R. A. Fisher, G. Ledyard
Stebbins, Th. Dobzhansky and E. Mayr, Simpson is regarded as one of the main architects of
the neo-Darwinian synthesis. An important kerygma in this programme was the book edited
by G. L. Jepsen, G. G. Simpson, and E. Mayr entitled Genetics, paleontology and evolution
(Princeton University Press, Princeton, 1949). Simpson himself published the very influential
Tempo and mode in evolution (Columbia University Press, New York, 1944) and its effective
revision, The major features of evolution (Columbia University Press, New York,
1953).
414 notes to pp. 229–230
4. In The meaning of evolution: a study of the history of life and of its significance for man
(New Haven, Yale University Press, revised edition, 1967) Simpson devoted a good part of
Chapter 12 to a discussion of convergences.
5. Very similar views are expressed by E. Mayr in his essay ‘The probability of extraterrestrial
intelligent life’, reprinted on pp. 67–74 (Essay four) of Toward a new philosophy of biology:
Observations of an evolutionist (Belknap, Cambridge, MA, 1988). Mayr’s view is that for all
intents and purposes intelligence is a fluke; see his article ‘Does it pay to acquire high
intelligence?’, in Perspectives in Biology and Medicine, vol. 37, pp. 337–338, 1994. Effectively
identical sentiments were expressed in his chapter (pp. 152–156) in Extraterrestrials: Where
are they?, edited by B. Zuckerman and M. H. Hart (Cambridge University Press, Cambridge,
1995 [Second edition]).
6. This being the title of the paper in Annals of the New York Academy of Sciences, vol. 950,
pp. 276–288, 2001. This article, however, seems over-concerned with the specifics of history
as against the emergence of general biological properties. This volume, entitled Cosmic
questions, and edited by J. B. Miller, has a number of other interesting chapters, notably those
by A. Case-Winters (pp. 154–168) and S. Via (pp. 225–240).
7. Simpson (1964a; see note 2 for citation), p. 774. Very similar sentiments are expressed by Carl
Sagan in his Cosmos (Macdonald, London, 1981), where he writes of those who try to imagine
extraterrestrial life that ‘They seem to me to rely too much on forms of life we already know.
Any given organism is the way it is because of a long series of individually unlikely steps. I do
not think life anywhere else would look very much like a reptile, or an insect or a human’
(p. 40). Subsequently, Sagan remarked in Scientific American, vol. 271 (4), pp. 70–77, 1994,
that given life on Earth had a single origin then ‘we have no way of knowing which aspects of
terrestrial life are necessary (required by all living things anywhere) and which are merely
contingent (the results of a particular sequence of happenstances that, had they gone
otherwise, might have led to organisms having very different properties),’ p. 71.
8. See George Beadle’s 11th Annual Arthur Dehon Little Memorial Lecture, entitled ‘The place
of genetics in modern biology’ (MIT Press, Boston, 1959).
9. G. Beadle (1959; citation is in note 8), p. 20.
10. The origins and genomic organization of the animal sodium channels are discussed by
J. D. Spafford et al. in Receptors and Channels, vol. 6, pp. 493–506, 1999.
11. Concerning the sodium pump see the paper by W. D. Stein in Philosophical Transactions of
the Royal Society of London B, vol. 349, pp. 263–269, 1995. See also R. D. Keynes and F.
Elinder in Proceedings of the Royal Society of London B, vol. 266, pp. 843–852,
1999.
12. See Sponges by P. R. Bergquist (London, Hutchinson, 1978), pp. 63, 67–69. Even so, sponges
possess contractile cells (the myocytes) and they evidently contain neurotransmitters; see for
example the paper by T. L. Lentz in Journal of Experimental Zoology, vol. 162, pp. 171–179,
1966. In the hexactinellids there is also indirect evidence for electrical conduction, albeit
non-nervous and probably controlled by calcium; see G. O. Mackie et al. in Philosophical
Transactions of the Royal Society of London B, vol. 301, pp. 401–418, 1983. For a more
general discussion of conductance in sponges see G. O. Mackie in American Zoologist,
vol. 30, pp. 907–920, 1990.
13. See the paper by C. Febvre-Chevalier et al. in Journal of Experimental Biology, vol. 122,
pp. 177–192, 1986. Although there is also a small calcium component, this heliozoan ‘is thus
the first exception to [the] generalization that sodium-requiring action potentials are
unknown outside the metazoan animals’ (p.189). See also the chapter by C. Febvre-Chevalier
et al. (pp. 237–253) in Evolution of the first nervous system, edited by P. A. V. Anderson
(Plenum, New York, 1990). Here the principal dependence of Actinocoryne on sodium is
regarded as ‘puzzling’ (p. 249), yet the authors also emphasize the ‘remarkable sensitivity and
contractility’ (p. 239) of this organism.
14. For an exhaustive review of ion transport mechanisms, including sodium, see M. H. Saier in
Microbiology and Molecular Biology Reviews, vol. 64, pp. 354–411, 2000.
notes to pp. 230–233 415
15. See the series of seven papers, forming a mini-review, on bacterial sodium energetics in
Journal of Bioenergetics and Biomembranes, vol. 21, pp. 633–740, 1989.
16. See the paper by D. Ren et al. in Science, vol. 294, pp. 2372–2375, 2001, and accompanying
commentary by W. A. Catterall on pp. 2306–2308.
17. See the helpful overview of ion channels and their possible phylogenies by P. A. V. Anderson
and R. M. Greenberg in Comparative Biochemistry and Physiology B, vol. 129, pp. 17–28,
2001. They support the idea that the sodium channel derived from a low voltage-activated
T-type calcium channel.
18. Concerning the origins of electrical excitability and the evolution of the sodium pump see the
article by M. Strong et al. in Molecular Biology and Evolution, vol. 10, pp. 221–242, 1993.
Interestingly, the bacterial sodium channel (note 16) also seems to have an evolutionary
connection to the calcium channel, but a convergent origin seems likely.
19. See the paper by S. H. Heinemann et al. in Nature, vol. 356, pp. 441–443, 1992. For other
ways of altering a sodium channel so that it allows the passage of calcium, see L. F. Santana
et al. in Science, vol. 279, pp. 1027–1033, 1998 (and the commentary by D. Hanck on p. 1004).
20. See J. D. Spafford et al. (1999; citation is in note 10), p. 494.
21. See Milton Saier’s paper in BioEssays, vol. 16, pp. 23–29, 1994.
22. M. Saier (1994; citation is in note 21), p. 28.
23. The Drake equation is given as N = R ∗ fp ne fl fi ft L, where N is the number of communicative
civilizations, R ∗ is the rate of star formation, fp is the fraction of stars possessing a planetary
system, ne is the number of planets suitable for habitation, fl is the number of planets on
which life arises, fi is the fraction of biospheres where intelligence emerges, ft is the number
of civilizations that develop advanced communication, and L is the lifetime of a civilization.
24. For an excellent account of this early attempt at SETI and much else concerning life in the
Universe see S. J. Dick’s The biological universe: the twentieth century extraterrestrial life
debate and the limits of science (Cambridge University Press, Cambridge, 1996).
25. Leonard Ornstein’s wide-ranging and stimulating paper was published in Physics Today,
vol. 35 (March), pp. 27–31, 1982.
26. See T. F. Smith and H. J. Morowitz in Journal of Molecular Evolution, vol. 18, pp. 265–282,
1982.
27. L. Ornstein (1982; citation is in note 25), p. 27 (his emphasis).
28. L. Ornstein (1982; citation is in note 25), p. 29.
29. The proposal put forward by Ornstein is biologically a bit muddled, but in essence argues that
both cephalopod and vertebrate eyes derive from a common ancestor, also with a camera-eye.
A corollary of this suggestion is that supposedly primitive eyes, as seen in various molluscs,
are really regressive. On present evidence this seems decidedly less likely.
30. Robert Bieri’s article ‘Huminoids on other planets?’ was published in American Scientist,
vol. 52, pp. 452–458, 1964. Bieri’s work has also caught the attention of others interested in
the wider issues of evolution and the probabilities of extraterrestrial life; see, for example,
R. Puccetti’s absorbing discussion of humanoid convergences in Persons: A study of possible
moral agents in the universe (Macmillan, London; 1968) and N. J. Berrill’s Worlds apart: a
reflection on planets, life, and time (Sidgwick & Jackson, London; 1966), especially
chapter 10 where he remarks ‘Given the right encouragement, brains and minds are as likely
a natural evolutionary creation as legs, hearts, or wings . . . Wherever life may be . . . mind will
follow.’ (p. 149). See also Michael Ruse’s Can a Darwinian be a Christian? The relationship
between science and religion (Cambridge, Cambridge University Press, 2001), pp. 145–147.
Ruse also addresses the ideas of Simpson, noting ‘Sounding like a forerunner of Gould,
Simpson argues that it is chance and contingency all the way’, p. 147.
31. R. Bieri (1964; citation is in note 30), p. 452.
32. R. Bieri (1964; citation is in note 30), p. 457.
33. See Philip Morrison’s chapter (pp. 571–583) in Astronomical and biochemical origins and the
search for life in the universe, edited by C. B. Cosmovici et al. (Proceedings of the 5th
International Conference on Bioastronomy: Editrice Compositori, Bologna, 1997).
416 notes to pp. 233–236
34. P. Morrison (1997; citation is in note 33), p. 572. See also T. J. Grove and B. D. Sidell (in
Canadian Journal of Zoology, vol. 80, pp. 893–901, 2002), who document myoglobin
deficiency in the hearts of a variety of benthic fish of sluggish habit.
35. Interestingly, when a modification of the photosynthetic process evolves, notably emergence
of the so-called the C4 mechanism, it too shows rampant convergence, a topic which is
returned to in Chapter 10.
36. See A. Eschenmoser in Science, vol. 284, pp. 2118–2124, 1999, as well as discussion in
Chapter 2.
37. See S. J. Freeland and L. D. Hurst in Journal of Molecular Evolution, vol. 47, pp. 238–248,
1998, as well as discussion in Chapter 1.
38. In the description of a primitive rhizodont fish from Australia, Z. Johanson and P. E. Ahlberg
(Nature, vol. 394, pp. 569–573, 1998) note how limb-like are the fins, especially in the case of
the pectoral pair, and further note ‘similarities between rhizodont fins and tetrapod limbs are
thus probably convergent’, p. 569. See also their subsequent discussion in Nature, vol. 395,
pp. 792–794, 1998, of how the tetrapods arose out of one of several similar evolutionary
experiments among the osteolepiform fish.
39. See note 81 of Chapter 6 concerning the conversion of a caecilian amphibian to the
hydrostatic equivalent of a worm.
40. Concerning the distribution of the haemoglobins see the papers by R. C. Hardison in
Proceedings of the National Academy of Sciences, USA, vol. 93, pp. 5675–5679, 1996 and
Journal of Experimental Biology, vol. 201, pp. 1099–1117, 1998, as well as an overview in
American Scientist, vol. 87, pp. 126–137, 1999. Although emphasizing invertebrates the
review by R. E. Weber and S. N. Vinogradov in Physiological Reviews, vol. 81, pp. 569–628,
2001 is also helpful with updates of the occurrences of haemoglobin in prokaryotes and
protistans. A recent report, by S. Hou et al. (in Proceedings of the National Academy of
Sciences, USA, vol. 98, pp. 9353–9358, 2001) explores the role of these globins in bacteria, in
the context of oxygen sensors.
41. See Our molecular nature: The body’s motors, machines and messages (Copernicus, New
York, 1996), pp. 76–77.
42. For an overview of acetylcholine as ‘a universal cell molecule’ see I. Wessler et al. in Clinical
and Experimental Pharmacology and Physiology, vol. 26, pp. 198–205, 1999.
43. Concerning the cholinergic system in the ciliate Paramecium see, for example, the papers by
F. Trielli et al. in Journal of Experimental Zoology, vol. 279, pp. 633–638, 1997 and M. U.
Delmonte Corrado et al. in Journal of Experimental Biology, vol. 204, pp. 1901–1907, 2001.
44. See D. LeRoith et al. in Proceedings of the National Academy of Sciences, USA, vol. 79,
pp. 2086–2090, 1982, as well as B. Zipser et al. in Brain Research, vol. 463, pp. 296–304, 1988,
for further information on β-endorphin in the ciliate Tetrahymena. Concerning dopamine see
the papers in Biochimica et Biophysica Acta by M. E. Goldman et al. and R. E. Gunderson
and G. A. Thompson in vol. 676, pp. 221–225, 1981 and vol. 755, p. 186–194, 1983
respectively. Another neuropeptide, a somatostatin-like molecule, has been described by M.
Berelowitz et al.; see Endocrinology, vol. 110, pp. 1939–1944, 1982. The glycoprotein
choriogonadotropin, a hormone associated with mammalian pregnancy, is also reported from
the womb-free bacteria, albeit showing a much lower activity; see T. Maruo et al. in
Proceedings of the National Academy of Sciences, USA, vol. 76, pp. 6622–6626, 1979. A
similar occurrence is reported in yeast, see E. Loumaye et al. in Science, vol. 218,
pp. 1323–1325, 1982; in this case it is intriguing to note the peptide in yeast is responsible for
mating and the formation of the zygote.
45. Thus, signalling mechanisms involving glutamate receptors evidently evolved before the
divergence of plants and animals; see J. Chiu et al. in Molecular Biology and Evolution,
vol. 16, pp. 826–838, 1999. Signalling for disease resistance is also very primitive, given the
shared occurrence of a zinc-binding protein in protistans, plants, and animals; see K. Shirasu
et al. in Cell, vol. 99, pp. 355–366, 1999.
notes to pp. 236–239 417
46. G. O. Mackie (1990; citation is in note 12), see p. 909.

47. See B. J. Crespi’s paper ‘The evolution of social behavior in microorganisms’, in Trends in
Ecology & Evolution, vol. 16, pp. 178–183, 2001.
48. See, for example, the review entitled ‘The languages of bacteria’ by S. Schauder and
B. L. Bassler, in Genes & Development, vol. 15, pp. 1468–1480, 2001.
49. B. J. Crespi (2001, citation is in note 47), p. 182.
50. See, for example, J. A. Shapiro’s article in Annual Review of Microbiology, vol. 52,
pp. 81–104, 1998.
51. Effectively the argument in Rare Earth: Why complex life is uncommon in the Universe by
P. D. Ward and D. Brownlee (Copernicus, New York, 2000).
52. As well as the overview by M. Syvanen in Annual Review of Genetics, vol. 28, pp. 237–261,
1994, other more recent papers include those by H. M. Robertson and D. J. Lampe (in
Molecular Biology and Evolution, vol. 12, pp. 850–862, 1995), W. F. Doolittle (in Trends in
Biochemical Sciences, vol. 24 (12), pp. M5–M8, 1999), R. Jain et al. (in Proceedings of the
National Academy of Sciences, USA, vol. 96, pp. 3801–3806, 1999), and M. J. Leaver (in Gene,
vol. 271, pp. 203–214, 2001).
53. This general problem is usually referred to as the C-value paradox, C being the amount of
DNA in a cell. The evolution of genome size, edited by T. Cavalier-Smith (Wiley, Chichester,
1985) provides an introduction; an overview is given by T. R. Gregory in Biological Reviews,
vol. 76, pp. 65–101, 2001.
54. These results have been widely reported: J. F. Y. Brookfield in Current Biology, vol. 10,
pp. R514–R515, 2000 gives a helpful overview.
55. Concerning the body axis, where the Hox genes play an important role, see, for example,
papers by M. Akam in Philosophical Transactions of the Royal Society of London B, vol. 349,
pp. 313–319, 1995; S. B. Carroll in Nature, vol. 376, pp. 479–485, 1995; J. K. Grenier et al. in
Current Biology, vol. 7, pp. 547–553, 1997; and B. Schierwater and K. Kuhn in Molecular
Phylogeny and Evolution, vol. 9, pp. 375–381, 1999.
56. The topic of dorso-ventrality and the developmental similarities between the dorsal side of
vertebrate and ventral side of arthropods (and vice versa), with the implication that among
other things the hollow neural tube of the former might be homologous with the nerve cord
of the fly has attracted much attention. See D. Arendt and K. Nübler-Jung in Roux’s Archive
for Developmental Biology, vol. 203, pp. 357–366, 1994, and Development, vol. 126,
pp. 2309–2325, 1999; E. M. De Robertis and Y. Sasai in Nature, vol. 380, pp. 37–40, 1996;
J. Schmidt et al. in Development, vol. 121, pp. 4319–4328, 1995; E. L. Ferguson in Current
Opinions in Genetics and Development, vol. 6, pp. 424–431, 1996; and S. A. Holley et al. in
Nature, vol. 376, pp. 249–253, 1995 and Cell, vol. 86, pp. 607–617, 1996.
57. The similarities between the genetic architecture of fly wings and vertebrate limbs are
reviewed by S. J. Gaunt in Nature, vol. 368, pp. 324–325, 1997; N. Shubin et al. in Nature,
vol. 388, pp. 639–648, 1997; and C. J. Tabin et al. in American Zoologist, vol. 39, pp. 650–663,
1999. For a wider review see G. Panganiban et al. in Proceedings of the National Academy of
58. Eyes are the classic story of genetic similarity, especially via the Pax-6 gene. See, so to speak,
papers by R. Quiring et al. in Science, vol. 265, pp. 785–789, 1994; G. Halder et al. in Science,
vol. 267, pp. 1788–1792, 1995; N. M. Bonini et al. in Development, vol. 124, pp. 4819–4826,
1997; S. I. Tomarev et al. in Proceedings of the National Academy of Sciences, USA, vol. 94,
pp. 2421–2426, 1997; R. J. Davis et al. in Development, Genes and Evolution, vol. 209,
pp. 526–536, 1999; and P.-X. Yu et al. in Development, vol. 126, pp. 383–395, 1999.
59. The gene known as tinman plays a key role in heart formation. See, for example, Y. Fu et al.
in Development, Genes and Evolution, vol. 207, pp. 352–358, 1997.
60. See P. W. H. Holland and J. Garcı́a-Fernández in Developmental Biology, vol. 173,
pp. 382–395, 1996, and A. P. Martin in American Naturalist, vol. 154, pp. 111–128,
1999.
418 notes to pp. 239–245
61. See her paper in Journal of Experimental Zoology (Molecular Development and Evolution),
vol. 285, pp. 19–26, 1999.
62. S. F. Gilbert et al. (in Developmental Biology, vol. 173, pp. 357–372, 1996) point out that
hitherto curious examples of linked malformations, such as those affecting both limbs and
kidneys, are now more comprehensible because of shared genetic proximity early in
development.
63. Relevant papers include those by M. Koyanagi et al. in FEBS Letters, vol. 436, pp. 323–328,
1998; K. Ono et al. and H. Suga et al. respectively in Journal of Molecular Evolution, vol. 48,
pp. 654–662 and pp. 646–653, 1999; H. Suga et al. in Gene, vol. 280, pp. 195–201, 2001; and J.
Zhang et al. in Development, vol. 128, pp. 1607–1615, 2001.
64. Citation is in note 316, Chapter 7.
65. See papers in Development by C. Walther and P. Gruss in vol. 113, pp. 1435–1449, 1991; and
T. Yamasaki et al. in vol. 128, pp. 3133–3144, 2001; see also note 317, Chapter 7.
67. See L-I. Larsson et al. in Mechanisms of Development, vol. 79, pp. 153–159, 1998.
69. See A. D. Chisholm and H. R. Horvitz (pp. 52–54) and Y. Zhang and S. W. Emmons (pp. 55–59)
in vol. 377 of Nature (1995).
70. See E. Saló et al. in Gene, vol. 287, pp. 67–74, 2002. See also note 320, Chapter 7.
71. See A. Cvekl and J. Piatigorsky in BioEssays, vol. 18, pp. 621–630, 1996.
72. See G. Sheng et al. in Genes & Development, vol. 11, pp. 1122–1131, 1997.
73. See, for example, G. R. Merlo et al. in International Journal of Development Biology, vol. 44,
pp. 619–626, 2000, and R. I. Schwartz and E. N. Olson in Development, vol. 126,
pp. 4187–4192, 1999.
74. See the paper by F. Loosli et al. in Mechanisms of Development, vol. 74, pp. 159–164;
1998.
75. See, for example, C. Desplan in Cell, vol. 91, pp. 861–864, 1997.
76. This unease is well articulated by E. H. Davidson and G. Ruvkun in Journal of Experimental
Zoology (Molecular Development and Evolution), vol. 285, pp. 104–115, 1999, and is taken a
bit further in a review I wrote in Cell, vol. 100, pp. 1–11, 2000.
77. See Lisa Nagy’s article in American Zoologist, vol. 38, pp. 818–828, 1998.
78. L. Nagy (1998; citation is in note 77), pp. 818–819.
79. See B. Mittmann and G. Schultz in Development, Genes and Evolution, vol. 211,
pp. 232–243, 2001.
80. See, for example, A. Simeone et al. in Current Opinion in Genetics & Development, vol. 12,
pp. 409–415, 2002.
81. See Nic Williams and Peter Holland’s article in Molecular Biology and Evolution, vol. 15,
pp. 600–607, 1998.
82. N. Williams and P. Holland (1998; citation is in note 81), p. 606.
83. Again and again we see that a trivial change, sometimes a single substitution of an amino
acid at a key site in a protein, can have major consequences for its function. Not surprisingly,
given this sensitivity examples of molecular convergence are more frequent than sometimes
realized. This topic is returned to in Chapter 10.
84. See the two papers in Science dealing with chimp/human immunochemistry and genes
respectively by V. M. Sarich and A. C. Wilson (vol. 158, pp. 1200–1203, 1967) and M. C. King
and A. C. Wilson (vol. 188, pp. 107–116, 1975). An overview is given by P. Gagneaux and
A. Varki in Molecular Phylogenetics and Evolution, vol. 18, pp. 2–13, 2001.
85. Appropriately entitled ‘Jack of all trades’, further details of this amazing story can be found in
Nature, vol. 347, p. 704, 1990. I am most grateful also to Euan for a copy of an original
newspaper clipping from South Africa, from which some of the information is taken.
86. For an overview see the chapter (pp. 216–244) by H. J. Jerison in Handbook of intelligence,
edited by R. J. Sternberg (Cambridge, Cambridge University Press, 2000).
notes to pp. 246–249 419
87. See H. J. Jerison’s Evolution of the brain and intelligence (Academic Press, New York, 1973),
especially his Fig. 3.6 (p. 7) and Table 10.3 (p. 219), and chapter 21 (pp. 275–283) and
Appendix 6 of J. F. Eisenberg’s The mammalian radiations: An analysis of trends in
evolution, adaptation, and behaviour (Athlone Press, London, 1981).
88. His article entitled ‘Fast tracks to intelligence (considerations from neurobiology and
evolutionary biology)’ is on pp. 237–245 of Bioastronomy – The next steps, edited by G. Marx
(Kluwer, Dordrecht, 1988). See also note 99.
89. Concerning the primates see S. M. Reader and K. N. Laland in Proceedings of the National
Academy of Sciences, USA, vol. 99, pp. 4436–4441, 2002. The evidence for a correlation
between brain size and behavioural flexibility in birds is given by the group led by L. Lefebvre,
see articles in Oikos, vol. 90, pp. 599–605, 2000; Behaviour, vol. 139, pp. 939–973, 2002; and
vol. 137, pp. 1415–1429, 2000; and Animal Behaviour, vol. 63, pp. 495–502, 2002.
90. Although the following section emphasizes the dolphins, it is worth mentioning that there
are other fascinating convergences. One such involves an extinct Peruvian whale, probably
related to the monodontids (beluga, narwhal), that has become strikingly convergent on the
Northern Hemisphere walrus, referred to as ‘startling’ by C. de Muizon in Nature, vol. 365,
pp. 745–748, 1993; see also C. de Muizon et al. in Comptes Rendus de l’Académie des
Sciences, Paris; Sciences de la Terre et des Planètes, vol. 329, pp. 449–455, 1999, and C. de
Muizon and D. P. Domning in Zoological Journal of the Linnean Society, vol. 134,
pp. 423–454, 2002.
91. See L. Marino in Brain, Behavior and Evolution, vol. 51, pp. 230–238, 1998.
92. See the article by L. Marino in Evolutionary Anthropology, vol. 5, pp. 81–85, 1996.
93. The section that follows owes much to L. Marino’s paper ‘Convergence of complex cognitive
abilities in cetaceans and primates’ in Brain, Behavior and Evolution, vol. 59, pp. 21–32,
2002.
94. See L. Marino et al. in Journal of Mammalian Evolution, vol. 7, pp. 81–94, 2000.
95. See, for example, the papers by J. G. M. Thewissen et al. in Nature, vol. 413, pp. 277–281,
2001, with accompanying commentaries by C. de Muizon and H. Gee on pp. 259–260 of the
same issue, and J. Gatesy and M. A. O’Leary in Trends in Ecology & Evolution, vol. 16,
pp. 562–570, 2001. Also of direct relevance is The emergence of whales: Evolutionary
patterns in the origin of Cetacea, edited by J. G. M. Thewissen (Plenum, New York, 1998).
96. Lori Marino, personal communication (03/01/02).
97. In humans the brain accounts for about 2% of the body weight, but consumes 20% of the
total metabolic energy; see J. W. Mink et al. in American Journal of Physiology, vol. 241,
pp. R202–R212, 1981 and E. Armstrong in Science, vol. 220, pp. 1302–1304, 1983.
98. In a molecular phylogeny of the cetaceans M. Nikaido et al. in Proceedings of the National
Academy of Sciences, USA, vol. 98, pp. 7384–7389, 2001, identify a rapid radiation of this
group at about 30 Ma, with the dolphins diversifying at about 20 Ma.
99. Thus, in sceptical vein, Calvin (1988; see note 88 for citation) remarks of the standard
Darwinian view how it is that it continues to bask in optimism, ‘And thus, in this caricature
of evolutionary thought, progress towards intelligence seems inevitable (and indeed a
soothing counter to the inevitable disorder predicted by entropy) as long as variations keep
exploring the possibilities.,’ p. 238.
100. See R. E. Fordyce in Palaeogeography, Palaeoclimatology, Palaeoecology, vol. 31,
pp. 319–336, 1980, and his chapter (pp. 368–381) on cetacean evolution in Eocene–Oligocene
climatic and biotic evolution, edited by D. R. Prothero and W. A. Berggren (Princeton
University Press, Princeton, 1992).
101. The process by which deeper, nutrient-rich waters are brought to the surface. There are
several mechanisms by which this can occur, although the most familiar is persistent wind
shear that displaces surface waters, so allowing the ascent of the deeper waters.
102. This possibility is reviewed by several authors in Paleoclimate and evolution, with emphasis
on human origins, edited by E. S. Vrba et al. (Yale University Press, New Haven, 1995). See
420 notes to pp. 249–251
also J. Chaline et al. in Special Publications of the Geological Society of London vol. 181,
pp. 185–198, 2000.
103. See the chapter by R. C. Connor et al. (pp. 415–443) in Coalitions and alliances in humans
and other animals, edited by A. H. Harcourt and F. B. M. de Waal (Oxford University Press,
Oxford, 1992), and Cetacean societies: Field studies of dolphins and whales, edited by J.
Mann et al. (University of Chicago Press, Chicago, 2000).
104. D. Reiss et al., in Trends in Cognitive Science, vol. 1, pp. 140–145, 1997, p. 141. The extent of
the convergence between dolphin and chimp societies is reviewed by R. C. Connor et al. in
their chapters (pp. 91–126 and 247–269) in Mann et al. (2000).
105. See p. 91 of R. C. Connor et al. in J. Mann et al. (2000; citation is in note 103).
106. A similar fission–fusion society is also found in the northern bottlenose whale; see p. 217 of
R. C. Connor (in J. Mann et al. 2000; citation is in note 103).
107. L. Marino (2002; citation is in note 93), p. 27.
108. See, for example, the paper by R. C. Connor et al. in Behaviour, vol. 133, pp. 37–69, 1996.
109. See B. Würsig in Biological Bulletin, vol. 154, pp. 348–359, 1978, p. 355.
110. See the paper by R. C. Connor et al. in Proceedings of the Royal Society of London B,
vol. 268, 263–267, 2001.
111. Thus R. C. Connor et al. (in J. Mann et al. 2000; citation is in note 103) remark that at one
spot ‘in southern Brazil fishermen line up in murky thigh-deep water, holding weighted
throw nets, in pursuit of fish they cannot see. One or two dolphins, facing seaward several
metres offshore of the fisherman . . . come to an abrupt halt 5–7 m from the fisherman, diving
with a surging roll that cues the fisherman to toss their nets’ (p. 98). On pp. 211–212 R. C.
Connor notes similar cooperation on the coast of Mauritania.
112. See the paper by L. Weilgart et al. appropriately entitled ‘A colossal convergence’ in American
Scientist, vol. 84, pp. 278–287, 1996; see also R. C. Connor et al. in Trends in Ecology &
Evolution, vol. 13, pp. 228–232, 1998, and H. Whitehead and L. Weilgart (in J. Mann et al.
2000; citation is in note 103). This example has attracted the attention of those interested in
extraterrestrial intelligences; see P. Morrison on p. 573 of his chapter in Astronomical and
biochemical origins and the search for life in the Universe, pp. 571–583, edited by C. B.
Cosmovici et al. (Editrice Compositori, Bologna, 1997). Morrison here emphasizes the
similarities in their long-distance communication.
113. L. Weilgart et al. (1996; citation is in note 112), p. 278.
114. Concerning the long-distance communication by elephants (and some other large mammals)
see C. E. O’Connell-Rodwell et al. in American Zoologist, vol. 41, pp. 1157–1170, 2001. See
also discussion by B. T. Arnason et al. of elephants’ response to various geophysical fields in
Journal of Comparative Psychology, vol. 116, pp. 123–132, 2002.
115. H. P. Whitehead in Behavioral Ecology and Sociobiology, vol. 38, pp. 237–244, 1996.
116. See B. McCowan and D. Reiss on pp. 178–207 of Social influences on vocal development,
edited by C. Snowdon and M. Hausberger (Cambridge University Press, Cambridge, 1997).
The likely importance of communication in the fission–fusion communities is emphasized
by R. A. Smolker et al. in Behavioral Ecology and Sociobiology, vol. 33, pp. 393–402, 1992,
and L. S. Sayigh et al. in Animal Behaviour, vol. 57, pp. 41–50, 1998, who remark ‘Overall,
what is known of bottlenose dolphin behaviour and social structure supports the idea that
individuals do have concepts of one another as individuals and that they track the history of
their individual relationships,’ p. 48.
117. See D. Reiss and B. McCowan in Journal of Comparative Psychology, vol. 107, pp. 301–312,
1993, as well as P. L. Tyack in Behavioral Ecology and Sociobiology, vol. 18, pp. 251–257,
1986, and D. G. Richards et al. in Journal of Comparative Psychology, vol. 98, pp. 10–28, 1984.
118. See V. M. Janik in Science, vol. 289, pp. 1355–1357, 2000, and commentary by P. L. Tyack on
pp. 1310–1311.
119. Their paper on ‘The fallacy of the signature whistle hypothesis’ (p. 1159) is published in
Animal Behaviour, vol. 62, pp. 1151–1162, 2001.
notes to pp. 251–252 421
120. This is briefly reviewed by P. L. Tyack et al. in their chapter (pp. 333–339) in J. Mann et al.
(2000; citation is in note 103).
121. Thus Tayler and Saayman in Annals of the Cape Provincial Museums (Natural History),
vol. 9, pp. 11–49, 1972 note ‘It is remarkable that our dolphins [held in the Port Elizabeth
Oceanarium] have not learned to receive or to repeat – and the latter is possible by means of
their sonar system, there being no vocal chords – the complex harmonic composition of
human Afrikaans words,’ p. 46.
122. See K. Ralls et al. in Canadian Journal of Zoology, vol. 63, pp. 1050–1056, 1985.
123. K. Ralls et al. (1985; citation is in note 122), p. 1051.
124. See the article ‘A Beluga whale imitates human speech’ by R. L. Eaton in Carnivore, vol. 2,
pp. 22–23, 1979.
125. Among aquatic mammals at least two whales (beluga, humpback) and the Harbour seal show
evidence of vocal learning; see P. L. Tyack’s chapter (pp. 270–307) on cetacean communication
in J. Mann et al. (2000; citation is in note 103). He remarks, ‘Evidence for vocal learning
among seals . . . is particularly interesting from an evolutionary perspective, because the
pinnipeds evolved from a different terrestrial ancestor than the cetaceans. This suggests that
there were at least two independent origins of vocal learning among marine mammals,’ p. 306.
126. See B. McCowan and D. Reiss in Journal of Comparative Psychology, vol. 109, pp. 242–260,
1995, and B. McCowan et al. in Animal Behaviour, vol. 57, pp. 409–419, 1999.
127. B. McCowan and D. Reiss in Ethology, vol. 100, pp. 194–209, 1995, as well as comparisons to
Zipf’s Law by McCowan et al. (1999). See also the chapter by R. C. Connor et al. (pp. 91–126)
in J. Mann et al. 2000; citation is in note 103.
128. Thus in the pygmy marmoset, a New World monkey, the young also engage in a similar
‘babbling’. See A. M. Elowson et al. and C. T. Snowdon and A. M. Elowson in Behaviour,
vol. 135, pp. 643–664, 1998, and vol. 138, pp. 1235–1248, 2001 respectively. In an overview in
Trends in Cognitive Sciences, vol. 2, pp. 31–37, 1998 A. M. Elowson et al. remark ‘social
parallels make the Callitrichids [the group that includes the pygmy marmoset] a more
compelling analogous group [to humans] to study than the phylogenetically closer, but
socially dissimilar, apes,’ p. 32.
129. See their chapter (pp. 663–680) in Phonological development: Models, research, implications,
edited by C. A. Ferguson et al. (York Press, Timonium, MA, 1992); quotation is on p. 669.
130. Quotation is on p. 253 of the chapter (pp. 253–264) by D. Reiss in Marx (1988; citation is in
note 88) entitled ‘Can we communicate with other species on this planet? (Pragmatics of
communication between humanoid and non-humanoid species)’. See also B. McCowan et al.
1999 (citation is in note 126).
131. These famous ‘songs’ are produced only by the male. P. M. Gray et al. (2001; citation is in
note 171, Chapter 8) remark, ‘The undersea songs of humpback whales are similar in
structure to bird and human songs and prove that these marine mammals are inveterate
composers . . . humpback whale songs are constructed according to laws that are strikingly
similar to those adopted by human composers,’ p. 52.
132. Thus P. M. Gray et al. (2001; citation is in note 171, Chapter 8) comment ‘Is there a universal
music [as part of Mathematical Platonism] awaiting discovery, or is all music just a construct
of whatever mind is making it, human, bird, whale? The similarities among human music,
bird song, and whale song tempt one to speculate that the Platonic alternative may exist –
that there is a universal music awaiting discovery,’ p. 54.
133. The paper on ‘Comprehension of sentences by bottle-nosed dolphins’ is by L. M. Herman
et al., and is published in Cognition, vol. 16, pp. 129–219, 1984, while in Journal of
Experimental Psychology: General, vol. 122, pp. 184–194, 1993, L. M. Herman et al. report
responses by dolphins to syntactically anomalous sequences. In the same year L. M. Herman
et al. published a parallel paper, as Chapter 20 (pp. 403–442) in Language and
communication: Comparative perspectives, edited by H. L. Roitblat et al. (Lawrence Erlbaum
Associates, Hillsdale, NJ, 1993), where they remarked that it appeared that dolphins can
422 notes to pp. 252–254
generate ‘a mental representation of the grammatical structure of their language’ (p. 408) in a
way directly analogous to humans who when presented with grammatical novelties in an
artificial language are then in a position to decode other novel strings to at least a reasonable
degree of accuracy.
134. See L. M. Herman et al. in Animal Learning and Behavior, vol. 29, pp. 250–264, 2001.
135. L. Marino (2002, citation is in note 93), p. 28.
136. In a more general review on the evolution of intelligence E. MacPhail and J. J. Bolhuis, in
Biological Reviews, vol. 76, pp. 341–364, 2001, remark ‘that there are no qualitative
differences in cognition between animal species in the processes of learning and memory’
(p. 341). Interestingly, they also conclude that human language is a unique attribute,
concluding, ‘Unlike Darwin, we believe that there is a qualitative difference between human
and non-human intelligence; but we know of no convincing evidence that language evolved
to “solve” some novel “problem” posed by a niche invaded by an initially pre-linguistic
ancestor, nor any analysis of that niche’s demands that promises to throw light on the
mechanisms of language,’ p. 361.
137. See the discussion by E. Kako in Animal Behavior and Learning, vol. 27, pp. 1–14, 1999,
where he also addresses the question, in a moderately sceptical tone, of similar
comprehension by African Grey parrots and chimps. The response, on pp. 18–23, by L. M.
Herman and R. K. Uyeyama on the grammatical competency of bottlenose dolphins is an apt
counterpoint to Kako’s paper, to which he replies (pp. 26–27). The cases for African Grey
parrots and chimps are put forward respectively by I. M. Pepperberg (pp. 15–17) and S. G.
Shanker et al. (pp. 24–25).
138. Nor should we forget other attributes in birds such as tool-making (notes 176–178), and
remarkable accounts of coordinated hunting, such as in pied currawongs summoning
assistance to dispatch a rat; see M. G. O’Neill and R. J. Taylor in Corella, vol. 8l, pp. 95–96,
1984.
139. A point stressed by both I. M. Pepperberg (1999) and L. M. Herman and R. K. Uyeyama
(1999).
140. See the paper ‘Evolution of a universal grammar’ by M. A. Nowak et al., in Science, vol. 291,
pp. 114–118, 2001. B. McCowan et al. (in Journal of Comparative Psychology, vol. 116,
pp. 166–172, 2002) emphasize basic similarities in Zipf’s coefficient (see note 5, Chapter 1)
and Shannon’s measure of entropy between human words, bottlenose whistles, and squirrel
monkey calls at different stages of development. In each case communication has arisen for
specific adaptational reasons, but it is also convergent.
141. See R. K. Thompson and L. M. Herman in Science, vol. 195, pp. 501–503, 1977.
142. See E. Mercado et al. in Animal Learning & Behavior, vol. 26, pp. 210–218, 1998.
143. Their paper is in Journal of Experiment Psychology: General, vol. 124, pp. 391–408, 1995;
quotation is on p. 391.
144. ‘Half-awake to the risk of predation’ is the title of the paper by N. C. Rattenborg et al. in
Nature, vol. 397, pp. 397–398, 1999; a more general review of unihemispheric sleep in birds
(as well as cetaceans and some other marine mammals) is given by the same group in
Neuroscience and Biobehavioural Reviews, vol. 24, pp. 817–842, 2000. The origins of such
sleep may lie deeper than the birds, as some evidence suggests equivalent states in the
reptiles. Its evolution in two or three groups of marine mammals (and possibly the elephants,
p. 833), however, is almost certainly independent and regained from primitive Mesozoic
mammals that may have lost this ability when their mode of life was largely nocturnal and/or
burrowing.
145. See the paper by P. D. Goley in Marine Mammal Sciences, vol. 15, pp. 1054–1064, 1999, as
well as Rattenborg et al. (2000).
146. As reported by H. Elias and P. Swartz in Science, vol. 166, pp. 111–113, 1969; see also Jerison
(2000; citation is in note 86, pp. 222–223).
147. See L. Marino et al. in Brain, Behavior and Evolution, vol. 56, pp. 204–211, 2000.
notes to pp. 256–259 423
148. Thus in assessing the question of convergences in the dolphin brain, with the subtitle ‘Does
evolutionary history repeat itself?’ P. Manger et al. (in Journal of Cognitive Neuroscience,
vol. 10, pp. 153–166, 1998) emphasize how both the modularity of the brain and the
underlying constraints of neural architecture may lead to similar end points.
149. Helpful summaries are given by L. Marino (1996, 2002; see notes 92 and 93 for citations). See
also L. Marino et al. in Anatomical Record, vol. 264, pp. 397–414, 2001. Other key references
are by P. J. Morgane et al. in Brain Research Bulletin, vol. 5 (Supplement 3), pp. 1–107, 1980,
and I. I. Glezer et al. in Behavioral and Brain Sciences, vol. 11, pp. 75–89 (and succeeding
discussion on pp. 89–116), 1988.
151. See L. Marino (1996; citation is in note 92), pp. 84–85.
152. L. Marino (2002; citation is in note 93), p. 25.
153. See the paper on the scalable architecture of mammal brains by D. A. Clark et al. in Nature,
vol. 411, pp. 189–193, 2001, as well as the commentary by J. H. Kaas and C. E. Collins on
pp. 141–142.
155. See, for example, the paper on mosaic evolution of brain structure in mammals by R. A.
Barton and P. H. Harvey in Nature, vol. 405, pp. 1055–1058, 2000. The thrust of their
argument is that the mosaic evolution of the brain, with emphasis on those sections, e.g.
neocortex, that are functionally significant, will transcend developmental constraints.
156. See I. I. Glezer et al. on pp. 1–38 of Marine mammal sensory systems, edited by J. A. Thomas
et al. (Plenum, New York, 1992).
157. See the paper by B. McCowan et al. in Journal of Comparative Psychology, vol. 114,
pp. 98–106, 2000.
158. See the two chapters in the book Self-awareness in animals and humans: Developmental
perspectives, edited by S. T. Parker et al. (Cambridge University Press, Cambridge, 1994),
respectively by K. Marten and S. Psarakos (pp. 361–379) and L. Marino et al. (pp. 380–391).
159. See D. Reiss and L. Marino in Proceedings of the National Academy of Sciences, USA,
vol. 98, pp. 5937–5942, 2001.
160. D. Reiss and L. Marino (2001; citation is in note 159), p. 5942.
161. See, for example, L. von Ferson et al. on pp. 753–762 in Thomas et al. (1992; see note 156 for
citation), and also B. McCowan et al. (2000; citation is in note 157).
162. Particular attention has been paid to the killer whales (see, for example, J. K. B. Ford in
Canadian Journal of Zoology, vol. 69, pp. 1454–1483, 1991, and more recently P. J. O. Miller
and D. E. Bain in Animal Behaviour, vol. 60, pp. 617–628, 2000), as well as the sperm-whales
(see L. Weilgart and H. Whitehead in Behavioral Ecology and Sociobiology, vol. 40,
pp. 277–285, 1997). Information on elephant seals and Weddell seals can be found in papers in
Animal Behaviour, vol. 22, pp. 656–663, 1974 by B. J. Le Boeuf and L. F. Petrinovich (elephant
seals) and Canadian Journal of Zoology, vol. 61, pp. 2203–2214, 1984 by J. A. Thomas and I.
Stirling (Weddell seals).
163. The extent to which birds can be said to have cultural attributes is another fascinating (and
contentious) topic; the brief remarks (pp. 341–342) by S. K. Lynn and I. M. Pepperberg in L.
Rendell and H. Whitehead (2001, citation is in 164) are a forceful plea to take some birds as
seriously as killer whales and humans. The existence of dialects is also well known. T. F.
Wright and G. S. Wilkinson, in Proceedings of the Royal Society of London B, vol. 268,
pp. 609–616, 2001, present evidence in a parrot, the Yellow-naped Amazon, for emergence of
dialects despite gene flow between populations. See also P. Enggist-Dueblin and U. Pfister in
Animal Behaviour, vol. 64, pp. 831–841, 2002.
164. Key papers are those by H. Whitehead in Science, vol. 282, pp. 1708–1711, 1998 (with a
commentary by G. Vogel on p. 1616); V.-B. Deecke et al. in Animal Behaviour, vol. 60,
pp. 629–638, 2000, and L. Rendell and H. Whitehead in Behavioral and Brain Sciences,
vol. 24, pp. 309–382, 2001 (which includes a series of commentaries and critiques) and
424 notes to pp. 259–262
Proceedings of the Royal Society of London B, vol. 270, p. 225–231, 2003. Vertical cultural
transmission in stable matrilineal groups is reviewed by H. Whitehead and J. Mann in their
chapter (pp. 219–246) on female reproductive strategies in J. Mann et al. (2000; citation is in
note 103) where they comment: ‘it seems possible that cetacean societies . . . such as those of
killer, sperm, and pilot whales, contain cultures that are qualitatively more similar to human
cultures than are those of most terrestrial mammals . . . [and] may therefore explain curious
attributes of these species, such as nonadaptive mass strandings, low genetic diversity, and
menopause,’ pp. 243–244.
165. See H. Whitehead and J. Mann in J. Mann et al. (2000; citation is in note 103), pp. 219–246,
who remark, ‘It is notable that the cetacean species in which menopause has been found,
pilot and killer whales, possess matrilineally based social systems. There is some evidence
that cultural processes are very important within the matrilineal groups of these animals . . .
If an older female’s role as a source of information significantly increases her descendants’
fitness, and reproduction towards the end of her life decreases survival, menopause could be
adaptive,’ p. 233.
166. See K. McComb et al. in Science, vol. 292, pp. 491–494, 2001, as well as the commentary by
E. Pennisi on pp. 417–418.
167. Moreover, as Whitehead and Weilgart (2000; citation is in note 112) note, ‘Evidence is
beginning to accumulate that sperm whale units have distinctive vertically transmitted
cultures: functionally important information learned maternally or within matrilineal units.
Units have characteristic coda repertoires, which are unlikely to be genetically determined,’
p. 169.
168. Whitehead and Weilgart (2000) remind us the sperm whale ‘was the subject of two massive
hunts. The first of these, peaking in about 1840, greatly reduced sperm whale populations
throughout the world . . . while providing the oil that lubricated the Industrial Revolution’
(p. 155), while in the 1960s more than 20 000 a year were killed.
169. Their paper is published in Animal Behaviour, vol. 60, pp. 617–628, 2000.
170. Miller and Bain (2000), p. 626.
171. ‘Cultural revolution in whale songs’ is the title of the paper by M. J. Noad et al. in Nature,
vol. 408, p. 537, 2000.
172. For this, the first report of tool use by dolphins, see the paper by R. Smolker et al. in Ethology,
vol. 103, pp. 454–464, 1997.
173. This ‘bubble-feeding’ is reviewed by P. J. Clapham in his chapter (pp. 173–196) on the
humpback whale in J. Mann et al. (2000; citation is in note 103). There is also some
suggestion that unweaned calves blowing small bubble-clouds next to their feeding mothers
may be learning the technique.
174. His book is entitled Nature’s destiny: How the laws of biology reveal purpose in the Universe
(The Free Press, New York, 1998); see his Chapter 11. Denton’s work shows a number of
valuable parallels to this book but Nature’s destiny has, however, a distinct anti-evolutionary
bias (see my review ‘Awe, wonder and gene committees’ in the Times Higher Education
Supplement, p. 25 of the December 24/31, 1999 issue).
175. See, for example, J. Boswall in Avicultural Magazine, vol. 89, pp. 94–108 and 170–181, 1983.
176. See G. R. Hunt in Nature, vol. 379, pp. 249–251, 1996 (with commentary by C. Boesch on
pp. 207–208), and Emu, vol. 100, pp. 109–114, 2000 and vol. 102, pp. 349–353, 2002.
177. See also the report of hook manufacture in laboratory-kept New Caledonian crows by
A. A. S. Weir et al. in Science, vol. 297, p. 981, 2002.
178. Concerning this, see G. R. Hunt in Proceedings of the Royal Society of London B, vol. 267,
pp. 403–413, 1999, and G. R. Hunt et al. in Nature, vol. 414, p. 707, 2001.
179. G. R. Hunt (1999; citation is in note 178), p. 412. See also G. R. Hunt and R. D. Gray in
Proceedings of the Royal Society of London B, vol. 270, pp. 867–874, 2003.
180. See I. M. Pepperberg’s The Alex studies: cognitive and communicative abilities of Grey
Parrots (Harvard University Press, Cambridge, MA, 1999).
notes to pp. 262–263 425
181. See C. Caffrey in The Wilson Bulletin, vol. 112, pp. 283–284, 2000 and vol. 113, pp. 114–115,
2001.
182. See The use of tools by human and non-human primates, edited by A. Berthelet and J.
Chavaillon (Clarendon Press, Oxford, 1993), especially the chapters by C. Boesch and H.
Boesch (pp. 158–174) and Y. Sugiyama (pp. 175–190), and also W. M. McGrew’s Chimpanzee
material culture: Implications for human evolution (Cambridge University Press,
Cambridge, 1992).
183. See N. Toth et al. in Journal of Archaeological Science, vol. 20, pp. 81–91, 1993.
184. N. Toth et al. (1993; citation is in note 183), p. 89.
185. See W. C. McGrew in Science, vol. 288, p. 1747, 2000.
186. See T. S. Stoinski and B. B. Beck in Primates, vol. 42, pp. 319–326, 2001.
187. See, for example, G. C. Westergaard and S. J. Suomi in Current Anthropology, vol. 35,
pp. 75–79 and pp. 468–470, 1994; International Journal of Primatology, vol. 16,
pp. 1017–1024, 1995; and Journal of Archaeological Sciences, vol. 22, pp. 677–681, 1995; as
well as E. Visalberghi on pp. 118–135 of A. Berthelet and J. Chavaillon (1993; citation is in
note 182). In addition, as an extension of S. M. Reader and K. N. Laland (2002; citation is in
note 89), S. M. Reader has looked at plots of behavioural plasticity among the primates,
including tool use. Here too, with chimp and bonobo Cebus stands out. See his paper in
Towards a biology of traditions: models and evidence, edited by D. Fragaszy and S. Perry
(Cambridge University Press, Cambridge, in press).
188. See S. Chevalier-Skolnikoff in Behavioral and Brain Sciences, vol. 12, pp. 561–627,
1989.
189. See G. C. Westergaard and S. J. Suomi in Journal of Human Evolution, vol. 27, pp. 399–404,
1994.
190. G. C. Westergaard and S. J. Suomi (1994; citation is in note 189), p. 403.
191. See their paper in Journal of Human Evolution, vol. 6, pp. 623–641, 1977, see p. 634.
192. See Gregory Westergaard and colleagues’ paper in International Journal of Primatology,
vol. 20, pp. 153–162, 1999.
193. G. C. Westergaard et al. (1999; citation is in note 192), p. 161.
194. See S. Chevalier-Skolnikoff in Primates, vol. 31, pp. 375–383, 1990. K. A. Phillips (in
American Journal of Primatology, vol. 46, pp. 259–261, 1998) reports the use of ‘leaves as
cups to retrieve water from tree cavities’, p. 259.
195. E. Visalberghi (1993; citation is in note 187; see also note 182), see pp. 127–128.
196. The nature of the food resources may also be important; see M. A. Panger in American
Journal of Physical Anthropology, vol. 106, pp. 311–321, 1998.
197. See G. C. Westergaard and S. J. Suomi in Journal of Human Evolution, vol. 30, pp. 291–298,
1996.
198. G. C. Westergaard and S. J. Suomi (1996; citation is in note 197), p. 296.
199. See E. Visalberghi and L. Limongelli in their chapter (pp. 57–79) of Reaching into thought:
The minds of the great apes, edited by A. E. Russon et al. (Cambridge University Press,
Cambridge, 1996). Capuchins in the context of primate mentalities are reviewed by R. Byrne
in his chapter (pp. 110–124) in Creativity in human evolution and prehistory, edited by S.
Mithen (Routledge, London, 1998).
200. E. Jalles-Filho et al., in Journal of Human Evolution, vol. 40, pp. 365–377, 2001. They point
out that there may be cognitive limits that militate against such key activities as carrying the
tools to the food, although this may also be a consequence of a social structure whereby
access to food is largely controlled by aggression. Hence it makes little sense to wander off
looking for a suitable tool and leave the food unattended. However, food and tool exchanges
are known; see G. C. Westergaard and S. J. Suomi in American Journal of Primatology,
vol. 43, pp. 33–41, 1997, where they note, ‘we propose that tool- and food-sharing came into
existence through convergent evolution in large-brained, extractive foraging primate genera,
including Cebus, Pan and Homo’ (p. 40).
426 notes to pp. 263–267
201. See S. W. Williston, in Entomological News and Proceedings of the Entomological Section of
the Academy of Natural Sciences of Philadelphia, vol. 3, pp. 85–86, 1892.
202. S. W. Williston (1892; citation is in note 201), pp. 85–86.
203. For a review of parallel homoplasies in the mammalian neocortex, see R. G. Northcutt and
J. H. Kaas in Trends in Neurosciences, vol. 18, pp. 373–379, 1995.
204. Willem de Winter and Charles Oxnard’s paper is published in Nature, vol. 409, pp. 710–714,
2001.
205. The essay, by Thomas Nagel, entitled ‘What it is like to be a bat’, may be found in The
Philosophical Review, vol. 83, pp. 433–450, 1974.
206. T. Nagel (1974; citation is in note 205), p. 438.
207. T. Nagel (1974; citation is in note 204), pp. 439–440.
208. A similar point is made by R. Dawkins on pp. 33–35 The blind watchmaker (Norton, New
York, 1987) The emergence of novel functions in brain evolution in the context of
convergence is well addressed by K. C. Nishikawa (in Bioscience, vol. 47, pp. 341–354, 1997).
Thus he writes, ‘Convergent evolution of neural circuits that serve similar functions may
provide insights into the functional architecture of nervous systems. For example, features
such as parallel, distributed processing and population coding (in which a signal is encoded in
the activity of a population of neurons instead of a single cell) have evolved convergently in
distantly related species throughout the animal kingdom; these features likely represent
analogous solutions to similar problems in different animals,’ p. 342.
209. T. Nagel (1974; citation is in note 204), p. 441.
210. See U. M. Norberg and M. B. Fenton in Biological Journal of the Linnean Society, vol. 33,
pp. 383–394, 1988.
211. Concerning this, see the convergent evolution of trichromacy (Chapter 7, notes 122–124) and
the enzyme lysozyme (Chapter 10, notes 88, 89).
212. W. de Winter and C. Oxnard (2001, citation in note 204), pp. 713–714. Similar conclusions
were reached by F-J. Lapointe et al. (in Brain, Behavior and Evolution, vol. 54, pp. 119–126,
1999) where they noted ‘obvious adaptive convergences of the brain to trophic niches . . . one
could clearly distinguish a folivorous, an insectivorous/carnivorous, and a frugivorous/
nectarivorous clade based on size-corrected brain data,’ p. 122.
213. See also the example of convergence in echolocation between a Neotropical species of Myotis
and the temperate pipistrellids by M. Siemers et al. in Behavioural Ecology and Sociobiology,
vol. 50, pp. 317–328, 2001, who emphasize evolutionary flexibility over phylogenetic
constraint in response to particular ecological challenges.
214. See L. Krubitzer’s discussion of convergences in the mammalian neocortex in Trends in
Neurosciences, vol. 18, pp. 408–417, 1995.
215. L. Krubitzer (1995; citation is in note 214), p. 416.
216. The convergence between locomotory gaits of primates and the South American didelphid
Caluromys (the woolly opossum) is addressed by D. Schmitt and P. Lemelin in American
Journal of Physical Anthropology, vol. 118, pp. 231–238, 2002, and thus reinforces the various
other convergences already noted between this marsupial and the prosimian primates; see, for
example, P. Lemelin in Journal of Zoology, vol. 247, pp. 165–175, 1999 and D. T. Rasmussen
in American Journal of Primatology, vol. 22, pp. 263–277, 1990.
217. See the paper by J. G. Fleagle in Folia Primatologica, vol. 26, pp. 245–269, 1976.
218. J. C. Fleagle (1976; citation is in note 217), p. 264.
219. These are the platyrrhines, and include the capuchins (Cebus), the spider monkeys (Ateles),
the howlers (Alouatta), marmosets (Callithrix), and owl monkeys (Aotus); see M. Moynihan’s
The New World primates; Adaptive radiation and the evolution of social behavior,
languages, and intelligence (Princeton University Press, Princeton, 1976). These monkeys
evidently represent invaders from the Old World, arriving about 30 Ma ago (see R. F. Kay et al.
in Journal of Vertebrate Paleontology, vol. 18, pp. 189–199, 1998), most probably by rafting
across the Atlantic from Africa; see A. Houle in American Journal of Physical Anthropology,
notes to pp. 267–269 427
vol. 109, pp. 541–559, 1999. Concerning their interrelationships, see I. Horowitz and A.
Meyer on pp. 189–224 of Molecular evolution and adaptive radiation, edited by T. J. Givnish
and K. J. Sytsma (Cambridge University Press, Cambridge, 1997) and I. Horovitz in American
Museum Novitates, no. 3269, pp. 1–40, 1999, with inclusion of more fossil taxa. The
present-day fauna may be markedly impoverished in comparison to the fossil diversity. W. C.
Hartwig and C. Cartelle describe (in Nature, vol. 381, pp. 307–311, 1996) a giant Pleistocene
Brazilian monkey (Protopithecus) with a rather puzzling mixture of howler monkey-like
(cranial) and spider monkey-like (postcranial) anatomy.
220. See the papers in Folia Primatologica by R. A. Mittermeier in vol. 30, pp. 161–193, 1978, and
J. G. H. Cant in vol. 46, pp. 1–14, 1986. Concerning convergent evolution within the ateline
monkeys see C. A. Lockwood in American Journal of Physical Anthropology, vol. 108,
pp. 459–482, 1999. This paper is important, not only because of its implications for
convergence with the Old World apes, but also because of Lockwood’s emphases on the
importance of adaptation (Chapter 10, note 1) and the difficulty of discerning a reliable
phylogenetic signal in the face of rampant convergence. The evolutionary position of the New
World monkeys (platyrrhines) in the general scheme of primate phylogenies is addressed by
S. M. Ford on pp. 595–673 of Anthropoid origins, edited by J. G. Fleagle et al. (Plenum, New
York; 1994). She notes that the ancestral form was very generalized, readily diversified in a
number of directions, and that ‘The hominoids, in particular, and Pliopithecus [a Miocene
primate], demonstrate numerous convergences with atelines’ (p. 656); see also D. R. Begun in
Journal of Human Evolution, vol. 24, pp. 373–402, 1993.
221. See D. A. Clark et al. (2001; citation is in note 153), where they note that ‘neocortical volume
fractions have become successively larger in lemurs and lorises, New World monkeys, Old
World monkeys, and hominoids, lending support to the idea that primate brain architecture
has been driven by directed selective pressure,’ p.189.
222. Details of the convergences (and divergences) between atelines (and capuchins and squirrel
monkeys) with Old World primates is given by J. G. Robinson and C. H. Janson (pp. 69–82) in
Primate Societies, edited by B. B. Smuts et al. (University of Chicago Press, Chicago; 1987),
see pp. 80–82. A more general overview of convergences and divergences is in the chapter
(pp. 158–170) by P. M. Kappeler in Primate communities, edited by J. G. Fleagle et al.
223. See the paper by C. A. Chapman et al. in Behavioral Ecology and Sociobiology, vol. 36,
pp. 59–70, 1995. Quotation is on p. 59. These workers are careful to point out that the
similarities, while strong, are not precise.
224. M. Moynihan (1976, citation is in note 219); quotations on pp. 107 and 108
respectively.
225. Thus G. C. Westergaard and S. J. Suomi (1999; citation is in note 192) draw attention to
bipedalism in capuchin monkeys, interestingly in connection with food loads.
226. One of the curiosities of tool use in the primates is that given they are all pretty intelligent,
then why is the employment of tools so sporadic? Thus W. C. McGrew has observed that ‘All
great apes are smart enough to use tools but they only do so in useful circumstances’, see
p. 470 of his chapter (pp. 457–472) in Comparative socioecology: The behavioural ecology of
humans and other mammals, edited by V. Standen and R. A. Foley (Blackwell, Oxford, 1989).
And these ‘useful circumstances’ seem to be best correlated with access to animal material, a
point returned to in his chapter (pp. 143–157) in A. Berthelet and J. Chavaillon (1993; citation
is in note 182).
227. See the paper by F. Spoor et al. in Nature, vol. 369, pp. 645–648, 1994, as well as further
discussion in Journal of Human Evolution, vol. 30, pp. 183–187, 1996. The crux of their
argument involves the structure of the vestibular system (semicircular canals) of the inner
ear, which in the obligatory bipeds has to be very finely tuned to permit balance (not least on
a bicycle). They conclude that bipedality was less well developed in early hominids, but
certainly present by Homo erectus times.
428 notes to pp. 269–270
228. See P. E. Wheeler in Journal of Human Evolution, vol. 13, pp. 91–98, 1984, vol. 14, pp. 23–28,
1985, vol. 21, pp. 107–116 and 117–136, 1991, and vol. 23, pp. 379–388, 1992. A critique of
some of these ideas is presented by L. Q. Amaral in vol. 30, pp. 357–366, 1996.
229. For an overview, see A. Gibbons in Science, vol. 295, pp. 1214–1219, 2002.
230. B. Wood in Nature, vol. 418, pp. 145–151, 2002.
231. B. Wood (2002; citation is in note 230), p. 134.
232. Dating of the beds that yield Oreopithecus is reported by L. Rook et al. in Journal of Human
Evolution, vol. 39, pp. 577–582, 2000.
233. The phylogenetic position of this animal is difficult to resolve on account of its peculiar
anatomy. S. Moyà Solà and M. Köhler, in Comptes Rendus de l’Académie des Sciences, Paris,
vol. 324 (Ser. IIa), pp. 141–148, 1997, place Oreopithecus close to such apes as Dryopithecus.
On the other hand T. Harrison and L. Rook in their chapter (pp. 327–362) in Function,
phylogeny and fossils: Miocene hominoid evolution and adaptations, edited by D. R. Begun
et al. (Plenum, New York, 1997), tentatively conclude that Oreopithecus is very close to, if
not actually, a hominid, but they emphasize also specific features that make convergence an
alternative possibility; see also E. E. Sarmiento in American Museum Novitates, no. 2881,
pp. 1–44, 1987.
234. See R. L. Bernor et al. in Bollettino della Società Paleontologia Italiana, vol. 40, pp. 139–148,
2001.
235. See M. Köhler and S. Moyà-Solà in Proceedings of the National Academy of Sciences, USA,
vol. 94, pp. 11747–11750, 1997.
236. M. Köhler and S. Moyà-Solà (1997; citation is in note 235), p. 11747.
237. See L. Rook et al. in Proceedings of the National Academy of Sciences, USA, vol. 96,
pp. 8795–8799, 1999, which examines the bone structure of the hip bone known as the iliac,
and specifically the cancellous bone architecture that orientates itself according to the
prevailing stress regime, which is characteristic for bipedal activity.
238. See S. Moyà-Solà et al. in Proceedings of the National Academy of Sciences, USA, vol. 96,
pp. 313–317, 1999, where they remark ‘The functional resemblances with the
australopithecine pattern suggests for Oreopithecus similar manipulative skills, with
improved finger control and the capability to hold objects securely and steadily between the
pads of thumb and index finger . . . This strongly suggests that the hand morphology of
Oreopithecus is derived for apes and convergent with that of early hominids,’ p. 316.
239. See M. Köhler and S. Moyà-Solà (1997, citation is in note 235). They remark ‘insular
ecosystems . . . are characterized by lack of predators and limitation of space and thus of
trophic resources . . . Whereas the absence of predation removes the need for adaptations
related to predator avoidance, intraspecific and interspecific competition for food resources
increases . . . Both factors impose specific selective pressures that favor . . . adaptations . . .
related to energetically less expensive locomotor activities . . . and to reduction of bone mass
in the locomotor apparatus at the expense of mobility and speed. On the other hand they
select for feeding strategies that increase the efficiency of resource utilization . . . These
adaptations are universally found in all mammal faunas of small islands. These selective
pressures probably played a crucial role in the evolution of Oreopithecus . . . the lack of
predators may have led to a decrease of energetically expensive . . . and risky . . . climbing
activities, while favoring significant terrestriality’ (pp. 11749–11750). Some support for this
also comes from the reduction in canine size. As D. M. Alba et al. (in Journal of Human
Evolution, vol. 40, pp. 1–16, 2001) discuss whether this decrease could represent diminished
levels of agonistic behaviour, especially in the males, itself a reflection of the relaxation of
selective pressures that result from insularity and the absence of predators.
240. Being marooned on islands has other evolutionary consequences. Such environments, most
famously in the case of the Galápagos and Hawaiian island chain, have long been famous as
natural evolutionary ‘laboratories’. Indeed, while the island habitat is a rich source of
evolutionary novelty, so, too, are they of convergence. The mammal-like kiwis and mice-like
notes to pp. 270–273 429
wetas of New Zealand (Chapter 8) are splendid examples. Island faunas show other types of
convergence, of which one of the most fascinating examples is how big animals get small, and
small animals big. Thus, in the various Mediterranean islands, for example, the fossil record
shows dwarf elephants (a metre high and weighing little more than a large human) and
hippopotamuses, along with mice the size of rabbits. For a further discussion of this topic, see
for example, the papers by P. Y. Sondaar on pp. 671–707 of Major patterns of vertebrate
evolution, edited by M. K. Hecht et al. (New York, Plenum, 1977); L. R. Heaney in Evolution,
vol. 32, pp. 29–44, 1978; M.V. Lomolino in American Naturalist, vol. 125, pp. 310–316, 1985;
V. L. Roth in Oxford Surveys in Evolutionary Biology, vol. 8, pp. 259–288, 1992; and J.
Damuth in Nature, vol. 365, pp. 748–750, 1993.
241. Part of the problem of assigning an appropriate phylogenetic position (citation is in note 233)
is the substantial modification of the skull that shows Oreopithecus had a jaw structure and
teeth consistent with it being a dynamic vegetarian.
242. The origin of skilled forelimb movement is reviewed by A. N. Iwaniuk and I. Q. Whishaw in
Trends in Neurosciences, vol. 23, pp. 372–376, 2000. They produce an impressive catalogue of
examples, and argue that these manipulative abilities are primitive and drawn upon by
various tetrapod vertebrates as and when appropriate.
243. Enthusiasts for the intriguing ‘aquatic ape’ hypothesis, originally put forward by Alister
Hardy and revived by Elaine Morgan, may find an echo in this idea of an insular origination.
244. See M. G. Leakey et al. in Nature, vol. 410, pp. 433–440, 2001, with a thoughtful commentary
on pp. 419–420 by D. E. Lieberman. Interestingly, in discussing where best to place this fossil,
Lieberman remarks ‘At present, it is hard to believe any reconstruction of hominin
relationships because of the abundance of independently evolved similarities in the hominin
fossil record. The complex mosaic of features seen in the new fossil will only exacerbate the
problem’ (p. 419).
245. See M. Brunet et al. in Nature, vol. 418, pp. 145–151, 2002; see also note 229.
246. See B. Senut et al. in Comptes Rendus de l’Académie des Sciences, Paris, Sciences de la Terre
et des Planètes, vol. 332, pp. 137–144, 2001. See also the commentary on both Kenyanthropus
and Orrorin by L. E. Aiello and M. Collard in Nature, vol. 410, pp. 526–527, 2001.
247. The importance of homoplasy in hominid evolution is addressed by H. M. McHenry and L. R.
Berger in Journal of Human Evolution, vol. 35, pp. 1–22, 1998.
248. Robust convergence is documented by R. R. Skelton and H. M. McHenry in Journal of
Human Evolution, vol. 23, pp. 309–349, 1992. They remark that a robust chewing habit also
emerged in a number of extinct primates, including the Madagascan Hadropithecus, and
Gigantopithecus, apparently as adaptations to xeric (i.e. desert) vegetation.
249. See Rob Foley’s commentary on Skelton and McHenry (1992), published in Trends in
Evolution & Ecology, vol. 8, pp. 196–197, 1993.
250. R. Foley (1993; citation is in note 249), p. 196.
251. See pp. 31–42 of Structure and contingency: Evolutionary processes in life and human
society, edited by J. Bintliff (Leicester University Press, London, 1999).
254. R. Foley (1999; citation is in note 249), pp. 40–41.
255. See the paper by S. Semaw et al. in Nature, vol. 385, pp. 333–336, 1997 with accompanying
commentary by B. A. Wood on pp. 292–293, as well as S. Semaw in Journal of Archaeological
Science, vol. 27, pp. 1197–1214, 2000, where he suggests that the oldest of the artefacts may
have been made by an australopithecine species (A. garhi).
256. See H. Roche et al. in Nature, vol. 399, pp. 57–60, 1999 (with an accompanying review by J.
Steele on pp. 24–25). Although written before these important discoveries were made an
interesting overview of the earliest hominid stone cultures is given by J. W. K. Harris and
S. D. Capaldo on pp. 196–220 of Berthelet and Chavaillon (1993; citation is in note 182).
While the vast bulk of our understanding of early hominid technology is based on lithic
430 notes to pp. 273–274
material, evidence dating back to c. 1.8 Ma suggests that bone tools were being used for
termite foraging; see L. R. Backwell and F. d’Errico in Proceedings of the National Academy
of Sciences, USA, vol. 98, pp. 1358–1363, 2001, with commentary by P. Shipman on pp. 1335–
1337.
257. Fragmentary remains of early hominids and frequently complex cave stratigraphies with
recurrent roof falls make definitive conclusions difficult, not to mention that rather too
seldom is a skeleton found grasping a stone tool. Even so, K. Kuman and R. J. Clarke, in
Journal of Human Evolution, vol. 38, pp. 827–847, 2000, present evidence for tool-making by
late (c. 2 Ma) Paranthropus. Somewhat older evidence for use of tools is given by T. R.
Pickering et al. in American Journal of Physical Anthropology, vol. 111, pp. 579–584, 2000.
Substantially older finds, from about 2.5 Ma ago, include bones with cut marks and percussive
damage that are almost certainly the product of intentional fleshing and marrow extraction.
Whether, however, these activities were at the hands of early Homo or australopithecines is
not known; see J. de Heinzelin et al. in Science, vol. 284, pp. 625–629, 1999.
258. See S. Elton et al. in Journal of Human Evolution, vol. 41, pp. 1–27, 2001.
259. S. Elton et al. (2001; citation is in note 258), pp. 23–24.
260. D. E. Lieberman et al. in Journal of Human Evolution, vol. 30, pp. 97–120, 1996.
261. D. E. Lieberman et al. (1996; citation is in note 260), p. 115.
262. See Gerrit van Vark’s paper in Perspectives in Human Biology, vol. 4, pp. 237–243, 1999.
263. G. van Vark (1999; citation is in note 262), p. 241.
264. See, however, H. Roche et al. (1999; citation is in note 256), who question the notion of
technological stasis in very early tool material; see also Semaw (2000; citation is in note 255)
for a somewhat different view (p. 1209).
265. See H. Thieme in Nature, vol. 385, pp. 807–810, 1997; and commentary by R. Dennell on
pp. 767–768.
266. A useful overview concerning the difficulties in establishing the use of fire by ancient
hominids is given by S. J. James in Current Anthropology, vol. 30, pp. 1–26, 1989. It is
important also to distinguish between levels of utility in fire use, and not to assume that fire
indicates spit-roasts, gravy pans, and a neat stack of logs.
267. Evidence from the famous Swartkrans Cave in South Africa is used by C. K. Brain and A.
Sillen in Nature, vol. 336, pp. 464–466, 1988 for use of fire as far back as 1–1.5 Ma. Similarly
R. V. Bellomo, in Journal of Human Evolution, vol. 27, pp. 173–195, 1994, identifies the
controlled use of fire by hominids from 1.6 Ma sediments in Kenya, but notes that they ‘did
not use fire for the purpose of hunting, cooking, preserving food, intentional plant selection,
vegetation clearing, or improving the flaking characteristics of lithic materials . . . [but]
primarily as a source of protection against predators, as a source of light, and/or a source of
heat’ (p. 173). However, S. Weiner et al. argue against the controlled use of fire by the
half-million-year-old Homo erectus inhabitants of the Zhoukoudian caves near Beijing in
Science, vol. 281, pp. 251–253, 1998; see also the accompanying commentary by B. Wuethrich
on pp. 165–166, as well as the interesting article by N. T. Boaz and R. L. Ciochon in Natural
History, vol. 110(2), pp. 46–51, 2001.
268. See A. Walker et al. in Nature vol. 296, pp. 248–250, 1982. An accessible account is also given
in The wisdom of the bones: In search of human origins by A. Walker and P. Shipman (Knopf,
New York, 1996); see pp. 158–167.
269. O. Bar-Yosef, in Cambridge Archaeological Journal, vol. 8, pp. 141–163; 1998, provides a
useful overview. The best evidence for this dramatic event is from Europe, but its origins
were possibly in either East Africa or the Levant, unless these are independent?
270. These, of course, are almost entirely lithic, but rare, and to some extent enigmatic, use of
bone tools is also known; see S. Gaudzinski in Journal of Archaeological Science, vol. 26,
pp. 125–141, 1999.
271. See the report by I. Turk et al. report (in L’Anthropologie, vol. 101, pp. 531–540, 1997) of a
possible flute from a Mousterian culture in Slovenia dated at c. 45 000 years bp. A common
notes to pp. 274–276 431
explanation, that the perforations, in this case in the femur of a cave bear, are the product of a
chomping carnivore, are dismissed by these researchers; see also M. Otte in Current
Anthropology, vol. 41, pp. 271–272, 2000, who supports the musical interpretation.
272. P. M. Gray et al. (2001; citation is in note 171, Chapter 8), p. 53. Recall also that the parallels
in the songs of birds and humpback whales suggest convergence, and possibly access to a
universal music.
273. Evidence for rather advanced cultures, including the manufacture of bone points, and possibly
dating back to c. 90 000 years bp, from Katanda in Zaire is reported by A. S. Brooks et al. and J.
F. Yellen et al. in Science, vol. 268, pp. 548–553 and pp. 553–556 respectively, 1995 (see also
accompanying commentary by A. Gibbons on pp. 495–496, which casts some doubts on the
reliability of these dates). Certainly the appearance of some of the artefacts, such as the
well-crafted harpoon points, is astonishingly modern, and J. E. Yellen (in African
Archaeological Review, vol. 15, pp. 173–198, 1998) argues for an effective continuity in an
African context, albeit not in terms of a specific culture.
274. See the lengthy arguments by S. McBrearty and A. S. Brooks in Journal of Human Evolution,
vol. 39, pp. 453–563, 2000. See also the overview by R. Foley and M. M. Lahr in Cambridge
Archaeological Journal, vol. 7, pp. 3–36, 1997, with an emphasis on the so-called Mode 3
(c. 250 000 years ago) technology as a key step in hominid behaviour.
275. Sites in southern Africa, such as Blombos Cave, located east of Cape Town, are of particular
importance in providing evidence for the emergence of modern human behaviour, and by
implication language, c. 70 000 years ago. See C. S. Henshilwood et al. in Journal of Human
Evolution, vol. 41, pp. 631–678, 2001.
276. A. Marshack in Current Anthropology, vol. 37, pp. 357–364, 1996, describes an intriguing find
from the Quneitra Mousterian site on the Golan Heights, consisting of a flat flint cortex with
regular incised lines, including nested semicircles. And its significance? Removing our
preconceptions is practically impossible, but Marshack tentatively suggests the pattern
represents a rainbow, but of ‘spiritual’ rather than pictorial significance.
277. See F. d’Errico et al. in Antiquity, vol. 75, pp. 309–315, 2001). These authors argue that the
striations were not the result of butchery, but rather were constructed as a series of
intentional strokes.
278. See C. S. Henshilwood et al. in Science, vol. 295, pp. 1278–1280, 2002.
279. Excellent introductions are: The Neandertals: Changing the image of mankind by E.
Trinkaus and P. Shipman (Pimlico, London, 1994); The Neanderthal legacy: an archaeological
perspective from western Europe by P. Mellars (Princeton University Press, Princeton, 1996);
and In search of the Neanderthals: solving the puzzle of human origins by C. Stringer and
C. Gamble (Thames and Hudson, London, 1993).
280. See, for example, J. H. Schwartz and I. Tattersall, in Proceedings of the National Academy of
Sciences, USA, vol. 93, pp. 10852–10854, 1996, on a peculiarity of the nasal structure in the
form of a prominent bony projection extending into the cavity; a commentary by J. T.
Laitman et al. on pp. 10543–10545, focuses on implications of air-flow through the nose, and
what this might tell us about respiratory adaptations and vocalizations.
281. See M. S. Ponce de Leon and C. P. E. Zollikofer in Nature, vol. 412, pp. 534–538, 2001; as well
as remarks on cranial form by D. E. Liebermann et al. in Proceedings of the National
282. See M. Krings et al. in Proceedings of the National Academy of Sciences, USA, vol. 96,
pp. 5581–5585, 1999, and Nature Genetics, vol. 26, pp. 144–146, 2000, as well as I. V.
Ovchinnikov et al. in Nature, vol. 404, pp. 490–493, 2000, and R. W. Schmitz et al. in
Proceedings of the National Academy of Sciences, USA, vol. 99, pp. 13342–13347, 2002. A
somewhat critical review of the present state of play, however, is offered by N. Caldararo and
S. Gabow in Ancient Biomolecules, vol. 3, pp. 135–158, 2000, and G. Gutiérrez et al. in
Molecular Biology and Evolution, vol. 19, pp. 1359–1366, 2002. Rob Foley, who agrees that
Neanderthals are probably a separate species, reminds me that the DNA evidence is equivocal
432 notes to pp. 276–278
inasmuch as the comparison refers to that of modern human genetic variation, which is
remarkably limited and probably reflects an evolution bottleneck (‘mitochondrial Eve’),
c. 150 000 years ago.
283. C. W. Marean in Journal of Human Evolution, vol. 35, pp. 111–136, 1998 questions the widely
held view that these hominids were principally scavengers, a view broadly shared by J. J.
Shea, who argues in Current Anthropology, vol. 39 (Supplement), S45–S78, 1998, that the
Neanderthals had a variety of hunting strategies but were possibly greater carnivores than
modern humans, albeit replying on close-quarter intercepts (see also note 285). Their role as
top-level carnivores is reinforced by isotopic study of their bones, especially with respect to
nitrogen (δ 15 N), see M. P. Richards et al. in Proceedings of the National Academy of Sciences,
USA, vol. 97, pp. 7663–7666, 2000; see also H. Bocheren et al. in Journal of Archaeological
Science, vol. 26, pp. 599–607, 1999, and V. Balter in Comptes Rendus de l’Académie des
Sciences, Paris: Sciences de la Terre et des Planètes, vol. 332, pp. 59–65, 2001.
284. A. Defleur et al. present convincing evidence from a site in southern France, reported in
Science, vol. 286, pp. 128–131, 1999; see also the accompanying commentary on pp. 18–19 by
E. Culotta.
285. For an engaging account of the excavations at Shanidar, see R. S. Solecki’s Shanidar: The
humanity of Neanderthal man (Penguin [Allen Lane], London, 1972), as well as the more
technical account in E. Trinkaus’s The Shanidar Neandertals (Academic Press, New York,
1983).
286. See E. Trinkaus and M. R. Zimmerman in American Journal of Physical Anthropology,
vol. 57, pp. 61–76, 1982. Evidence for care in more primitive Neanderthals, dating back to
about 170 000 years, is given by S. Lebel et al. in Proceedings of the National Academy of
Sciences, USA, vol. 98, pp. 11097–11102, 2001, where the discovery of a mandible with
evidence for an abscess may indicate preparation of softer food for the sufferer, a reasonable
assumption given that Neanderthal teeth typically show extensive wear suggesting heavy-
duty use in daily activities.
287. See T. D. Berger and E. Trinkaus in Journal of Archaeological Science, vol. 22, pp. 841–852,
1995.
288. T. D. Berger and E. Trinkaus (1995; citation is in note 287); both quotations are on p. 849.
289. For evidence of interpersonal violence and the likelihood of subsequent assistance, see C. P. E.
Zollikofer et al. in Proceedings of the National Academy of Sciences, USA, vol. 99, pp. 6444–
6448, 2002.
290. The paper, whose full title is ‘Grave shortcomings: the evidence for Neandertal burial’, by
R. H. Gargett in Current Anthropology, vol. 30, pp. 157–190, 1989. Gargett remounts his
sceptical attack in Journal of Human Evolution, vol. 37, pp. 27–90, 1999. See also note 291.
291. The claim for flowers in a Neanderthal burial in the Shanidar caves in Iraq is made by A.
Leroi-Gourhan in Science, vol. 190, pp. 562–565, 1975, but the evidence is based on the
abundance of flower pollen in the associated sediments, for which there could be alternative
explanations. The description of the excavation of the skeleton (Shanidar IV) is well described
in R. S. Solecki (1972; citation is in note 285), see pp. 173–178. Scepticism of this claim can be
found in Gargett (1989), p. 176, although in the discussion Leroi-Gourhan provides a short
response (p. 182), to which Gargett replies on p. 185.
292. B. Vandermeersch in Comptes Rendus de l’Académie des Sciences, Paris D, vol. 270,
pp. 298–301, 1970, however, interprets deer bones in association with a Neanderthal child at
Qafzeh as evidence of grave goods.
293. This includes the famous Kebara burial in Israel, see O. Bar-Yosef et al. in Current
Anthropology, vol. 27, pp. 63–64, 1986, and H. Valladas et al. in Nature, vol. 330, pp. 159–160,
1987. The latter paper provides an estimated age of c. 60 000 years. Other reports include
those by L. V. Golovanova et al. from the Mezmaiskaya cave in the northern Caucasus,
reported in Current Anthropology, vol. 40, pp. 77–86, 1999.
294. The hyoid is located in front of the larynx, and muscles attached to it help to control the
shape of this part of the vocal tract and thereby sound production. The Neanderthal hyoid is
notes to pp. 278–279 433
very similar to our own, leading some, e.g. A. B. Arensburg et al. in American Journal of
Physical Anthropology, vol. 83, pp. 137–146, 1990, to argue that their speech was effectively
the same as ours. In addition to the hyoid bone, much has been made of the so-called
‘descended larynx’ and its importance in the production of articulate speech. Interestingly,
this feature, thought to be unique to humans, is now known to occur in such mammals as the
red deer, famous for the stag’s roaring, and as such is another example of independent
evolution; see W. T. Fitch and D. Reby in Proceedings of the Royal Society of London B,
vol. 268, pp. 1669–1675, 2001. This is not to say that one day deer will speak, but as Fitch and
Reby point out it is possible that originally the human descended larynx was associated with
sexual display, specifically so-called size exaggeration (the bull-roarer effect), even though we
now associated this feature with the pleasures of hearing the male baritone.
295. See for example D. Lieberman et al. in Journal of Human Evolution, vol. 23, pp. 447–467,
1992. For an accessible and more popular introduction to the topic, inter alia, of Neanderthal
fluency, see P. Lieberman’s Eve spoke: Human language and human evolution (Macmillan
[Picador], London, 1998).
296. See R. F. Kay et al. in Proceedings of the National Academy of Sciences, USA, vol. 95,
pp. 5417–5419, 1998.
297. See A. M. MacLarnon and G. P. Hewitt in American Journal of Physical Anthropology,
vol. 109, pp. 341–363, 1999.
298. That may be true to the first approximation, but still requires qualification. Marshack (1996;
citation is in note 276), for example, draws attention to the example of a Neanderthal carving
of ‘an exquisite nonutilitarian oval plaque from a lamella of a compound mammoth molar at
Tata, Hungary, dated ca. 100,000 B.P.’ (p. 361); see also the article in Yearbook of Physical
Anthropology, vol. 32, pp. 1–34, 1989.
299. While taking a somewhat different slant on various matters the review on the origins of the
Aurignacian by S. E. Churchill and F. H. Smith in Yearbook of Physical Anthropology, vol. 43,
pp. 61–115, 2000 is of great value. In addition, what appear to be shell beads dating back to
about 40 000 years have been described from caves near the Mediterranean coast of Turkey
and Lebanon; see S. L. Kuhn et al. in Proceedings of the National Academy of Sciences, USA,
vol. 98, pp. 7641–7646, 2001.
300. The youngest Neanderthals are dated at about 28 000 years bp, see F. H. Smith et al. in
Proceedings of the National Academy of Sciences, USA, vol. 96, pp. 12281–12286,
1999.
301. For a helpful introduction see S. E. Churchill and F. H. Smith (2000; citation is in note 299),
pp. 76–77.
302. These spectacular examples of Châtelperronian expertise are discussed by A. Leroi-Gourhan
and A. Leroi-Gourhan in Gallia Préhistoire, vol. 7, pp. 1–64, 1965, and Marshack (1989;
citation is in note 298); other occurrences from Lot and Dordogne are reviewed by J. Pelegrin
in Cahiers du Quaternaire 20 (Technologie lithique le Châtelperronien de Roc-de-Combe
(Lot) et de la Côte (Dordogne) (CNRS Editions, Paris, 1995)). Unequivocal evidence of the
association of the Châtelperronian culture with Neandertal skeletal material is given by J-J.
Hublin et al. in Nature, vol. 381, pp. 224–226, 1996.
303. See J. J. Hublin et al. in Comptes Rendus d’Académie des Sciences, Paris, vol. 321 (IIa),
pp. 931–937, 1995. R. N. E. Barton et al. in Antiquity, vol. 73, pp. 13–23, 1999 provide a
preliminary report on the Neanderthals from Gibraltar, with dates as young as c. 32 000 years
BP, but still showing Mousterian technology. An overall review of the Iberian occurrences is
also given by F. d’Errico et al. in Current Anthropology, vol. 39 (Supplement), 1998,
pp. S19–S20 in the context of the identification of the Châtelperronian cultures. It is evident
that in Iberia the newly arriving Upper Palaeolithic industries of Homo sapiens made little, if
any impact, and d’Errico et al. use this finding to support their thesis (citation is in note 308)
that elsewhere in Europe the shift by Neanderthals to the Châtelperronian cultures was
independent of our species’ technological progression. In passing one should also remember
that occasional Châtelperronian artefacts have been found in Iberia, and as Hublin et al.
434 notes to pp. 279–281
(1995) remark ‘The occurrence of these artefacts is still puzzling’ (p. 934); perhaps
Neanderthal trade?
304. As suggested, for example, by R. White (in Annual Review of Anthropology, vol. 21,
pp. 537–564, 1992).
305. In his response as part of the discussion concerning Neanderthal acculturation in Current
Anthropology, vol. 39 (Supplement), pp. S1–S44, 1998, where Mellars remarks, ‘no one has
ever suggested that the copying of airplane forms in the New Guinea cargo cults implied a
knowledge of aeronautics or international travel’ (p. S26), a point he reiterates in a
continuation of the debate in a subsequent issue, vol. 40, pp. 341–364, 1999; see p. 350.
306. For an engaging and intriguing account of the John Frum cult in Tanna see Chapter 13 of
David Attenborough’s Quest in paradise (Lutterworth, London, 1960), with an illustration of
a cultic turboprop opposite p. 154. A useful overview of cargo cults is given by I. C. Jarvie on
pp. 133–137 of Encyclopedia of Papua and New Guinea (Melbourne University Press,
Melbourne, 1972).
307. From my chapter (pp. 329–347) in God and design: The teleological argument and modern
science, edited by N. A. Manson (Routledge, London, 2003), p. 340.
308. The key paper is by J. Zilhâo and F. d’Errico in Journal of World Prehistory, vol. 13, pp. 1–60,
1999; see also the short article by J. Zilhâo in the July/August 2000 issue of Archaeology on
pp. 24–31.
309. See, for example, the extended debates in two issues of Current Anthropology, vol. 39
(Supplement), pp. S1–S44, 1998 and vol. 40, pp. 341–364, 1999. See also D. Richter et al. in
Journal of Archaeological Science, vol. 27, pp. 71–89, 2000 (see pp. 84–86), and F. B. Harrold in
Journal of Anthropological Research, vol. 56, pp. 59–75, 2000, who provides a historical
overview (see especially pp. 68–69).
310. See E. Trinkhaus et al. in Journal of Archaeological Science, vol. 26, pp. 753–773, 1999, who
suggest that the massive proportions of the body are a reflection of a continuing existence in a
near-Arctic environment rather than a phylogenetic ‘baggage’.
311. This point is emphasized by d’Errico et al. in Current Anthropology, vol. 39 (Supplement);
1998, see pp. S11–S13; see also note 314.
312. See I. Karavanić and F. H. Smith in Journal of Human Evolution, vol. 34, pp. 223–248, 1998.
Subsequently these authors, with colleagues (citation is in note 264) obtained new, and
younger, dates. These indicated that in this region of Europe (Croatia) Neanderthals and
modern humans had not overlapped. Moreover, re-dating of an associated bone of a modern
human (a mere 5000 years bp) together with consideration of other dates, suggest that
modern humans arrived even later (c. 32 000 years bp) in Europe than thought. While this
does not rule out trade, it certainly strengthens the case for cultural independence.
313. A view in support of Neanderthal genetic assimilation into European populations is given by
S. E. Churchill and F. H. Smith (2000; citation is in note 299).
314. D. Kaufman weighs up the possibilities of contacts between H. sapiens and Neanderthals,
especially in the Levant, in Oxford Journal of Archaeology, vol. 20, pp. 219–240, 2001. He
notes that the Châtelperronian culture has certain hallmarks such as abundant use of
colouring and particular types of incised stones and pendants that were possibly adopted by
our ancestors.
315. S. E. Churchill and F. H. Smith (2000; citation is in note 299), pp. 105–106.
316. What L. G. Straus tartly refers to as ‘A review of reality’ concerning the likelihood that
Solutrean people crossed to North America is reviewed in American Antiquity, vol. 65,
pp. 219–226, 2000.
317. This strange occurrence from Tecaxic-Calixtlahuaca has been known for many years (see the
chapter (pp. 5–53) by S. C. Jeff in Man across the sea: Problems of Pre-Columbian contacts,
edited by C. L. Riley et al. (University of Texas Press, Austin, 1971)), and was brought back to
prominence by R. Hristov and S. Genovés, in Ancient Mesoamerica, vol. 10, pp. 207–213,
1999. Questions of the dating, using the techniques of thermoluminescence (TL), were
notes to pp. 281–284 435
subsequently raised by P. Schaaf and G. A. Wagner in the same journal (vol. 12, pp. 79–81,
2001), to which the original authors robustly reply (pp. 83–86). The principal point is that the
TL date is of relatively little use in establishing the authenticity of this find, but it still does
not support a colonial, i.e. a post 1492 ad date. Hristov and Genovés emphatically reject the
suggestion that the head was ‘planted’ and remind us that, given that the Canary Islands were
colonized by at least the fifth century bc, a chance crossing of the Atlantic more than a
thousand years in advance of Columbus is not inconceivable.
318. Interestingly, this is not the only possible evidence for transatlantic pre-Columbian contact
(excluding of course the well-known Viking forays). Jeff (1971) reviews a number of other
possible examples. The whole question of cultural contacts, ‘Diffusion versus independent
development’ to use Jeff’s chapter title, is, of course, relevant to cultural convergences. Do the
many striking similarities found between far-flung societies indicate independent innovation,
which in turn might reflect a ‘limitation of possibilities’ (p. 32), or the sharing of ideas?
319. In his justly famous introduction to English literature in the sixteenth century excluding
drama (Clarendon Press, Oxford, 1954), C. S. Lewis recalls how the great essayist Montaigne
‘passionately asks why so noble a discovery [of the Americas] could not have fallen to the
Ancients who might have spread civility where we have spread only corruption’ (p. 17).
10. evolution bound: the ubiquity of convergence

1. Now is the time to avoid that old chestnut of whether it is convergent evolution as against
parallel evolution. As W. R. Scott, in his prescient discussion of mammal evolution (in
Journal of Morphology, vol. 5, pp. 301–406, 1891) remarked ‘The problems of parallelism and
convergence . . . open up a discussion of far-reaching extent and importance . . . The distinction
between the two classes of phenomena is obviously one of degree rather than of kind, and it
will therefore be convenient to consider them together’ (p. 363). For a level-headed
comparison of convergent and parallel evolution based on taxonomy, see J. Kaster and J.
Berger in BioSystems, vol. 9, pp. 195–200, 1977. C. A. Lockwood and J. G. Fleagle (in Yearbook
of Physical Anthropology, vol. 42, pp. 189–232, 1999) offer a valuable review of this topic.
2. R. Bieri (1964, citation is in note 30, Chapter 9), p. 457.
3. Within the arthropods themselves arthropodization from a soft-bodied onychophore-like
ancestor may have occurred twice; see G. E. Budd in Transactions of the Royal Society of
Edinburgh: Earth Sciences, vol. 89, pp. 249–290, 1999; see p. 286; see also G. Fryer in
Biological Journal of the Linnean Society, vol. 58, pp. 1–55, 1996. Insects may be convergent;
see F. Nardi et al. in Science, vol. 299, pp. 1887–1889, 2002. Arthropods belong to a superclade
known as the Ecdysozoa, of which the other principal phyla are the priapulids and nematodes.
The rather obscure kinorhynchs, related to the priapulids have also developed
arthropodization, presumably independently. Finally an extinct group of Cambrian fossils,
the vetulicolians that are believed to be related to the chordates, show what may be a trend
towards a sort of arthropodization; see D. G. Shu et al. in Nature, vol. 414, pp. 419–424, 2001
(and commentary by H. Gee on pp. 407, 409).
4. See, for example, W. D. I. Rolfe’s paper on pp. 117–152 of The terrestrial environment and the
origin of land vertebrates, edited by A. L. Panchen (Academic Press, London, 1980), where he
writes ‘Hibbertopterus . . . has short, stubby limb segments and is noticeably hexapodous: it
would seem well adapted for movement on land despite its bulky opithosoma,’
p. 145.
5. Tracheal systems have evolved independently in winged insects, isopod crustaceans
(wood-lice and their relatives), arachnids (spiders), and onychophorans; see p.14 of J. Moore
and P. Willmer’s outstanding paper on convergence in Biological Reviews, vol. 72, pp. 1–60,
1997. See also W. Ripper in Zeitschrift f ür Wissenschaftliche Zoologie, vol. 138, pp. 303–369,
1931, and G. Pritchard et al. in Biological Journal of the Linnean Society, vol. 49, pp. 31–44,
1993.
6. See C. J. Klok et al. in Journal of Experimental Biology, vol. 205, pp. 1019–1029, 2002.
436 notes to pp. 285–288
7. See L. M. Meffert et al. in Journal of Evolutionary Biology, vol. 12, pp. 859–868, 1999. Their
principal conclusion is also of wider interest, because it appears that the convergence
overrides the founder-flush effect, whereby tiny populations pass through evolutionary
‘bottlenecks’ before rediversifying on the basis of a restricted genetic diversity.
8. See C. S. Henry et al. in Evolution, vol. 53, pp. 1165–1179, 1999, where they demonstrate
‘songs that are strikingly similar’ (p. 1165) in species of the green lacewing Chrysoperla in
western North America and Kyrgyzstan in central Asia. The degree of similarity is such that
North American representatives can be fooled by an Asian song.
9. For a possible example in the Hawaiian cricket Laupala, which like many groups in this
mid-ocean archipelago demonstrates a major adaptive radiation, see K. L. Shaw in Evolution,
vol. 50, pp. 237–255, 1996. Convergence of song in the more familiar field-cricket Gryllus is
discussed by R. G. Harrison in Evolution, vol. 33, pp. 1009–1023, 1979; see also R. G. Harrison
and S. M. Bogdanowicz (in Journal of Evolutionary Biology, vol. 8, pp. 209–232, 1995).
10. See R. Kusmierski et al. in Proceedings of the Royal Society of London B, vol. 264,
pp. 307–313, 1997, with an emphasis on the lifting of phylogenetic constraints.
11. A helpful overview of the bowers is given by M. Hansell in Bird nests and construction
behaviour (Cambridge University Press, Cambridge, 2000). Hansell draws attention
(pp. 212–213) to the independent development of elaborate display and so-called court objects
in other birds, as well as cichlid fish. In this context it is also interesting to note the positive
correlation between bower complexity and brain size; see J. Madden in Proceedings of the
Royal Society of London B, vol. 268, pp. 833–838, 2001.
12. See Chapter 9 of The rise and fall of the third chimpanzee (Radius, London, 1991).
13. See C. Sturmbauer et al. in Proceedings of the National Academy of Sciences, USA, vol. 93,
pp. 10855–10857, 1996. They remark that ‘adaptations to higher intertidal life, such as
excellent vision, deep burrowing, rapid locomotion, and water retention preceded and allowed
the rise of extensive periods of subaerial reproductive displays. Sexual selection on
morphology and behavior may also have been directional, resulting in strikingly similar
solutions irrespective of the phylogenetic position’ (p. 10857).
14. See G. F. Striedter and R. G. Northcutt in Brain, Behavior and Evolution, vol. 38,
pp. 177–189, 1991.
15. A. M. Paterson et al. in Evolution, vol. 49, pp. 974–989, 1995, take a positive, if not upbeat,
view of the use of behavioural data in helping to establish bird phylogenies. So, too, do M.
Kennedy et al. in Animal Behaviour, vol. 51, pp. 273–291, 1996, although some of their
results are qualified by subsequent work; see Kennedy et al. in Molecular Phylogenetics and
Evolution, vol. 17, pp. 345–359, 2000.
16. ‘Why are some protein structures so common?’ ask S. Govindarajan and R. A. Goldstein in
Proceedings of the National Academy of Sciences, USA, vol. 93, pp. 3341–3345, 1996, and
provide an answer in the context of highly optimizable structures.
17. See, for example, C-I. Brändén in Current Opinion in Structural Biology, vol. 1, pp. 978–983,
1991, as well as G. K. Farber in the same journal, vol. 3, pp. 409–412, 1993.
18. See, for example, the essay by J-F. Gibrat et al. in Current Opinion in Structural Biology,
vol. 6, pp. 377–385, 1996.
19. See, for example, the review by T. Takagi in Current Opinion in Structural Biology, vol. 3,
pp. 413–418, 1993. See also R. A. Watts et al. in Proceedings of the National Academy of
Sciences, USA, vol. 98, pp. 10119–10124, 2001, who document a haemoglobin common to
plants and various microorganisms. This protein possesses unique biochemical properties,
and is absent from animals (and fungi).
20. See the paper by L. Moens et al. in Molecular Biology and Evolution, vol. 13, pp. 324–333,
1996.
21. See A. Pesce et al. in EMBO Journal, vol. 19, pp. 2424–2434, 2000, p. 2432.
22. See T. Burmester in Journal of Comparative Physiology, vol. 172B, pp. 95–107, 2002.
Although the arthropod haemocyanin appears to have a single origin, there is a diversity of
notes to pp. 288–290 437
sub-unit forms and evidently these have emerged by convergence; see J. Markl and H. Decker
in Advances in Comparative and Environmental Physiology, vol. 13, pp. 325–376, 1992.
23. See C. P. Mangum in Proceedings of the Biological Society of Washington, vol. 103,
pp. 235–247, 1990; K. E. van Holde and K. I. Miller in Advances in Protein Chemistry, vol. 47,
pp. 1–81, 1995; and Burmester (2002). In their review K. E. van Holde et al. (in Journal of
Biological Chemistry, vol. 276, pp. 15563–15566, 2001) acknowledge the differences between
arthropod and molluscan haemocyanins, but suggest that despite their independent
evolution they may both derive from a primitive copper protein (see also A. Volbeda and W.
G. M. Hol in Journal of Molecular Biology, vol. 206, pp. 531–546, 1989).
24. For a review of the evolution of myoglobin see the paper by T. Suzuki and K. Imai in Cellular
and Molecular Life Sciences, vol. 54, pp. 979–1004, 1998.
25. See, for example, the paper by S-G. Hou et al. in Nature, vol. 403, pp. 540–544, 2000, where
they report the presence of myoglobin-like proteins in both the Archaea and Eubacteria that
serve to monitor oxygen levels.
26. The report of myoglobin in cyanobacteria is by M. Potts et al. in Science, vol. 256,
pp. 1690–1692, 1992. For its occurrence in the protistan Tetrahymena, and an overview of the
earlier literature, see S. Korenaga et al. in Biochimica et Cosmochimica Acta, vol. 1543,
pp. 131–145, 2000. Korenaga et al. conclude ‘that the contracted or truncated globins from
various types of unicellular organisms have a separate, distinct origin from conventional
globins’, p. 143.
27. See, for example, C. Busch in Comparative Biochemistry and Physiology, vol. 86A,
pp. 461–463, 1987. A. J. Lechner (in Journal of Applied Physiology, vol. 41, pp. 168–173, 1976)
also documents high levels of muscle myoglobin in the burrowing pocket gophers, and
remarks on the parallels with the deep-diving mammals.
28. See K. A. Joysey et al. on pp. 167–178 of Myoglobin, edited by A. G. Schnek and C.
Vandecasserie (Editions de l’Université de Bruxelles, Brussels, 1977); A. M. Gurnett et al. in
Journal of Protein Chemistry, vol. 3, pp. 445–454, 1984; and K. A. Joysey on pp. 34–48 of
Molecular evolution and the fossil record; Short Courses in Paleontology 1, edited by B.
Runnegar and J. W. Schopf (Paleontological Society, Knoxville, TN, 1988). These data are
preliminary, but we should note Joysey’s (1988) prescient remark, ‘it is my view that many
other examples of adaptation in molecular evolution will emerge in due course’, p. 47.
29. See T. Suzuki and T. Takagi in Journal of Molecular Biology, vol. 228, pp. 698–700, 1992; T.
Suzuki and K. Imai in Comparative Biochemistry and Physiology, vol. 117B, pp. 599–604,
1997; and the review by T. Suzuki et al. in Comparative Biochemistry and Physiology B,
vol. 121, pp. 117–128, 1998.
30. See T. Shimizu et al. in Journal of Biological Chemistry, vol. 253, pp. 4700–4706, 1978.
31. See F. Hirata et al. in Journal of Biological Chemistry, vol. 252, pp. 4637–4642, 1977.
32. The details of the gene structure of the IDO are reviewed by T. Suzuki et al. in Biochimica et
Biophysica Acta, vol. 1308, pp. 41–48, 1996. They posit a gene duplication in the IDO gene
that allowed the protein to be recruited for oxygen carrying.
33. See several papers by T. Suzuki et al. in Journal of Protein Chemistry, vol. 14, pp. 9–13, 1994,
and vol. 17, pp. 651–656 and 817–826, 1998 respectively.
34. See N. Maeda and W. M. Fitch in Journal of Biological Chemistry, vol. 257, pp. 2806–2815,
1982. See also T. J. Grove and B. D. Sidell in Canadian Journal of Zoology, vol. 80, pp.
893–901, 2002.
35. The knock-out experiments in mice are described by D. J. Garry et al. in Nature, vol. 395,
pp. 905–908, 1998.
36. See J. Gatesy et al. in Science, vol. 291, pp. 2603–2605, 2001.
37. See, for example, K. B. Storey and J. M. Storey in Annual Review of Physiology, vol. 54,
pp. 619–637, 1992.
38. Notothenioid fish show spectacular evidence for an adaptive radiation, as reviewed by J.
Montgomery and K. Clements in Trends in Ecology & Evolution, vol. 15, pp. 267–271, 2000.
438 notes to pp. 290–294
These authors also touch on such topics as loss of haemoglobin and myoglobin (see above
p. 289).
39. See L. Chen et al. in Proceedings of the National Academy of Sciences, USA, vol. 94,
pp. 3817–3822, 1997. This example of molecular convergence has received justifiably wide
attention, and J. M. Logsdon and W. F. Doolittle provide a thoughtful commentary of what
they describe as ‘a cool tale’, on pp. 3485–3487.
40. Details of this gene are given by L. Chen et al. in the paper preceding L. Chen et al. (1997), on
pp. 3811–3816. Further information on the glycoprotein itself is provided by A. N. Lane et al.
in Biophysics Journal, vol. 78, pp. 3195–3207, 2000.
41. Further details of the evolutionary steps from a trypsinogen-like protein to one with an
antifreeze function are given by C. H. C. Cheng and L. Chen in Nature, vol. 401, pp. 443–444,
1999.
42. G. L. Fletcher et al. in Annual Review of Physiology, vol. 63, pp. 359–390, 2001. Fletcher
et al. also suggest another example of molecular convergence among the antifreeze proteins,
this time involving C-type lectins in sea raven, herring, and smelt.
43. See P. L. Davies et al. in Philosophical Transactions of the Royal Society of London B,
vol. 357, pp. 927–935, 2002. As part of a general review Davies et al. draw attention to a
striking convergence in spruce budworm and mealworm beetle antifreeze proteins,
specifically at the ice-binding sites. Concerning antifreeze proteins in insects, with
comments on the equivalents in the fish, see also N. Li et al. in Journal of Experimental
Biology, vol. 201, pp. 2243–2251, 1998.
44. See W. J. Swanson and C. F. Aquadro in Journal of Molecular Evolution, vol. 54, pp. 403–410,
2002.
45. See M. Pitts and M. Roberts’s Fairweather Eden: Life in Britain half a million years ago as
revealed by the excavations at Boxgrove (Century, London, 1997).
46. See, for example, the papers by J. R. Petit et al. and K. M. Cuffey and F. Vimeux in Nature,
47. See, for example, the papers by S. Bains et al. in Science, vol. 285, pp. 724–727, 1999; G. R.
Dickens et al. in Geology, vol. 25, pp. 259–262, 1997; and M. E. Katz et al. in Science,
vol. 286, pp. 1531–1533, 1999. It is sobering to realize that the recovery time of the planet
after the LPTM was about 100 000 years. Massive methane release has also been implicated in
Quaternary events (see J. P. Kennett et al. in Science, vol. 288, pp. 128–133, 2000 (with
commentary by T. Blunier on pp. 68–69)) and the end-Cretaceous firestorm (see M. D. Max
et al. in Geo-Marine Letters, vol. 18, pp. 285–291, 1999).
48. See C4 plant biology, edited by R. F. Sage and R. K. Monson (Academic Press, San Diego,
1999), especially the introductory chapter (pp. 3–16) by the senior editor, and the overview by
R. F. Sage in Plant Biology, vol. 3, pp. 203–213, 2001.
49. The expansion of C4 ecosystems is documented by T. E. Cerling et al. in Nature, vol. 361,
pp. 344–345, 1993 and vol. 389, pp. 153–158, 1997, and Journal of Vertebrate Paleontology,
vol. 16, pp. 103–115, 1996.
50. See R. F. Sage (2001; citation is in note 48), and Y. Huang et al. in Science, vol. 293,
pp. 1647–1651, 2001, with commentary by R. A. Kerr on pp. 1572–1573.
51. See N. R. Sinha and E. A. Kellogg, in American Journal of Botany, vol. 83, pp. 1458–1470,
1996, where they remark, ‘Such complexity should be difficult to evolve, yet the [C4 ] pathway
has evolved multiple times in the history of the flowering plants’ (p. 1458). See also Kellogg’s
chapter on pp. 411–444 in R. F. Sage and R. K. Monson (1999; citation is in note 48).
52. See E. V. Voznesenskaya et al., in Nature, vol. 414, pp. 543–546, 2001; and H. Freitag and
W. Stichler in Plant Biology, vol. 2, pp. 154–160, 2000, as well as an overview by R. F. Sage in
Trends in Plant Science, vol. 7, pp. 283–285, 2002.
53. See the chapter by R. K. Monson, on pp. 377–410 of R. F. Sage and R. K. Monson (1999;
citation is in note 48). Some of the necessary antecedents to C4 photosynthesis are discussed
by J. Hibberd and W. P. Quick in Nature, vol. 415, pp. 451–453, 2002 (with commentary by
J. A. Raven on pp. 375 and 377).
notes to pp. 294–295 439
54. See R. F. Sage (2001; citation is in note 48).

55. See J. R. Reinfelder et al. in Nature, vol. 407, pp. 996–999, 2000, as well as the preceding paper
by U. Riebesell on pp. 959–960. Scepticism was expressed by A. M. Johnston et al. in a
subsequent issue of Nature (vol. 412, pp. 40–41, 2001), to which Reinfelder offered a reply
(p. 41).
56. See M. M. M. Kuypers et al. in Nature, vol. 399, pp. 342–345, 1999. They also note that
increasing aridity may have played a part in this shift to C4 photosynthesis (see also note
50).
57. See T. E. Cerling et al. (1997; citation is in note 49) where in passing the authors note that
although the history of terrestrial plants has been dominated by C3 photosynthesizers ‘A
possible exception to this could have been during the late Carboniferous to Permian
glaciation if PCO2 levels were low enough and some plants independently evolved the C4
pathway, which was subsequently lost in the Mesozoic when CO2 levels were again high’,
p. 157.
58. See N. R. Sinha and E. A. Kellogg (1996; citation is in note 51); see also L. M. Giussani et al.
in American Journal of Botany, vol. 88, pp. 1993–2012, 2001.
59. See J. Kraut in Annual Review of Biochemistry, vol. 46, pp. 331–358, 1977, and I-M. Frick
et al. in Proceedings of the National Academy of Sciences USA, vol. 89, pp. 8532–8536, 1992.
Kraut notes the independent evolution of serine proteases in eukaryotes (e.g. trypsin) and
bacteria (subtilisin) (see also D. L. Ollis et al. in Protein Engineering, vol. 5, pp. 197–211,
1992). Their function is to break the peptide and ester bonds of other compounds, so making
them available for digestion. Concerning these two families Kraut remarks that ‘the two
enzyme families have entirely different overall three-dimensional structures and are therefore
probably descended from unrelated ancestral enzymes. Thus nature appears to have invented
the same biochemical mechanism on at least two separate occasions’ (p. 332). In a similar
vein Ollis et al. (1992) comment, ‘It is remarkable that there are now four different examples
of the catalytic triad which are related by convergent, not divergent evolution . . . We believe
that this reflects the primordial nature of hydrolysis . . . It also reflects how central hydrolysis
is to biochemical pathways, and how few solutions are possible . . . to the problem of
hydrolyzing esters and amides’ (p. 210). See also R. M. Garavito et al. in Biochemistry, vol. 16,
pp. 5065–5071, 1977, who also draw attention to convergences in terms of the active sites of
various enzymes, including lactate dehydrogenase, glyceraldehyde-3-phosphate
dehydrogenase, and papain.
60. N. D. Rawlings and A. J. Barrett in Biochemical Journal, vol. 290, pp. 205–218, 1993 remark
that the peptidases that catalyse the hydrolysis of peptide bonds, could be represented by up
to 60 evolutionary lines and that there may have been ‘many separate evolutionary origins of
peptidases,’ p. 216.
61. See T. Terada et al. in Nature Structural Biology, vol. 9, pp. 257–262, 2002. The details
concern two lysyl-tRNA synthetases which in an archael bacteria has a completely different
protein architecture from the ‘normal’ synthetase, but converges in terms of strategies for
substrate recognition.
62. See A. Beschin et al. in Nature, vol. 400, pp. 627–628, 1999.
63. See K. J. H. Robson et al. in Nature, vol. 335, pp. 79–82, 1988. R. F. Doolittle, in Trends in
Biochemical Sciences, vol. 19, pp. 15–18, 1994, expressed some caution about this example,
but conceded that if the sequences are ‘truly unrelated, this case would have to rate as the
nearest thing to sequence convergence yet reported’, p. 18.
64. See Y. Cao et al. in Journal of Molecular Evolution, vol. 47, pp. 307–322, 1998.
65. See G. Wu et al. in Proceedings of the National Academy of Sciences, USA, vol. 96,
pp. 6285–6290, 1999.
66. See S. Schenk and K. Decker in Journal of Molecular Evolution, vol. 48, pp. 178–186,
1999.
67. See S. J. Charnock et al. in Proceedings of the National Academy of Sciences, USA, vol. 99,
pp. 12067–12072, 2002, a convergence that the authors regard as ‘stunning’.
440 notes to pp. 295–298
68. See J. La Roche et al. in Proceedings of the National Academy of Sciences, USA, vol. 93,
pp. 15244–15248, 1996.
69. See P. Robson et al. in Molecular Biology and Evolution, vol. 17, pp. 1739–1752, 2000.
70. See Z. C. Shen and M. Jacobs-Lorena in Journal of Molecular Evolution, vol. 48, pp. 341–347,
1999.
71. See K. A. Crandall et al. in Molecular Biology and Evolution, vol. 16, pp. 372–382, 1999.
72. See K. H. Roux et al. in Proceedings of the National Academy of Sciences, USA, vol. 95,
pp. 11804–11809, 1998.
73. See G. E. Schulz in Current Opinion in Structural Biology, vol. 2, pp. 61–67, 1992; see also
the discussion of convergence among nucleic acid binding molecules by P. Graumann and M.
A. Maraherl in BioEssays, vol. 18, pp. 309–315, 1996.
74. See J. Hodgkin in Genes & Development, vol. 16, pp. 2322–2326, 2002, forming a
commentary on the paper by R. Lints and S. W. Emmons on pp. 2390–2407.
75. See C. S. Thummel and J. Chory in Genes & Development, vol. 16, pp. 3113–3129, 2002.
76. See E. Stebbins and J. E. Galán in Nature, vol. 412, pp. 701–705, 2001.
77. This is just one example, referring to the pathogen Yersinia pseudotuberculosis. In the case of
another bacterium, Salmonella, not only does the pathogen use a molecular convergence to
allow it to enter through the cell wall, but once inside the host cell the bacterium helps to
repair the damage to protect its new niche, again using methods of molecular mimicry.
78. E. Stebbins and J. E. Galán (2001; citation is in note 76), p. 703.
79. E. Stebbins and J. E. Galán (2001; citation is in note 76), p. 705.
80. See Y. Nakamura in Journal of Molecular Evolution, vol. 53, pp. 282–289, 2001. He
comments, ‘What’s so remarkable in molecular mimicry is the fact that the three proteins
structurally known as a tRNA mimic possesses completely different protein folds with
unrelated primary and secondary structures of protein’ (p. 287). Protein-DNA mimics have
also been recognized, see for example, the papers in Cell by D. C. Mol et al., vol. 82,
pp. 701–708, 1995 and D. Liu et al., vol. 94, pp. 573–583, 1998.
81. See K. Salehi-Ashtiani and J. W. Szostak in Nature, vol. 414, pp. 82–84, 2001. (See also note
46, Chapter 4).
82. K. Salehi-Ashtiani and J. W. Szostak (2001; citation is in note 81), p. 84.
83. See, in particular, the article by Doolittle (1994; citation is in note 63), with the title
‘Convergent evolution – the need to be explicit’.
84. See T. Bauchop and R. W. Martucci in Science, vol. 161, pp. 698–700, 1968. Ruminant-like
digestion is also known in the macropod marsupials, e.g. kangaroos (see D. W. Dellow et al. in
Australian Journal of Zoology, vol. 31, pp. 433–443, 1983), and possibly the sloths (see the
chapter (pp. 329–359) by G. G. Montgomery and M. E. Sunquist in The ecology of arboreal
folivores, edited by G. G. Montgomery (Smithsonian Institution Press, Washington, DC,
1978)). So far as I am aware there is no evidence whether the digestive system of at least the
macropods shows the molecular convergence of lysozymes.
85. The hoatzin is evidently fairly closely related to the cuckoos; see S. B. Hedges et al. in
Proceedings of the National Academy of Sciences, USA, vol. 92, pp. 11662–11665, 1995. E. S.
Morton, in his chapter (pp. 123–130) in Montgomery (1978) also draws attention to various
similarities between the hoatzin and New Zealand Owl Parrot, remarking ‘Both species are
highly convergent in ways directly attributable to their leaf-eating habits,’ p. 125.
86. Concerning its fore-gut fermentation systems, see A. Grajal et al. in Science, vol. 245,
pp. 1236–1238, 1989. The parallels with the system in ruminant mammals is emphasized by
A. Grajal in The Auk, vol. 112, pp. 20–28, 1995, and his remark that ‘Fore-gut fermentation in
a 680-g flying endotherm is theoretically unexpected’ (p. 26) is an important reminder of how
principles of convergence leap both phylogenetic barriers and sometimes our expectations.
87. These similarities include functional wing claws in the juveniles that enable them to climb
trees. It now seems that the limited flight abilities of the hoatzin are more to do with the need
to accommodate the massive fermentation chambers at the expense of the flight muscles.
notes to pp. 298–303 441
88. See C. B. Stewart and A. C. Wilson in Cold Spring Harbor Symposia on Quantitative Biology,
vol. 52, pp. 891–899, 1987.
89. See J. R. Kornegay et al. in Molecular Biology and Evolution, vol. 11, pp. 921–928, 1994.
90. See R. F. Doolittle (1994; citation is in note 63).
91. See C. B. Stewart and A. C. Wilson (1987; citation is in note 88), and also further evidence for
adaptive evolution of these langur monkey lysozymes in the paper by W. Messier and C. B.
Stewart in Nature, vol. 385, pp. 151–154, 1997 (with a commentary by P. M. Sharp on
pp. 111–112) and (with certain provisos) by Z. Yang in Molecular Biology and Evolution,
vol. 15, pp. 568–573, 1998 (see also the subsequent paper by Z. Yang and R. Nielsen in the
same journal, vol. 19, pp. 908–917, 2002). In addition, evidence exists in the langur monkey
for adaptive evolution in the ribonuclease enzyme, produced in the pancreas and involved
with digestion of the fore-gut bacteria. See J. Zhang et al. in Nature Genetics, vol. 30,
pp. 411–415, 2002, and commentary by S. Yokoyama on pp. 350–351. The evolution of the
enzyme is due to a few critical amino acid substitutions that allow the ribonuclease to
process large quantities of bacterial RNA in a lower pH environment. It will be interesting to
see if parallel substitutions occur in other ruminant species, in a way analogous to the ‘five
site rule’ of colour vision (note 168, Chapter 7).
92. See, for example, N. G. C. Smith and A. Eyre-Walker and J. C. Fay et al. in Nature, vol. 415,
pp. 1022–1024 and pp. 1024–1026 respectively, 2002, as well as overviews by Z. Yang and J. P.
Bielawksi in Trends in Ecology & Evolution, vol. 15, pp. 496–507, 2000 and M. Kreitman and
H. Akashi in Annual Review of Ecology and Systematics, vol. 26, pp. 402–422, 1995.
93. See M. Harry et al. in Molecular Phylogenetics and Evolution, vol. 9, pp. 542–551,
1998.
94. See T. B. Patterson and T. J. Givnish in Evolution, vol. 56, pp. 233–252, 2002.
95. Their paper is published in American Naturalist, vol. 146, pp. 349–364, 1995, where they
specifically address the topic of seed size.
96. See S. J. Gould and R. C. Lewontin in Proceedings of the Royal Society of London B, vol. 205,
pp. 581–598, 1979.
97. Not that every example of convergence is necessarily adaptive; see for example the important
paper by D. B. Wake in American Naturalist, vol. 138, pp. 543–567, 1991.
98. See Kirk Winemiller’s paper in Ecological Monographs, vol. 61, pp. 343–365, 1991; See also
note 134; Chapter 6.
99. K. Winemiller (1991; citation is in note 98), p. 361.
100. See Rob Foley’s chapter Pattern and process in hominid evolution, on pp. 31–42 of Structure
and contingency: Evolutionary processes in life and human society, edited by J. Bintliff
(Leicester University Press, London, 1999).
101. R. Foley (1999; citation is in note 100), p. 40. Oddly, in the same volume S. J. Gould writes of
Foley’s work that it ‘suggests a much chancier, much less guaranteed, much less repeatable
story replete with dominating contingency,’ p. xx. Did Gould actually read what Foley wrote,
I wonder?
102. See Eviator Nevo’s Mosaic evolution of subterranean mammals: Regression, progression and
global convergence (Oxford University Press, Oxford, 1999).
103. E. Nevo (1999), p. 208.
104. Notably his The blind watchmaker: why the evidence of evolution reveals a universe
without design (Norton, New York, 1987). Given Dawkins’s emphasis on adaptation and a
specific section on convergence, it is not surprising that a number of the examples given here
are also discussed in his book.
105. See for example, the papers by B. D. Patterson (in Journal of Mammalogy, vol. 80,
pp. 345–360, 1999) and S. B. Emerson (in Biological Journal of the Linnean Society, vol. 73,
pp. 139–151, 2001).
106. See, for example, G. Balavoine in Compte Rendu d’Académie des Sciences, Paris, Science de
la Vie, vol. 320, pp. 83–94, 1997. It should be noted, however, that the acoels are probably
442 notes to pp. 303–307
genuinely primitive within the triploblastic metazoans; see I. Ruiz-Trillo et al. in Proceedings
of the National Academy of Sciences, USA, vol. 99, pp. 11246–11251, 2002.
107. See the paper in Nature, vol. 401, p. 762, 1999, by M. Kobayashi et al.
108. See note 113, chapter 7.
109. See, for example, M. H. Tai et al. in Proceedings of the National Academy of Sciences, USA,
vol. 90, pp. 1852–1856, 1993. Similarly, in reviewing the Wilms’ tumour suppressor gene
(WT1), N. D. Hastie (in Cell, vol. 106, pp. 391–394, 2001) noted that the two principal
isoforms (−KTS, +KTS) have very distinct functions, but differ by only three amino
acids.
110. See M. Rosenquist et al. in Journal of Molecular Evolution, vol. 51, pp. 446–458, 2000.
111. See the overview by G. B. Golding and A. M. Dean in Molecular Biology and Evolution,
vol. 15, pp. 355–369, 1998, and more specifically papers by S. Brogna et al. in the same
journal, vol. 18, pp. 322–329, 2001; Y-H. Lee and V. D. Vacquier in Biological Bulletin,
vol. 182, pp. 97–104, 1992; W. J. Swanson and V. D. Vacquier in Proceedings of the National
Academy of Sciences, USA, vol. 92, pp. 4957–4961, 1995 and Science, vol. 281, pp. 710–712,
1998; J. Vieira and B. Charlesworth in Genetics, vol. 155, pp. 1701–1709, 2000; J. Vieira et al.
in Genetics, vol. 158, pp. 279–290, 2001; X. Gu et al. in Genetica, vol. 102/103, pp. 383–391,
1998; L. S. Jermiin et al. in Molecular Biology and Evolution, vol. 12, pp. 558–563, 1995; N.
A. Singhania et al. in Journal of Molecular Evolution, vol. 49, pp. 721–728, 1999; C. S. Willett
in Molecular Biology and Evolution, vol. 17, pp. 552–562, 2000. See also note 92.
112. See, for example, Evolutionary trends, edited by K. J. McNamara (Belhaven, London, 1990),
where a variety of views are expressed, as well as specific examples such as those concerning
the hinge mechanism of articulate brachiopods (S. J. Carlson in Paleobiology, vol. 18,
pp. 344–366, 1992) and suture complexity in Palaeozoic ammonoids (W. B. Saunders et al. in
Science, vol. 286, pp. 760–763, 1999). Also germane to this area is G. J. Vermeij’s interesting
essay in American Naturalist, vol. 153, pp. 243–253, 1999.
113. Janis and Damuth in K. J. McNamara (1990; citation is in note 112), p. 313.
114. Janis and Damuth in K. J. McNamara (1990; citation is in note 112), p. 337.
115. See his paper in Journal of Paleontology, vol. 62, pp. 319–329, 1988; see also S. C. Wang in
Evolution, vol. 55, pp. 849–858, 2001 for an analysis of passive and driven evolutionary
trends. A. H. Knoll and R. K. Bambach, in Paleobiology, vol. 26 (Supplement), pp. 1–14, 2000,
offer an outstanding critique on this issue.
116. See his paper in Nature, vol. 385, pp. 250–252, 1997, as well as the chapter (pp. 256–289) in
Evolutionary biology: in honor of James W. Valentine, edited by D. Jablonksi et al.
(University of Chicago Press, Chicago, 1996).
117. See J. Trammer and A. Kaim in Historical Biology, vol. 13, pp. 113–125, 1999.
118. See J. Alroy in Science, vol. 280, pp. 731–734, 1998.
119. J. Alroy (1998; citation is in note 118), p. 732.
120. See B. A. Maurer in Evolutionary Ecology, vol. 12, pp. 925–934, 1998.
121. B. A. Maurer (1998; citation is in note 120), p. 925.
122. Their paper is published in Evolution, vol. 46, pp. 939–953, 1992.
123. Maurer et al., p. 951; on p. 949 they argue that ‘Contrary to Gould’s (1988; citation is in note
115) assertion, more than random cladogenetic events are required to account for
evolutionary trends such as Cope’s rule.’
124. See Peter Wagner’s paper in Evolution, vol. 54, pp. 365–386, 2000. Also directly relevant is the
exploration by C. K. Boyce and A. H. Knoll (in Paleobiology, vol. 28, pp. 70–100, 2002) of
plant leaf form and the exhaustion of potentiality, leading to widespread convergence.
125. P. Wagner (2000; citation is in note 124), p. 382.
126. See S. J. Gould’s chapter (pp. 319–338) in Evolutionary progress, edited by M. H. Nitecki
(University of Chicago Press, Chicago, 1988).
127. Or more specifically ‘Progress is a noxious, culturally embedded, untestable, nonoperational,
intractable idea that must be replaced if we wish to understand the patterns of history,’
notes to pp. 307–313 443
p. 319. For a polite, short, and devastating demolition of this canard of Gould, see the letter by
F. K. McKinney in Science, vol. 237, p. 575, 1987.
128. In this context see the discussion by D. H. Geary et al. (in Paleobiology, vol. 28, pp. 208–221,
2002) discussion of iterative evolution in gastropods, with a wide-ranging and intelligent
analysis of the alternative explanations.
129. See, for example, G. B. West et al. in Science, vol. 284, pp. 1677–1679, 1999, and papers by K.
J. Niklas and B. J. Enquist and J. H. Marden and L. R. Allen, in Proceedings of the National
Academy of Sciences, USA, vol. 98, pp. 2922–2927; 2001 and vol. 99, pp. 4161–4166, 2002
respectively.
130. See also my paper in Astronomical Society of the Pacific Conference Series, (IAU Symposium
213) (in press).
131. See Lee Cronk That complex whole: Culture and the evolution of human behaviour
(Westview, Boulder, CO, 1999).
132. L. Cronk (1999; citation is in note 131), p. 26.
133. The question of such ‘navigation’ is not, of course, new; see, for example, K. J. Niklas in
Proceedings of the National Academy of Sciences, USA, vol. 91, pp. 6772–6779, 1994, and a
series of papers by S. Gavrilets, e.g. in Trends in Ecology & Evolution, vol. 12, pp. 307–312,
1997; American Naturalist, vol. 154, pp. 1–22, 1999; and Proceedings of the Royal Society of
London B, vol. 266, pp. 817–824, 1999.
134. L. Cronk (1999; citation is in note 131) invokes such an analogy when he writes, ‘we are in
search of the Great Attractor of human culture, the unseen mass that pulls human cultures
toward it and so limits their diversity,’ p. 26.
135. So L. Cronk (1999; citation is in note 131) writes ‘Maybe the problem is that only certain
pathways through ethnographic hyperspace are actually possible,’ p. 26.
11. towards a theology of evolution?

1. The predicament is clear for all to see. As R. Helms in Tolkien’s world (Thames and Hudson,
London, 1974) remarks, ‘From the end of the Middle Ages to the first nuclear explosion (to be
overly precise) our deepest spiritual urges have been Faustian, directing our emotional and
intellectual energies in an endless quest for knowledge of and power over nature, over our
world. Now we have come like Sauron; we can control nature, but we find in the process that
every controlling touch spoils and corrupts. Like Sauron, we can darken the sky, blast the
vegetation, pervert and control even the minds of men; and again like Sauron, we remain the
prisoners of our own assumptions, seeing no alternative to the constant expansion of our
corrupting control,’ p. 68.
2. The curious thing is that while animals, be they termites or cows, rely on symbionts to break
down the refractory cellulose, a few animals (arthropods) actually possess enzymes
(cellulases) that can break down cellulose. Most probably this ability was acquired by gene
transfer from bacteria, and it would be interesting to know why this apparently has not
happened more often; see H. Watanabe and G. Tokuda in Cellular and Molecular Life
Sciences, vol. 58, pp. 1167–1178, 2001.
3. S. Garcia Vallvé et al. (in Molecular Biology and Evolution, vol. 17, pp. 352–361, 2000),
however, report evidence for the lateral transfer of a gene involved with degradation of
cellulose, from a bacterium to a fungus. Both these organisms occur in mammalian rumens,
such as those of the cow.
4. See M. Carrier’s chapter (pp. 83–97) in Concepts, theories, and rationality in the biological
sciences, edited by G. Wolters and J. G. Lennox (Universitätsverlag and University of
Pittsburgh Press, Konstanz and Pittsburgh, 1995).
5. M. Carrier (1995; citation is in note 4), p. 95.
6. See D. Duboule and A. Wilkins in Trends in Genetics, vol. 14, pp. 54–59, 1998.
7. See R. N. Brandon’s essay on biological teleology in Studies in the History and Philosophy of
Science, vol. 12, pp. 91–105, 1981.
444 notes to pp. 313–322
8. The blue cross, the first story in The innocence of Father Brown (Cassell, London, 1940),
p. 22.
9. See A. Peacocke’s chapter (pp. 101–130) of Darwinism and divinity: Essays on evolution and
religious belief, edited by J. Durant (Blackwell, Oxford, 1985).
10. A. Peacocke (1985; citation is in note 9), p. 123
11. An example comes from a review (published in the Times Literary Supplement, 24 December,
1999) by Philip Kitcher of Matt Ridley’s book Genome where he remarks, ‘Ridley obviously
has a fine time sharing his delight [of the genome]. Indeed, perhaps he has too good a time. For
the booming voice of conviction that sounds through the chapters, from the initial discussion
of the origins of life to the philosophically limp conclusions about free will, is utterly certain
about everything. Like the village squire to the Victorian parson with “doubts”, Ridley
prescribes fresh air and exercise. He seems quite unworried by the thought that some of the
scientific claims he reports might be controversial or even unfounded, and even less
disconcerted by the possibility that . . . scientific truths might lead to social harm,’
p. 24.
12. J. C. Greene’s Debating Darwin: Adventures of a scholar (Regina Books, Claremont, 1999).
13. J. C. Greene (1999; citation is in note 12), p. 42.
15. See, in particular, J. C. Greene (1999; citation is in note 12) and his essay ‘Huxley to Huxley’
in his Science, ideology, and world view. Essays in the history of evolutionary ideas
(University of California Press, Berkeley, 1981).
16. Published by Penguin (London, 1998).
17. See my review in Geological Magazine, vol. 136, pp. 601–603, 1999.
18. Julian Huxley Evolution in action: Based on the Patten Foundation Lectures delivered at
Indiana University in 1951 (Chatto & Windus, London, 1953).
19. J. Huxley (1953; citation is in note 15), p.12; for its discussion in the context of process
theology, see J. C. Greene (1981; citation is in note 15), pp. 164–165.
20. J. Huxley (1953; citation is in note 19), pp. 152–153.
21. J. C. Greene (1981; citation is in note 15); both quotations are on p. 165.
22. J. C. Greene (1981; citation is in note 15), p. 168; see also in note 1.
23. Daniel Gasman The scientific origins of national socialism: Social Darwinism in Ernst
Haeckel and the German Monist League (Macdonald and American Elsevier, London and
New York, 1971).
24. D. Gasman (1971; citation is in note 23), explanation to Plate I, facing p. 8.
25. The close connection between Haeckel’s monistical phantasies and the rise of European
fascism, including not only Germany but Italy and France, is returned to by D. Gasman in
Haeckel’s Monism and the birth of fascist ideology, Studies in modern European history,
vol. 33, edited by F. J. Coppa (Peter Lang, New York, 1998).
26. Chapter 1 (pp. 9–32) of Structure: In science and art, edited by W. Pullan and H. Bhadeshia
27. D. Gasman (1971; citation is in note 23), p. 61.
28. S. Conway Morris (2001; citation is in note 26), pp. 30–31.
29. Concerning this trial see K. Tierney’s Darrow: A biography (Crowell, New York, 1979) and
E. J. Larson’s Summer for the gods: The Scopes trial and America’s continuing debate over
science and religion (Harvard University Press, Cambridge, 1997).
30. K. Tierney (1979; citation is in note 29), p. 341.
34. K. Tierney (1979; citation is in note 29), p. 358, taking a quote from W. Herberg.
35. K. Tierney (1979; citation is in note 29), p. 365
notes to pp. 323–330 445
37. See Susan Oyama’s article ‘The accidental chordate: contingency in developmental systems’
in South Atlantic Quarterly, vol. 94, pp. 509–526, 1995.
38. S. Oyama (1995; citation is in note 37), p. 512.
39. See Peter Koslowski’s chapter on pp. 301–328 in Sociobiology and bioeconomics: The theory
of evolution in biological and economic theory, edited by P. Koslowski (Springer, Berlin,
1999).
40. P. Koslowski (1999; citation is in note 39), p. 308.
41. But see W. J. Alonso and C. Shuck-Paim in Proceedings of the National Academy of Sciences,
USA, vol. 99, pp. 6843–6847, 2002.
42. See Eva Neumann-Held’s article on pp. 105–137 of Kowslowski (1999; citation is in note 39).
43. Consilience: The unity of knowledge (Knopf, New York, 1998).
44. As one reviewer of Consilience, the philosopher John Dupré, wrote, ‘the central thesis of the
book is vague, and many of the opinions expressed are quite eccentric’; see Science, vol. 280,
pp. 1395–1396, 1998.
45. See his Beyond evolution: Human nature and the limits of evolutionary explanation
(Clarendon Press, Oxford, 1997); in particular pp. 156–158.
46. The title of one of C. S. Lewis’s most influential books, first published in 1943.
47. See J. C. Greene’s article on Darwin and religion in Proceedings of the American
Philosophical Society, vol. 103, pp. 716–725, 1959.
50. See T. F. Smith and H. J. Morowitz in Journal of Molecular Evolution, vol. 18, pp. 265–282,
1982.
51. T. F. Smith and H. J. Morowitz (1982; citation is in note 50), p. 281.
52. C. S. Lewis’s That hideous strength: a modern fairy tale for grown-ups (Bodley Head, London,
1946).
53. C. S. Lewis (1946; citation is in note 52), pp. 46–48.
54. The title of his extraordinary book, the full title of which is Forbidden knowledge: From
Prometheus to pornography (Harcourt, Brace [Harvest], San Diego, 1997).
55. See R. S. Noel’s The mythology of Middle-Earth (Thames and Hudson, London, 1977), and
recall the power of the One Ring, whose property is to enthrall its owner and to lead him or
her to a damnation that ultimately may not be of their own choosing.
56. See Howard Van Till’s contribution on pp. 188–194 in Science & Christianity: Four views,
edited by R. F. Carlson (InterVarsity Press, Downers Grove, 2000).
57. H. Van Till (2000; citation is in note 56), p. 192 (his emphases).
58. Michael Polanyi Personal knowledge: Towards a post-critical philosophy (Routledge and
Kegan Paul, London, 1962).
59. M. Polanyi (1962; citation is in note 58), pp. 284–285.
60. See my book The crucible of creation: The Burgess Shale and the rise of animals (Oxford
University Press, Oxford, 1998), p. 223, and S. R. Taylor’s Destiny or chance: our solar system
and its place in the cosmos (Cambridge University Press, Cambridge, 1998), p. 204.
61. See notably D. A. Griffin’s Animal minds: Beyond cognition to consciousness (Chicago,
University of Chicago Press, 2001), as well as the discussion in Chapter 9.
62. See G. K. Chesterton’s ‘A defence of humility’ in The defendant (Dent, London, 1922).
63. G. K. Chesterton (1922; citation is in note 62), pp. 134–135.
64. Asked what most struck him of God’s creation, J. B. S. Haldane is said to have replied: an
inordinate fondness for beetles.
General index
Page numbers in italics indicate Figures.
α-helices 167, 170, 230, 357n55 343n61; proline 13; racemic mixture
α-proteobacteria 352n6 40–41, 60; serine 16; tryptophan 288,
abalone 289 n.12/1
absorption spectra 109, 289, 110 ammonia (NH3 ) 25, 62
acetylcholine 235–236 ammonites n.35/6
ACLU see American Civil Liberties Union amphibians 120, 190, 216, 220, 270,
Acromyrmex 198 289
Actinocoryne 230 amphioxus 6, 7, 180, 7
adenine 4, 13, 28, 30, 63, 64 Amyciaea 212–213
African golden moles 141 angiosperms 135–138
African Grey parrot 225–6, 253 annelids 154, 193; eyes 154
agriculture 2, 197–200, 205–207, 208–209, anthropic principles 327
210, 275, 401n25; ‘agro-predation’ 200, antibiotics 200, 207
207; ant ‘cartons’ 210; ‘bucket brigade’ ants 2, 8, 116–117, 143, 197–201, 208, 210,
198, 400n6; caches 198; climate change 212–213, 284, 388n201
206; cropping 199; division of labour aphids 114
198; dump-pits 205, 206; exchange 207; apinid honey bees 203
fertilizer 199, 210; fungus gardens 116, Apollo 11 70
197–200, 205–207, 206; herbicide 200; aporrhaids 157
Mesoamerica 206; monocultures 207, Arachnocampa 116
210; mulch 199; pathogens 200; Archaea 170, 171
pruning 199; waste disposal 205–206; archaeogastropods 289
weeding 199, 209, 210 Archaeopteryx 298
alanine 13, 16, 115, 340n32 Ardipithecus 271
albatrosses 2, 14, 3 argyrolagids 145–146
algae 62, 95, 107, 173, 294 army ants 116, 200–201, 203–205, 207,
Allan Hills meteorite (ALH 84001) 33–34, 210–211, 202, 406n98
75 arthropods 116, 193, 284, 288
ambrosia beetles 207–208 Asian swiftlets 181–182
American Civil Liberties Union asteroids 95, 96, 97
(ACLU) 320, 321 atmosphere 61–62, 72, 76, 107, 109;
amino acids 4, 5, 8, 9, 10, 13–14, 16, 18, 23, clouds 66
27, 35, 36, 40–41, 46, 53, 57, 59, 60, 61, 63, atonal gene 193
77, 115, 167, 168, 171, 287, 296, 298; ATP (adenosine triphosphate) 25, 49, 113,
α-amino isobutyric acid 37; alanine 13, 170
16, 115, 340n32; arginine 13, 288, 296; Atta 198
artificial 335n22; aspartic acid 41, 296; attine ants 116, 198–200, 205–207, 208,
dextral (d-) amino acids 340n30; 198
enantiomorphs 39–41, 60, 65; glycine Aurignacian culture 275, 278, 279, 280
13, 16, 33, 60, 115; impact origin 339n25; Australopithecus 269, 271
isovaline 37; left-handedness 40;
lysine 126; oligomerization 53–54, baboons 242–243, 244
general index 447
bacteria 1, 2, 18, 26, 34, 56, 74, 94, 99, 107, 256; Pax-6 gene 193, 240; scombrid fish
113, 114, 121, 124, 166, 170, 200, 230, 235, 223; sharks 399n327; size 245–248,
236, 288, 307; bacteriome 114; flagellum 245; size increases 248, 273–274, 436n11,
111, 230; genome 238; motor 111, 230 245; structural organization 265,
bacteriorhodopsin 170–172 266–268; teleost 194; temporal 256;
Barbourofelis 130 tenrec 246; toothed whales 246–247;
barnacles 119 torus semicircularis 196; valvula
bats 173, 181, 182, 192, 218, 256, 257, cerebelli 194; vertebrate 6, 179–180
265–268 Branchiostoma 6
bees 143, 202, 205, 207, 284 brittle-stars 155, 159–160, 406n75
beetles 143, 147, 149, 292, 372n4 bryozoans 126
beluga whale 251 bubble-rings 257–258, 257
bennettitaleans 137 budgerigars 219
biosynthetic pathways 16, 18, 336n39 butterflies 169
birds (see also cultures, sounds) 96, byssus threads 289
138–139, 190, 194, 223, 225–226, 246, 254,
269, 306, 387n170 C-value paradox 417n53
bivalves 158–159, 289, 159 C3 109, 293–294
bladder kelp 355n37 C4 109, 293–294
Blainville’s beaked whale 15 cactus 134–135, 134
Bombardier beetle 372n4 caddis-flies 116
Borszczowia 294 Caenorhabditis 4, 238
bottlenose whale 420n106 calcite 41, 159
bowerbirds 285–286 Callisto 348n49
Boxgrove Man 291, 292 camouflage 168, 215
brain (see also mental functions) 243, Campylomormyrus 184
244–249; amphioxus 6; bird 226, cancers 173, 240
412n180; blood–brain barrier 214; carbohydrates 9, 24, 27, 135
cephalopod 214; cerebellum 194, 195, carbon dioxide 55, 62, 76, 99, 108, 109, 216,
254, 256; cetacean 254; chimp 247; 292–293, 294, 295
cortical cytoarchitecture 256; cortical carbon monoxide 62, 63, 113, 356n38
structure 267; dolphin 247–249, 291, carrot 141
247, 255; dorsal cochlear nucleus castes 199
396n284; electrosensory line lobe 189; cats 147–148, 177–178, 181
encephalization 248, 273–274; Cebus 262, 263, 268
encephalization quotient (EQ) 245, 247; cells (see also ion channels) 38, 46, 47;
environmental triggers 248–249, 291; endosymbiosis 352n6; secondary
folding 195, 256; frontal 256; great endosymbiosis 352n6; sodium transport
apes 248; gymnotids 195–196; 230
hippocampus 265; hominids 273; cellulose 27, 135, 199, 298, 311, 373n5
Homo erectus 249; humans 246, 254, cephalopods 150, 151–154, 169, 214, 303
255; insects 179–180, 407n106; Cephalorhynchus 251
inter-hemispheric independence 254; cerebellum 194, 195, 254, 256
laterality 254; mammals 226; Chalcides 221
metabolic costs 195, 419n97; Chamaesyce 294
microchiropteran bats 256; mormyrid chameleons 164
fish 194–196, 245, 256, 195; Châtelperronian culture 278–281, 279,
multi-layered cortex 226; neocortex 434n314
195, 254, 256, 265, 267; occipital 256; chenopodiacean 294
olfactory bulbs 265; optic tectum chimpanzees 5–6, 242, 247, 249, 262
n.284/7; orthodenticle (otd) 241; otx chironomids 116
gene 242; paralimbic 254; parietal chitin 27
448 general index
Chlamydomonas 165, 171 fences 214; ‘fire zone’ 210; plugs 214;
chlorophyll 106–111, 135, 233, 298, 110 rafts 401n32; sheds 406n98; tents
chloroplasts 352n6 116–117, 212, 358n70; termite nests
cichlids 133, 215–216, 224, 364n136, 208; tracks 199; tunnels 208
436n11 convergences see Convergences index
ciliates 150, 235, 236, 287, 288, 356n44, Cope’s rule 304–306
374n17 corals 112, 154
circadian rhythms 172–173, 385n138 cows 113, 298
circulatory systems 215–216 crab spider 212
Circumstellar Habitable Zone (CHZ) 83, crabs 138, 150–151, 214
99–100, 100 crayfish 169
citric acid cycle 57 creation ‘scientists’ xv, 301, 316, 322
cladistics 299, 353n21, 367n169 crested auklet 139
clover 235 crickets 192, 218, 285
Clovis culture 281 crocodiles 227
coccoideans 149 crows 246
cockroaches 208 crustaceans 126, 130, 143–144, 158, 180,
comets 36, 42, 43, 94–99 214, 217, 129
commelinoids 365n142 cryptochromes 11, 172–173
common dolphin 247; Pacific white-sided crystallins 163, 166–167, 172, 173, 240, 312
dolphin 247; Tucuxi dolphin 247 cubozoans 154–155, 157, 160
communication (see also mental functions, cultures 258–260; apprenticeships 286;
sounds); ant 203–204; artificial language Aurignacian 275, 278, 279; beads 274;
252–253; babbling 251–252; bacterial bicycles 111, 112, 270, 289, n.227/9;
236; categorical perception 194; birds’ nest soup 181; bowers 436n11;
categories 259; chemical signals 236; burial 276, 278, 277; cargo cults 279;
chromatic signals 215; courtship 190; carvings 275, 280; Châtelperronian
dialects 258, 423n163; distress cries 278–281, 279; Clovis 281; ‘cultural
210; electrical 184–188, 190; gestures revolution’ 260; destruction 359n83;
252; grammar 253; head drumming figurines 275; gin and tonic 193,
209; language 278; ‘language space’ 253; 334n12; instruction 259, 260; John Frum
learned signals 252; microbial 236; cult 279; language 278; Mousterian
mimics 225, 250–251; pheromones 179, 275, 278; musical instruments 274–275;
203–204, 209; re-amplification 210; necklaces 279, 280, 279; ochre 275,
rule-based systems 253; seismic 141, 276; paintings 275, 281; Roman 281;
144, 190, 209–210; semantics 252; Solutrean 281; symbolic compositions
‘signature whistles’ 251; speech 194, 275; Szeletian 280; terracotta head 281;
251–252, 278; steroid signalling 295; transmission 258, 259–260, 261, 286;
subterranean mammals 141, 190–191; Uluzzian 280
symbolic 252–253, 258; syntax cuttlefish 214–215
252–253; termite 209–210; tutors 260; cyanobacteria 107, 113, 235, 287, 288,
vocal capabilities 278; vocalizations 352n6
223, 225–226, 249, 250–252, 258, n. 140/9; cytosine 13, 28, 29
whistles 250–251; ‘whistle matching’
251; Zipf’s law 334n5, 407n107, Darwinian selection 385n149, 441n91
421n127, 422n140 Deimos 349n72
conifers 137 developmental biology 238–242;
constructions (see also cultures, tools); appendages 179, 239; atonal gene 193;
ant-gardens 210; attine ant nests cervical vertebrae 239–240; distal-less
205–206, 206; birds’ nests 358n72; gene 179, 241–242; dorso-ventrality
bivouacs 201, 202; bowers 285–286; 239; eyes 239; heart 239; Hox gene
bridges 201, 401n32; ‘cartons’ 210, 214; 239; Math 1 gene 193; otd gene 242; otx
general index 449
gene 242; Pax-6 gene 193, 240–241; Erythropsidinium 165

star-nosed mole 179 Escherichia coli 121, 124, 126
Dictyostelium 236 Escovopsis 200
dicyemids 303 Eubacteria 170
Dinopis 156 eugenics 173, 325–326
dinosaurs 1, 96, 115, 137, 228, 269 eukaryote 4, 94
dipterans 148, 191, 192 euphausiids 158
distal-less gene 179, 241–242 Euphorbia 134, 134
DNA (deoxyribonucleic acid) 4, 13, 18, Europa 86
23–24, 25, 27–31, 50, 52, 53, 54, 173, 233, eurypterids 284
234, 237, 296, 298, 300; alternatives eusociality 141–144, 177, 180, 185, 203,
30–31, 338n41; ancient DNA 276; 224, 284–285, 403n43
Borromean rings 28; C-value paradox evolution; adaptation 96, 121, 122, 124,
417n53; ‘junk’ DNA 4, 237, 324; Zipf’s 125, 213, 224, 236, 249, 253, 256, 267, 272,
law 334n5 300, 301–302, 313, 388n192, 392n235;
dogs 34, 173, 265 ‘arms race’ 201; attractors 127, 233,
dolphins (see also common dolphin, whales) 309, 443n134; ‘bricolage’ 312; cladistics
234, 248–258, 260, 331, 257; bottlenose 299, 353n21, 367n169; co-option 8, 166,
dolphin 247, 249 193, 238, 241, 290, 312, 397n302;
Dorylus 201 combinatorial space 117–119, 127, 118;
Drake equation 231 competition 96, 125, 203; constraints
Drosophila 4, 124–125, 169, 238 239–240, 300, 302, 353n21, 423n155,
duck-billed platypus 219, 388n195, 426n213, 440n86; Cope’s rule 304–306;
392n235, 395n268 ‘dead ends’ 360n90; homoplasies 297,
299; hyperspace 10–11, 12, 16, 19–21,
earthworm 120, 154 107, 119, 126, 127–128, 232, 266, 287, 296,
echidnas 388n195, 392n235 308–310, 327, 334n17; immense numbers
echinoderms 155, 159–160 16, 18, 334n11; inherency 5, 6, 166, 196,
Eciton 204, 202 234–235, 236, 237, 238, 242–243, 261;
eclipses 93 ‘intelligent design’ 111; macroevolution
Eimer’s organ 175, 178 12, 358n64; navigation 11, 13, 19–21, 66,
electrogeneration 182–190; 127–128, 232, 233, 253–254, 287, 289–290,
communication 185; defence 185; 296, 297, 301, 308, 309–310, 327, 334n17,
electric organ discharge (EOD) 185; 443n133, 443n135; optimization 17, 31,
gymnotids 184–190; jamming avoidance 121, 253, 436n16; phylogenies 286, 297,
response (JAR) 186; mormyrids 299–300; progress 301, 307, 397n70;
184–190 radiations 96, 243, 358n74, 371n2;
electroreception 184–190, 265; gymnotids regression 6, 302–303, 382n113;
184–190, n.259/7; interference 186; saturation 301, 306–307; selection 1,
monotremes 183, 395n268; mormyrids 46, 204, 253, 266, 272, 290, 296, 307, 312,
184–90, 394n259; sharks 187 385n149, 427n221; self-organization 46,
electrotherapy 183, 393n244 48, 57, 66, 201, 204, 341n9; stabilizing
elephant seals 258 selection 115, 240; trajectories 24, 121,
elephants 246, 250, 259 144, 145, 156, 222, 229, 234, 246, 253, 294,
Eligmodontia 146 301, 302, 304, 308, 361n103, 403n43;
Elvis taxa 126 trends 120–121, 213, 222, 268, 300, 301,
encephalization quotient (EQ) 245, 247 304–306, 359n87
endosymbiosis 352n6 extraterrestrial (see also meteorites, Solar
enzymes 9, 26, 44, 108, 113, 126, 288, 295, System); 55 Cancri 78, 83, 79, 80;
334n12 aliens 158, 160, 162, 167, 172, 193, 229,
Epoicotherium 141 265, 288, 331–332; amino acids 8, 16, 33,
EQ see encephalization quotient 35, 36–37, 39–41, 43, 60, 339n11;
450 general index
extraterrestrial (cont.) 383n119; cubozoans 154; dichromacy

β-Pictoris 77, 98; biochemistry 76–77; 168; dinoflagellates 165–166;
biospheres 16, 90, 110–111, 285; echinoderms 159–160; extra-ocular
carbaquist 24–25; civilizations 22, 96, muscles 153–154; eye spots 154, 165,
100, 104, 226; communication 107, 252; 171; ‘five-site rule’ 169; fovea 177;
Copernican principle of mediocrity 103, heteropods 154; lens 153, 154; mirrors
105; Drake equation 231; extra-solar 164, 167, 355n33; mormyrids 187, 188,
comets 97–98; extra-solar hot ‘Jupiters’ 188; mysids 160, 162, 161; oil droplets
78–79, 82, 86, 101–103, 104; extra-solar 170; Pax-6 gene 193, 240–241;
moons 86, 92; extra-solar planets phospholipase C n.219/7; polarization
77–87, 100–103; Fermi paradox xiii, 104; vision 170; protistans 165; red–green
‘floaters’ 354n29; galactic niches 308; vision 168, 385n138; refractive index
galactic zone of habitability 104–105; 153; retina 151, 153, 154, 157, 170, 172,
gamma-ray bursts 103; gravity 79, 92; 187, 223; retinal 167, 171–172; ‘reverse
habitable zones 43, 104–105; eye’ 167; rhodopsin 153, 167, 377n41,
high-resolution stellar spectroscopy 380n75; schizochroal 159, 160; snails
77–78; intelligence 90, 103, 104, 156, 154; spectral sensitivity 157, 168–170;
229, 420n112; interstellar clouds 25, 27, spherical aberration 153, 156, 160;
33; interstellar planets 86, 178; life 16, spiders 156–157; stomatopods 170;
22, 25, 26, 29, 31, 34, 66, 79, 86, 95, 106, tapetum 164; telephoto system 157;
178, 229–230, 354n29; microbial transduction 180–181; trichromacy
cul-de-sac 237; music 421n132; 167–168; trilobite 159–160; ultraviolet
nervous systems 230; neutron stars vision 168, 169, 170, 386n152;
103; orbital eccentricity 82, 86; organic vertebrates 151–154; visual streak
chemistry 23, 32, 33–43; Orion 378n62; visual tracking 163; wasps
388n199; OZMA project 231; 373n6
Panspermia 26, 337n16; photosynthesis
353n14; planetary engulfment 85; fatty acids 27
planetary migration 85–86; planetary Fermi paradox xiii, 104
obliquity 90, 90; polycyclic aromatic fiddler crabs 189, 214, 286, 378n62
hydrocarbons (PAHs) 33, 43; proteins finches 125
288, 289; red giant 337n16; SETI fish 113, 133, 216, 234, 239, 133; alopiids
231–232, 234, 351n4; silicon-based life 224; Arctic cod 290; cichlids 133,
24–25; snowline 42, 86, 96–97; songs 215–216, 224, 364n136, 436n11; conger
226, 252; stars (HD 12661, 37124, 38529, eels 112; deep-sea 154, 164, 169;
49674, 76700, 80606, 162020, 168443, electric eel 184; gymnotids 184–190,
202206) 78, 82, 80; stellar luminosity 195–6; lamnids 224; lamprey 150;
99; stellar metallicity 101–103, 104; minnows 189; mormyrids 184–190,
sugars 36; supernovae 103; supply of 216, 245, 184; notothenioids 290; pike
organics 45, 62; Threga IX 23, 136, 138, 133, 133; rays 119, 222, 223; sandlance
172, 228, 259, 266, 289, 336n6; Upsilon 164; scombrids 223; sea-bream 133;
Andromedae 78, 80, 82; vocalizations sharks 119, 222, 223, 224, 363n126;
226; volatiles 42–43, 45, 87, 96–97, 98–99 spade-nose shark 221; swim-bladder
eyes 151–170; accessory retina 154; acute 112–113; teleosts 194, 223; torpedo ray
zone 378n62; binocular 162; 183; tuna 223–224
brittle-stars 159–160; calcitic 159; flatworms 240, 303
camera 148, 151–158, 162, 172, 193, 214, flies 148–149, 169
152; cephalopods 151–154; chromophore flowers 117, 135–138, 278
167, 172; cnidarians 154; colour vision formaldehyde 32, 45, 50, 64
157, 167–170, 172; compound 147–148, Fortean bladders 112–115, 116, 117, 164,
151, 158–163, 193, 204, 149, 159; copepod 216
374n21; crystallins 163, 166–167, Fortean Times 356n49
general index 451
frogs 120, 139, 270, 391n.196 hedgehogs 257

fruit-flies 237–238, 299 hemipterans 149
fungi 171; ascomycetes 207; Heterocephalus 142, 142
basidomycetes 208; chytridiomycetes heteropod 154, 157, 193
171, 356n44; fungicides 200; mushrooms heteropterans 139
208–209; ophiostomatoid 207; zoospores Hibbertopterus 284
171 hippopotamus 291
fungus-gnat 116 hoatzin 298
hominids (see also cultures, sounds) 203,
galatheid crabs 158 222, 248, 249, 262, 263, 269, 270–281, 247;
gall thrips 143 Arcy-sur-Cure 279, 280; Aurignacian
geckos 383n119 275, 278, 279, 280; australopithecines 268,
genetic animism 323; code 13, 15–19, 52, 270, 271, 273; bipedality 269; Blombos
59, 107, 233, 336n40, 17; duplication cave 275; Boxgrove Man 291, 292; brain
6, 168, 239, 324, 437nn32,34; sizes 273, 276; burial 276, 278, 277;
fundamentalism 237, 238, 322–325; cannibalism 276; care 274, 276;
lateral transfer 237, 287, 443nn2,3; Châtelperronian 278–281, 434n314, 279;
manipulation 311–312, 315; networks Clovis 281; encephalization 273–274;
239; transfer 237, 287; triplet code evolution 269, 270–271; Grotte du Renne
(codons) 13, 16, 336n40 280; ‘hominization’ 274; hunting
genome 122, 237–238 patterns 276; Kebara 277; language
giraffe 239 278; Les Chauvet 275; locomotion 267;
global warming 292–293, 295 Mousterian 275, 278, 281; murder 276;
glucose 121–122, 290 Neanderthals 276–281; Oldowan 262,
glycine 13, 16, 33, 60, 115 263, 273; robust australopithecines 272;
Gnetales 135–138 Shanidar Caves 276; Solutrean 281;
golden lion tamarin 262 Szeletian 280; tools 273–274; Uluzzian
gorilla 247 280; Upper Palaeolithic Revolution
graphite 59, 72, 108, 341n30 274–275
grasses 293–294 Homo 5, 203, 269, 270, 271, 273
grasshoppers 286 Homo erectus 247, 249, 274, 291
great apes 2, 180, 181, 182, 187, 234, 246, Homo neanderthalensis 271, 276
247, 262 Homo sapiens 271, 273, 276, 279, 280, 281,
grebes 138 284
guanine 13, 28, 29, 112–113, 114, 164 horses 293
gymnosperms 136 houseflies 285
gymnotids 184–190, 195–6 hoverflies 163
Hox gene 239
haemocyanin 235, 288, 289, 408n117, humans (see also Homo) 1, 15, 16, 26, 95,
436n22 150, 160, 162, 194, 201, 206, 207, 209, 224,
haemoglobin 5, 8, 235, 287–288, 331 225–226, 237, 239, 246, 250, 251–252, 253,
halteres 148–150, 159, 149 254, 259, 270, 294–295, 308, 314, 325;
Harbour seal 251 accidents 387n175; blind 340n222;
hearing (see also brain, sound) 2, 148, 186, fingerprints 174; hunting 409n134,
189, 190–193, 398n305; bulla 371n207; 424n168; language 422n136; Maoris
chordotonal organ 191, 193; cochlea 409n134; Polynesians 19–21, 218,
190; ear 190–193; Johnston’s organ 148, 363n123
191, 192, 193; mormyrids 189–190; otic humming birds 138, 368n171
capsule 190; owls 2, 395nn265,266; Hydra 8
pinna 191; semicircular canals 150, hydrocarbons 27, 37, 38, 43, 46, 61, 62–63
191, 427n227; tympanic membrane 190, hydrogen cyanide 32, 45, 61, 62, 64
191–192, 193 hydrogen sulphide 55
452 general index
hydrogenase 113, 114, 356nn45,46 life, origin see origin of life

hydrogenosomes 113–114, 356n44 lipids 24, 27, 38, 170
hydrolysis 53, 54, 439n59 Littorina 154, 157, 158
hydrothermal systems 54–57, 62, 63, 66, lizards 2, 125, 144–145, 164, 220, 227,
74, 113, 114 362n116
hyenas 274 lobster 119, 130
hymenopterans 143 locomotion (see also wheels); arboreal 267;
hyperthermophiles 56–57, 74, 350n107 bipedal 268–270; knuckle-walking 269;
hypervitaminosis 274 quadrupedal gait 267
locusts 357n51
Ibla 119 loons 138
ice 25, 42, 290–291, 327, 352n7 luminescence 116, 164, 167, 383n117,
ice ages 96, 99, 222, 248, 275, 291, 292, 294 385n150
impacts 43, 56, 62, 70–74, 87–89, 90,
94–95, 71 Mabuya 221
Indian Red ant 212 macroevolution 12, 358n64
insectivores 265 macrotermitinids 208, 209
insects 114, 115, 116, 128, 130, 179–180, magnetic fields 41, 86, 170
191–193, 194, 223, 284 maize 294, 325
intelligence (see also mental states and malignant hyperthermia 224
functions) 66, 89, 90, 141, 156, 229, 230, mammals 96, 223, 240, 305–306;
246; abstractions 253; chimps 246; bathyergids 143, 143; bonobo 253;
collective 201, 203–204; comprehension chrysochlorids 141; flying pigs xii, 114,
252; crows 246, 262; dolphins 246–253, 356n49; necrolostid 139; palaeanodont
256–258; gestures 252; great apes 246; 141; thylacosmilid 130, 227, 131
human xi; monkeys 215; octopus 215; manatee 216–217
parrots 246; primates 246, 262; Mantispa 128, 129
problem-solving 203 Maoris 409n134
ion channels 193; bacterial 230; calcium Mare Imbrium 70
230–231; evolution 230–231; heliozoan Mare Tranquillitatis 70
230; jellyfish 230; potassium 230; maria 69, 70
sodium 138–139, 230–231 marmoset 168
Mars 27, 34, 42, 43, 58, 73, 75–76, 83, 90,
jerboas 146 92, 96, 99
Johnston’s organ 148, 191, 192, 193 marsupial mole 139
‘junk’ DNA 4, 237, 324 marsupials 130, 132, 139, 145–146, 267,
Jupiter 42, 78, 79, 83, 88, 93–97, 354n29 270, 372n3, 412n168, 426n216
mass extinctions 94–95; ecosystem
Kaali event 35 recovery 94–95; Elvis taxa 126;
kakapo 218 end-Cretaceous (K/T) 94–95, 96, 206,
kangaroo rats 145–146 222, 339n25; end-Permian 94–95, 126;
kangaroos 15, 130 Lazarus taxa 126; postponed, not
kea 14 cancelled 96, 222
Kenyanthropus 271 Math 1 gene 193
kiwi 218–220, 222, 227, 219 meliponinid bees 203
Kuiper Belt 97–98 memory 187, 215, 226, 250, 253, 254, 256,
402n37, 407n108
lacewings 285 mental functions (see also cultures,
Lake Tagish 35 intelligence); abstractions 258, 420n116;
leaf-cutting ants 197–200, 209 animal navigation 178, 215, 252, 3;
learning 187, 215, 226, 246, 259, 402n37, baboon 242–243; beauty, sense of 286;
407n108, 421n125, 424n173 bees 402n37; categorical perception
Les Chauvet culture 275 194; cognitive abilities 252–253;
general index 453
collective decisions 402n36; cultural Mononykus 141, 144,

transmission 258, 286; curiosity 268; monotreme 183, 219
electrical world 187; equivalences 189, Moon 69–72, 87–92, 275
391n223; feeding behaviours 259; mormyrids 184–190, 216, 245, 184
flexibility 215; hunting 422n138; mosquitoes 147–148, 191, 352n20
hydrodynamic image 216; innovations moths 138, 168, 181, 191, 394n256
246; internal representations 254; mouse 289
learning 187, 215, 226, 246, 259, 402n37, Mousterian culture 275, 278, 281
407n108, 424n173; maps 178, 187, 216, Murchison meteorite 35, 36
266, 388n201, 394n258; memes 324; murinine bat 267
memory 187, 215, 226, 250, 253, 254, mynahs 225
256, 402n37, 407n108; memory Myrmarachne 212
consolidation 202; mind of a bat myrmicine ants 200
265–266; neophobias 246; personality mysids 160, 162, 161
215; plasticity 187, 189, 202, 252, Mystacina 218
407n106, 425n187; qualia 266, 396n281; Myxobacteria 236, 333n1(ch.1)
self-recognition 258; sentience 329; myxozoans 303, n.113/7
sleep 202, 254, 256; social knowledge
259; synaesthesia 396n281; temperament naked mole rat 142–143, 177–178, 285, 142
215; ultimate limitations 266; Nakhla 34, 75
uncertainty response 254; naphthalene 33, 340n38
unihemispheric sleep 254; vocal Neanderthals 276–281
production 226 nematode 4, 238, 240
Mercury 82, 83, 92 Nematodinium 165
Mesoplodon densirostris 15 neocortex 195, 254, 256, 265, 267
meteorites (see also extraterrestrial) 34–38, Neptune 88, 97
75, 95; Allan Hills (ALH 84001) 33–34, nerve net 155, 160
75; carbon 345n11; carbonaceous nervous systems (see also brain, ion
34–38, 42–43, 50, 60; contamination 35, channels, intelligence); acetylcholine
38, 339n11, 339n21, 340n32; fatalities 235–236; action potential 230; glomeruli
34–35; Kaali event 35; Lake Tagish 35; 179–180, 331; nerve net 155, 160;
Murchison 35, 36; Nakhla 34, 75 neuropeptides 236; neurotrophins 214;
methane 47, 62, 292, 295, 352n7, 438n47 synapse 235
methanogens 352n10 neuropterans 128
microchiropteran bat 256 New Caledonian crows 261, 263
midges 116 nucleotides (and related compounds; see also
Miller-Urey experiments 58, 60–63 DNA, RNA) 4, 23, 24, 27, 53, 56, 59, 63;
millipedes 372n4 adenine 4, 13, 28, 30, 63, 64; cytosine
mites 211–212, 368n171 13, 28, 29; guanine 13, 28, 29, 112–113,
mitochondria 4, 300, 352n6, 353n19, 114, 164; isocytidine (iso-C) 29;
356n44 isoguanosine (iso-G) 29;
moa 218 nucleotide-binding 295; oligomerization
mole 139, 178, 239, 140 53–54; oligonucleotides 27–28, 65;
mole rat 141–143, 178, 142, 143; Zambian purines 37, 59, 77; pyrimidines 37, 59,
common mole rat 141 77; synthesis 341n27; thermal stability
molluscs 151, 288, 305; eyes 151–154, 342n37; thymine 13, 28, 29; uracil 13,
158–159, 169, 159 29
monkey, ateline 249, 267; callitrichid
262; capuchin 262, 263; colobine 298; octopus 113, 151, 169, 170, 215
howler 168; langur 298; New World Ocythoe 113
168, 234, 262, 268, 426n219; Old World Oecophylla 116, 212
167–168, 262, 268, 298; Spider monkey ogre-faced spiders 115, 156–157
249, 267 oilbirds 181–182
454 general index
Oldowan culture 262, 263, 273 plankton 95, 114

olfaction 173–181, 265; glomeruli Planodiscus 211–212
179–180, 331, 389n208; kiwi 219–220; plants (see also chlorophyll, photosynthesis)
nautilus 375n23; nervous system 193; 107, 134–138, 235, 366n151, 404n67;
Pax-6 gene 193; phospholipase C bundle-sheath cells 293; cryptochromes
390n219; proteins 180–181; transduction 173; double fertilization 136–137;
180–181 endosperm 137; Kranz anatomy 293;
Oort Clouds 94, 97–98, 104 leguminous 235; mesophyll 293; resins
orang-utan 247 208; stomata 294
Oreopithecus 269–270 Plectrocnemia 116
origin of life (see also RNA, sugars) 44–68; Pollimyrus 190
catalysts 52, 53, 57, 59, 63; clays 53, 54; polychaetes; alciopid 154, 157, 193;
cyanoactylene 61; failures 49–52; sabellid 158
Fischer-Tropsch reaction 62; formose Polynesians 19–21, 218, 363n123
reaction 50–51, 51; hydrothermal porpoise 248, 251
systems 54–57, 62, 63, 66; hypercycles Portuguese man-of-war 113
46; Miller-Urey experiments 58, 60–63; praying mantis 128, 129
monolayers 59; oligomerization 30, primate 267
53–54, 58, 59, 65; ribose problem 50–52, prosimian 168
51; time required 73–74 proteins 4, 9, 10, 11, 24, 111, 115, 126, 165,
Ormia 192 166, 168, 171, 179, 193, 194, 234, 287–291,
ormiinids 192 295, 303; α-helices 167, 170, 230,
Orrorin 271 357n55; α/β barrels 287; antifreezes
orthodenticle (otd) 241, 242 290–291; arrestins 180; β-endorphin
orthopterans 192 236; β-pleated sheet 357n55;
orypodid crabs 214 bacteriorhodopsin 170–172; cellulases
ostracods 158 443n2; chlamyrhodopsin 171; collagen
otd gene see orthodenticle 8, 13; cryptochromes 11, 172–173;
otx gene 242 crystallins 163, 166–167, 172, 173, 240,
owls 2, 186, 395nn265,266 312; Darwinian selection 385n149,
oyster 289 441n91; dopamine 236; elastic 215,
295, 408n113; F1 -ATPase 353n19;
pancreas 193, 240, 290 fibroins 115; haemocyanin 235, 288,
pangolin 112 289, 408n117, 436n22; haemoglobin 5, 8,
Paranthropus 271, 272, 273 235, 287–288, 331; heat-shock 166, 312;
parrots 14–15, 218, 225, 246, 253 hyperspace 10–11, 16, 66, 287, 295, 296,
pathogens 11, 200, 207, 210, 295–296 308, 20; indoleamine 2,3-dioxygenase
Pax-6 gene 193, 240–241 (IDO) 288–289; isoforms 303, 324;
penaeids 217 lysozyme 297–298; membrane-spanning
Peniocereus 134, 134 167, 170, 230, 390n217; myoglobin 235,
Phobos 349n72 288–289, 414n34; opsins 31; PEP
phorid flies 199, 208 carboxylase 293–294; photolyase 173;
photosynthesis (see also chlorophyll, plants) proteorhodopsin 171; rhodopsin 11,
73, 107–111, 135, 171, 293–294; C3 109, 163, 165, 167, 170–172, 173, 180, 241;
293–294; C4 109, 293–294; C4 pathways RuBisCo 108–109, 293; sensitivity to
294; light-harvesting proteins 295; change 172, 287, 303, 313, 442n109;
oxaloacetate 293; phosphoglycerate trypsin 290, 334n12
293; red shift 109 protistans 236; ciliate 150, 235, 236, 287,
pigs 224, 325 288, 356n44, 374n17; diatom 294;
pines 138 dinoflagellate 165; foraminiferans
Pitohui 138 358n75; heliozoan 230, 231; slime mould
pituitary 193, 236, 240 236
general index 455
proton pump 170, 172 plasticity 189; transduction 193, 194;

pyrite 54–55, 65, 108 transplants 189
SETI 231–232, 234, 351n4
quinones 340n38, 351n5; red shift sharks 119, 222, 223, 224, 363n126
109 Shoemaker-Levy, comet 93
shrimps 112, 119, 160, 162
reproduction; aggressions in 402n40; silk 13, 14, 115–117, 156, 213, 289; ant
conjugation 236; dogfish egg case 116–117; insect 116; spider 115–116
357n59; gestation 221; hormones silk moth 116
416n44; maternal care 412n168; siphonophores 113
matrotrophy 221–223; menopause 259; skeletons (see also teeth) 117–120, 276;
ovoviviparity 219–223; ovum 220, 221; baculum 280, 359n78; bone 15, 119,
penis 119; placenta 220–221; 189; cartilage 119, 295; foot 269; hand
promiscuous mating 249; sex 270; hip 269; hydrostatic 119–120;
determination 219–220, 295; sexual hyoid 278; hypoglossal canal 278;
dimorphism 226, 395n263; sexual pre-pollex (radial sesamoid), 333n9;
display 432n294; sexual signalling skull 248, 274; vertebrae 239–240
394nn256,261; solitary males 250; sloth 240
temperature 222 snail 120, 126, 154, 157, 288–289, 406n97
rhachiberothidids 128 snakes 144, 178, 220–221
rhodopsin 11, 163, 165, 167, 170–172, 173, social systems 2, 11, 116, 142–144, 186,
180, 241, 402n41, 380n75 187, 201, 202–203, 205, 207, 210–211, 236,
ribose 30, 50–52, 58, 63, 64 249–250, 259–260; ‘babysitting’ 250;
ribozymes 44, 297 eusocial 141–144, 177, 180, 185,
RNA (ribonucleic acid) 4, 13, 25, 27, 29, 30, 202–203, 284–285, 323; fission-fusion
31, 44, 50, 52, 65, 297, 336n35; folding 249–250, 285; ‘hyperspaces’ 308–309;
343n48; hammerhead ribozyme 297; killer whales 260; lekking 214; male
ribozymes 44, 297; sequence space apprenticeship 286; matrilineal 250,
334n17; world 29, 44, 56, 296–297, 259; monkeys 268; ‘super-alliances’
343n50 250
Roche limit 89 social vole 143
rodents 125, 145–146, 270, 335n30 Solutrean culture 281
RuBisCo 108–109, 293 sounds (see also communication) 189–190;
babbling 251–252; drumming 225;
sabre-toothed cat 130, 132, 227, 131, echolocation 181–182, 256, 265,
363n122 390n220, 391n233, 426n213; flutes 275;
Sahelanthropus 269, 271 improvisation 260; music 225,
sarcophagids 192 274–275; song 192, 225–226, 252, 260,
Saturn 83, 88, 95 286, 421n131, 436n8–9; stridulations
scallop 164, 167 209; ultrasound 181
Scapanus 178 spiders 14, 112, 115–116, 120, 156–157,
sea anemones 112, 154 160, 212–213, 381n99
sea urchin 155, 337n30 spinosaurs 227–228
seal 169 sponges 8, 230, 260
Search for Extraterrestrial Intelligence see spurge 134–135, 134
SETI squid 151, 157, 167, 169, 170
seaweeds 107, 111 staphylinid beetles 210–211, 213, 211
sensory systems (see also electrogeneration, star-nosed mole 174–179, 175, 176
electroreception, eyes, hearing, olfaction, starfish 155
sounds); electrical ‘colour’ 188; Stephens Island wren 218
equivalents 174, 177, 180–181, 182, 186, stomatopods 112, 170
188–189, 193, 265–266; exchange 178; storks 99
456 general index
strepsipterans 149–150, 159, 149 tulips 2, 316

stromatolites 94 Tycho 70
Strombus 154, 157
sugars and related compounds 8, 23, 27, 30, ultraviolet radiation 40, 61, 100, 173,
36, 46, 50, 52; α-threofuranosyl 31, 52; 334n16, 338n7, 351n6; reflectance 115;
aldopentoses 50; glucose 121–122, 290; vision 168, 169, 170
hexose 31; maltose 121–122; ribose Uluzzian culture 280
30, 50–52, 58, 63, 64; threose 31, 52 ungulate 276
Sulculus 289 Upper Palaeolithic Revolution 274–275
swallows 115, 331 uracil 13, 29
symbiosis (see also cells, endosymbiosis) Uranus 348n68
12, 107, 114, 207–208, 335n18
symbiotic bacteria 352n6 vangids 147
Synalpheus 144 Venus 43, 72, 83, 89, 90, 92, 97, 99, 331
Szeletian culture 280 vestimentiferans 114
vitalism 5, 48
tachinids 192 vitamin A 167, 171, 274; B2 172
teeth (see also skeletons) 133, 227, viviparity 220–223
271–272, 273; canines 130, 227, 279; Volvox 165, 171
enamel 413n184; fish 363n126;
hypocone 413n183; incisors 177, 227, walrus 419n90
335n30; megadonty 272; molars 227, warmbloodedness (see also
272; sea urchin 444n30; tribosphenic thermoregulation) 96, 223–225
227; weak dentition 179, 187 wasps 143, 205, 208, 263, 264, 284, 373n6,
teiids 227 402n40
teleology 2, 5, 60, 111, 128, 297, 313–316 weaver ants 116–117, 212
tenrec 246 Weddell seal 258
termites 143, 208–210, 284 wetas 218
Termitomyces 208–209 whales (see also communication, dolphins,
Thermopalia 113 social systems, sounds); ancestors 248;
thermoregulation 201, 269 beluga whale 251; Blainville’s beaked
TNA (threo-nucleic acid) 52–53 whale 15; bottlenose whale 420n106;
tools (see also constructions, cultures) 246, humpback whales 252, 260, 261; killer
261–264, 292, 427n226; Aurignacian 275, whales 247, 258, 259–260; monodontids
278; birds 261; bonobo 262; bubble 419n90; odontocetes 246; sperm whales
nets 261; callitrichids 262; capuchins 250; toothed whales 169, 246, 249, 258,
262–263; chimpanzee 261, 262; crows 260
262; dolphin 260; fire 274, 275, 276; wheels 111–112; microtubules 353n15;
gloves 260; golden lion tamarins 262; molecular motors 111; mosquitoes
hooks 261; laterality 261, 263; 353n20; stomatopods 112
lateralization 261; lithic 262, 273; winkle 154, 158
Mousterian 275, 278; New Caledonian woodpecker 147, 371n1
crow 261–263; Oldowan Industrial woolly mammoth 291
Complex 263, 273; spears 274, 275,
276; stasis 274, 275; tool-kits 274, 278; Xenocranium 141
wasps 263–264
Townsend’s mole 178 Zipf’s law 334n5, 407n107, 421n127,
trichopterans 116 422n140
trilobites 126, 155, 159, 362n107 ziphodonts 227
Convergences index
Page numbers in italics indicate Figures.
accessory retina 154 balance 150–151, 153

acute visual zone 378n62 barnacles 358n77
African golden mole 141 Batesian mimicry 213
Afrotheria 132 bathyerids 143, 143
afterbirth 221 bats 181, 265–267, 390n221
aggression 141, 220 bat carnivory 266
agriculture 2, 198–200, 205–207, 208–209, bees 143, 402n40
275, 401n25 bee eusociality 203
alciopids 154 beetles 143, 149, 211, 360n90, 362n116,
aldolase 126 372n4
alkaloids 138–139 behaviours 139, 141, 213, 214, 236,
allosaurids 130 285–286
α/β barrels 287 binocular vision 162–163, 376n36
ambrosia beetles 207–208 bipedalism 144, 145, 269–270
aminoacyl-tRNA synthetases 295 bird songs 226
amphibians 120, 220 bird vocalizations 226, 251–252
annelids 154, 158–159 birds 125, 138–139, 162, 181–182, 190,
anolids 125 225–226, 254, 368n170, 412n180
anteaters 132, 353n21 bivalve molluscs 158, 159
antifreeze proteins 290–291, 438nn42,43 bladders 112–113, 355nn35,36,37
antigen receptors 295 blood-brain barrier 214
ants 135, 143 see also army ants blood rete 224
aorta 215, 408n114 blue vision 169
apatemyids 372n3 blue-sensitive vision, loss 169
appendages 241 Bombardier beetle 372n4
aquatic birds 138 bone 119
arboreal gait 267 bottlenose dolphins 249
argyrolagids 145–146 brain structure 132, 153, 158, 266–268,
armadillos 354n26 364n128, 371n203
army ants 200–201, 204, 406n98 brains 196, 267, 406n102, 412n180,
arthropodization 284 426n212
arthropods 159, 284 brain-size increase 248, 273–274, 427n221
artiodactyls 132 brittle-stars 155, 159–160
atelines 249, 267, 427n220 bryophytes 366n151
attachment pads 128, 130, 362n116 bryozoans 126, 362n108
auditory mechanisms 191 ‘bucket brigades’ 400n6
aye-aye 372n3 burrowing 120, 132, 139–145, 189–190, 266
burrowing cycle 359n79
babbling 251–252, 421n128 butterflies 169, 368n174
bacteria 230
bacterial symbionts 114 C4 photosynthesis 109, 293–294
bacteriorhodopsin 171–172 cactus 134–135, 134
458 convergences index
caddisfly 406n97 DM domain factors 295

camera eyes 148, 151–158, 216, 232 dolphins 181, 248–258
canines 130, 132, 131 double-fertilization 137
cartilage 216 duck-billed platypus 388n195, 392n235
cephalopods 150, 154, 158, 214, 215–216
chameleons 164 E. coli 124
Châtelperronian 280–281 ears 190–191
chemical 138–139, 360n90, 368n174 echidna 388n195, 392n235
chemosensory 139 echolocation 181–182, 256, 390n220,
chimps 249 391n233, 426n213
chitin-binding proteins 295 ecomorphological 125, 133, 302, 361n104,
chrysochlorids 141 390n221
cichlids 133, 364n136, 436n11 eel 133
circulatory systems 215–216, 224 elastic proteins 295, 408n113
cnidocyst 382n113 elastic release 128
coccoideans 149–150 electric organs 11, 183–190
cochlea 190 electrical signals 186
cognitive processes 254, 256, 258 electrocytes 184–185
collembolans 360n90 electrogeneration 182–190
colobine monkeys 294 electroreception 184–190, 392n235,
commelinoids 365n142 394n259
communication 141, 252–253, 402n41 elephants 250
communities 125, 361n102 Elvis taxa 126
compound eyes 148, 155, 158–163, 284, encephalization 273, 274
149, 159 endosymbiosis 352n6, 382n109
concerted 299–300 endothermy 223–225
constructions 406n98 enzyme 126, 343n50, 439n59
counter-current systems 355n34 Ernanodon 353n21
courtship behaviour 285 euphasiids 158
crab spider 212–213 Eurotamandua 353n21
crabs 130, 150 eusociality 141–144, 177, 180, 185, 203,
crested auklets 139 224, 284–285, 403n43
crickets 285 expressions of surprise 128, 134, 297, 304,
crustaceans 126, 130, 143 359n78, 362n111, 368n172, 376n32,
cryptochromes 173 382n108, 382n113, 393n245, 394n259,
crystallins 167, 383n119 395n267, 398n305, 398n309, 401n27,
cubozoans 154–155 411n145, 419n90
cultural transmission 258, 412n175, extra-embryonic membranes 410n144
423nn163,164 extra-ocular muscles 153–154, 375n30
culture 280–281, 423n164, 424n165, eye spots 165, 381n102
435n318, 436n11 eyes 151–170, 175, 177, 193
cyanogenic glycosides 139
cytokinases 295 farming 206
fibre-optics 158, 379n66
defence 372n4 fiddler crabs 214, 286, 378n62
descended larynx 432n294 finches 125
development 241–242 fish 112–113, 133, 154, 220, 223, 133,
dialects 423n163 363n126
dichromacy 168 fission-fusion society 249, 268, 420n106
dinoflagellates 165 five-site rule 169
dipterans 148–149 flies 148–149, 163, 285
disgust 404n55 flightlessness 218, 409n133
convergences index 459
flowers 135–138 kiwi 219–220, 219

fore-limb/hind-limb separation 267
fore-limb manipulation 270 lacewings 285
fovea 177 lactate dehydrogenases 295
frogs 139, 391n.196, 381n92 lamprins 295
fruit-flies 299 langur 298
lateral-line systems 216–217
gall-thrips 143 latex 404n67
gas-filled bladders 112–113 Laurasiatheria 132
geckos 383n119 leaf-eating 266, 440n85
grebes 138 leaves 135, 442n124
gymnotids 184–190 lekking 214
gyroscopes 148–150 lens 153, 156
light-harvesting proteins 295
haemocyanin 235, 288, 436n22 lilies 299
haemoglobin 235, 287–288 limb loss 145
halteres 148–150, 149 limbs 416n38
hearing 148, 189, 190–193, 371n207, lizards 125, 144–145, 220, 227, 362n116
398n305 locomotion 267
heat-sensitive organ 389n203 loons 138
helicases 343n50 luminescence 383n117
hemipterans 149 lysozyme 297–298
heteropods 154
heteropterans 139 malignant hyperthermia 224
hexapod gait 284 ‘mammal-ness’ 222–223, 228, 234, 308
high-altitude birds 138 mammals 139–144, 158, 162, 190, 214,
HIV proteases 295 219, 221–222, 304
hoatzin 298 manipulative skills 428n238
hominid cultures 280–281 mantids 128, 163
hominids 269, 271–274 mantispids 128, 129
hooves 227 marsupial mole 139
houseflies 285 marsupials 130, 132, 139, 145–146, 267,
humming birds 138, 368n171 372n3, 412n168, 426n216
hyaluronan synthase 373n5 marsupium 363n124
hydrogenosomes 356n44 maternal care 412n168
hydrostatic skeleton 120 matrilineal social systems 250
hymenopterans 143 matrotrophy 221–222
hypocone 413n183 Mediterranean-style floras 135, 365n144
megadonty 272
Indian Red ant 212–213 memory 254, 407n108
ink clouds 406n101 menopause 259
insect-bacterial association 356n48 mice 218
insectivores 132 migratory birds 138
insectivorous birds 138 millipedes 372n4
insects 128, 130, 138, 284, 435n3 mimicry 211–214, 406n94
intelligences 156, 256–257 mites 211–212, 368n171
island faunas 428n240 molars 227
mole 139, 140
jamming avoidance response (JAR) mole rat 141–143, 142, 143
186 molecular 10, 241, 287–298, 373n5, 436n77
jellyfish 150 molluscs 151–154
jerboas 146 mormyrids 184–190
460 convergences index
morphological 12–13 pre-pollex (Panda’s thumb) xii, 334n9

moths 138, 394n256 proteases 295
Müllerian mimicry 213 proteinase inhibitors 139
multicellularity 237 proteins 126, 165, 289, 295–296
music 225–226 push-rod mechanoreceptors 388n195
myoglobin 235, 288
quorum sensing 401n29
NADH dehydrogenase 1 295
naked mole rats 142–143, 177–178, 142 raptorial 128
necrolestids 139 red vision 169
nectar-feeding bats 266–267, 391n233 red-green vision 168, 385n149
nectar-feeding birds 138 reflective structures 164
neural algorithms 186, 285, 395n265 Rensch’s rule 368n171
neural networks 155 reproduction 220–223
neural processing 163, 189, 375n28, resin 404n67
381n92, 388n191, 395n266, 396n284, rhachiberothidids 128
426n208 rhodopsin 165, 171–172
nicotine oxidases 295 RNA-protein 296
nucleotide binding proteins 295 rodents 125, 132, 145–146, 335n30
rumen 298
octopus 113, 215
oil droplets, in eye 170 sabellids 158–159
olfaction 148, 179–180, 369n180 sabre-tooth cats 130, 139, 131, 363n122
orypodid crabs 214 sandlance 164
ormiinids 192 sarcophagids 192
ostracods 158 sea-bream 133
ovoviviparity 220 sea urchin 335n30
ovum, minute 221 seed dispersal 135
owls 186 seismic communication 141, 144, 190, 209
self-organization, role of 205
palaeanodonts 141 self-recognition (mirrors) 258
pangolins 353n21 semicircular canals 150–151
parallel, versus 435n1 sharks 222, 223, 224, 295, 363n126
parental care 225 shrimps 144
pathogen–host relationship 295–296 sieve tubes 355n37
pectinate claws 130 silk 115–117, 289
penis 119 silk nets 116
peptidases 295 skeletons 119
personalities 215 sleep 202, 422n144
phorid flies 208 snails 126, 154, 406n97
phylogenies 127 snakes 144, 220
physiology 139, 141, 224 social systems 205, 233, 236, 249–250, 268,
pike 133, 133 285
placenta 220–221, 365n145, 411n151 social voles 143
placentals 130, 132 sodium channels 230–231, 400n18
plants 134–138, 366n151, 404n67 somatosensory systems 177
plant eating 139 song 225–226, 421n131, 436nn8,9
polarization vision 170 spectral sensitivities 168–169
polysaccharide lyases 295 speech 226
porpoise 251 sperm whales 250
praying mantis 128, 129 spherical aberration (eyes) 153, 156, 160
precision grip 270 sphinx moths 138
convergences index 461
spider-monkey 249, 267, 268 ultrasound detection 391n225, 394n256

spiders 156–157 uncertainty response 254
spurge 134–135, 134 ungulate 132
staphylinids 211 unihemispheric sleep 254
star-nosed moles 141, 175, 177, 265, 266
statocysts 150, 153 vangids 147
steroid signalling 295 vermivores 179
strepsipterans 148–149, 159, 149 vertebrates 150–154, 190
structural mimicry 295–296 vestibulo-oculomotor reflex 153
super-efficient transport 201 vision 148
surprise, expressions of see expressions of visual stimuli 180
surprise visual streak 378n62
visual systems 177
teeth 133, 227, 272, 335n30, 363n126, visual tracking 163
412n181, 413nn182,183,184, 429n248 viviparity 221–223
teiids 227 vocal control systems 226
temperaments 215 vocal learning 412n180, 421n125
tents 116–117 vocalizations 225–226, 251–252, 412n175,
termite eaters 179 423n163
termites 143, 208–210
terrestrialization 284, 286 walking 148
threads, adhesive 166 walrus 419n90
thylacosmilids 130, 132, 139, 131 warmbloodedness 223–225
toe-fringes 145 wasps 143, 402n40
tools 233, 261–263, 425n200 waste management 205, 208, 404n55
topoisomerases 343n50 water-vascular system 366n151
trabecular bone 119 weaver ants 212
trachea 284 wetas 218
tribosphenic 227 winkle 154, 158
trichromacy 167–168 woodpecker 147, 371n2, 372n3
trilobites 126, 155, 159, 362n107 ‘worker policing’ 402n40
tropical seabirds 138 worm 120, 234
tuna 223–224
turret eyes 164 xenarthran 353n21
tympanic, ear 192 xerophyte 135
Index of authors and other
individuals
Page numbers in italics indicate Figures. [Notes] are listed by [page number].
Alexander, R. D. 142 Dawkins, R. 302, 315, 323

Alroy, J. 305 de Duve, C. 22, 32
Annis, J. 103–104 Denton, E. 112
Arnold, N. 145 Denton, M. 11, 261
Desmond, A. 317
Bain, D. 259 DeVore, I. 229
Bakker, R. 130, 132 Diamond, J. 14, 218, 286
Bartel, D. 44 Dioscorides 183
Beadle, G. 230, 231, 232 Dose, K. 48
Beaven, C. 174 Doupe, A. 226
Benz, W. 89 Drake, F. 231
Berger, T. 276 Duboule, D. 312
Bernal, J. D. 54 Dulvy, N. 223
Bieri, R. 232–233, 284
Blackburn, D. 220–222 Edey, M. 23
Block, B. 223, 224 Eigen, M. 46, 66
Brownlee, D. xiv Ellington, A. 55, 66, 67, 296, 297
Bryan, W. J. 319–321 Elton, S. 273
Butlerov, A. M. 50 Engel, M. 203
Eschenmoser, A. 30–31, 233
Cairns-Smith, G. 54
Calvin, W. 246 Faraday, M. 183–184
Cameron, A. G. W. 89 Farmer, C. G. 224–225
Carrier, M. 312 Faulds, H. 174
Catania, K. 175, 177, 179 Feinberg, G. 25
Cellini, Benvenuto 74 Fernald, R. 170
Chambers, J. 83 Foley, R. 272, 302
Chesterton, G. K. 173, 313, 329 Fort, C. 112
Chevalier-Skolnikoff, S. 262 Fortey, R. A. 112
Churchill, S. 281 Franck, S. 100
Comins, N. 90 Franks, N. 203–205
Craig, C. 115 Freeland, S. 15, 16–18
Crespi, B. 236 Fry, I. 24–25, 66, 67
Crick, F. 26, 67
Cronk, L. 308 Galán, J. 296
Currey, J. 15 Galen 183
Galis, F. 239
Damuth, J. 304 Galton, F. 173–174, 325
Darlington J. 209 Gans, C. 14
Darrow, C. 319–322 Gasman, D. 319
Darwin, C. 1, 136, 225, 314, 316–317, 318, Gatesy, J. 289
325 Gervase of Canterbury 70
index of authors and other individuals 463
Gibson, K. 263 Lenski, R. 121

Gonzalez, G. 104 Lewis, C. S. 326
Goodsell, D. 235 Lewontin, R. C. xi
Göpfert, M. 191, 193 Lieberman, D. 273–274
Gould, S. J. xi, xv, 66, 106, 283, 297, 301, Lissauer, J. 89, 92
304, 306, 307 Liu, R. 54
Gray, P. 275 Lord, J. 300
Greene, J. 315, 316, 318, 325 Losos, J. 125
Griffin, D. 182 Lovelock, J. 99
Haeckel, E. 316, 318–319 McCowan, B. 251

Haldane, J. B. S. 48, 330 McEwan, M. R. 187
Hansell, M. 286 Mackie, G. 236
Hart, M. 99 Mares, M. 145–146
Hauser, M. 252 Marino, L. 247, 249, 251, 252, 254, 256
Heim, R. 209 Marler, P. 252
Hildebrand, J. 179 Marshall, C. 11
Hitler, A. 94, 283, 319 Marzari, F. 86
Holland, P. 242 Maurer, B. 306
Hoover, the Harbour seal 251 Maynard Smith, J. xi, xii, 46
Horgan, J. 48 Mayr, E. xi, 164, 314
Hoyle, F. xv, 309 Mellars, P. 279
Humboldt, A. von 182, 225 Miller, P. 259
Hunt, G. 261 Miller, S. 16, 48, 49, 58, 60–62
Hurst, L. 15, 18 Modlin, R. 162
Huxley, J. 315, 316, 317–318 Morowitz, H. 9–10, 12, 16, 232, 325
Huxley, T. H. 314, 316, 317, 325 Morrison, P. 233
Moynihan, M. 268
Irwin, G. 19 Mueller, U. 207
Jablonski, D. 305 Nagel, T. 265–266

Jack, the baboon 242–243 Nagy, L. 241, 242
James, K. 66, 67, 297 Nakamura, Y. 296
Janis, C. 304 Neumann-Held, E. 322
Jeans, C. 53 Nevo, E. 302
Johanson, D. 23 Nicholson, C. 194
Johnson, J. 132 Nieuwenhuys, R. 194
Joshi, A. 124 Nilsson, D. 162
Joyce, G. 65 Nisbet, E. 242–243
Nolen, T. 155
Kanzi, the bonobo 262 Norris, R. 103
Karlstrom, A. 126
Kay, R. 278 O’Hear, A. 324
Keefe, A. 49 Omodeo, P. 165
Kirsch, J. 132 Oparin, A. 48
Kistner, D. 211, 212 Orgel, L. 26, 45, 54, 57, 59, 65, 67
Koslowski, P. 323 Ornstein, L. 232
Krabitzer, L. 267 Oxnard, C. 265, 266, 267
Kuhl, P. 226 Oyama, S. 323
LaBarbera, M. 112 Pace, N. 77

Land, M. 157, 164 Parker, S. 262
464 index of authors and other individuals
Paul the Apostle 281 Taylor, S. R. xiv

Peacocke, A. R. 314 Thomas, R. 117
Petroski, H. 14 Tierney, K. 320, 321, 322
Piccinni, E. 165 Till, H. van 327
Piccirilli, J. 29 Toth, N. 262
Pliny the Younger 281 Travisano, M. 121, 122
Polanyi, M. 327 Trinkhaus, E. 276
Pond, C. 225
Ponnamperuma, C. 32 Unrau, P. 44
Pytheas 35 Urey, H. 58, 60–61
Reif, W-E. 117 Vark, G. van 274

Reiss, D. 251, 252 Veitch, N. 39
Reynolds, J. 223 Visalberghi, E. 263
Robert, D. 191, 193 Vogel, S. 114
Roberts, T. 187
Roy, K. 157 Wächtershäuser, G. 55, 57, 67
Russell, M. D. 252 Wagner, P. 306–307
Wald, G. 68, 76, 109–110, 233,
Saier, M. 231 294
Salehi-Ashtiani, K. 297 Ward, P. xiv
Salvini-Plawen, L. v. 164 Weber, A. 15
Sandeman, D. 148, 151 Weber, N. 206
Satterlie, R. 155 Weidenschilling, S. 86
Schluter, D. 125 Weilgart, L. 250
Scopes, J. T. 321, 322 Westergaard, G. 262, 263
Scribonius Largus 183 Wetherill, G. 94
Shapiro, R. 25, 33, 45, 48, 50, 55, 63–65 Wide, J. E. 242–243, 244
Shattuck, R. 326 Wilkins, A. 312
Shine, R. 220 Williams, N. 242
Simpson, G. G. 229, 230, 231, 232, 234, Williston, S. W. 263–264
246, 314 Wilson, E. O. 324
Smith, F. 281 Winemiller, K. 133, 302
Smith, T. 9–10, 12, 16, 232, 325 Winter, W. de 265, 266, 267
Spudich, J. 171 Wood, B. 269
Stebbins, E. 296 Wyttenbach, R. 194
Stebbins, G. L. 117
Strausfeld, N. 179 Yau, K. 34
Suomi, S. 262
Switzer, C. 27–28, 29 Zhou, Z-H. 144
Szostak, J. 297 Zubay, G. 52

Simon Conway Morris-Life's Solution-Cambridge University Press (2003) PDF

Uploaded by

Copyright:

Available Formats

Simon Conway Morris-Life's Solution-Cambridge University Press (2003) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Simon Conway Morris-Life's Solution-Cambridge University Press (2003) PDF

Uploaded by

Copyright:

Available Formats

Life’s Solution

Inevitable Humans in a Lonely Universe

The assassin’s bullet misses, the Archduke’s carriage moves forward,

s i m o n c o n w a y m o r r i s is Professor of Evolutionary Palaeobiology at

‘Are human beings the insignificant products of countless quirky

simon conway morris

Cambridge University Press

© Simon Conway Morris 2003

This book is in copyright. Subject to statutory exception and to the provision of

First published in print format 2003

- ---- eBook (NetLibrary)

- ---- hardback

Cambridge University Press has no responsibility for the persistence or accuracy of

Preface. The Cambridge sandwich page xi

1. Looking for Easter Island 1

2. Can we break the great code? 22

3. Universal goo: life as a cosmic principle? 32

4. The origin of life: straining the soup or our

5. Uniquely lucky? The strangeness of Earth 69

Making the Solar System 77

6. Converging on the extreme 106

7. Seeing convergence 147

8. Alien convergences? 197

9. The non-prevalence of humanoids? 229

10. Evolution bound: the ubiquity of convergence 283

11. Towards a theology of evolution 311

12. Last word 331

Writing in the New York Review of Books,1 John Maynard Smith,

usually been arrived at multiple times, meaning that the emergence

material, from immense interstellar gas clouds rich in carbon

reaffirms our one-ness with the rest of Creation. Nevertheless, the

‘To copy one paper is plagiarism, to copy many is scholarship’: few

Melville, Eviator Nevo, Dan Nilsson, Euan Nisbet, David Norman,

RNA ribonucleic acid, a polynucleotide that conveys genetic

some abbreviations for units

I am a bipedal hominid, of average cranial capacity, write my

figure 1.1 Two trackways, obtained by satellite monitoring, of the

One such ambiguity is how life itself may have originated.

of creation ‘science’, the reality of evolution as a historical process is

inherency: where is the ground plan

vital functions, although ones no longer specifically connected to the

and its molecular inherency in any way unusual. Equally instructive

the navigation of protein hyperspace

which this protein could be assembled is absolutely gigantic. Specifi-

One inference that might be immediately drawn from this is

the game of life

respiration), or are sensitive to light (rhodopsin and cryptochromes),

If forms as complex as the protein folds are intrinsic features of

(1) Hindsight and foresight are strictly forbidden. Of course we are

in certain fish (a topic I return to in Chapter 7), but we can only

In this way Smith and Morowitz neatly encapsulate what evo-

There is at least one major evolutionary trend not immediately

This prescient statement prefigures the main purpose of this book,

genetic code is excellently adapted to the needs of reliably providing

are universal they necessarily depend on the terrestrial mechanism of

15 000 Natural code

figure 1.3 Eerie perfection. The relative efficiency of randomized

remarkably effective. Very few of the alternatives are really impres-

as already noted, they estimate to be about 1018 . On this basis there