Sim Scientific-American 1986 1 PDF
Sim Scientific-American 1986 1 PDF
Sim Scientific-American 1986 1 PDF
ERIC
� /98(j
© 1985 SCIENTIFIC AMERICAN, INC
FORD TAURUS.
engineered to work well, but
also to be easy to use.
This means making sure
Taurus is a car that is easy to
live with day to day. That is satis
fying to own, maintain and
operate. And a car that accom
modates the needs of the driver,
the needs of the passengers,
that provides ample room for Ford makes the best-built Amer
luggage.
BEST, BUILT ican cars. This is based on an
AMER ICAN CARS. average of problems reported
THE LIFETIME by owners in a six month
At Ford Quality is Job 1. A
SERV ICE GUARANTEE. 1985 survey established that period on 1981-1984 models
Participating Ford Dealers designed and built in the U.S.
stand behind their work, in
Have you driven a Ford ...
writing, with a free Lifetime
lately?
Service Guarantee for as long
as you own your Ford car. Ask
to see this guarantee when
you visit your participating
Ford Dealer.
ARTICLES
48 THE STRUCTURE OF COMET TAILS, by John C. Brandt and Malcolm B. Niedner, Jr.
Observations of Halley should illuminate the interrelations of the plasma tail and the solar wind.
DEPARTMENTS
8 LETTERS
14 THE AUTHORS
16 COMPUTER RECREATIONS
27 BOOKS
126 BIBLIOGRAPHY
BOARD OF EDITORS Armand Schwab, Jr. (Associate Editor), Timothy Appenzeller, John M. Benditt, Peter G. Brown, David L. Cooke, Jr.,
Ari W. Epstein, Michael Feirtag, Gregory R. Greenwell, Robert Kunzig, Philip Morrison (Book Editor), James T. Rogers
ART DEPARTMENT Samuel L. Howard (Art Director), Steven R. Black (Assistant Art Director), Ilil Arbel, Edward Bell
PRODUCTION DEPARTMENT Richard Sasso (Production Manager), Carol Eisler and Leo J. Petruzzi (Assistants to the Production Manager),
Carol Hansen (Electronic Composition Manager), Carol Albert, Karen Friedman, Nancy Mongelli, William Sherman,
Julio E. Xavier
COpy DEPARTMENT Sally Porter Jenks (Copy Chief), Debra Q. Bennett, M. Knight, Dorothy R. Patterson
SCIENTIFIC AMERICAN ( ISSN 0036-8733), PUBLISHED MONTHLY BY SCIENTIFIC AMERICAN, INC., 415 MADISON AVENUE, NEW YORK, N.Y. 10017. COPYRIGHT @ 1985 BY SCIENTIFIC AMER
ICAN, INC. ALL RIGHTS RESERVED. PRINTED IN THE U.s,A. NO PART OF THIS ISSUE MAY BE REPRODUCED BY ANY MECHANICAL, PHOTOGRAPHIC OR ELECTRONIC PROCESS,
OR IN THE FORM OF A PHONOGRAPHIC RECORDING. NOR MAY IT BE STORED IN A RETRIEVAL SYSTEM, TRANSMITTED OR OTHERWISE COPIED FOR PUBLIC OR PRIVATE USE WITH
OUT WRITTEN PERMISSION OF THE PUBLISHER. SECOND·CLASS POSTAGE PAID AT NEW YORK. N.Y.. AND AT ADDITIONAL MAILING OFFICES. AUTHORIZED AS SECOND·
CLASS MAIL BY THE POST OFFICE DEPARTMENT, OTTAWA, CANADA, AND FOR PAYMENT OF POSTAGE IN CASH. SUBSCRIPTION RATE: $24 PER YEAR IN THE U.S. AND ITS
POSSESSIONS: $33 PER YEAR IN ALL OTHER COUNTRIES. POSTMASTER: SEND ADDRESS CHANGES TO SCIENTIFIC AMERICAN, 415 MADISON AVENUE, NEW YORK, N.Y. 10017.
5
© 1985 SCIENTIFIC AMERICAN, INC
6
© 1985 SCIENTIFIC AMERICAN, INC
Enemy submarines have nowhere to lurk now that the us. Navy has deployed a totally new passive
sonar system. The Surveillance Towed Array Sensor System (SURTASS), now operational, is an array
of miniaturized hydrophone listening devices towed behind a dedicated T-AGOS ship. It acquires and
transmits acoustic information to shipboard processors, while shore stations analyze the data to detect
and classify targets. A SURTASS preproduction development program is under way at Hughes Aircraft
Company to replace the present large array with one having a smaller diameter. This new version will
simplify storage and handling, as well as allow for a faster towing speed.
Advanced satellites will provide communications to the world's shipping and offshore industries later
this decade. The International Maritime Satellite Organization (INMARSAT), a group of 43 countries,
plans to launch the first of the spacecraft in 198 8. The new series will accommodate the increasing
demand for services, which is growing as fast as 60 % a year. Each spacecraft will be able to carry at
least 125 simultaneous transmissions. More than 3, 300 vessels are equipped to use the INMARSAT
satellite system. Users include operators of oil tankers, liquid natural gas carriers, off-shore drilling
rigs, seismic sur vey ships, fishing boats, passenger liners, and tug boats. British Aerospace will build
three satellites, with INMARSAT having an option to purchase six more. Hughes, which in 1976 built
the world's first maritime communications satellite, will provide the communications electronics for
the second-generation spacecraft.
An Amraam missile bored through radar clutter to intercept a drone aircraft target in the second guided
launch of the full-scale development program. The test firing was the third consecutive launch of the
advanced medium-range air-to-air missile, under development by Hughes for the US. Air Force and
Navy. An F-15 launched the missile in a "look-down, shoot-down" tail-aspect attack while flying at
Mach 0.9 approximately 16,000 feet above the desert floor at White Sands Missile Range. The QF-lOO
target flew at Mach 0.7 only 1,000 feet above the ground. The Amraam flew the first part of its flight
under control of its on-board inertial reference unit, using target coordinates provided in prelaunch by
the F-15's Hughes APG-63 radar. The missile then switched to its own active radar for guidance and
tracked the drone through the heavy ground clutter to intercept.
Controller-thrust aircraft engines will be tested at NASA's Lewis Research Center with three heat
®
monitors. Three Probeye thermal video systems will monitor, measure, and analyze heat from the
engines while they are operated within the environmental test chamber at the research center. Each
Probeye system consists of an infrared imager, image processor, and TV display, all of which interface
with a computer. Hughes designed and produces the Probeye system.
Hughes Missile Systems Group, in Canoga Park, California, an attractive suburb of Los Angeles, offers
engineers and scientists a unique environment that features challenging projects and professional recog
nition. Our primary work is in advanced missile and missile systems design. Openings are in analogi
digital circuit design, software engineering; high-voltage power supply design; microwave process,
transmitter and radome design; electro-optical design; IR imaging sensors; focal plane arrays; quality
assurance engineering; operations analysis; mechanical engineering; and production design and
systems engineering. Send resume to Hughes Missile Systems, Engineering Employment, Dept. S2,
8433 Fallbrook Ave., Canoga Park, CA 91304. Equal opportunity employer. US. citizenship required.
For more information write to: PO. Box 45068, Dept. 76-6, Los Angeles, CA 90045-0068
HUGHES
A I RCR AFT COMPANY
© 1986 Hughes Aircraft Company
8
© 1985 SCIENTIFIC AMERICAN, INC
PONTIAC GRANDAM
WE BUILD EXCITEMENT
LETS GET IT TOGETHEII � IJUCKLE UP
10
© 1985 SCIENTIFIC AMERICAN, INC
*!
CORVUS
Corporate Headquarters: 2100 Corvus Drive,
San Jose, CA 95124. (408) 559-7000. Telex: 278976.
European Offices: Corvus (U. K.), Ltd.,;;7 Falrmile, Henley·on·Thames,
Oxfordshlre, England RG9 2JR. Phone: 0491 571100.Telex: 847282.
Geneva Corvus Systems, S.A.,47A, Rue de Lausanne, CH-1201 Geneva,
Switzerland. Phone: 327289.Telex: 27699.
YEARSAGO
dredfold in less than 30 years." "An interesting paper has been com
municated to one of the California
scientific societies on the fossil wood
which is found in different localities
SCIENTIFIC
throughout the State. This silicified
wood is stated to be a variety of quartz;
13
© 1985 SCIENTIFIC AMERICAN, INC
14
© 1985 SCIENTIFIC AMERICAN, INC
Or call Toll-free:
800-356-4444
RECREATIONS
seems capable of simulating almost
the entire range of cluster behavior.
Three sets of arrays are used. The first
set keeps track of the accelerations
currently experienced by the stars in
D
of arrays record positions: x(i), y(i)
forms a cosmic dance to the of the distance between the two stars. and z(i) are the ith star's x, y and z co
tune of gravity. During a hu The sum total of all such paired forces ordinates of position. The starting val
man lifetime the stars barely move; acting through time presumably deter ues for the arrays x, y, z and VX, vy,
over a longer span, in which years are mines the pattern of movement within vz must be initialized at the head of
equivalent to seconds, they trace out the cluster. A program, called CLUSTER, the program.
a tangled figure of orbits. Occasionally computes the sum of the forces for The main body of the CLUSTER pro
a single star encounters a neighbor in a each star and moves the sum from its gram follows the initialization seg
pas de deux that hurls it out into space. present position to a new one nearby. It ment. The double loop can be en
If such escapes are more than occa does this repeatedly during centuries tered and reentered endlessly, or the
sional, the cluster gradually shrinks of simulated time. programmer can establish the specif
and the core begins to collapse. A certain tedium attends typing in ic conditions that control reentry. The
A powerful telescope can reveal the the coordinates and velocities of many outer loop considers each star in turn
structure of some clusters in our gal stars, but once this is done an armchair and sets the acceleration components
axy but it cannot compress years into universe unfolds on the display screen. to zero. After this has been done the
seconds-only a computer is able to Stars at the center of the cluster follow inner loop computes the forces pro
do that. A computer can also be pro wobbly, erratic courses; those at the duced on each star by its companions
grammed to serve as a kind of tele periphery drift away, stop and then in the cluster.
scope for viewing hypothetical clus glide back. The most interesting events For example, let us assume that the
ters. At cosmic speed one can watch include close encounters and escapes. index of the outer loop is i and that
the movement of the members of a When two stars approach each oth the inner-loop index isj. The inner loop
cluster as a succession of snapshots in er closely, they impart a tremendous first checks to determine whether i is
which each star leaves a dotted trail gravitational boost to each other and equal to j. If it is, the program does not
that weaves through the cluster [see il speed apart. Escapes are usually the invoke the force computation: a star
lustration on opposite page]. result of one or more close encounters. does not attract itself. In any event, to
Do gravitational forces alone ac When a star speeds away from its clus compute force under the circumstance
count for the evolution that astrono ter, there are only two possibilities: ei would cause the computer to attempt
mers infer from observed clusters? ther the star returns or it does not. An division by zero. (This is the only situa
Computers help to find answers to this astronomical body has an escape ve tion that can actually make me feel
question and related ones. A confer locity that depends on its mass and on sorry for a computer.) When i andj are
ence of simulators and theoreticians the mass of the body or object from not equal, CLUSTER uses Euclid's for
met at Princeton University in May, which it escapes. If the velocity is at mula for distance between the stars:
1984, to discuss the consistency of hy tained by a star moving outward from the differences of thc x, y and z coordi
pothetical and actual star clusters. It its cluster, it will never return. Inexpe nates are squared and added together.
was the 113th symposium of the Inter rienced cluster buffs are likely to wit The result, of course, is the square of
national Astronomical Union; the en ness frequent escapes from the config the distance. Next, the inner loop tests
tire symposium was devoted to the dy urations they design. In fact, a com whether this number is O. If it is, an
namics of star clusters. mon initial experience is to see one's alarm of some kind should be raised
It is relatively easy to choreograph hoped-for dance disintegrate. It is wise because the computer is about to be
a cosmic ballet. In principle the stel to practice by building a system of two asked to divide by zero. My version of
lar interactions within a cluster are or three stars. the program prints "Collision!"
classically simple: both members of a The structure of the CLUSTER pro If nothing is amiss, the inner loop
pair of stars experience a gravitation gram is simple. It consists of an initiali computes the distance between the
al force that is proportional to the in zation loop followed by a double loop. stars by taking the sq uare root d of the
verse square of the distance between Within the double loop the accelera squared distance computed earlier. It
them. The force is also proportional tion, velocity and position of each star then divides 1,000 by the square of the
to the product of the two stellar mass are updated according to the summed distance, a calculation that yields the
es. Such a formula is easy to compute: attractions of the other stars. I shall force. The final task to be performed
multiply the masses together; then mul describe a particularly simple version within the inner loop is to determine
tiply the product by a constant of pro- of the program in which the time incre- the acceleration components of the
16
Four stars put Oil a cosmic ballet for a few years alld thell leave the stage
17
People who aren't all that crazy about Field HaMe STATE
�;j;ji4a pr ;::t"
Operator Matches Ho COMbination
ntlEx ess Ion COMbine with .AHD.
now provides them with new easy-to-use Line I Field Operator Constant/Expression I Connect
I "ffV" !
pull-down menus for creating, using and
( STATE Matches .OR.
Z I STATE Matches "liE" ) .AHD.
3 j PROD_DESC Matches "LM Bass Lures" I .A"D.
More than or equa I 11/81185
[
OJlDElI-DATE
ASHTON ·TATE
22
7
/
----- -
/ /
/ /
/
_ -1 _
/
/ /
/ /
/ /
/
/ /
/ //
/ D
--
"
-
"
"
"
"
X Y Z VX VY VZ
A table listillg all but three stars ill the Ileighborhood of our solar system
24
© 1985 SCIENTIFIC AMERICAN, INC
I
n last October's column I described
three puzzles: Bill's baffling burr,
Coffin's cornucopia and Engel's enig
ma. Hundreds of readers have tackled
the puzzles. While some seek the mag
ic combination of moves that disas
semble the burr, others scratch their
head over the placement of polyomi
noes in a tray. Members of this group
will have to get by without help from
their friends: each puzzle is unique.
Still other readers keep rotating the
wheels of Engel's enigma in a vain at
tempt to unscramble it. Some of the
devotees are succeeding, at least on an
abstract plane: claims of solutions to
the enigma have started to come in.
A call for two-dimensional burrs
brought in a number of designs. The
most charming design received so far
is shown at the top of this page. The
problem is to remove the four pieces
from the tray symbolized by the rec
tangular outline. The pieces can only
be moved in four directions confined
to the plane of the page: up, down, left
and right. The four corner squares are
regarded as immovable. Which piece
must be moved first? Jeffrey R. Carter
of Littleton, Colo., designed this two
dimensional tour de force. Our three
dimensionality confers the advantage
Globular cluster Messier 13 ill the constellation Hercules
25
27
© 1985 SCIENTIFIC AMERICAN, INC
28
© 1985 SCIENTIFIC AMERICAN, INC
J ust published,
FROM FALLING BODIES TO RADIO WAVES
traces the development of scientific thought from the
works of the "founding fathers" to the more recent
discoveries of Maxwell, Boltzmann, and Gibbs. In QUAN. ISBN TIT LE PRICE
Segre's whimsical prelude, he asks us to imagine him 1482 From Failing Bodies to Radio Waves $13.95
conversing with these scientific giants, moving within (paper)
the times in which they lived. We peer through Galileo's 1147 From X·Rays to Quarks (paper) $12.95
telescope, call on an inhospitable Isaac Newton, and
SPECIAL TWO·VOLUME SET PRICE $19.95
discuss the enigma of magnetism with Michael Faraday.
Postage and Handlin g $ 1.50
New York, California, & Utah residents
In FROM X-RAYS TO QUARKS, Segre describes $
add sales tax.
the evolution of atomic and subatomic physics, inter $
TOTAL
woven with the personal histories of the leading re
searchers, and set against the political and cultural D I enclose a check made payable to
W. H. Freeman and Company.
backgrounds of those various times and places. As The
American Journal of Physics said, "Segre illuminates D MasterCard D VISA
this vast and cluttered field with mastery of both detail Account #
and perspective. This excellent book is one every /
Exp. Date Signature
physicist will want. It will also be a fine gift for non (All credit card orders must be signed)
communicates the inspiration, creative effort, and true Address ______________________ _____________
panying coupon to order the two-volume set and save Mail to: W. H. FREEMAN AND COMPANY
more than 25070! 4419 West 1980 South, Salt Lake City, Utah 84104
0221
31
© 1985 SCIENTIFIC AMERICAN, INC
T
here is something about the topic With the space station in place, the vated in some quarters to the quasi-re
of outer space that induces hy National Commission on Space envi ligious belief that space is a natural
perbolic expectations. With no sions a number of options for building habitat of human beings. According to
difficulty at all I can think of a billion what it calls the "infrastructure re this belief, the real goal of the space
dollar space mission before breakfast quired for the initial exploration and program is to establish "man's per
any day of the week and a multibillion occupation of the inner solar system." manent presence in space," a slogan
dollar mission on Sunday. Ordinarily The options include the construction that does not respond to the simple
I do not inflict such visions on my fel of three more space stations, one in question: "For what purpose?" Cou
low citizens, but I note that proposals high earth orbit, one in lunar orbit and pled with the public acclaim for the
of comparable or lesser merit and of one in orbit around the planet Mars; manned Apollo missions to the moon,
much greater cost receive public atten the deployment of additional space this kind of advocacy has committed
tion, and some are influential in high stations in orbits around the earth NASA to an overriding emphasis on the
circles of government. I submit that moon system or the earth-Mars sys development of manned space flight:
the proposed permanently manned tem, to serve as long-range "buses" for roughly two-thirds of the agency's
space station is in this category. earth-moon or interplanetary trans funding is allocated to that objective.
A National Commission on Space, port, and the construction of several The directions embodied in NASA'S
mandated by Congress and appointed vehicles to shuttle astronauts among budgetary policy ignore the basic his
by the president, has bravely under the various space stations, moon and tory of space flight: in the more than
taken to foresee the course of the U.S. planets. The concept of a joint U.S. 28 years since the launching of Sput
space effort over the next 50 years. The U.S.S.R. manned mission to land on nik !the overwhelming majority of sci
commission's final report, to be re Mars has been endorsed by many offi entific and utilitarian achievements
leased in March, will take it for grant cials both in and out of NASA. The pres in space have come from unmanned,
ed that the space station will be operat ence of people living and working in automated and commandable space
ing in orbit within a decade, as Presi space, with necessarily elaborate pro craft. For example, the program of
dent Reagan announced in his 1984 visions made for their health and well unmanned planetary exploration has
State of the Union message. According being, is common to all the major op been brilliantly successful and has
to the timetable of the National Aer tions being considered for recommen made immense contributions to hu
onautics and Space Administration, dation to the president. man knowledge. Robot satellites in
the initial operations capability of the earth orbit have revolutionized glob
T
space station is to be achieved by 1993. he acceptance of such grandiose al communications and navigation,
Official estimates set its development proposals by otherwise rational in and they have yielded fundamental ad
costs at $8 billion in constant 1984 dol dividuals stems from the mystique of vances in our understanding of the at
lars, but the true costs will probably be space flight, as nurtured over many mosphere, the oceans, the weather and
many times that preliminary figure. centuries by early writers of science the distribution of natural resources.
There have been no announcements fiction and their present-day counter Finally, they have enhanced national
about the costs of operating and main parts. Indeed, to the ordinary person security by making it possible to moni
taining the station in orbit or about the space flight is synonymous with the tor military activities abroad.
costs of the equipment needed to make flight of human beings. The simple Let me make it clear that I have no
the station a useful facility for scientif taste for adventure and fantasy ex hesitation in granting the technical
ic and technical purpose. pressed in that sentiment has been ele- feasibility of a space station or of a
32
© 1985 SCIENTIFIC AMERICAN, INC
DETAILED IMAGES of the distant planets betoken the accom ring in the Encke division in the outer main ring (ring A) of Saturn;
plishments of the unmanned, scientific space program. The image at the data for the computer-generated image were gathered by a pho
the upper left shows a storm on Mars; it was transmitted by the topolarimeter aboard the Voyager 2 spacecraft, which recorded the
Viking Orbiter 1 spacecraft and processed in false colors to high occultation of starlight passing through the rings. In the image at
light the details of the storm. At the upper right is an image of the lower right the theta-aurora of the earth is shown as a yellow
Jupiter, which has been constructed by a computer from data trans ring and crossbar, on which the outline of Antarctica has been su
mitted by the Voyager 1 spacecraft to show the planet as it would perposed. The image was transmitted by the Dynamics Explorer 1
appear from directly above its south pole; no spacecraft has ever satellite. The first three images were prepared by the Jet Propul
made a real photograph of Jupiter from that vantage. There is no sion Laboratory and are shown courtesy of the IBM Gallery of Sci
photographic data from the black, irregular region at the pole. The ence and Art in New York City. The image of aurora over Antarc
bright red band in the false-color image at the lower left is a thin tica is shown courtesy of Louis A. Frank of the University of Iowa.
33
© 1985 SCIENTIFIC AMERICAN, INC
__
ests with which it competes will re
OL-____ � ________ �
________ �________ � ________ _L________ � � ceive any significant subsidies from
1960 1965 1970 1975 1980 1985 the Defense Department budget.
One is left, therefore, to consider the
ANNUAL BUDGET for the National Aeronautics and Space Administration in constant
objectives in space that are not overtly
1982 dollars is plotted on the graph in black; the budget for space science and applications
military in nature. The history of the
is superposed on the graph in color. NASA'S greatest spending took place in the mid-1960's,
during the development of the manned Apollo missions to the moon. The growth of space civil space program in the U.S. shows
science and applications in that period did not keep pace with the growth in the manned that following the peak in funding
space program. Since then the overall NASA budget has fallen to about a third of its peak generated by the Apollo program in
value, and about 15 percent of the total has been allocated to science and applications. the mid- 1960's, appropriations fell by
34
© 1985 SCIENTIFIC AMERICAN, INC
other projects? Have we become so Precise characterization of external magnetic fields of the moon, the earth, Mercury, Venus,
enamored of the capabilities of com Mars, Jupiter and Saturn
mandable spacecraft that we have ig Detailed study of the structure, composition and dynamics of the earth's atmosphere and
nored the fact that a man in space can exploratory study of the atmospheres of Venus, Mars, Jupiter, 10, Saturn and Titan
carry out these tasks more efficiently Precise characterization of the external gravitational fields of the moon and the earth
and with less effort? Comprehensive observation of the solar wind and of shock waves, energetic solar particles and
To answer these questions the histo galactic cosmic rays in interplanetary space out to a distance of 3.4 billion miles from the sun and
ry of the space program, and particu continuing outwards
larly that of the Space Transportation Comprehensive surveys of stellar and planetary sources of gamma rays, X rays and ultraviolet,
System, would seem to be a more reli infrared and radio-frequency radiation and the detailed spectral study of selected sources
able guide than the promises and fore Marked advances in understanding the origin and evolution of the solar system and of stars and
casts made by interested parties. The galaxies
present Space Transportation System Significant contributions to fundamental plasma physics and its role in planetary and astrophysi
includes a fleet of four manned, orbit cal systems
ing space shuttles, each of which is in Study of ocean currents and the global dynamics of the oceans
essence a high-velocity aircraft and Negative evidence on the past or present existence of living organisms on the surface of Mars
spacecraft that is launched by rockets,
flies in low-altitude orbit about the
earth, reenters the earth's atmosphere MAJOR ACCOMPLISHMENTS of the unmanned space program are snmmarized in the
on command and lands on a very long table. They inclnde commercial and military applications of space technology, civil appli
airstrip. The development and initial cations that are partly public and partly commercial and many scientific accomplishments.
35
© 1985 SCIENTIFIC AMERICAN, INC
36
© 1985 SCIENTIFIC AMERICAN, INC
37
© 1985 SCIENTIFIC AMERICAN, INC
38
© 1985 SCIENTIFIC AMERICAN, INC
-.. --
SPACE STATION, if it were constructed, could resemble the design At this stage the two designs and, indeed, several others can be
shown here, but fundamental decisions about the design are stilI quickly interchanged on the color monitor of a computer-aided de
pending. The latest version of the basic structural design differs sign system, such as the one responsible for the image shown, made
from a previous one in baving two main "towers" instead of one. In by the McDonnell Douglas Corporation. Nevertheless, the current
tbis computer-generated image of the current version only the shut NASA schedule for development calls for tbe electronic image to be
tle-docking area of the space station is shown, and the scale of the translated into a real device in earth orbit by the year 1993. The cost
structure is indicated by the human figures near one of the towers. of that effort may be as high as $30 billion in constant 1984 dollars.
39
© 1985 SCIENTIFIC AMERICAN, INC
by Leo Sachs
T
he cells in the body descend the various white blood cells and the their cell membrane; there the foreign
from precursors known as stem red blood cells. As a model system I molecules are "recognized" by other
cells. Stem cells can multiply studied myeloid stem cells, the mature kinds of white blood cells, which then
rapidly; their progeny, after matur white blood cells to which myeloid mount an immune response to the in
ing and differentiating into specialized stem cells give rise and myeloid leuke truder. James E. Till and Ernest A.
forms, generally stop growing. Dur mic cells. The cells in myeloid leuke McCulloch of the Ontario Cancer
ing normal infancy and adulthood the mia appear to be frozen in their pre Treatment and Research Foundation
processes of multiplication and differ cursor form. Ordinarily they do not have shown that in the body macro
entiation are in harmony: the growth differentiate into mature cells, and so phages and granulocytes develop from
of stem cells provides new tissue and they continue to multiply. a common precursor form, the mye
replaces dying cells, while the cessa In trying to establish general prin loid stem cells. The cells in our culture
tion of growth after cells reach their ciples governing cell growth and dif therefore followed the same pattern of
final form keeps cell multiplication in ferentiation and the reversibility of development as they do in the body.
check. In cancer the harmony breaks malignancy I had to establish a cell What was the role of the fibroblasts,
down: there are too many immature, culture system that made it possible which had to be present in the culture
multiplying cells. to observe these processes in vitro. In as a "feeder layer" if the myeloid cells
An understanding of how the proc 1963, working with my graduate stu were to grow and differentiate? I had
esses of growth and differentiation are dent Haim Ginsburg at the Weizmann shown in 1965 that the cells of the
regulated in normal cells makes it pos Institute of Science in Israel, I found feeder layer secrete substances induc
sible to answer several questions about that when normal blood-cell precur ing growth and differentiation. The
cancer. Have all the cellular mecha sors are placed in a liquid culture me following year, working with my stu
nisms controlling growth and differen dium containing fibroblasts, or con dents Pluznik and Yasuo Ichikawa, I
tiation gone awry in malignant cells, or nective-tissue cells, the precursors showed that the inducers are detect
do some of the controls still operate? multiply into clones (descendants of a able in the culture medium; moreover,
If some of the mechanisms are intact, single cell) and ultimately differenti the inducers that acted on clones con
can they be reactivated to make can ate. In 1965, working with another taining different types of cells seemed
cer cells differentiate and stop grow graduate student, Dov H. Pluznik, I to differ. The inducers were purified
ing? I have concluded that malignant showed that clones can also develop and shown to be proteins, either glyco
cells can retain the genetic bases for and differentiate in a semisolid medi proteins (proteins with attached sug
differentiation. Appropriately stimu um containing agar. The gel inhibits ars) or proteins without detectable sug
lated, they can complete the normal the mobility of the cells, making it eas ars. The sugars do not seem to be
sequence of growth, differentiation ier to distinguish separate clones. In necessary to the inducers' activity. Us
and cessation of growth. These find 1966Thomas R. Bradley and Don ing the same cell-culture system, other
ings have opened new possibilities for ald Metcalf of the University of Mel laboratories have isolated inducer pro
cancer therapy. bourne and the Walter and Eliza Hall teins for all the many kinds of blood
I studied the processes of growth Institute of Medical Research in Mel cells, including the various lympho
and differentiation in normal and leu bourne also cloned precursor cells in cytes. A range of cell types can secrete
kemic blood cells. In hematopoiesis, agar. The first normal precursors for inducer proteins, both in culture and in
the process by which the body gen which the method was used were mye the body.
erates new blood cells, a single kind loid stem cells, which gave rise to mac
of precursor (the multipotential stem
cells in the bone marrow) develops
rophages and granulocytes.
Macrophages and granulocytes are T he development of a clone of ma
ture macrophages or granulocytes
inte a number of more specialized the body's phagocytic cells: they en requires the multiplication of myeloid
precursor cells, among them myeloid gulf and dispose of foreign materi precursors, followed by their differen
stem cells and lymphocyte precursors. al, including invading microorganisms. tiation. It seemed plausible that a spe
These more specialized stem cells in Macrophages also display molecules cialized protein induces each of the
turn mature and differentiate into all from the engulfed foreign material on processes. Eitan Fibach and I found
40
MYELOID LEUKEMIC CELLS are shown in stages of develop cytes, one of the cell types they can give rise to. Leukemic cells
ment from an undifferentiated, rapidly growing form (top left) to a normally do not differentiate, but in this case the cells were induced
mature granulocyte, which does not divide (bottom right). Normal to do so: they were incubated with differentiation factor, a pro
myeloid precursor cells, a variety of immature, undifferentiated tein made by normal cells. Once the leukemic cells had assumed a
blood cells, pass through these stages as they mature into granulo- mature form they stopped growing and hence were not malignant.
41
42
@ @ ©
that the source of the differentiation
inducer was the myeloid cells them LEUKEMIC CELLS
1 11
selves. Growth factor, then, has two
effects: it causes the precursor cells to
grow and it induces in them the pro
duction of differentiation factor. When
the precursors have multiplied into a
large enough population, their com
* * @
bined production of differentiation
factor is sufficient to cause them to dif
ferentiate. This normal relation be PARTIALLY
DIFFERENTIATED
tween growth and differentiation can
CELLS
be short-circuited, however, by certain
compounds that cause cells other than
1
the precursors themselves to produce
/� !
and secrete differentiation factor.
The production of differentiation
factor may take place later in the
* ** ©1
growth of some precursors than it does
in others, allowing a larger population
MATURE
of cells to form before differentiation
CELLS
occurs. Specific growth factors may in
!
duce precursor cells to make specific
differentiation factors, thereby leading
to particular kinds of mature cell. This MACROPHAGE GRANULOCYTE
possibility might explain why each of
the four myeloid growth factors is as
sociated with a different set of mature
� @
cells; it offers a possible alternative to
the explanation that different growth
factors act at different stages of cell
P
development.
The normal coupling between cell
growth and differentiation is accompa
RECEPTOR PROTEINS
nied by a second linkage, in the cells of
the blood and other tissues: between
differentiation and the cessation of RESPONSE TO DIFFERENTIATION FACTORS varies among clones of leukemic cells.
growth. Why do mature cells stop mul Some clones are differentiation positive (D+): the addition of differentiation factors isolat
-
ed from cnltnres of normal cells causes them to develop normally. First they assume an
tiplying? Mature red blood cells in hu
intermediate stage of differentiation, in which their surface membrane displays receptor
man beings and other mammals elimi
proteins characteristic of mature cells, and then they take on the appearance of mature,
nate their nucleus and so are unable to nondividing macrophages or granulocytes. Other clones are differentiation defective (D ).
divide. Other mature cells retain their Under the influence of differentiation factors they may reach an intermediate stage of dif
nucleus but do not grow; it has been ferentiation, in which their growth is merely slowed, or they may be entirely unaffected and
suggested that such terminally differ continue to multiply as before. Other compounds can make D- cells differentiate, however.
entiated cells produce growth-inhibit
ing compounds, which block the cells'
own multiplication. obtain cloned lines of malignant cells cancer cells can metastasize to sites
that could be grown in culture and where the growth factors their nor
T eukemic cells have escaped from the were not contaminated with normal mal counterparts need to survive and
L constraints governing the growth cells. In my laboratory and in others grow are lacking. Conversely, the pref
of normal myeloid cells. One limita several such lines of myeloid leukemic erence of some kinds of metastatic
tion on normal growth is the supply of cells have been isolated. Their study cells for particular organs may indi
growth factor. Normal myeloid pre revealed that the cells circumvent the cate that they still require a small ex
cursors must depend on other cells to growth limitation ordinarily imposed ternal supply of growth factor, which
produce the factor, and the supply can by the supply of growth factor in two their preferred organs supply.
be sporadic. During an infection, for ways. Certain strains of leukemic cells To qualify as malignant a cell must
example, fibroblasts and other kinds simply need less growth factor than also escape the second constraint limit
of cells secrete large quantities of the normal cells, and the amount they re ing the growth of normal cells: the cou
factor, thereby increasing the body's quire decreases as the cells grow in cul pling of growth and differentiation,
population of white cells, but at oth ture until they have no need of any which regulates the balance between
er times the level of growth factor de inducer. Another strain produces its multiplying cells and cells that have
clines. Does the continued growth of own growth factor. stopped growing. If growth factor in
leukemic cells indicate that they have Either alteration makes the cells ca duced the production of differentia
escaped from the normal requirement pable of multiplying steadily rather tion factor in leukemic cells, then ma
for growth factor? than intermittently, as normal cells do. lignant cells producing growth factor
To answer the question we had to These changes can also explain how on their own or exposed to it from oth-
43
44
of mature macrophages and granulo changes produce asynchrony and prevent D - cells from responding to differentiation factor.
45
W
e found that the various sub nant character. We found, for exam than induced, the necessary coordina
stances capable of inducing my ple, that in normal myeloid precursors tion of gene expression fails. The asyn
eloid leukemic cells to differentiate growth factor induces the synthesis of chrony can block the normal differ
work in diverse ways. Some of the a number of proteins and stops the syn entiation program, resulting in a cell
compounds that cause D+ cells to dif thesis of others. In leukemic cells, how that never differentiates fully and that
ferentiate induce the production of ever, these protein changes do not need therefore continues to multiply. Geof
normal differentiation factor by turn to be induced: the gels showed that the frey Symonds and I found that cer
ing on the genes that encode it. Other cells had already undergone the chang tain serums cause constitutive protein
compounds, such as the steroid hor es in protein synthesis. For the leuke changes that are characteristic of D
mones, act on the genome without the mic cells the protein changes were leukemic cells to revert, so that their
intermediary of normal differentiation "constitutive." Presumably the leuke expression has to be induced. After
factor. In some D- clones differentia mic cells had no need of externally treatment with these serums the cells
tion requires combined treatment with supplied growth factor because the became capable of maturing under
several compounds, each of which genes that ordinarily are affected by the influence of normal differentiation
turns on some of the genes needed growth factor had already been turned factor. We had in effect unblocked the
for cell differentiation. Together the on or off. genetic program for differentiation.
compounds activate a combination of We also identified some of the pro The finding that various compounds
genes that is sufficient to result in dif tein changes induced in normal cells can stimulate D- cells to mature in
ferentiation. by differentiation factor. D+ leukemic spite of constitutive protein changes
A study of cellular proteins in nor cells, like normal cells, only showed that make the cells resistant to normal
mal and leukemic cells revealed some the protein changes after exposure to differentiation factor indicates that
of the genetic bases for the varying differentiation factor: the changes had there are several genetic programs for
responses of leukemic strains to dif to be induced. Many of the changes in differentiation. Lydia Cohen and I
ferentiation-inducing substances. Dan protein synthesis were constitutive in used two-dimensional gel electropho
Liebermann, Barbara Hoffman-Lie- D- clones, however. The more resis- resis to examine the cellular proteins
of two kinds of myeloid leukemic cells:
those that could be induced to differen
100
tiate by normal differentiation factor
I I
·
but not by the steroid hormone dexa
90 1 - ,
- _.....
1 - -_ ... ·_-t· methasone and those that responded to
dexamethasone but not to differen
80 - .. -.. tiation factor. The clones displayed
different sets of constitutive protein
i=' 70 changes. It appears that normal differ
. .....-
z entiation factor and some other induc
w
I
C,) ers of differentiation activate distinct
a::
w 60 sets of genes, although the net effect
e:-
I of the genes-cell maturation-is the
w
C,)
50 i __ .. . ..
I-- .
.
�
same. It is likely that cells in other
�
CJ I cancers can also follow a variety of
z genetic pathways to differentiation.
:> 40
:>
M
a:: y colleagues and I have found
:::J I
(fJ 30 that none of the cells in our mye
loid leukemic clones have a normal ar
.-
20 . ray of chromosomes. The abnormali
ties we observed include changes in the
10 I -
I--
I number of chromosomes as well as re
I
arrangements and deletions of chro
i I mosome segments. We noted consis
5 10 15 20 25 30 35 40 45 50 55 60 65 tent differences in the clones that can
DAYS be induced to mature by differentia
tion factor and those that cannot. The
SURVIVAL OF MICE injected with myeloid leukemic cells was longer (color) when they
were given differentiation factor than when they were not treated (black). The lenkemic chromosome changes that can be ob
cells were from D+ clones, which matnre and stop growing in culture when they are ex served in leukemic cells presumably
posed to differentiation factor. The extended survival of the mice that were treated with are responsible for the genetic abnor
differentiation factor shows that it inhibited the development of leukemia in the animals. malities in the control of cell growth
46
© 1985 SCIENTIFIC AMERICAN, INC
[IJJ
cells malignant.
One way that malignancy can be
reversed is by restoring the chromo
/�
somes of malignant cells to a more
normal pattern. In 1968 Zelig Rabi
nowitz and I showed that clones of sar
coma cells (malignant cells arising in
connective tissue) cultured under cer
tain conditions often yielded cells that
had reverted to a nonmalignant state. CHROMOSOME CHANGES DIFFERENTIATION
/ \ �
In sarcomas that Carmia Borek and I
induced in cultured normal fibroblasts
by means of chemical carcinogens or
X rays, the revertant cells had the lim
ited life span of normal fibroblasts.
CD IIIII [ill
They also had a chromosome compo GENEW NE GAIN
sition different from that of their ma
lignant progenitors: one more closely
resembling the chromosomes of nor
mal cells. NONMALIGNANT NORMALLY NONMALIGNANT
The chromosomes whose alteration
DIVIDING CELL NONDIVIDING CELL
was crucial to the reversion of malig REVERSAL OF MALIGNANCY can be achieved in two ways. Cells become malignant as
nant cells carried genes that at the time a result of genetic changes disturbing the normal balance between genes that can cause
I called E genes (genes for the expres malignancy (color), generally known as oncogenes, and genes that suppress oncogene ex
sion of malignancy) and S genes (genes pression (gray). In culture malignant cells have been observed to resume normal growth
following chromosome changes that restored the balance between oncogenes and their sup
for the suppression of malignancy).
pressors (left). Arresting malignancy through cell differentiation, in contrast, bypasses the
Chromosome changes that acted to re
genetic changes (right). A malignant cell's genetic makeup remains abnormal, but after
move E genes from the influence of S differentiating the cell assumes a mature form, stops dividing and is no longer malignant.
genes caused cancer; those that re
stored the gene balance led to rever
sion. The E genes were later isolated in
various laboratories and named onco tion factor or a compound that causes the varying susceptibility of malignant
gel)es; the S genes can be called sonco normal cells in the body to increase cells to the differentiation-inducing
genes or anti-oncogenes. the production of differentiation fac compounds.
We asked whether a similar restora tor. It may be that similar effects can Differentiation therapy for leuke
tion of a more normal chromosome be achieved in human beings, provid mia could use the same substances in
makeup takes place when leukemic ing an alternative to the cytotoxic much lower doses, perhaps in combi
cells are made to differentiate and stop drugs now employed in cancer chemo nation with normal differentiation fac
growing. Are the genetic abnormalities therapy, which kill many normal cells tor. Some of the compounds could be
corrected or does the reversion of ma as well as cancer cells. given in small doses to induce differen
lignancy by differentiation bypass the In another therapeutic use, normal tiation while others are administered
genetic changes that made the cells macrophage and granulocyte induc in larger quantities for their cytotox
malignant in the first place? We exam ers-both growth and differentiation ic effects, thereby reducing the total
ined the chromosomes of myeloid leu factors-might be administered to pa number of leukemic cells. In order ·to
kemic cells that had been made to tients undergoing conventional cyto choose the best combination of com
differentiate and found their chromo toxic therapy for other cancers. The pounds and approaches before treat
some composition was still abnormal. factors might bolster their macro ment, leukemic cells from the patient
It appears that arresting multiplication phage and granulocyte popUlations, could be tested outside the body in or
by inducing differentiation bypasses which chemotherapy depletes. The der to determine their susceptibility
the genetic abnormalities that original factors might also be given to alleviate to various substances.
ly disrupted the normal sequence of nonmalignant deficiencies in macro My suggestions have already led
growth and differentiation. phages or granulocytes. Factors that several workers to attempt clinical tri
act on other cell types could be put to als of differentiation inducers in mye
47
T
he years 1985-86 will one day tember 11, 1985, and relayed a mass curved and fuzzy, usually displaying
be regarded as a golden age for of data that are still being evaluat little or no internal structure. Plas
cometary astronomy. Indeed, if ed. Probes from the European Space ma tails are quite different. They are
we had been allowed to choose two Agency, Japan and the U.S.S.R. will made of molecules that have been ion
years in which to be active as cometary venture to Halley in March, trans ized in the atmosphere by solar radia
scientists, these would have been our mitting valuable measurements bear tion, trapped on interplanetary mag
clear choices. Two important comets, ing on the structure, composition netic fields generated by the sun, and
Giacobini-Zinner and Halley;have ap and physical conditions of cometary wrapped around the comet so that the
proached within range of observat'ion atmospheres, as well as providing the ions form a long, hairpin structure,
as they orbit around the sun. Giacobi first images of a cometary nucleus. which often exhibits threadlike for
ni-Zinner has already yielded a boun The missions will be supported by net mations, knots and large-scale distur
ty of information as a result of being works of ground-based observations; bances. The gases in the coma and the
the first comet to be visited by a space concurrent observations will be made tail fluoresce, that is, they absorb sun
'
craft, and astronomers have deployed by the crew of NASA S Astro 1 space light and reradiate it.
an unprecedented array of resources shuttle mission. According to a generally accepted
to examine Comet Halley. Data are be model proposed by Fred L. Whipple in
ing gathered by observatories on the
earth's surface, by spacecraft orbit T he current missions extend a rich
history of comet observation that
1950,the source of all cometary ma
terial is the nucleus, which is inside
ing the earth, by vehicles in space and has stretched across the centuries. The the atmosphere but is never observed
in orbit around other planets and by word "comet" comes from the ter through telescopes because it is too
six spacecraft that will fly near or into minology of Greek astronomers, who small. Whipple compared the nucleus
the comet's atmosphere. first named these solar visitors "aster to a dirty snowball; it consists of ices
It is fortunate that Giacobini-Zinner kometes," or long-haired star. It is now of water and other molecules. Dust
and Halley-among the few known known that comets are composed of grains, and possibly rocky material,
periodic comets that are sufficiently three main pa�ts: the atmosphere, the are interspersed more or less uniform
bright and are also nearly complete in tail and the nucleus. The visual part ly throughout the icy matrix. The nu
their range of cometary features-have of the atmosphere is often called the cleus is often considered to pe some
come under close scrutiny at virtually coma or head; it is an essentially spher what spherical, measuring several kil
the same time. Astronomers will com ical cloud of gas and dust. The atmos ometers in diameter. How can such a
pare findings derived from similar phere may range from less than 1,000 minor body give rise to plasma tails
observational techniques for the two kilometers in diameter to several mil that are sometimes 50 million kilome
rather different comets. The massive lion kilometers, depending on the spe ters long?
efforts organized to study Giacobini cies of gas. One or more tails, which The answer lies in the dynamics of
Zinner and Halley promise to provide are directed away from the sun, extend the material the nucleus releases into
direct evidence for theories about the behind the atmosphere. There are two interplanetary space. As a comet ap
origin, composition and dynamics of principal types of tails: dust tails and proaches the sun the nucleus absorbs
comets and their tails; we also expect plasma tails. sunlight and heats up until it reaches
them to raise many new questions. Dust tails typically consist of sol the temperature at which the ices sub
The National Aeronautics and id, micrometer-size particles that have limate, or go directly from the solid
Space Administration's International been pushed away from the coma by phase into the gaseous one: The escap
Cometary Explorer (ICE) sped through the pressure force resulting from sun ing gases leave the nucleus at an initial
the tail of Giacobini-Zinner on Sep- light striking the dust grains. They are speed of several tenths of a kilometer
per second; as they move outward they
undergo many chemical reactions. In
addition fleeing gas molecules collide
COMET WEST, photographed on March 9, 1976, was characterized by two types of com
with newly liberated dust grains and
et tail. The dust tail is the wide, diffuse tail to the left, consisting of three broad bands with
propel them outward. Many of the gas
faint emissions between them. It is composed of solid particles that have been pushed away
from the comet's atmosphere by the pressure force resulting from sunlight striking the dust
molecules absorb ultraviolet photons
grains. The plasma tail is the narrower one to the right, which is threaded with intricate of sunlight and gain kinetic energy as
streamers. It is composed of molecules that have been trapped by interplanetary magnetic they break into smaller molecules-a
field lines (carried by particles of the solar wind) that have been wrapped around the comet. process called photodissociation. This
49
© 1985 SCIENTIFIC AMERICAN, INC
50
© 1985 SCIENTIFIC AMERICAN, INC
ORBIT OF
COMET
GIACOBINI
ZINNER
ORBIT OF
COMET
HALLEY
GIACOBINI-ZINNER AND HALLEY are periodic comets, which proximately once every 76 years and Giacobini-Zinner approach
visit the inner solar system in the course of their regular journeys es the earth every 6.5 years. The colored dots on the three orbits in
around the sun. Halley's comet is observable from the earth ap- dicate the position of the two comets and the earth as of January 15.
51
© 1985 SCIENTIFIC AMERICAN, INC
� "
"
"
"
"
"
" TAIL
"
,
"
"
_COMA
TO SUN
MAGNETIC FIELD LINES from the incident solar wind are com become very large, the pressure of outwardly flowing ions matches
pressed in the comet's atmosphere, whose ions become captured by the pressure of the inwardly flowing solar wind. This balance makes
the magnetic field lines as the solar wind streams into the cometary it impossible for the solar-wind plasma, and the lines of force it
atmosphere. The compression is caused by the increasing mass of carries, to penetrate any farther into the atmosphere; the result is a
the solar wind as a result of picking up ions; conservation of mo magnetic-free region in the comet. Off to the sides the magnetic
mentum dictates that the flow must slow down. At some point deep field lines WC2p around and behind the comet. These lines drag ions
inside the atmosphere, where the concentration of captured ions has in the atmosphere along with the solar wind, away from the sun.
DIRECTION OF
MAGNETIC
FIELD LINES �
�
S
N
\ AREA OF MAGNETIC
FIELD REVERSAL
IONS IN COMET TAIL are channeled along the magnetic field the supersonic solar wind has captured a large enough mass of com
lines of the solar wind as it wraps around the comet. A sheet of etary ions, it is abruptly slowed as it passes through a "bow shock,"
electric current (vertical arrow) and a magnetic field of Idw strength which is analogous to the shock wave caused by a supersonic air
separate lobes of opposite magnetic field polarity in the tail. When craft or the bow wave caused by a boat as it moves through water.
52
© 1985 SCIENTIFIC AMERICAN, INC
53
© 1985 SCIENTIFIC AMERICAN, INC
T hePhenomena
coverage by the Large-Scale
Network will be sup
with Giacobini-Zinner. The other five
will encounter Halley in March of this
shape of the captured magnetic field,
the two-lobed tail and the flow of elec
plemented by photography of Halley year, and ICE will be an upstream mon tric current in the tail-features that
by the wide-field cameras on board itor. Prior to these missions there have follow from Alfven's 1957 model. ICE
the Astro 1 space shuttle mission. Dur- been no in situ measurements of any measured a striking amount of turbu-
1 2 3 4
" E " " E E " E
EO " E E E " E E
EO " EO EO EO EO " EO
EO EO EO EO
0
z EO EO EO EO
� EO " � EO
a:
<l: EO EO EO EO
....J
0
(f) EO EO EO EO
'7 " EO EO EO
V
" EO EO " " EO
EO " EO EO EO EO
�
EO EO " EO
" EO EO EO NEW
TAIL
EO " EO EO
" " EO EO
EO EO
E
EO EO
" EO
EO "
" EO
EO EO
MAGNETIC SECTOR BOUNDARY
) )
PLASMA TAIL DISCONNECTS when the comet crosses a sector lines retain ions they have captured, but they are no longer bound
boundary, moving from a sector where the magnetic field has the to the comet. When all the old magnetic fields have been ejected
same polarity as the tail to a sector where the field has opposite from the atmosphere, the newly created ions in the atmosphere have
polarity. When the new field enters the coma and is pressed into the no magnetic connection to the plasma tail. The tail appears to dis
oppositely directed old fields threading the tail, the old field lines connect from the coma. The atmosphere immediately supplies ions
are cut by a process known as magnetic reconnection. The old field for a new tail having the polarity of the new magnetic field sector.
54
© 1985 SCIENTIFIC AMERICAN, INC
55
© 1985 SCIENTIFIC AMERICAN, INC
�:����
ENCOUNTER: 7
;��
DATE SEPTEMBER 11, 1985 MARCH 6, 1986 MARCH 9, 1986 MARCH 13, 1986 MARCH 8, 1986 MARCH , 1986
. .. .. . . ._----+---_...- .... -1-... .. ..
�)
..._-_ . .._._. . ... - ._... .. .. . . . .. . .. .
DISTANCE TO NUCLEUS
-- .. .. ... _--_._ .....
8,000 (TAILWARD) 10,000 (SUNWARD) 3,000 (SUNWARD) 500 (SUNWARD) 4 x 106 (SUNWARD) 2 x 105(SUNWARD)
.
... .... ..
. . . .
77 7 7
21 80 69 4 5
. .
_
(KILOMETERS) -
DISTANCE FRO:-:-M-:-::c S, -;
U,- ,- -1-
N - - X- - -S-- -- X S . X S X
........ . - . ............. .. . .. _ . . .
COMET EXPERIMENTS 8 13 13 10 3
SPACECRAFT ENCOUNTERS with two important comets are (ICE was launched into orbit between the earth and the sun in Au
listed in this table. The encounter with Giacobini-Zinner provid gust, 1978, and was retargeted in December, 1983, for the encoun
ed valuable data on the structure of comet tails when the Nation ter with Giacobini-Zinner.) The missions that will intercept Halley
al Aeronautics and Space Administration's Illtematiollal Cometary in March should provide a more detailed understanding of the prop
Explorer (ICE) passed through the plasma tail last September 1 1. erties of comet tails by supplying information from the sunward side.
of cometary physics, but many ques changes over time will be needed to et Wild II in 1995 and record valuable
tions, including those that emerge deepen understanding of comet tails. data as it travels with the comet along
' its orbital path for approximately two
from the new data, will remain. The NASA S Comet Rendezvous and Aster
direct exploratory missions provide in oid Flyby mission, planned for launch and a half years. If the mission is suc
formation based on a series of snap ing in' the early 1990's, should pro cessful, it will mark the next logical
shots taken along single trajectory vide this important information. The step in attempts to explore and under
lines. Global data recording cometary vehicle is expected to approach Com- stand the nature of comets.
GREAT COMET OF 1843 dominated the northern skies. It was million kilometers long, or longer than the distance from the sun to
depicted in this lithograph as it was seen over Paris on March 19. the orbit of Mars. The artist recorded a bright tail that is probably of
The comet, perhaps the brightest one of the past two centuries, has the dust variety. Although there must have been a plasma tail, it was
not lJeen seen since the 1843 appearance. The tail was about 300 probably embedded in the dust tail and was too faint to be observed.
56
© 1985 SCIENTIFIC AMERICAN, INC
You' re driving 55 MPH on a rain-slick curve. Suddenly the unexpected: You stand
on the brake pedal and steer to stay in your lane. You might expect Europe's most
exotic cars to handle such a crisis effortlessly. Yet for all its awesome straight-line
braking ability, Ferrari 308 GT Si failed to negotiate a 150-foot radius curve at
maximum braking in USAC-certified testing. Lamborghini Countach failed. Lotus
Esprit Turbo failed. Porsche 944 failed. Only the 1986 Corvette demonstrated the
ability to steer and stop in these conditions at the same time. Only Corvette made
the turn while coming to a controlled stop. When conditions turn foul, Corvette's
new computerized Bosch ABS II anti-lock braking sy stem is designed to help
improve a driver's ability to simultaneously brake and steer out of trouble.
Why does the Corvette feature the world's most advanced braking technology?
Because a world-class champion should give y ou the edge in an emergency.
Corvette. A world-class champion.
G)
'
But training a computer to respond information it needs
results in several areas ofknowl
to analogous or unexpected situations to mimic the experts'
edge-based systems. In addition,
-teaching it to learn-is a very differ • E responses to an
research is under way to teach
ent challenge. And this is one of our immense variety of
computers to learn by them S
long-range programs in AI research. problems.
selves, much as humans do.
COMPASS (Cen
The ultimate brain-picking. tral Office Maintenance Printout
It's extremely tedious and difficult to
The Expert-Systems version of AI
teach a computer to respond to spe Analysis and Suggestion System) is an
is literally the result of programming
cific problems in an intelligent way. Expert System we devised for tele
the experiences of experts into a com
communications. It is being phased
puter.
into field use to monitor switch per-
(iji�
an immense variety of industrial and
human problems.
The box lists some of the pertinent
papers GTE personnel have pub
lished on various aspects of Artificial
60
© 1985 SCIENTIFIC AMERICAN, INC
A breakthrough
in portable
computing power
•
and versatility.
Zip.
I
ent files on the pocket size disk- You'll be amazed at what it can do!
ettes. W ith Telecom and the Tandy Visit a Radio Shack Computer
I Phone I
600's built-in modem, you're able Center today. ._----_.
'RAM upgrades are in banks of 96K each (26-3910. $399.95). Prices apply at Radio Shack Computer Centers and participating stores and dealers. Mu�iplan and MS·WordITM Microsoft Corp.
F lographers
or a little more than a year crystal d ifferent set. For such symmetry to be chance agreement," he estimates, "is
and solid-state physi possible, however, the atoms cannot be surely less than one in 1, 000."
cists have been faced with the possibil arranged on the planes in a regular Investigators trying to determine the
ity that there is a previously unknown way, and the distances between adja possible structure of quasicrystals con
"quasicrystalline" form of ordered cent parallel planes cannot vary peri tend, however, that the hypothesis of
matter. Unlike a crystal, such matter odically. It is therefore not clear how twinning does not fit all the available
cannot be described in terms of a re to describe a quasicrystal in a way that data. Indeed, Cahn recalls that "when
peated lattice. Unlike an amorphous indicates the positions of every atom. Shechtman first showed me the d iffrac
solid, on the other hand, it does have a John W. Cahn of the N B S , one of tion pattern, I said, 'Go away, Dany,
long-range orientational order. Shechtman's coauthors, points out that that's just twinning: We then spent
The patterns of speculation over the "in crystallography you can say 'Here more than two years, before we pub
nature of this new state of matter have is what the unit cell looks like; repeat lished, ruling out just such conven
been disturbed by Linus Pauling. He that and you have the crystal. ' But how tional crystallographic explanations. "
says the data on which the hypothesis much must one say about a quasicrys Cahn and other solid-state physicists,
is based can be explained, without pro talline structure in order to define it including David R. N elson of Harvard
posing such a quasicrystalline state, uniquely?" University, insist that results obtained
by a well-known crystallographic phe Pauling contended, in a talk deliv by the techniques called microdiffrac
nomenon called directed twinning, in ered in August to a session on "crystal tion and high-resolution microscopy
which several identical crystals grow lography applied and misapplied" at eliminate Pauling's hypothesis from
along shared faces. a meeting of the American Crystallo consideration.
The original evidence for the exis graphic Association, that it is not nec In microdiffraction a d iffraction pat
tence of quasicrystals was an electron essary to consider such exotic struc tern is made of an area as small as 100
d iffraction pattern published in N o tures. In his talk, a revised version of angstroms in d iameter. Such patterns
vember, 1984, in Physical Review Let which was later p ublished in Nature, he have been made from many locations
ters by Dany Shechtman of the Israel suggested that the same fivefold sym on the same sample of aluminum
Institute of Technology-Technion and metry could be caused by twinning: by manganese alloy; every microdiffrac
colleagues at the Technion, at the a growing together of 20 identical crys tion pattern made by an electron beam
French N ational Center for Scientific tals from a single seed. aimed at the appropriate angle, re
Research and at the U.S. N ational Bu In Pauling's model each crystal gardless of position within the sample,
reau of Standards (N BS). The pattern, grows as a compressed trigonal pyra shows fivefold symmetry. Cahn argues
which was made by aiming a beam of mid: a squat pyramid with a triangular that if the material were composed of
electrons at a rapidly cooled alloy of base. The 20 pyramids grow with their twins, not all the microdiffraction pat
aluminum and manganese, had five tips pointing in toward the seed. Each terns would include every one of the
fold rotational symmetry; in other crystal fits together with the neighbor twinned crystals; some would there
words, when the pattern was rotated ing pyramids, and so as they grow their fore consist of only the d iffraction pat
by one-fifth of a full circle, it looked bases fit together as the faces of an ico- terns due to a few of the crystals and
62
© 1985 SCIENTIFIC AMERICAN, INC
$6.95
with a few carefully chosen words. It trains you to de
C
ould a tabletop experiment shed fend yourself with a simple eloquence that will sub
the U.S., including verbal self-defense
sessions for doctors, lawyers, and
light on the birth of galaxies? It due your verbal opponents. other professionals.
very well might, according to Woj Keep Cool During Arguments SPECIAL SALE PRICE AVAILABLE BY MAIL ONLY
ciech H. Zurek, a theoretical physicist
i--.-n;s&NObie-
The Gentle Art of Verbal Self-Defense also helps you
working at the Los Alamos N ational avoid the self-defeating, overly emotional attitudes that
Laboratory. many people assume during arguments. Never again
lose an argument by being sullen, uncontrollably I BoolcsJellenJ Since 1873
Zurek bases his bold proposal on I DEPT. V165, 126 FIFTH AVE, NEW YORK, NY. 10011
I
angry, peevishly defensive, or apologetic.
the universal behavior of matter near 1295690. Please rush me__copies of The Gentle Art
Learn to Handle AD Attacks I of Verbal Self-Defense a your special sale price of $6.95
phase transitions. A phase transition I t
(Plus $1.00 per bookshipping and insurance per book). NY
takes place when a material changes Learn how to deal with all types of people, including I and N.J. residents: Please add sales tax.
bullying bosses, backbiting co-workers, guilt-produc I
from one state to another. Perhaps the I
I
ing mothers, nagging wives, condescending husbands,
best-known phase transition is the and many others, with helpful features like these: Name
8
melting of ice to water. Remarkably,
the laws of physics that describe this
•
•
The 1YPes of Verbal Attacks
The 4 Principles of Verbal Self-Defense
I Address
common example also apply to phase • The 5 Personalities & How They Communicate I Cily
j
which helium liquefies and at which
the universe was born.
Writing in Nature, Zurek proposes
that phase transitions in liquid helium
ordlc rack
can serve as models of phase tran Jarless Total Body
sitions in the universe soon after its
birth. Specifically, Zurek's proposal fo
Cardiovascular Exerciser
cuses on the transition from normal Duplicates X-C Skiing for the
liquid helium to superfluid. The prop Best way to Fitness
erties of the superfluid state are exotic
indeed. As a superfluid, for instance, formJy exercises the large leg muscles and also adds
important upper body exercise. Higher pulse rates,
liquid helium seems to flow without necessary for building fitness, seem easier to attain
friction. In addition its thermal con because the work is shared by more muscle mass.
ductivity, which is a million times Even Better Than Swimming
greater than that of the normal fluid, NordiCfrack more effectively exercises the largest
muscles in the body, those located in the legs and
is better than the conductivity of buttocks. When swimming, the body is supported
any metal. M oreover, quantized vor Cross.country skiing is often cited by phYSiologists by the water, thus preventing these muscles from
as the most perfect form of cardiovascular exercise being effectively exercised. T he stand up exercising
tex lines, or long-lived whirlpools, are for both men and women. Its smooth, fluid, total pOSition on the NordiCfrack much more effectively
known to exist in the superfluid. body motion uniformly exercises more muscles so exercises these muscles.
higher heart rates seem easier to attain than when
Such vortex lines are analogous to A Proven, High Quality Durable Product
jogging or cycling. NordiCfrack closely simulates the
cosmic strings, or loops of very dense pleasant X·( skiing motion and provides the same NordicTracks have been in prodllction since 1976.
cardiovascular endurance· building benefits-right NordicTrack is quiet, motorless and has separately
material, that correspond in a mathe in the convenience of your home, year 'round. adjustable arm and leg resistances. We manu
matical sense to stable solutions of cer Eliminates the usual barriers of time, weather, facture and sell direct. fuo year warrantee,
30 day Irial period with return privilege.
'
tain fundamental equations of high chance of injury, etc. Also highly effective for weight
control.
energy physics. Phase transitions in the Folds and stands on end
early universe are thought to have re
More Complete Than Running
to require only 15" x 17"
NordicTrack gives you a more complete work out- storage space. '
sulted in the copious production of conditions both upper body and lower body I
these postulated but as yet undetected muscles a1the same time. Fluid, jarless motion does Callor u'rile for. . . 1" ,
FREE BROCHURE
entities. Cosmic strings may be as long
not cause joint or back problems
63
© 1985 SCIENTIFIC AMERICAN, INC
A even
one might infer the density of cosmic thick shell of the earth, perhaps (whose locations coincide with the sta
strings formed in the early universe. including part of its very ble spin axis of the earth). It soon be
The experiment he puts forward is core, has slipped erratically around came clear, however, that the distances
simple. A ringlike tube containing the earth's center in the course of mil traveled by the paleomagnetic p oles
liquid helium in its normal state at lions of years. The evidence for this were greater than could be allowed by
about two degrees K. is subjected to proposition is presented in Journal the prevailing theory of the earth's
high pressure. By rapidly removing of Geophysical Research by J ean A. magnetism. Moreover, the apparent
the pressure a transition to the super Andrews of Columbia University's paths traced by the wandering magnet
fluid state is induced. If Kibble's sce Lamont-Doherty Geological Obser ic poles varied according to the conti
nario is correct, the forming of vor- vatory. Andrews' findings, startling nent where the data were collected.
For some time these discrepancies
seemed to be explained by the theory
of plate tectonics: the drift of conti
nents on great plates of the earth's sol
id outer layer, the lithosphere, had
jumbled the paleomagnetic data by re
arranging the positions of the rocks
from which the data had been collect
ed. The apparently excessive distances
between past magnetic and present ge
ographic poles could therefore be
viewed as the combined result of plate
motions and geomagnetic instabilities.
Yet calculations of ancient pole loca
tions that took these two factors into
account still left a substantial amount
of the p olar wandering unexplained.
Because the geomagnetic poles are
linked to the geographic ones and
could not have wandered freely over
the surface of the earth, Andrews
reasoned that the entire surface of
the earth must have wandered over
the poles instead: a major part of the
earth's thick plastic mantle, which un
derlies the lithosphere, must have
shifted about the center of the earth,
carrying the entire lithosphere with it.
Such a large-scale motion of the man
tle and lithosphere, when added to the
relative motions of the lithospheric
plates and the vagrant motion of the
geomagnetic poles, could fully explain
APPARENT WANDERING of the North Pole is due to slippage of the mantle and litho
polar wandering.
sphere around the center of the earth; the pole has actually remained stationary in space.
To prove this hypothesis Andrews
.The endpoints of the line segments designate positions of the pole on the surface of the had to select well-defined p oints in the
earth at particular times (millions of years ago) in the past. Each point is encircled hy its earth's mantle from which she could
range of locational error. Landmasses are shown at their present latitudes and longitudes. measure any change in position with
64
© 1985 SCIENTIFIC AMERICAN, INC
N1CDONNEL
OUGLAS
68
© 1985 SCIENTIFIC AMERICAN, INC
Tailored Photosynthesis
69
© 1985 SCIENTIFIC AMERICAN, INC
S
A solution of hexachloroplatinate ome bacteria, including those re phage lambda, a virus that infects bac
ions did yield photoactivity. (That sponsible for typhoid fever and tu teria, and proceeded to parcel out the
is, the thylakoid-platinum composite, berculosis, invade animal cells and total genetic "library" of Y. pseudotu
trapped on filter paper, responded to survive there; other infectious bacteria berculosis in different E. coli cells.
light by simultaneously generating hy are content to make their living outside The E. coli were applied to a mono
drogen and oxygen from water. ) Each the cell. Now two investigators have layer of some 10 million HEp-2 cells,
hexachloroplatinate ion, in which a isolated a gene responsible for the in a laboratory cell line derived from a
platinum atom is surrounded by six vasion process. Understanding of such human lung carcinoma. After three
chlorine atoms, has a net charge of genes could lead to the development ho urs the monolayer was washed re
-2. Greenbaum proposes that the of alternatives to antibiotics for com peatedly. The only E. coli remaining
successful precipitation results from bating harmful invasive bacteria. would be those that somehow had
linked themselves to the cells of the
monolayer. Some might have got in
side the cells. The survivors of the
washing were cloned (that is, allowed
to generate colonies of identical off
spring) and the clones were given a sec
ond opportunity to enter HEp-2 cells.
Then the HEp-2 cells were bathed in
the antibiotic gentamycin, which can
not penetrate cells. The surviving bac
teria would have to be inside the cells.
Twelve clones survived. The cells were
examined under the electron micro
scope. E. coli indeed were inside them.
The genetic material from Y. pseudo
tuberculosis that confers the ability to
invade proves to be a DNA strand no
longer than 3,200 bases; Isberg and
Falkow call it the inv locus. It is in fact
a single gene, which specifies the struc
ture of a single large protein with a
molecular weight of 108,000. In the
case of Y. pseudotuberculosis. at least,
the genetic basis of the ability to in
vade cells is strikingly uncomplicated.
Isberg and Falkow observe that their
work "may be a first step towards de
fining a previously undefined class of
proteins encoded by many invasive
BACTERIAL INVADERS of a mammalian cell are the five dark bodies toward the top of
this electron micrograph of the interior of a lung-carcinoma cell. The bacteria are Esche disease-producing microorganisms as
richia coli, which normally cannot invade; a single gene, designated if/v, was transferred to part of their strategy of pathogenesis."
the E. coli to change them into invaders. Each bacterium is about two micrometers long. J ust what this protein d oes to promote
The micrograph was made by Stanley Falkow of the Stanford University Medical Center. invasion remains to be learned. When
70
© 1985 SCIENTIFIC AMERICAN, INC
ferries that carry proteins or DNA into classroom. And Just look at what the Webster's contains:
I n an ecologically conscious era bi We s er ,s New Umversal Unabridged DlctJOnary continues that great
!J<l �
dltlon. The ebster's defines common and obscure words, words used
ological methods for the control 10 othe.r . centunes and words newly created in the sciences and the
humanities.
of pests and weeds have strong appeal. t�adition of quality also extends to its format and organiza
More Than 3,000 Illustrations tion: The pages are thumb-indexed for handy reference.
The release of sterile adult insects, for While other large dictionaries use eye-straining type, the
!
The Webster'S is strong on visuals, too. Finely detailed pen drawings ap Webster's features good-sized print, with defined words stan
instance, may d iminish screwworm pear throughout the text. These illustrations help make the Webster's a ding out c e�rly in .boldface. To make reading easier still,
. .
each def1OltlOn IS listed on a separate line or set of lines .
and mosquito populations, thereby fun, visually exciting learning experience, especially for children.
71
© 1985 SCIENTIFIC AMERICAN, INC
Complex
Fluid Development
To solve the p roblem, Exxon soug ht
to c reate a unique f l u i d that cou ld be
E))(ON
su rfactants and polymers . Tog ether, they reseNes around the worl d . For more
ref ined the formulation through a series i nformation on shear-thickening f l u i d ,
of tests i n fu l l-scale su rface equi pment, write Exxon Production Research Com
) then i n wel l s provided by Exxon Com- pany, po. Box 21 89, Houston, TX
o pany, U . S.A. , using materials supplied 77252-21 89.
by EXxon Chemical Company. The f inal
result was successfu l ly f ield-tested , solv-
Exxon Research and
Engineering Company
ER&E is also a whol ly owned subsid
iary of Exxon Corporation . ER&E scien
tists and eng i neers are involved i n the
by David M. Pepper
B
oris Va. Zel'dovich and his col phase conjugation [see "Optical Phase Amnon Yariv of the California Insti
leagues observed a curious phe Conjugation," by Vladimir V. Shku tute of Technology and me (then at
nomenon as they did an experi nov and Boris Va. Zel'dovich; SCIEN Caltech), and by David M. Bloom,
ment at the P. N. Lebedev Physical TIFIC AMERICAN, December, 1985]. Gary C. Bjorklund and Paul F. Liao of
Institute in Moscow in 1972. The in Phase-conjugate-wave technology the AT&T Bell Laboratories. The tech
vestigators intentionally distorted an has many intriguing applications. A nique involves the interaction in a non
intense beam of red light from a pulsed high-quality optical beam can, for in linear medium of four optical beams:
ruby laser by directing it through a stance, be transmitted through a turbu three input beams and one outPlJt
frosted glass plate. Then they aimed lent atmosphere and, after generation beam. The three input beams consist of
the smeared beam down a long tube of its phase-conjugate beam, be made a probe beam, whose "time-reversed"
containing high-pressure methane gas. to retrace its path exactly. When the replica is sought, and two counter
In accordance with a well-known ef beam returns to its point of origin, it propagating pump beams that are
fect called stimulated Brillouin scat will therefore be free of degradation. necessary to "sensitize" the nonlinear
tering the beam interacted with the Such beams can be used in the pointing medium. The steps leading to the pro
molecules of the gas and was reflected and tracking of moving objects, the duction of the fourth beam, the phase
backward; the gas acted as if it were a processing of images, optical comput conjugate output beam, are analogous
mirror, but a very unusual one. What ing, interferometry, laser gyroscopes, to those of conventional holography.
surprised the investigators was that af fiber and satellite communication sys In holography a photographic emul
ter the reflected wave passed back tems, laser weapon systems and photo sion is illuminated with light from an
through the same piece of frosted glass lithography. The "mirrors" that gen object and a reference beam. The ref
a nearly perfect, undistorted optical erate phase-conjugate beams also erence beam and the light from the ob
beam emerged. In other words, the make possible novel laser resonators. ject interact to produce a hologram,
distortions introduced during the first or three-dimensional interference pat
passage through the glass had been un Generating "Time-reversed" Light tern, in the emulsion. After the film is
done. (Reflection from a convention developed the fixed, or static, holo
al flat mirror, in contrast, would have Two standard methods of producing gram can be read by illuminating it
increased the distortions.) The back phase-conjugate waves are stimulat with the same reference beam. The re
ward-traveling wave could therefore ed Brillouin scattering and four-wave sult is a three-dimensional reconstruc
be loosely thought of as the "time-re mixing. In the years following the tion of the object's image.
versed" replica of the incident wave. landmark experiment by Zel'dovich In four-wave mixing the nonlin
The phrase "time-reversed" is in investigators have found that phase ear medium acts as the photograph
tended to imply that the beam reflected conjugation by stimulated Brillouin ic emulsion. The interaction of the
back by the gas faithfully carried all scattering can occur in a variety of probe, or object, beam with one of the
the distortions introduced by the frost substances other than compressed gas. pump, or reference, beams produces a
ed glass plate, but in a reversed sense. Such substances are called nonlinear wave pattern of reinforcement and
When the beam returned through the mediums. The term is used here in a cancellation (constructive and destruc
glass, the distorting properties of the special sense: the optical properties of tive interference) in the medium that
plate therefore canceled the effects a nonlinear medium, in contrast to amounts to a real-time, or dynamic,
they had originally produced. In other those of a linear medium, are affected hologram. The phase-conjugate out
words, if it were possible to make a by light. Examples of nonlinear medi put beam is generated when the other
motion picture of the incident beam, ums include semiconductors, crystals, counterpropagating pump beam is re
the "time-reversed" beam would be liquids, plasmas, liquid crystals, aero flected from the hologram. Actually
portrayed by playing the same film sols and atomic vapors. two holograms are created: each pump
backward. "Time-reversed" waves are Nonlinear mediums are also utilized beam, in conjunction with the probe
more accurately known as phase-con in four-wave mixing, a scheme that beam, yields a hologram from which
jugate waves, or wave-front-reversed was proposed in 1977 by Robert W. the other reference beam is reflected.
replicas. The technology by which Hellwarth of the University of South Although each method-stimulat
they are generated is known as optical ern California, followed by studies by ed Brillouin scattering and four-wave
74
© 1985 SCIENTIFIC AMERICAN, INC
Adaptive Optics
76
© 1985 SCIENTIFIC AMERICAN, INC
MASK SEMITRANSPARENT
MIRROR
PHASE-CONJUGATE
MIRROR
PHOTOGRAPHIC
EMULSION SEMICONDUCTING
CHIP
PHOTOLITHOGRAPHY, the transfer of a two-dimensional pat mogeneities in the amplifying medium distort its qnality. After the
tern from one plane in space to another, can be done without beam is phase-conjugated (red) it propagates back through the am
optical lenses by employing a phase-conjugate mirror. The process plifier, undoing the distortions and emerging with the original beam
can be used to transfer a mask pattern containing a microelectron information restored. On reflection from the semitransparent mir
ic circuit layout onto a semiconducting chip coated with a pho ror, an intensified image of the mask pattern exposes the emulsion.
tographic emulsion. A beam of light (blue) from a laser passes The system compensates for optical aberrations, has high resolu
through the mask, a semitransparent mirror and then a laser ampli tion over a large field of view, eliminates laser speckle, minimiz
fier. Although the amplifier increases the power of the beam, inho- es beam spreading and avoids physical contact with the substrate.
_••"'......_ UNRECOGNIZABLE
IMAGE
INPUT
IMAGE
/
PHASE-CONJUGATE
MIRROR
ORDINARY
MIRROR
DISTORTIONS DUE TO FIBER-OPTIC CABLES can be re fiber (blue). Since all the modes traverse different paths, by the
moved by a phase-conjugate mirror. The distortions arise because time they reach the end of the fiber they are out of step, producing
a spatial image (left) sent down a cable travels in many optical a scrambled image (center). The image can be unscrambled if it is
modes: each mode corresponds to a given ray that zigzags down the phase-conjugated (red) and sent down an identical cable (right).
77
© 1985 SCIENTIFIC AMERICAN, INC
78
© 1985 SCIENTIFIC AMERICAN, INC
HOLOGRAPHIC
EMULSION
BELOW
THRESHOLD
ASSOCIATIVE MEMORY can recognize two images that share square matches a ray from one of the holograms stored in the film a
common features with each other. A memory-storage element is strong output ray is produced (heavy lille); if the rays do not match,
formed by illuminating a photographic emulsion with a reference a weaker output is produced (medium alld thill lilies). Strong and
beam and an image of a square (top left), creating a hologram, or an medium output rays are reflected by the phase-conjugate mirror at
interference pattern, in the film. The process is repeated on the the right (red). Weak output rays, however, do not have enough
same emulsion for a triangle and a circle. The incident angle of the energy to cause conjugation to occur and therefore pass through the
reference beam is changed in each case so that the images can phase-conjugate mirror (blue ray at right). In this way the qualities
be distinguished. The photographic emulsion-now a holographic of "squareness" are selected. After the strong and medium output
memory element-is developed and placed between two phase-con rays are conjugated a second time by the phase-conjugate mirror at
jugate mirrors (bottom). The image to be categorized (imperfect the left the process is repeated. Successive iterations yield a square
square showll here) is reflected by a semitransparent mirror onto as an output. In other words, the system "recognized" that the input
the emulsion (blue). Every time a ray of light from the imperfect image was an imperfect square, not an imperfect circle or triangle.
79
© 1985 SCIENTIFIC AMERICAN, INC
f\
INPUT
DATA
1 \
-----7�
OPTICAL
FIBER
\
SEMI TRANSPARENT
MIRROR�
PHASE·
CONJUGATE
MIRROR
OUTPUT
DATA
ORDINARY/
-----)�I OP � I----?)(\�
AL
FIBER
MIRROR
TEMPORALLY BROADENED PULSE can be reconstituted by a optic cable transmitting the pulse will rearrange the pulse in such
phase-conjugate mirror. The broadening occurs because the various a way that the slower traveling components enter the second half
frequency components (shown here as different colors) that consti of the cable first. As the conjugated pulse propagates through the
tute a pulse propagate at slightly different speeds through space. A remainder of the cable the faster traveling components will there
four-wave-mixing phase-conjugate mirror (pumped at the central fore catch up with the leading (that is, slower) part of the pulse: the
frequency component of the pulse) placed midway along a fiber- pulse is renarrowed and the cable can transmit more information.
80
© 1985 SCIENTIFIC AMERICAN, INC
1
~ ��I
OFF
MIRROR
------:;..
PUMP BEAMS
t t t t t
I�I
OUTPUT
PULSE
1
OUTPUT
PULSE
t t t t t
PUMP BEAMS
� �I
OFF
~
TEMPORAL REVERSAL of a pulse sequence is achieved by a gate mirror (right) will reverse the temporal ordering if the mirror
phase-conjugate mirror. An ordinary mirror (left) preserves the is turned on when the entire pulse is inside the mirror. The mirror is
temporal ordering of a pulse sequence on reflection. A phase-conju- actuated by applying two reference, or pump, beams to the mirror.
set in different typefaces (Gothic and memory information and phase-con lasers (one to provide the pump beams
Roman, say) or that one or both are jugate mirrors to provide the neces and the other to provide the probe
blurred or even have missing areas. sary image feedback [see bottom illustra beam) that have slightly different fre
These examples are of particular inter tion on page 79]. Physically the device quencies may be involved in a given
'
est because they correspond to daily is simply a resonator capped with a system; a pulse of light, which con
experience: one recalls an incomplete phase-conjugate mirror at each end. sists of a continuous band of fre
image through a complicated process The holographic memory element is quencies, could be incident on a phase
known as associative memory. positioned within the resonator cavi conjugate mirror.
Teuvo K. Kohonen of the Universi ty. One of the phase-conjugate mirrors As Yariv, Fekete and I suggested in
ty of Helsinki and more recently John has a threshold characteristic that en 1979, a phase-conjugate mirror could
J. Hopfield of Caltech have developed ables it to select the features stored in possibly be exploited to compensate
models of associative memory, var memory that have the most in com for pulse-spreading effects in fiber-op
iations of which can be realized by mon with those of an input image for tic cables. Such effects limit the infor
means of a system containing phase the next iteration. mation-handling rate of a cable since
conjugate mirrors. The model might Although digital computers can exe densely spaced pulses that spread ap
provide insight into how neural net cute image processing and associative preciably can spill into one another,
works within the brain function. The memory functions, they do the job in rendering the data string unrecogniz
basic idea consists of a mathematical discrete steps. The optical approach able. The spreading occurs because the
algorithm that, in essence, compares offers an advantage: image informa frequency components of a pulse trav
a given input data string with one in tion can be processed in a parallel el at slightly different speeds through
memory, generates a new version of manner. In other words, the entire pic a given material. If a four-wave-mix
the data string and then iterates, or re ture field is processed at the same time. ing phase-conjugate mirror is placed
peats, the procedure with the new ver at the midpoint of a fiber-optic cable
sion as the input. The comparison, se Frequency Effects that is transmitting a pulse, the fre
lection and feedback process repeats quency components will be reordered
until the input and output data strings Throughout the discussion of the ap so that the slower-propagating frequen
cor;"'erge, or agree. Many research plications of four-wave mixing the im cies leave the mirror before the faster
grouA induding those led by H. John plicit assumption has been made that propagating frequencies do. The fast
Caulfield of the University of Ala the three input beams and the one out er-propagating ones will therefore
bama, Demetri Psaltis of Caltech, put beam have the same frequency, or catch up with slower ones on transit
A. D. Fisher of the U.S. Naval Re optical wavelength. Such a scenario through the second half of the cable,
search Laboratories and A. A. Vede is known as the "degenerate" case. In restoring the quality of the pulse by re
nov of the I. V. Kurchatov Institute of 1978 Richard L. Abrams of Hughes narrowing it to its original state.
Atomic Energy in the U.S.S.R., have and I analyzed the case of "nearly de How does a phase-conjugate mirror
been exploiting electronic techniques generate" four-wave mixing, an exam reorder the frequency components of a
(computers) and electro-optical tech ple of which occurs when the frequen pulse? In the degenerate case the probe
niques to simulate the process. cy of the incident probe is slightly beam and one of the two counter
More recently Bernard H. Soffer, shifted with respect to the frequency of propagating pump beams generate a
Dunning, Yuri Owechko and Emanuel the two identical pump beams. stationary optical interference pattern
Marom, all at Hughes, and, indepen There are several situations in which within the medium of the four-wave
dently, Dana Z. Anderson of the Joint nearly degenerate four-wave mixing mixer. The other pump beam is reflect
Institute for Laboratory Astrophysics occurs: the frequency of a probe beam ed from the pattern and produces the
in Colorado are developing architec reflected from a moving satellite phase-conjugate beam. In the nearly
tures for all-optical associative memo would be Doppler-shifted with respect degenerate case the frequency shift
ries. The Hughes scheme uses a con to the pump beams in a ground-based causes the interference pattern to
ventional fixed hologram to store the phase-conjugate mirror; two separate move through space, creating a mov-
81
© 1985 SCIENTIFIC AMERICAN, INC
ORDINARY
MIRROR
(
(
)
OFF
PHASE-CONJUGATE
MIRROR
OFF
ON
PARALLEL AMPLIFYING SYSTEM could be used to initiate mirror. By the time they reach the mirror they are out of step
fusion. A fusion pellet is illuminated by a pulsed low-power la because they have traveled different distances. When all the pulses
ser (blue). Pulses scattered from the pellet pass through the three are inside the mirror, it is turned on, conjugating each pulse and re
parallel laser amplifiers shown here. The intensity of the individual versing their temporal order with respect to one another (red). On
pulses is increased, but at the expense of introducing distortions. the return trip the distortions are removed and the pulses become
The pulses are then directed to a four-wave-mixing phase-conjugate synchronized, so that an intense pulse of radiation hits the pellet.
82
© 1985 SCIENTIFIC AMERICAN, INC
83
© 1985 SCIENTIFIC AMERICAN, INC
by Peter. A. Rona
hen the theory of plate tec that is capable of concentrating metal Seawater and fractured oceanic crust
T chambers
oceans, primarily by rivers. This mod he principal sources of heat are materials is attributable to cooling by
el accounted for the only deposits of containing magma, or the convective circulation of seawater.
minerals on the sea floor then known. molten rock. Heat also originates from They estimated that the entire mass
They consist of gold, tin, titanium, dia chemical reactions between seawater of the oceans circulates through the
monds and other heavy materials me and certain constituents of the man crust at oceanic ridges every 10 mil
chanically eroded from exposed rocks tle rock. An additional heat source is lion years. On an annual basis the mass
on land and concentrated in pockets on the geothermal gradient: the increase flow rate is comparable to that of the
the sea floor by flowing water, and also in the temperature of the earth with Amazon River.
of phosphorite and manganese chemi depth, primarily due to the dissipation A two-way chemical exchange oc
cally precipitated in the form of nod of heat from the radioactive decay of curs as seawater convectively circu
ules and crusts from materials dis elements in the crust and mantle. lates through oceanic crust at high
solved in seawater.
As the dynamic-earth model of plate
tectonics developed, it became evident
SEA-FLOOR CHIMNEY several meters high on the East Pacific Rise vents hot water,
that the ocean basins were themselves precipitating particles of iron-copper-zinc sulfides assimilated in chemical reactions with
a source of heat and metals. The met fractured rocks in the oceanic crust. The investigators who found such chimneys in 1979
als result from an exchange process be called them black smokers because they resemble smokestacks. The photograph was made
tween the ocean and the oceanic crust by Kyung-Ryul Kim from the submersible Alvin. Part of the sampling equipment is visible.
84
\
\
\
\
I
I
I
\
\ NOR T H
AMERICAN
P LATE
HAWAIIAN
ISLANDS
PACIFIC
PLATE
]
)
AMERICAN
PLATE
MINERAL DEPOSITS formed by sea-floor hot springs are identi id, fractured volcanic rock or sediment as a permeable medium and
fied by the colored circles on this map. The conditions necessary for a source of metals, and a magma chamber as a source of heat. Fa
the formation of such a deposit are seawater as an ore-forming flu- vorable conditions occur at sites along the boundaries of the plates
86
E
is more than 2,000 meters and the wa xplorations by oceanographic re
ter temperature is 350 degrees C. The search vessels of many nations are
result is a quenching of the solution, developing a clear picture of sea-floor
lowering its temperature and changing mineralization in different settings of
its composition from acidic to alka sea-floor spreading. The mineralizing
line. Metallic minerals precipitate rap process exhibits many common fea
idly. Under reducing conditions (the tures and some differences that depend
ANATOLIAN PLATE absence of free oxygen) the metals on the setting. Sites in the Red Sea rep
combine with sulfur derived from sea resent mineralization in one setting: an
water and dissolved from the crust early stage of the opening of an ocean
to deposit polymetallic sulfides. Under basin, associated with a slow spreading
oxidizing conditions the metals com center that bisects the Red Sea. There
bine with oxygen, hydroxide, silica sul hot, metal-rich brines fill a number of
fate or carbonate to deposit metallic basins along the axial zone. The brines
oxides, hydroxides, silicates, sulfates are produced by ore-forming hydro
or carbonates. thermal convection systems that oper
Boiling takes place when the confin ate in the basins. Cold, dense seawater
: RIC A N
ing pressure of the overlying seawater that penetrates down through volcanic
decreases to the critical point, which rocks in marginal zones becomes un
usually occurs at water depths of less usually salty because it passes through
than 2,000 meters for a solution at 350 thick beds of rock salt (sodium chlo
P L ATE
degrees C. A liquid phase containing ride) buried in the crust. Such beds are
the sodium chloride separates from a characteristic of an early stage of the
vapor phase that retains the hydrogen opening of an ocean basin, when un
sulfide. Boiling may result in the pre der dry climatic conditions evapora
cipitation of polymetallic sulfide de tion exceeds the replenishment of sea
posits under the sea floor or in the dis water because the inflow from exterior
charge of a hypersaline brine denser oceans is restricted by the surrounding
than normal seawater. landmasses.
The anomalous physical and chem As the salinity increases, salt crystals
ical conditions for ore-forming hy precipitate, settle and accumulate in
drothermal convection systems appear thick beds on the sea floor. The high
in sea-floor spreading centers that ex salinity of the circulating solutions en
hibit both common features and dif hances their capacity to transport dis
ferences. All spreading centers have a solved metals as complexes with chlo
segmented structure along their axes, ride from the salt. It also decreases
consisting of linear segments some 10 their density, so that the heated solu
kilometers long that alternate with tions collect as density-stratified brines
transverse offsets at fracture zones when they discharge from the floors
ranging from a few kilometers to of the basins. Metals precipitate from
that make up the lithosphere (the earth's hundreds of kilometers in length. All the hot brines as particulate mineral
rigid outer shell) and in volcanic regions spreading centers have an axial zone of phases that settle in the basins and
within plates, as in the Hawaiian Islands. volcanic activity about one kilometer are trapped, forming layered deposits
87
© 1985 SCIENTIFIC AMERICAN, INC
1 J
Atlantis II Deep in the axial zone of the
Red Sea, in water 2,000 meters deep
due west of Mecca. The deposit is esti
COPPER-IRON-ZINC SULFI DES mated to contain a salt-free bulk dry
COPPER weight of 100 million metric tons and
COPPER
to consist of 29 percent iron, 2 to 5
percent zinc, .3 to .9 percent copper, 60
parts per million of silver (6,000 met
ric tons at 1.8 ounces per ton) and .5
part per million of gold (50 metric tons
at .015 ounce per ton).
This deposit ranks with the larger
ancient sulfide deposits on land. The
Saudi-Sudanese Red Sea Commission,
which represents the adjacent coastal
states, sponsored a mining-feasibility
test there in 1979. The commission
plans to undertake a full-scale pilot
mining operation when the world met
al market improves.
HELIUM , METHANE (?) HELIUM, METHANE (?) As an ocean basin widens by slow
.£\. spreading from the early stage
MAGMA CHAMBER represented by the Red Sea to the ad
vanced stage of opening represented
by the Atlantic, several changes take
MINERALIZATION PROCESS is shown in a cross-sectional view of two symmetrical
place. The oceanic circulation changes
hydrothermal convective circulation systems at a sea-floor spreading center. Seawater pen
etrates downward to a depth of several kilometers through fractured crust containing vari
from restricted to unrestricted; more
ous metals. Heat from an underlying magma chamber causes the water to expand and to oxygen is present, so that conditions
rise convectively. The heated water dissolves a number of elements, including metals, from shift from reducing to oxidizing, and
the rock and loses a few elements. Additional metals may come from the magma chamber the salinity of hydrothermal solutions
and the underlying mantle, along with helium and possibly methane gas. In "leaky" systems generally decreases to that of nor
fresh seawater penetrates downward and mixes with the upwelling hot, metal-rich solutions, mal seawater as the rock-salt beds are
triggering the precipitation of metals as sulfides in the crust; metals remaining in solution moved away from the axial zone by
are deposited on the sea floor from hot springs. In "tight" systems the solutions discharge
the spreading sea floor. Any sulfide de
directly into hot springs that deposit sulfides on the sea floor. The diagram is schematic.
posits exposed on the sea floor would
be decomposed by oxidation unless
they were preserved by an imperme
CONSTITUENT BASALT SEAWATER RIDGE/RIVER
able cover of sediment or lava. Depos
COPPER (Cu) -
+ its such as those of the Atlantis II Deep
IRON (Fe) -
+ may lie under such layers at sites along
MANGANESE (Mn) -
+ 1 the margins of the Atlantic. Indeed, a
ZINC (Zn) -
+ strip of deposits may be concentrated
POTASSIUM (K) -
+ 2/3 in oceanic crust along the flow lines of
LI THIUM (Li) -
+ 10 sea-floor spreading from an enduring
RUBIDIUM (Rb) -
+
hydrothermal system such as the one
BARIUM (Ba) -
+ 2/3
at the Atlantis II Deep.
CALCIUM (Ca) -
+ 1
Evidence that deposits of this type
SILICA (SiO,) -
+ 2/3
exist was obtained by the ship Discover
MAGNESIUM (Mg) + -
1
1 er of the National Oceanic and Atmos
SULFATE (SO,) + -
SODIUM (Na) + -
VARIABLE
pheric Administration in 1972. The
CHLORINE (CI) ? ? VARIABLE ship was making a transect to charac
terize the sea floor across the Mid-At
lantic Ridge at the latitude of Miami.
EXCHANGE OF CHEMICALS between circulating seawater and the basaltic rock of the
On the east wall of the axial valley
mantle is charted. The temperature is assumed to be between 200 and 400 degrees Celsius.
A minus sign means the basalt or the seawater is a source of the element, a plus sign that it
of the Mid-Atlantic Ridge the ship un
is a sink. The column at the right shows the ratio of the amount of the material introduced expectedly dredged a black, layered,
to the oceans by hot springs at oceanic ridges to the amount introduced by rivers flow crumbly specimen of rock. It was ex
ing off the continents. With lithium, for example, ridges deliver 10 times more than rivers. tremely pure manganese oxide (50 per-
88
© 1985 SCIENTIFIC AMERICAN, INC
M
posits produced by the discharge of ore bodies, called massive sulfide bod ineralization in the setting of an
low-temperature hot springs at the ies, on land. early stage of the opening of
TAG field lie on top of the volcan The combination of such deposits an ocean basin associated with an in
ic rocks of oceanic crust. The walls below the sea floor and layered manga termediate-to-fast spreading center is
of large fracture zones that offset seg- nese oxide deposits on the sea floor represented by sites in the axial zone
1.5
(j)
a:
w
f
W
::;;:
o
...J
g
I
f
a.
w
o
2
BASALT
t
0 2 22
DISTANCE (KILOMETERS)
1,700
RED SEA sea-floor mineralization typifies
AXIAL BASIN
one kind of setting in which the process
takes place: an early stage of the opening of
MnO" FE(OH), an ocean basin, associated with a slow sea
1,800 floor spreading center. The crust of the Red
1't Sea basin contains thick beds of rock salt
CaSO.
formed under earlier conditions. Seawater
(j)
a:
w
1,900 '--tf penetrating down through the crust becomes
unusually salty; the salt both enhances the
f-
W DENSITY-STRATIFIED HOT BRINE ability of the solution to transport metals as
6 complexes with chloride from the salt and
I increases the density of the solution. The
f-
a. 2,000 result is a density-stratified, metal-rich hot
w
0 brine that forms ponds in axial basins as is
shown at the left. The reactions represented
are the precipitation of metals as copper
2, 100 iron-lead-zinc sulfide particles under reduc
ing (no free oxygen) conditions in the lower
layers of the hot brines and as sulfate parti
cles under oxidizing conditions where the
2,200 upper layers of the brine mix with seawater.
9 10 11 12 13 14 15 16 The particles settle to the sea floor to form
DISTANCE (KILOMETERS) layered deposits of metalliferous sediment.
89
© 1985 SCIENTIFIC AMERICAN, INC
METALS, METALS,
METALS HYDROCARBONS METALS HYDROCARBONS
if)
IT:
W
f-
w
::?:
o
....J
g
�
0..
200
W
o
300
SETTINGS OF SEA-FLOOR HOT SPRINGS are shown in this spreading center (the East Pacific Rise), a volcanic island arc above
schematic cross section of the Pacific Ocean. The settings are a a subdnction zone at a convergent plate boundary and a marginal
90
91
ANCIENT MINERAL DEPOSIT" originally formed by sea-floor in sediment that originally formed in a sea-floor setting. The gray
hot springs and now on land account for a number of major sour lines indicate ophiolite belts: material that was generated at for
ces of ore (colored circles). They include deposits in volcanic rock or mer sea-floor spreading sites and was later lifted onto a landmass.
92
by Gerald A. Rosenthal
ven a casual consideration of the able or unsuitable as a source of food. Adult P. polyxenes avoid plants of
94
© 1985 SCIENTIFIC AMERICAN, INC
ABNORMAL PUPA with three heads develops when the larva of val exoskeleton, each time developing a new head capsule. The re
the fall armyworm eats an extract of Ajuga remota, a bugleweed. sulting extra head capsules block the function of the mouthparts,
The plant produces phytoecdysones: compounds that mimic normal and the larvae starve. Mimicking a predator's natural hormones is
growth hormones, called ecdysteroids, of the larva. Phytoecdysones one of the more sophisticated forms of chemical defense practiced
cause the insect to undergo the cellular events that normally pre by plants. Other methods include poisoning the herbivore or making
cede metamorphosis several times without actually shedding the lar- substances that can deter it from feeding or from depositing its eggs.
95
© 1985 SCIENTIFIC AMERICAN, INC
a b
c d
I/'
�
FLOSSFLOWER
PREMATURE
PUPATION
e
>
>
>
>
>
>
HIGHER PLANTS' CHEMICAL DEFENSE METHODS may usually flee from a pheromone-releasing plant as they would from a
range from the simple to the elaborate in their mode of action. predator. Other plants defend themselves by blocking the function
Some plants, such as the thistle Parthellium hysterophorus (a), man of an insect's biochemicals. For example, the flossflower, Ageratum
ufacture repellent chemicals that discourage predators from feeding houstoniallum (d ), manufactures a substance that blocks insects' ju
or laying eggs. Others, such as the bird's-foot trefoil (b), manufac venile hormone, killing larvae by forcing them to molt prematurely.
ture lethal toxins; some bird's-foot trefoil plants produce cyano Many plants do not keep permanent stores of defensive chemicals
genic glycosides (compounds consisting of sugars bound to cyanide but manufacture them only in response to predation. When the to
complexes) and enzymes that liberate cyanide from the cyanogenic mato (e), for example, is attacked by a chewing insect, it is stimulat
glycosides. More complicated defensive strategies consist in pro ed to manufacture proteinase inhibitors, macromolecules that inhib
ducing natural substances that mimic substances normally produced it the insect's digestion of many plant proteins. A more exotic kind
by a herbivore, as A. remota mimics natural growth hormones (see of inducible defense may also occur. When a Sitka willow (J), is at
illustration 011 preceding page). Another such plant is the wild pota tacked by certain caterpillars, the nutritional quality of the leaves
to Solallum berthaulthii (c), which synthesizes an ingredient of the of neighboring willows-even those that have not been attacked
aphid's alarm pheromone (the substance an aphid can release when deteriorates. Perhaps the tree emits a substance that acts like a pher
under attack to warn others that an insectivore is present). Aphids omone, warning other willows to prepare for an attacking herbivore.
96
© 1985 SCIENTIFIC AMERICAN, INC
97
© 1985 SCIENTIFIC AMERICAN, INC
98
© 1985 SCIENTIFIC AMERICAN, INC
�:��
win, then at Dartmouth College. It has
been established that when the Sitka
willow, Salix sitchellsis, is attacked by
insects, its leaf quality (a measure of
its suitability as a food resource for MONARCH BUTTERFLY BLUE JAY
insects) deteriorates. Rhoades noted,
however, that the leaf quality of near
by willows-even willows that had not
been attacked-also seemed to dete
riorate. Perhaps, Rhoades suggested,
the attacked tree produces a signal,
analogous to an insect's alarm phero VICEROY BUTTERFLY
mone, that travels through the air to MILKWEED PLANT
induce defensive responses in neigh
boring, unattacked trees. b
Schultz and Balwin tested Rhoades's
hypothesis by planting seedlings of
the sugar maple, Acer saccharum, in
two separate growth chambers. They
found that plants that had been dam
aged intentionally, as well as plants in
the same growth chamber as the dam
aged ones, tended to produce great
er quantities of tannins and phenolics,
two defensive compounds, than plants
grown in a separate chamber. Al
though these studies have not defini
tively proved the existence of commu
nicative chemical defense by plants,
they have generated great interest in at
tempts to prove that pheromonal com
MILKWEED PLANT
munication between trees may actual
ly take place.
c
trol of herbivores. It should be possible Poekilocerus bufollius (b) incorporates the cardenolides it ingests when eating milkweeds
into the poisonous spray it uses to protect itself from predators. Larvae of the bruchid bee
to exploit natural compounds that de
tle Caryedes brasiliensis (c) feed on seeds of the leguminous vine Dioclea megacarpa. The
ter feeding or the deposition of eggs,
seeds contain large amounts of L-canavanine, an amino acid that is not incorporated into
or even to grow crops from which the proteins and that is often a potent insecticide. The beetle's larvae, which live inside the
natural substances that attract herbi seed and feed on nothing else throughout their larval life, are able to obtain ammonia from
vores or stimulate their feeding behav canavanine. They then incorporate the nitrogen of the ammonia (nitrogen that was origi
ior have been eliminated. nally part of the canavanine molecule) into amino acids to be incorporated into proteins.
99
© 1985 SCIENTIFIC AMERICAN, INC
diocarbon dating is the principal of modern man and was accompanied Fortunately the original percentage
100
SAMPLE WHEEL for the radiocarbon accelerator in the authors' placed in the groove at the outer end of each tab. The wheel (shown
laboratory carries 20 metal tabs, arranged like petals, that hold car
104).
at about its actual size) is loaded into the accelerator's ion source,
bon samples from specimens to be dated (and samples of known age where each sample is ionized in turn (see illustration on page
or carbon-14 content for calibration). The carbon is deposited as a Carbon-14 atoms are sorted out from among the ions thus produced
thin layer of graphite on a short length of tantalum wire, which is and are counted; their number is a function of the specimen's age.
10 1
OJ
PRESENT
OJ
5,000 10,000 20,000 30,000
AGE (YEARS)
100
f='
z
w
()
a:
w
�
CJ
�
z 50
�
�
w
a:
z
w
CJ
�
--'
--'
0
()
5 10 15 20 25 30
THOUSANDS OF YEARS
QUANTITY OF BONE needed for radiocarbon dating depends on bow deposited in tbe bone after it became buried), wbicb would re
tbe bone's actual age, its state of preservation and tbe dating metb sult in underestimation of ages, wbat is measured is tbe carbon-14
od employed. Tbe relative amount of mass required for dating by content of purified amino acids derived specifically from a single
accelerator mass spectrometry (colored cubes) is 1,000 times less component of bone: tbe protein collagen. Because collagen deterio
tban tbe mass needed for dating by conventional radiocarbon metb rates in tbe course of time (bottom), a larger sample is needed for
ods (larger cubes), regardless of tbe bone's age. To avoid measur older bones. Tbe curve's fuzziness reflects tbe fact tbat tbe percent
ing tbe carbon-14 content in "modern" carbon (carbon tbat is some- age of collagen can vary, depending on tbe condition of preservation.
100 �---
-,
-.-
--.-
--
NEUTRON + NITROGEN-14
t
CARBON-14 f='
z
w
()
t
CARBON DIOXIDE
a:
w
�
t CJ
z
z
------���r-------------------+-----------------�
� 50 r-------
�
w
a:
""
Z
o
tIl
a:
�
()
CARBON 14, a radioactive isotope of carbon, is produced in trace balanced by tbe continuous loss tbrougb radioactive decay (right),
amounts wben cosmic-ray neutrons interact witb nitrogen atoms in establisbes a fairly constant C-14 concentration in all plants and
tbe atmospbere (left). Tbe C-14 tben combines witb oxygen to form animals. Wben an organism dies, bowever, tbe C-14 in its tissues
carbon dioxide. Like normal carbon dioxide (containing carbon 12), can no longer be replenisbed and tbe radioactive disintegration
tbe radioactive kind enters tbe biospbere tbrougb plant respiration, dominates. Because tbe C-14 concentration in living organisms is
and it undergoes tbe same carbon cbemistry. By working its way up known, tbe age of organic remnants can be derived by measuring
tbe food cbain C-14 is incorporated into tbe tissues of bigber organ tbe C-14 content in a sample of tbe remnants and calculating bow
isms as well. Tbe continuous infusion of C-14 into tbe biospbere, mucb time must bave elapsed to account for tbe observed decline.
102
: .. ..
. . .
103
VELOCIT Y FILTER
MAGNETIC
LENS
ION SOURCE
HIGH
VOLTAGE
SOURCE
STRIPPER
CHARGED TO
+ 2.5 MILLION VOLTS
LENS
---� ELECTROSTATIC
LENS
\-Eo('---NEGATIVE
-- IONS ------�>\-E(---P
--- OSITIVE IONS--------:,..
COLLECTOR ELECTRIC x
/ PLATES
�
FIELD x
x
x
x
! �x
x
x
/
- 20,000 VOLTS
.. x
�x
+400 VOLTS x
x
� . .
.
..
.
. .. .
.
: G AS
x
�x
x
x
• •
. .. .
•
•-
..
• c
.,.
,.
x
�x
}i . x
PLASTIC WINDOW
NITROGEN-14 it CARBON-14
CARBON IONS VACUUM
RADIOCARBON ACCELERATOR (top) accelerates carbon ions cesium ions to the carbon sample and the carbon ions away from it.
by applying a positive charge of 2.5 million volts to the stripper. The C-14 ions are finally distinguished from the remaining contami
The imparted charge first "pulls" the negatively charged ions to the nating ions by a detector (bottom right). After passing through a
stripper and, after the stripper converts them into positive ions, plastic window the accelerated ions travel through a gas-filled
then "pushes" them the rest of the way. Most of the contaminating chamber, where they produce plumes of secondary ions. Because
ions are filtered out by two beam-bending magnets. These magnets the velocity and atomic charge of the incoming ions are similar, the
separate the ions by mass: lighter ions tend to turn more sharply rate of interaction between the ions and the gas molecules depends
(col
which have a nuclear charge of +6, interact at a lower
than heavier ones. Ions of atomic mass 14 can therefore be singled primarily on the nuclear charge of the ion. Hence carbon ions
out by placing a slit at the proper radius of curvature downstream ored dots),
from the magnet. In the ion source (bottom left) carbon atoms of the rate than nitrogen ions (black dots), which have a nuclear charge of
sample to be tested (deposited as a thin layer on a tantalum wire) +7, and typically travel farther down the length of the chamber
are bombarded with positive cesium ions to produce negatively before ionizing the gas. An electric field causes the ion plumes to
charged carbon ions. The electric charges applied to the metal plate drift over to collector plates, where their average lengths are then
and cylinder (shown here ill cross sectioll) in the device direct the measured. From these measurements the C-14 ions are identified.
104
© 1985 SCIENTIFIC AMERICAN, INC
most of the problematic nitrogen-14 a slit downstream of the magnet (dots) therefore exhibit a distribution of masses, charges
105
© 1985 SCIENTIFIC AMERICAN, INC
SOIL CRACK
ANT OR TERMITE
ACTIVITY
FLECKS OF CHARCOAL
..;= i�!g!fZ{SF\ ..
SEED OR GRAIN
I \
LOOSE SAND ANIMAL TEETH
CHARCOAL IN /
BAKED CLAY
()�
�;:.,�
-I.", ;(, ,.-
l 4 r' . �
INSECT REMAINS
SHERD SEALED IN RUBBISH PIT IN LAKE MARL
STRATIGRAPHIC CONTEXT in which ancient organic remains, such as those shown, is likely to have dispersed the organic material.
such as bone, teeth, charcoal or seeds, are embedded must be care· Sound contexts (bottom) include those where the material to be car
fully examined before the results of their carbon dating can be con bon-dated has been sealed in an immobilizing matrix. Because the
fidently generalized to the strata in which they lie. Contexts of sealed material is likely to be of small size, accelerator mass spec
doubtful validity (top) include those where one or another agent, trometry is the technique best suited to date the strata by context.
106
© 1985 SCIENTIFIC AMERICAN, INC
Another aspect of selectivity that ac atively minor improvements in sam • Excellent for weight
control and body
celerator dating offers is in the sample ple chemistry can lead to sharper and tone.
treatment. Because far less carbon is more extensive chronologies. For ex
needed for a measurement, one can af ample, accelerator dating of purified
ford to be much more'discriminating amino acids from bones more than
in choosing the carbon compound to 25,000 years old showed that their age
be tested. Parchment, like the collagen had previously been consistently un
of bone, can be made to yield an un derestimated by 1,000 or more years.
contaminated sample of carbon from The earlier, conventional measure
its amino acid constituents. A piece of ments had been based on whole colla
ivory can be similarly dated by meas gen, which cannot be guaranteed to be
uring the C-14 content of the amino free of modern carbon.
acids extracted from its dentine. In One controversy of long standing on
this way the extinction date of the which the new dating technique has al
mammoth in western Europe could be ready had a major impact concerns the
worked out much more closely from first human migrations to the New
small pieces of tusk preserved at Pa World. An accurate time scale for
I
the past 30,000 years in areas where such as Canada and Peru, that seemed
Audio-Forum's
I
You can learn to speak Japanese on your
organic preservation has been favor to be considerably older than expect own with self-instructional
I
Course consists of a series of cassettes and
mass spectrometry can push the thresh other such finds are made in the fu
old further back in time, however. ture, accelerator mass spectrometry I accompanying textbook. Your cassette player
becomes your "teaching machine:' You leam
At a horizon of 40,000 years the will quite probably be the dating tech
I
where and when you want to. and at your
amount of carbon 14 in a bone or nique called on to determine the skele own pace. Available in two parts. each is I
• I
a piece of charcoal can be truly mi ton's age directly. shipped in a handsome library binder.
nute: �uch a specimen may contain Accelerator dating may also prove D Beginning Japanese, Part J. 8 cassettes
(11 'Iz hr.) and 409,p. text. $135.
D Beginning Japanese, Part 2. 16 cassettes
only a few thousand C-14 atoms. Con to be invaluable in establishing an
(19 hr.) and 410-p. text. $195.
sequently equally small quantities of
modern carbon can severely skew the
authoritative chronology of Neander
thal man. The archaeological evidence I (CT residents add sales tax)
I
measurements. Contamination of this
kind amounting to 1 percent of the
available indicates a rather abrupt
disappearance of this human subspe •I Phone orders call toll-free: 1-800-243-12341 I
carbon in a sample 25,000 years old
would make it appear to be about
cies at the beginning of the Upper Pa
leolithic, about 35,000 years ago. Al
I
To order, clip this ad and send with your
name and address. and a check or money I
I
order - or charge to your credit card (VISA.
1,500 years younger than its actual
age. Such contamination would, how
though it is too early to be sure, dating
by accelerator mass spectrometry may
I MasterCard. AmEx. Diners) by enclosing
card number. expiration date. and your sig-
ever, reduce the apparent age of a reveal that the Upper Paleolithic and
I w=.Unconditionally guaranteed. Try it for
I
60,000-year-old object by almost 50
percent. Clearly proper sample-decon
the period preceding it, the Middle Pa
leolithic, were unduly compressed by I
three weeks - if you are not fully satisfied.
we'll refund every penny you 've paid.
I
tamination procedures are of partic
ular importance in the dating of very
the limitations of conventional dating
methods. If this was the case, there I
We ofTer 120 courses in 42 languages:
send for free catalog. I
iIIUCla·�alWm@
\ I
old artifacts. Unfortunately stringent would have been much more time for
specimen handling and treatment pro the Neanderthal's disappearance than Room X26. 96 Broad SI.. Guilford. CT 06437
cedures ultimately result in relative- has commonly been supposed. �-----.,
107
© 1985 SCIENTIFIC AMERICAN, INC
T
he members of many animal individuals (the workers) labor in be er in social circumstances, recognition
species form social groups. half of a single fertile individual (the can come about through the mecha
There are flocks of birds, schools queen) that is their genetic sister. Why, nism of social learning, or familiarity.
of fish, troops of apes. The question however, should animals favor their That is, individuals of the same fami
arises: May some of these groups be kin, considering that such behavior ly group learn to recognize one anoth
based on family ties? In other words, may often be costly to the individual? er early in life. A second mechanism,
are some animals able to recognize Much of the current understanding phenotype matching, occurs if an indi
kin? Do they capitalize on this ability of the evolution of altruism derives vidual learns and remembers a specif
in order to associate with kin rather from the work of W. D. Hamilton of ic characteristic of itself or of its rela
than with unrelated individuals and the University of Oxford. His theories tives. Such a characteristic might be an
also to avoid mating with close rela to explain altruistic behavior predict odor, a plumage color or a particular
tives? Kin recognition has recent that individuals will aid or cooperate marking. The individual can then iden
ly been the focus of a number of stud with close relatives rather than with tify others as kin or nonkin by noting
ies by animal behaviorists. The studies distant relatives or unrelated individu similarities and differences between
lead eventually to consideration of an als because to do so gives them an evo the characteristic it remembers and
evolutionary approach to understand lutionary advantage. The more closely that of an unfamiliar individual. Phe
ing cooperation and altruism among related individuals are to one another, notype matching is fundamentally dif
animals. Before these theoretical im the more genes they have in common. ferent from recognition based on fa
plications can be considered, however, An altruist will aid relatives, even if miliarity because itenables an individ
one needs to know whether members such behavior may be risky for the ual to recognize unfamiliar animals.
of a particular species do indeed recog individual, because by doing so the The third mechanism, which de
nize their kin. If they can recognize kin, altruist is increasing the probability pends on specific "recognition genes,"
how do they do it? that genes identical with its own will also enables an individual to recognize
For the past several years we have be passed on to future generations. unfamiliar animals. This hypothetical
been investigating kin-recognition be Hamilton's term for the reproduc mechanism of kin recognition is, how
havior in tadpoles of several anuran tive success an organism shares when ever, a purely genetic one; no learn
(frog and toad) species that abound a genetic relative survives and repro ing is involved. Recognition genes (or,
in the lakes and ponds of the Pacific duces is "inclusive fitness." If relatives more specifically, recognition alleles,
Northwest. We find that these tadpoles live nearby (perhaps because dispersal which are alternative forms of genes)
do recognize their siblings, and that from birthplaces or hatching sites is are expressed as a phenotypic charac
in one species in particular the ability low), specific recognition of kin may ter such as odor. Individuals carry
to distinguish between kin and nonkin not be necessary for what is in effect ing copies of such an allele recognize,
is extremely sensitive. Tadpoles of the kin-directed behavior to operate. If in and so tend to favor, other individuals
Cascades frog, Rana cascadae. are able teractions with nonkin are also likely, carrying the same allele. It is impor
to recognize their siblings even if they however, the ability to recognize kin tant to remember that any of these
have never seen them before. They could prevent the misdirection of help mechanisms of kin recognition may
prefer to associate with their siblings ing behavior or ensure that the cooper operate alone or in conjunction with
rather than with nonsiblings, and this ative effort is exerted in behalf of the one another.
preference persists through their meta proper individual, namely a close rela
108
© 1985 SCIENTIFIC AMERICAN, INC
I
uals. Moreover, because one of us animal behaviorists often ask when ap n the first regime individuals were
(O'Hara) had studied the natural histo proaching any behavioral study for the reared in aquariums with members
ry of these frogs, we knew that Cas first time: How does this behavior de of their own clutch only, so that they
cades frog tadpoles do not disperse velop? What is the sensory basis of this were exposed only to their siblings. In
very far from their hatching site. This behavior? What is its adaptive value? the second regime we reared an equal
suggested there must be ample oppor We began by collecting clutches, or number of tadpoles from each of two
tunity for kin to interact with one an egg masses, of Cascades frog eggs laid different clutches on opposite sides
other; it was possible that the aggrega by different females and fertilized of an aquarium partitioned by a
tions we saw were composed mostly of by different males and brought them central mesh screen. Water could flow
kin. We speculated that kin recogni into the laboratory. (A Cascades frog freely between the two groups, allow
tion might give Cascades frog tadpoles clutch contains between 400 and 800 ing members of both groups to ex
the ability to form and maintain such eggs; all tadpoles hatched from a single change chemical as well as visual cues.
cohesive groups. clutch are siblings.) We also collected We call this a mixed-rearing regime
For an inquiry into the nature and mating pairs of frogs and allowed egg because members of two different kin
the mechanism of kin recognition we laying and fertilization to take place in groups are essentially reared together.
needed to design an experimental pro the laboratory. We then subjected de In the third regime we removed new
tocol that would let tadpoles choose veloping embryos to one of four differ- ly fertilized eggs from a clutch and
TADPOLES of the frog species Raila cascadae form small, co they choose to associate with siblings. This preference, which re
hesive schools in a pond in the Cascade Range of Oregon. Appar mains with an individual through metamorphosis into a frog, means
ently these tadpoles can distinguish siblings from nonsiblings, and it is likely that the schools of these tadpoles consist mostly of kin.
"The difference (in this and Writing from their own distinguished contributions, the authors of
other hook series) comes in Scientific American Library-Steven Weinberg, Christian de Duve, Philip
the quality of the authors, who Morrison, Julian Schwinger, Solomon Snyder, and John Archibald
comhine eminence in their
W heeler, to name a few-offer you the understanding you will need to
.fields with prol'en g!fts of
follow the on-going work-the controversies and breakthroughs-across
e::tposition. With the handsome
the frontiers of science.
1'Olumes in front of us ...
what re�ponse can there he • Each volume covers a topiC of major interest with immediate relevance
hut immediate surrender to on-going research.
to this.firepower?"
• Each volume carries the authority of an author who has made a signifi
NIGEL CALDER, NATURE
cant personal contribution in his field.
• Each volume proves in its turn that there is no topic in any field of
science that cannot be explained to readers who want to understand.
No Prepayments R<:guired
Payment for each volume is due only after you have decided to retain
it for your library.
No Minimum Purchase
From year to year, you may choose to keep as many or as few volumes
as you like.
You may purchase any volume at the low members' price of $24.95
-well below the bookstore price.
No Strings Attached
You may cancel your membership without notice, for any reason, with
no penalty.
Enroll now!
SCIENTIFIC AMERICAN LIBRARY
PO. Box 646 Holmes, PA 19043
I
n his famous essay on the two cultures, C. P Snow made the
no better than to perose this fas Second Law of thermodynamics the litmus test of scientific liter
cinating new volume by Peter acy: Not to know the Second Law is the same as not haVing read a
Atkins .. .. Even those who feel work of Shakespeare.
they already know all about the This is the law, of course, that explains why hot objects grow cool
Second Law can hardly fail to whereas cool objects do not spontaneously become hot; why a
find new angles and fresh in bouncing ball must come to rest and a resting ball cannot, of its
sights into what is going on in own, begin to bounce. To some people, the Second Law conjures up
the world around them. I hope visions of clanking steam engines, intricate mathematics and in
that this remarkable book will comprehensible physical theory It is better known to many in its
he widely read by scientists as restatement as Murphy's second law: "Things, if left alone, will grad
well as by the non-scientists for ually go from bad to worse!"
whom it is primarily intended." That restatement captures, in its way, the universality and power of
NEW SCIENTIST the law that inspired Arthur Eddington to call it "the arrow of time:'
In The Second Law P W Atkins, of Oxford University, breaks
'1\ lovely hook, beautifully illus through the mathematical barriers to understanding of the law. By
trated and presented, and vivid example he shows it at work in heat engines, refrigerators, and
clearly commensurate with its heat pumps; the cosmos, in the ecosphere and in the living cell; in
companion volumes in the Sci the physical, chemical, and life processes that drive the changing
entific American Library. It world in all its richness and variety
should engender affection as a A special appendix equips readers who have computers with a kit
sophisticated latterday Hogben, of tools to put the Second Law to work.
complete with social allusions 230 pages, 104 illustrations
implicit in the call to conserve,
P.WATKINS is a lecturer in physical chemistry at the University of Oxford and
not energy, but entropy, and to a fellow of Lincoln College. He is the author of numerous distinguished
use the vir tuous entropy works in his field, including Physical Chemistry, Second Edition, The Crea
conserving heat pump rather tion, and Molecular Quantum Mechanics, Second Edition.
than crude and p r ofligate
combustion. ..."
NATURE
, ,
time). The number of tadpoles that
spent more of their time near one end
of the test tank than they did near the
other was also recorded as an addition
al measure of preference. Appropriate
control tests revealed no biases in our
procedures or apparatus.
Our first series of experiments estab
b MESH
lished the basic fact that Cascades frog
�f �
\ tadpoles can distinguish between sib
, 11 � lings and nonsiblings. Some 80 per
\ !
", J
cent of the tadpoles reared only with
\
full siblings spent more time-a statis
tically significant amount more-near
� "-
"... familiar siblings than they did near
'-
---....
� unfamiliar nonsiblings. Tadpoles test
ed both early (from 1 1 to 19 days af
,
)
"- ter hatching) and late (from 34 to 40
\1- \....
�'-' �h
days after hatching) in development
displayed this preference.
,y-
T the
he next step was to try to determine
d MESH CYLINDER influence of familiarity with
other tadpoles on kin-recognition be
havior. This might yield insights into
the development of such behavior in
the species and perhaps give us a
handle on the mechanism controlling
the behavior. What we did was to test
three groups of tadpoles: those from
mixed-rearing regimes, those reared
with nonsiblings only and those reared
in total isolation. After being reared in
FOUR REARING REGIMES expose developing tadpoles to different environments before one of these regimes for an average of
they are tested for the ability to recognize kin. In one case embryos taken from a single egg
four weeks, test tadpoles were given a
mass are raised together in an aquarium, so that the tadpoles are exposed only to siblings
choice of associating with either an un
(a). In a "mixed rearing" regime (b) embryos from two different egg masses are raised on
each side of a mesh screen. Members of the two groups can exchange visual and chemical
familiar full-sibling group or an unfa
cues and thus are exposed to both siblings and nonsiblings. In a third regime a fertilized egg
miliar nonsibling group.
is put in an opaque dish (c), so that the tadpole is raised in isolation. In the fourth regime The results from these experiments
a single fertilized egg is placed inside a mesh cylinder (d ), with 12 embryos from a dif showed us that, like tadpoles reared
ferent egg mass outside the cylinder: the enclosed tadpole is exposed only to nonsiblings. with full siblings only, tadpoles reared
in a mixed-rearing regime also spent
most of their time near their siblings
enclosed each egg in an individual ing regimes had on tadpole kin recog rather than near nonsiblings. What was
opaque container so that the tadpole nition we tested the four sets of tad particularly significant in this series
would be reared in total isolation from poles by allowing them to choose to of experiments was that test tadpoles
other tadpoles. In the fourth rearing associate with one or the other of two could distinguish between unfamiliar
regime we enclosed a single newly fer groups of stimulus tadpoles. The stim siblings and unfamiliar nonsiblings.
tilized egg in a mesh cylinder posi ulus animals varied in their relatedness Tests of tadpoles reared with only non
tioned at the center of a rearing aquari to and familiarity with the test individ siblings gave the same result: they pre
um. We then placed developing em uals. Our standard tests were carried ferred to associate with unfamiliar sib
bryos from another clutch in the water out in a rectangular tank divided into lings rather than with unfamiliar non
surrounding the test embryo. Individ two basins by a central watertight, siblings. It appears, then, that exposure
uals reared in this regime were thus opaque barrier. We could test two tad to unrelated tadpoles during develop
chemically and visually exposed only poles simultaneously but independent ment does not alter the preference for
to unrelated tadpoles prior to being ly on opposite sides of the barrier. The association with siblings. Moreover,
tested. In summary, the four regimes ends of the tank were partitioned with early familiarity with siblings is not a
allowed us to work with four different plastic mesh to create small chambers prerequisite for identifying them later
sets of tadpoles: those reared with sib in which stimulus tadpoles (usually 25 in life.
lings only, those reared with both sib of them) swam. The test tadpoles could Finally, there were the tadpoles
lings and nonsiblings, those reared in see the stimulus animals; exchange of reared in total isolation, which had
total isolation and those reared with chemical cues was possible, but there never associated with, or even seen,
nonsiblings only. could be no tactile contact. other free-swimming tadpoles before
To see what effects the various rear- In a series of repeated experiments we tested them. We found they too
1 12
© 1985 SCIENTIFIC AMERICAN, INC
[]
sensitive kin-recognition system. The
LJ
fact that test tadpoles preferred mater
nal to paternal half siblings suggests
that maternal cues exert a stronger
..
influence on recognition. Perhaps the
, ..
reason is that the mother contributes
. ....
[ill
more to the developing embryo (cer .
� -::.-:
tain cytoplasmic factors in the eggs,
for example) than the father, whose
contribution is primarily genetic infor
mation. Still, the fact that paternal half
siblings are distinguished from nonsib
lings suggests that at least some part of
TADPOLE PREFERENCE TESTS are done in an aquarium 1.2 meters long. In the stan
kin recognition in this species is related dard test (top) the aquarium is divided by an opaque central barrier so that two tadpoles can
to a genetic factor. be tested simultaneously but independently. Plastic-mesh partitions create two end cham
bers, each of which is occupied by 25 "stimulus" tadpoles: either siblings or nonsiblings of a
W
hat can be the value for Cas test animal. The individual being tested can choose to swim near either of the two groups of
cades frog tadpoles of any kin stimulus tadpoles. The amount of time it spends near each group is recorded. Most of the
recognition system at all, let alone tadpoles from all four of the rearing regimes chose to spend more time swimming near their
such a sensitive one? The most obvious siblings than near nonsiblings. The sensory basis of kin recognition is determined by modi
fying the standard test tank in two ways. Watertight glass panes that replace the mesh end
possibility is that kin recognition is the
partitions (Illiddle) allow the test tadpoles to see the two stimulus groups but prevent the
basis of these tadpoles' ability to form
exchange of chemical cues. Under these conditions there is no preference for siblings. When
small, dense schools. Some advantages the stimulus groups are enclosed in boxes fitted with an opaque front facing the test ani
of living in schools rather than alone mals but with perforated side� and tops (bottolll), the test tadpoles cannot see the stimulus
are an enhanced ability to find and groups, but they can still sense chemical cues. Under these conditions test tadpoles showed
share a limited food supply, to avoid a preference for their siblings, indicating that kin recognition is mediated by chemical cues.
113
© 1985 SCIENTIFIC AMERICAN, INC
700 60
600 - -
50
500 - - f---
(fJ
--'
(jJ
<l:
:::J
0
40 r----
0
6400 - - r----
'>
is
30 f---
() �
w
lJ..
�
w 300 - - r----
0
a:
:::; w
i= (])
:::; 20 r---- -
200 - - f--- :::J
Z
100 - - 10 r---- -
0
STANDARD CHEMICAL-CUE VISUAL-CUE STANDARD CHEMICAL-CUE VISUAL-CUE
TESTS TESTS TESTS TESTS TESTS TESTS
KIN-PREFERENCE BEHAVIOR of R. cascadae tadpoles was the standard and the chemical-cne tests the tadpoles spent an aver
studied by allowing tadpoles to choose to swim near siblings or near age of 660 seconds at the end of the tank close to kin, but when
nonsiblings. The average number of seconds tadpoles swam at the the tadpoles had to rely on visual cues, the time spent near siblings
end of the test tank nearest their siblings is shown (left) for the dropped to the random level. A total of 60 individuals were ob
standard test (in which the tadpoles have the full range of their served in each test; the number that spent more of their time near
senses) and for the chemical- and visual-cue tests. Testing was for a siblings than near nonsiblings is another measure of preference
total of 1,200 seconds (20 minutes), so that 600 seconds spent on (right). In the standard and the chemical-cue tests some 75 percent
each side of the tank would represent random behavior. In both of the individuals tested spent more of their time near their siblings.
DEVELOPMENT of tadpoles into frogs proceeds over a span of at which individual tadpoles reach a particular stage is extremely
from six weeks to more than eight weeks. Four representative stag variable.) R. cascadae displays kin-recognition behavior at each of
es (in a continuum of 26 stages) in this process are shown. (The age the four developmental stages, including the froglet stage (right).
PERFORATED PLATE
I
u��;�"�_
", _____________
MESH BARRIER
�
�
, c,,�
__ �__________________��m
t?"<.ll=ll
l _
TEST CHAMBER was designed to determine whether kin recogni mined by the amount of time the test frog spends near the end of
tion persists after metamorphosis. The chamber at one end of the the tube housing its siblings compared with the amount of time it
clear plastic tube holds siblings of the test frog in the tube; the spends near its nonsiblings. The test showed that R. cascadae con
chamber at the other end holds nonsiblings. Kin preference is deter- tinues to exhibit kin-preference behavior even after metamorphosis.
1J4
© 1985 SCIENTIFIC AMERICAN, INC
retaining it. It is also possible that the resulting schools was counted. When the results of a series of control runs were plotted,
they yielded a bell-shaped curve (top): most of the schools were roughly half blue and half
adults depend on kin-recognition abil
red, indicating that color in itself does not influence aggregation. For the actual tests,
ity either in reprod uctive behavior
members of one clutch were dyed blue and those of a different clutch were dyed red. When
(to avoid breeding with close relatives) the results of a number of tests were plotted, the color composition of the schools these
or in social organization, perhaps to tadpoles formed was seen to approach a "bimodal" distribution: whether the siblings had
find their home pond after having been been reared apart (middle) or together (bottom), most of the schools were predominantly
displaced. blue or predominantly red. In other words, most of these aggregations consisted of kin.
Laboratory experiments can only ap
proximate and simplify the complex
social and ecological conditions ani It was not an easy task to approach active kin-recognition system. Having
mals encounter in nature. For this and collect aggregations of tadpoles. If shown that to be true, we looked at
reason during three consecutive sum they detected us moving toward them, other anuran species displaying differ
mers we did experiments in natural they would disperse rapidly. Eventual ent behavioral characteristics.
ponds and lakes inhabited by Cascades ly, however, we did determine that red Red-legged frogs (Rana aurora) are
frog tadpoles. We got tadpoles for field and blue tadpoles from the same kin closely related to the Cascades frog.
experiments by collecting egg masses group aggregated randomly in control When we put them through a series of
from natural breeding ponds and rear ponds: dye color had no effect in 103 standard preference tests, however, the
ing the individuals in the laboratory aggregations we sampled. In contrast, results were very different. The red
either with siblings only or in mixed the distribution of red and blue tad legged tadpoles can identify kin only
rearing regimes. We marked tadpoles poles from different kin groups was early in the free-swimming stage, and
from two kin groups with different col not at all random: the 353 aggregations then only if they have been reared with
ors by immersing them in a harmless we sampled in the experimental ponds f ull siblings; they lose their ability to
red or blue dye. Then we intermingled were dominated by one of the two col recognize kin as they develop. Western
members of the two groups and re ors. The results of these field experi toad (Bulo boreas) tadpoles can rec
leased them in natural ponds in the ments lead us to believe that Cascades ognize kin through the tadpole stage,
Cascade Range. We considered the frog tadpoles can identify their kin, but only if they have not been ex
possibility that tadpoles might form and prefer to associate with them, in posed to nonkin. If they have been
groups based on their new color rather nature as they do in the laboratory. exposed to nonkin, they rapidly lose
than on kinship. As a control we there their ability to identify kin.
W certain
fore divided some single kin groups, e had originally speculated that Our results with Western toads are
dyed half of the members of each aspects of the behav similar to those reported by Bruce
group red and the other half blue and ior we had noted in Cascades frog Waldman of Cornell University, who
released the groups in control ponds. tadpoles might be connected with an investigated tadpoles of the closely re-
1 15
© 1985 SCIENTIFIC AMERICAN, INC
o RANA AURORA
nestmates. Richard H. Porter of Van
derbilt University has demonstrated
o BUFO BOREAS
that spiny mice (Acomys cahirinus) that
680 ,--
are artificially "odorized" prefer to
associate with unfamiliar individuals
that have been given the same odor.
660 - - Clear cases of a recognition-gene
mechanism are harder to establish, but
(jJ recent studies of mating preferences in
0
Z - the house mouse (Mus musculus) have
0
() 640 yielded strong evidence that is consis
w
� tent with a genetic recognition system
w [see "The Chemosensory Recognition
::0;;
i= of Genetic Individuality," by Gary K.
620 , r== Beauchamp, Kunio Yamazaki and Ed
ward A. Boyse; SCIENTIFIC AMERICAN,
July, 1985]. The results showed that
,-- mice can choose between potential
600 mates that differ at a single genetic site
in the Major Histocompatibility Com
plex (H-2), which specifies the antigens
580
responsible for distinguishing between
REARED WITH SIBLINGS REARED WITH SIBLINGS self and nonself. Male mice prefer to
AND NONSIBLINGS
mate with females whose H-2 type is
different from their own. The prefer
SPECIES DIFFERENCES in kin-recognition behavior are revealed when tadpoles of two
frog species and a toad species are given the standard kin-recognition tests. All three spe
ence is based on the genetically deter
cies prefer to associate with kin if they have been raised with siblings only; that is, they mined scent associated with particular
spend more than the random amonnt of 600 seconds at the end of the tank nearest siblings. H-2 alleles.
(This is true for early and late developmental stages of R. cascadae and the toad Bujo The sensitive kin-recognition system
boreas bnt is true only for the early larval stages of Raila aurora.) For tadpoles reared with of Cascades frog tadpoles may also be
both siblings and nonsiblings the amount of time spent with siblings drops to the random a genetic recognition system. One can
level in R. aurora and B. boreas; the preference for kin remains significant in R. cascadae. not, however, r ule o ut the possibility
that a Cascades frog tadpole, even one
reared in total isolation, may learn its
lated American toad, Bu/o americanus. izations. Olfactory cues are the basis of own genetically determined odor by
In laboratory experiments he found identifying kin in most species of in experiencing itself. It could then match
that American toad tadpoles general sects and mammals. It is now clear that the odors of two unfamiliar groups
ly cannot distinguish between kin and the commonest of the three mecha and compare them with its own odor;
nonkin if they are reared in mixed nisms for identifying kin is familiarity, the group having the more familiar
rearing regimes. It is likely that in which is efficient when there is a high odor would be taken to be more close
both toad species, and possibly in the probability that the individuals a par ly related. It may be impossible to de
red-legged frog, familiarity influences ticular animal interacts with are rela termine experimentally whether phe
kin identification. tives. Kin recognition is influenced by notype matching or an innate genetic
It was not surprising to learn that familiarity in the Western toads we ex recognition system is operating in Cas
red-legged frog and Western toad tad amined; familiarity has at least some cades tadpoles; it is conceivable that
poles do not have a well-developed role in the recognition abilities of most both are operating simultaneously. In
ability to recognize kin. The tadpoles of the birds and small mammals that fact, one proposition holds that the dif
of both species disperse rapidly from have been studied. Familiarity is prob ferences between the two mechanisms
their hatching sites. Moreover, red ably an important means of avoiding are trivial. Both mechanisms enable
legged frog tadpoles are not known to inbreeding in many species. an animal to identify an unfamiliar
aggregate, and Western toad tadpoles Although we are sure familiarity is relative, and so the evolutionary con
are usually found in huge schools (up not an essential part of the kin-recog seq uences are the same.
to a million tadpoles) composed of in nition mechanism in Cascades frog In addition to the theoretical impli
dividuals from hundreds of different tadpoles, we cannot be s ure whether cations bearing on familial altruism,
clutches. Probably tadpoles of these the mechanism is learned phenotype kin recognition may yield important
species have not had enough oppor matching or derives from innate recog practical benefits. Knowledge of how
tunity for interaction with kin during nition genes. It is often difficult to dis members of a particular species recog
development to have evolved a sensi tinguish between the two, although nize one another is critical in animal
tive kin-recognition system that allows there are some fairly clear-cut cases husbandry and in the propagation of
them to form cohesive groups. of phenotype matching. For example, certain endangered species, particular
Gregory R. Buckle and Les Greenberg ly where "familiarity" with potential
1 16
© 1985 SCIENTIFIC AMERICAN, INC
r----------------
Your subscription includes:
MEDICINE
SCIENTIFIC
the two-volume, 2,300-page loose-leaf text
.OIERICAN
and, each month, six to nine replacement chapters,
415 MADISON AVENUE, NEW YORK. N.Y. 10017
a newsletter, new references,
and a completely revised index.
Signature ___________________
Nmne ____________________
S�ml� ----
Addre� ___________________
Slale ---- Zip Code _____
--
Ci� ----
• Add sales tax for Calif.. III.. Mich., OJ' N.Y. Allow 8 weeks fOI" deliverv. Add US$lO
SCIENTIST
turned to block the light.
The direction of rotation of the
light's polarization is specified from
the perspective of the observer. The
S
olutions containing optically active inary line (the polarization axis) paral certain compound rotates the polari
compounds such as sugar rotate lel to the polarization of the light rep zation of light, one speaks of "specific
the polarization of light passing resents the effect of the filter. rotation." In Epstein's work this meas
through them. The rotation reveals If a second polarizing filter is placed ure is the angle through which the po
asymmetries in the construction of the in the path of the beam, the intensity of larization rotates when the light passes
compounds. This phenomenon also the light that passes through it depends through one decimeter (.1 meter) of a
has practical applications, such as con on the filter's orientation. If its polari solution in which the concentration is
trolling the concentration of sugar in zation axis is parallel to the oscilla 100 grams per 100 cubic centimeters.
food products and sugar refining. Sam tions of the incident light, the light (Some references define specific rota
Epstein of Los Angeles has designed an is fully transmitted. If the axis is per tion in terms of other units.) The wave
inexpensive polarimeter that measures pendicular to the oscillations, the light length of the light is usually taken to
optical activity. is fully blocked. Intermediate orienta be 589 nanometers, that of the yellow
A classical model of light describes tions of the filter pass intermediate in emission line of sodium. The temper
it as a moving wave of oscillating elec tensities of light. ature of the solution is usually 20 de
tric fields. Often the light is depicted as The filter closest to the light source grees Celsius.
a ray in order to indicate its direction is called the polarizer; the other filter is Epstein's polarimeter operates on
of travel. Superposed on the ray are the analyzer. What you see in using the light from a 60-watt bulb. The light
vectors representing the direction and instrument is the light emerging from passes through a color filter and a col
strength of the associated electric field. the analyzer as it is rotated about the limating lens and then into a hous
The vectors are always perpendicular original ray. When the axes of the two ing, where it is reflected upward from
to the ray, and they vary in direction filters are parallel, you see the brightest a mirror. In the housing it travels
and strength in such a way that the light. After a 90-degree rotation of the through a polarizer and a cell holding
composite resembles a wave. The elec analyzer you see no light. This position the solution of interest. Thereafter the
tric field appears to oscillate as the of the analyzer is called the endpoint. light proceeds through a condensing
light moves past a given point. Solutions of optically active com lens and an analyzer, finally reaching
When the light is not polarized, the pounds such as lactic acid, tartaric an eyepiece through which the end
field can oscillate in any direction per acid, nicotine, turpentine, amino acids point is determined.
pendicular to the ray. If the light passes and vitamins rotate the polarization of The bulb and its socket are mount
through a polarizing filter, the oscilla light passing through them. They are ed on a wood support and covered
tions are restricted in such a way as to distinguished from other compounds with an inverted fruit can. The can is
be parallel to a single axis perpendic by their three-dimensional structure. mounted about an inch above the sup
ular to the ray. The light is said to be An optically active compound has one port so that air can flow into the can.
polarized. Its polarization is represent or more carbon atoms, each of which Holes in the top of the can allow air
ed by a double-arrow vector. An imag- is attached to one of four different heated by the bulb to escape. Extend
types of atoms or groups of atoms. As ing from a hole punched in one side of
light passes a carbon atom and its at the can is a length of polyvinyl chlo
Polarization
tachments, the electric field of the light ride (PVC) pipe of one-inch internal
interacts with the atoms in a way that diameter. The outer end of the PVC
rotates the polarization of the light tube is covered with a thin plate of
about the ray. ground glass.
Polarization Suppose a cell that contains a solu Since most data on specific rotation
of the. light tion of an optically active compound for optically active compounds are
is placed between the filters. When the listed for the yellow emission line of
filtEr.}
the ray. Hence when it reaches the ana cost of a professional color filter by
lyzer, it has an orientation different making a filter. His rig is made with
Polariz'lng Ray from the one it had before the cell two microscope slides that serve as
was introduced. To block the light windows in a rectangular cell. The top
A filter's polarizing action the analyzer must be rotated to a new of the cell is made of plastic fitted
120
(
Q) <i:)
Inverted
can over Vent holes
light bulb glass
filter
Ql
<9
121
�
slide should be centered on the tube.
The glue should be epoxy and must
provide a watertight seal.
This design has two drawbacks. You
Eyepiece hOlder '
will need to modify the PVC holder
_
so that the cell fits into place. A more
serious problem is that the cell may
be ruined if you do experiments with
some types of organic solvents.
I "
Above the sample-cell holder is a
condensing lens that directs light from
the cell through the analyzer and the
eyepiece. The analyzer is a sandwich
made in the same way as the polarizer
is made with one important difference:
the filters are skewed. Begin with a
Condensing 1'/4-inch square of polarizing filter.
lens Cut a five-degree triangular segment
out of the center. Slide the remaining
Partition parts of the filter together. Trim them
so that they form a one-inch square.
Sandwich this arrangement between
the microscope slides.
The skewed arrangement of filters in
the analyzer makes the determination
of the endpoint easier. Otherwise you
Sample cell must guess what position of the ana
lyzer best eliminates the light passed
by the test solution. With the skewed
filters you merely compare the relative
brightness of the light passing through
each part of the filter arrangement.
Door for The endpoint is achieved when the
acce.ss to
sample cell
two parts are equally bright. If you
Sample-cell
holder
Microscope.
slides, 1" x 3" ---",L.---.----:=-
�ht 50lution �
Potassium
dichromate
from
color filte.r
Details of the apparatus holding the sample cell Epstein's color ji Iter
122
© 1985 SCIENTIFIC AMERICAN, INC
5'---
degrees ---fL...
re.move.. .----'-.... :III ----- -------:III
I
I
IIII III
I II
rotate the analyzer in either direction Place the cell in the polarimeter with the solution until it reaches 67 degrees.
from the endpoint, one part brightens the mark on its side aligned with the Keep it in the bath for five more min
and the other part darkens. Hence the mark on the cell holder. Rotate the utes. (Do not let the temperature of the
skewed arrangement of filters enables eyepiece to find the endpoint. Note solution exceed 69.5 degrees.)
you to fine-tune the determination of where the pointer lies on the circular Transfer the flask to another water
the endpoint. scale of degrees. bath at 20 degrees C. Remove the flask
The analyzer sandwich is glued to Exchange the water for 50 milliliters from the bath when the temperature of
the bottom of a holder for the eye of sugar solution at a concentration the solution falls to 35 degrees. When
piece. The holder passes through a of 20 grams per 100 milliliters. Again the temperature of the solution reach
hole in the housing. It is held in place check for the endpoint. Epstein finds es 20 degrees, rinse the thermometer
with two flanges but is still free to that it is rotated clockwise by about 17 with about 25 milliliters of water, let
turn in the hole. The eyepiece can be a degrees from its position on the scale ting the water run into the flask. Add
single lens or a low-power compound when only water is examined. enough water to bring the volume in
lens system. A scale marked in degrees The rotation can also be ascertained the flask to 100 milliliters. Again mix
surrounds the holder. by a calculation. The specific rotation the contents.
To set a pointer on the scale switch for sugar is 66.5 degrees clockwise. In Place the flask in the 20-degree bath
on the light bulb and turn the eyepiece Epstein's experiment the length of the for another 15 minutes. If necessary,
holder (and thus the analyzer), moni light path is 1.3 decimeters. To com again add enough water to the flask
toring the brightness of the light as you pute the expected rotation multiply to bring the volume to 100 milliliters.
look through the eyepiece. Find the the specific rotation, the concentration Mix the contents and pour 50 millili
position where the two parts of the an (grams per milliliter) and the length ters of it into a sample cell. Position the
alyzer are equally bright. Then attach (decimeters). The expected rotation cell in the polarimeter and measure
a pointer to the eyepiece holder so that is clockwise by about 17 degrees. You how much the solution rotates the po
it points to zero on the rotation scale. might like to determine how the rota larization of the light. Epstein meas
The bottom of a sample cell often tion depends on the sugar concentra ures a rotation of about 2.8 degrees
has optical imperfections that alter the tion. Begin with the strongest solution counterclockwise.
polarization of light. To eliminate this and gradually dilute it as you measure The breakdown of sucrose is cata
possibility fill each cell with 50 milli the rotation. lyzed by the hydrochloric acid, pro
liters of water and mount it in the ap Epstein also investigated a well ducing 10.5 grams each of dextrose
paratus. Rotate the cell in its holder, known reaction in which sucrose is and levulose. The specific rotation of
monitoring the light through the eye broken down into two simpler sugars, dextrose is 5 2. 5 degrees clockwise, the
piece, until you get the best overall im dextrose and levulose. (The process is specific rotation of levulose 93 degrees
age and endpoint. Permanently mark called inversion.) Dissolve 20 grams counterclockwise. Calculate the rota
the side of the cell holder. Make a cor of sucrose in 50 milliliters of water in tions created by the two products in
responding mark on the cell. Whenev a 100-milliliter flask. Mix it well. Pre the sample cell. The dextrose should
er you use that cell, place it in the hold pare 10 milliliters of a mixture of hy rotate the polarization of light by
er with the marks aligned. Repeat this drochloric acid in the ratio of one part about 7.16 degrees clockwise and the
procedure for each cell. acid to three parts water. Add the mix levulose should rotate it by about 12.7
Epstein suggests testing the optical ture to the flask and again mix the solu degrees counterclockwise. Since the
activity of a solution of sucrose (or tion. Put a thermometer into the flask rotations are in opposite directions,
dinary sugar). Begin with a sample and place the flask in a water bath at 70 the calculation of the average rotation
cell containing 50 milliliters of water. degrees C. Monitor the temperature of amounts to subtracting the two num-
123
© 1985 SCIENTIFIC AMERICAN, INC
Region empty
Rotation about
virtual mirror
Rotation about
real mirror ;t
!J 1�
{) ��
Q
Twe..\ve. fold 5ixfold
5ymmetry I!) � symmetry
Ijj;J CE
Fourtold symmetry
A kaleidoscope with two types of symmetry A kaleidoscope with three types of symmetry
124
Cosmic-
is independent of your angle of view.
What other mirror systems yield
unambiguous image fields? Initially I
thought the only requirement was that
an optical system should fill the im
Consequences
age field with nonoverlapping copies
�/�======�=
of the direct view, much as one might
fill a floor with identical tiles. I discov
ered that I was wrong by considering
Vertex
the hexagonal arrangement of mirrors
shown in the upper illustration on the Bvolution of
Mattet and Enezty
Vi:tual�Y
opposite page. The direct view and two
reflected hexagons appear in each part
of the illustration. An easy way to de
rive the reflected hexagons is to rotate mirrors �
the direct view about an edge until it
lies again in the image plane. Then ro
tate the reflected hexagon about one
of its edges to form a second reflect
ed hexagon.
The first part of the illustration indi
. fills a uniquely useful ecological
cates that I began with the direct view niche because of its marvelous connec
and proceeded clockwise to find the tions betweell aile subject and allother.
Determining the sectors of a cluster
two reflected hexagons. I rotated the I predict a warm welcome alld wide
direct view about real mirror A to form readership." John Archibald Wheeler
it
other polygons I finally understood metry. The right-triangle systems offer
'O�
what arrangements of mirrors yield clusters displaying more types of sym
unambiguous image fields. For any ar metry. In one of them the right-angle SPRINGER·VERLAG NEW YORK. INC.. Ann: K. Quinn
�
rangement the key is to examine each vertex produces a fourfold symmetry 175 Fifth Avenue. New York, NY 10010
vertex at which two mirrors meet. and the 45-degree angle produces an Please send me the booklsl checked below.
I enclose a D check 0 money order or credit card
Looking into the system, you see a eightfold symmetry. number 0 MC 0 VISA 0 AmEx for the total of
sector of the direct view; around the The system with 60- and 30-degree $ (please include $1.50 for postage and
vertex are reflections of that sector. angles represents the most beautiful handlingl
125
© 1985 SCIENTIFIC AMERICAN, INC
ttfl
Readers interested in further explana CONTROL GROWTH AND DIFFEREN
tion of the subjects covered by the articles TIATION IN MYELOID LEUKEMIA: A
in this issue may find the following lists of MODEL FOR THE ORIGIN AND PRO
publications helpful. GRESSION OF MALIGNANCY. Leo Sachs
in Proceedings of the National Academy
COMPUTER RECREATIONS of Sciences of the United States of Amer
ica. Vol. 77, No. 10, pages 6152-
They were never rich.
AN APL2 GALLERY OF MATHEMATICAL 6156; October, 1980.
And most were always poor.
PHYSICS-A COURSE OUTLINE. Grego NORMAL DEVELOPMENTAL PRO-
But they managed to live, ry J. Chaitin. IBM Research, P.O. GRAMMES IN MYELOID LEUKEMIA:
through their own hard labor. Box 218, Yorktown Heights, N.Y. REGULATORY PROTEINS IN THE CON
The children worked, too, 10598; 1985. TROL OF GROWTH AND DIFFERENTI
DIRECT N-BODY CALCULATIONS. ATION. Leo Sachs in Cancer Surveys,
beginning a lifetime of hard
Sverre J. Aarseth in Dynamics of Star Vol. 1, No. 2, pages 321-342; 1982.
work early, proud to share the
Clusters. edited by Jeremy Goodman
responsibility of family sur
and Piet Hut. D. Reidel Publishing THE STRUCTURE
viVal. Sometimes they went Company, 1985. OF COMET TAILS
hungry, but they survived. A UNIFIED N-BODY AND STATISTICAL
TREATMENT OF STELLAR DYNAMICS. INTERPLANETARY GAS. XXIII. PLAS
Then the drought came. Alan P. Lightman and Stephen L. W. MA TAIL DISCONNECTION EVENTS IN
The worst in years. Suddenly McMillan in Dynamics of Star Clus COMETS: EVIDENCE FOR MAGNETIC
they were facing death. ters, edited by Jeremy Goodman and FIELD LINE RECONNECTION AT IN
Many fled to flnd new work, Piet Hut. D. Reidel Publishing Com TERPLANETARY SECTOR BOUNDA
pany, 1985. RIES? M. B. Niedner, Jr., and J. C.
to flnd food and water. And in
Brandt in The Astrophysical Journal,
many regions, war blocked
SPACE SCIENCE, Vol. 223, No. 2, pages 655-670; July
ald that might have come. 15, 1978.
SPACE TECHNOLOGY
and added other terrors and AND THE SPACE STATION STRUCTURES FAR FROM THE HEAD OF
miseries. This is the crisis COMET KOHOUTEK, II: A DISCUSSION
now spreading throughout MISSION TO EARTH: LANDSAT VIEWS OF THE SWAN CLOUD OF JANUARY 11
THE WORLD. Nicholas M. Short, Paul AND OF THE GENERAL MORPHOLOGY
Africa-the most serious
D. Lowman, Jr., Stanley C. Freden OF COMETARY PLASMA TAILS. Mal
situation there in many years.
and William A. Finch, Jr. NASA colm B. Niedner, Jr., and John C.
Oxfam America is working to SP-360, U.S. Government Printing Brandt in Icarus, Vol. 42, No.2, pages
send urgently needed sup Office, 1976. 257-270; May, 1980.
plies to the hardest hit areas, SPACE SCIENCE COMES OF AGE: PER THE COMET BOOK. Robert D. Chap
and to expand our long-range SPECTIVES IN THE HISTORY OF THE man and John C. Brandt. Jones and
programs to prevent more SPACE SCIENCES. Edited by Paul A. Bartlett Publishers, Inc., 1984.
Hanle and Von Del Chamberlain.
suffering in the future.
National Air and Space Museum, APPLICATIONS OF OPTICAL
Smithsonian Institution Press, 1981. PHASE CONJUGATION
Help people survive and
SPACE RESEARCH IN THE ERA OF THE
bemme self-reliant again.
SPACE STATION. Kenneth J. Frost and NONLINEAR OPTICAL PHASE CONJUGA
Ask your church, school or Frank B. McDonald in Science, Vol. TION. Special issue of Optical Engi
group to join you in helping. 226, No. 4681, pages 1381-1385; De neering, Vol. 21, No. 2; Marchi April,
cember 21, 1984. 1982.
�i�
ASTROPHYSICS TODAY. Edited by A. G. OPTICAL PHASE CONJUGATION. Edit
W. Cameron. American Institute of ed by R. A. Fisher. Academic Press,
Physics, 1985. 1983.
NONLINEAR OPTICAL PHASE CONJUGA
Wrlre_ GROWTH, DIFFERENTIATION TION. David M. Pepper in The Laser
Handbook: Vol. 4, edited by M. L.
Oxfam �
AND THE REVERSAL
OF MALIGNANCY Stitch and M. Bass. North-Holland
115
BoxM·l
Broadway
��
�!I
Vol. 47, pages 419-448; 1978.
CONTROL OF NORMAL CELL DIFFER
lag, 1985.
�;
SPRINGS
fJ
ture, Vol. 274, No. 5671, pages
535-539; August 10, 1978. HYDROTHERMAL PROCESSES AT SEA
CONSTITUTIVE UNCOUPLING OF PATH FLOOR SPREADING CENTERS. Edited
l WAYS OF GENE EXPRESSION THAT by Peter A. Rona, Kurt Bostrom, Lu-
126
© 1985 SCIENTIFIC AMERICAN, INC
QUESTAR
GATE WITH SIBLINGS: AN EXPERI
MENTAL FIELD STUDY. Richard K.
O'Hara and Andrew R. Blaustein in
Oeco!ogia, Vol. 67, No. 1, pages 44-
51; 1985.
KIN RECOGNITION IN ANIMALS. Edited
by David J. C. Fletcher and Charles
D. Michener. John Wiley & Sons,
Inc., 1986.
127
© 1985 SCIENTIFIC AMERICAN, INC