Earth and Planetary Science Letters, 57 (1982) 251-262 251

Elsevier Scientific Publishing Company, Amsterdam - Printcd in The Netherlands

[?l

The role of episodic bombardment in geophysics

S.V.M. Clube and W.M. Napier
R~)'al Oh~ert alorv, l:'dmhur~h, St ollaml f U. K. j

Received August I. 1981

Revised version received November 9, 1981

It is shown that cometary or asteroidal impacts strongly perturb the boundary conditions at the core, outer mantle
and atmosphere of the Earth. Impact episodes occur following the capture or orbital disturbance of these bodies as the
Solar System passes through the spiral arms of the Galaxy. There is thus a tendency for magnetic field reversals,
principal plate movements, ice ages and mass extinctions to be correlated and to recur on time scales dictated by
galactic processes. It follows on the basis of a purely astrol,omical theory' that it may be possible to synthesise the major
phenomena of geophysics: knowledge of the continuous state of lhe Galaxy then becomes a prerequisite for the
understanding of terrestrial history.

1. Introduction -400 Myr. One is therefore led to consider the

possibility that the Galaxy is controlling the ter-
It has long been suspected that there is epi- restrial phenomena. But if so, how? Passage
sodicity, if not quasi-periodicity, in certain geo- through a gaseous nebula, for example, would
physical phenomena. Thus large-scale continental seem to be too gentle a process to give rise to
movements, often corresponding to the well-known worldwide vulcanism.
geological boundaries, appear tO have taken place In addition to the clustering of the geophysical
rather rapidly at intervals of - 5 0 Myr, the inter- phenomena, there is evidence that they are some-
vals themselves being relatively quiescent; pre- times correlated (e.g. [2]). This may be exemplified
Cambrian rock ages cluster worldwide around dis- by consideration of the great dinosaur extinction.
tinct epochs - 8 0 0 Myr apart; for the past 1000 Theories of this event have tended to polarise into
Myr ice ages of duration - 50-100 Myr have been the "catastrophic" or the "gradualistic". An exam-
separated by an average interval - 100 Myr; and ple of the former type is the "giant meteorite
there appear to have been four or five global mass impact" hypothesis proposed from time to time
extinction events with separation - 100-200 Myr (e.g. [3-6]) and revived recently by Alvarez et al.
within the Phanerozoic. The evidence for these and [7] on the basis of geochemical anomalies dis-
other periodicities is fully discussed by McCrea [1]. covered by them at the Cretaceous-Tertiary (KT)
The periodicities, if real, raise a problem, for boundary (see also Smit and Hertogen [8], Ganap-
there are no obvious internal or Solar System athy [9] and Kyte et al. [10]). However, McLean
mechanisms with time scales of this order. On the ([11,78]) has pointed out that one of the greatest
other hand the time scales find a very natural outpourings of lava in geological history occurred
counterpart in galactic phenomena; thus the inter- 65-60 Myr ago, and proposes a gradualistic hy-
val between passages of the Solar System through pothesis in which this created greenhouse condi-
the spiral arms of the Galaxy is ~ 50 Myr; that tions, exterminating the dinosaurs by heat-
between passages through dense nebulae (which infertility linkage, creating an extraordinary uni-
are probably confined to the spiral arms) is ~ 100 versal hiatus which causes the illusion of abrupt

0012-821 X / 8 2 / 0 0 0 0 - 0 0 0 0 / $ 0 2 . 7 5 '~ 1982 Elsevier Scientific Publishing Company

252

extinction, and giving rise to the various geochemi- (see also Clube and Napier [16], Napier and
cal disturbances. Associated with this event also is Staniucha [19]). In due course, a fraction of these
apparently an abrupt drop in sea level, one of bodies appears as short-period comets, some of
many which have been recorded [12] and which which eventually outgass and form Apollo (Earth-
occurred at 5- to 10-Myr intervals throughout the crossing) asteroids. The fluctuating comet and
Tertiary. Arguments in favour of a KT cosmic asteroid populations thus lead to episodes of ter-
missile are, first, the chondritic composition of the restrial bombardment. If n (m, t s) bodies are thrown
anomalous material at the boundary, and second, into the inner Solar System during each capture
the extreme rapidity of deposition (~< 100 years) of episode (time ts) then to a sufficient approxima-
the anomalous material, concurrent with a simi- tion the predicted bombardment rate by masses m
larly rapid planktonic extinction at the section within each inter-spiral-arm interval may be given
examined by Smit and Hertogen. But if so, then it by:
is possible that the impact was also somehow dn(m)
responsible for the simultaneous sea level drop - - odt
c ~ F( t - ts)n( m,ts) (1)
and for precipitating the global vulcanism. Other ts
such correlations have been claimed; at the termi- where:
nal Eocene in particular, a magnetic field reversal
is associated not only with drastic radiolarian ex. n( m ) oc ( m//~ ) -17 (2)
tinctions but also with the bediasite tektites and a m - 1018g (3)
profound climatic change of ~> 1 Myr duration
[13,14]. The Popigai crater of 100 km diameter and the function F is given in Fig. 1.
appears to be contemporaneous with these phe- A rule of thumb summarising these relations to
nomena. We assume here that tektites and crater order of magnitude gives the average time interval
are indicative of impact, but note that this is not 8t between impacts of energy ~>E due to mass
universally accepted [75]. ~>m as:
These qualitative considerations lead naturally 8t (in years) --- E (in Mt)
to the question whether impacts control mass ex-
tinctions and geophysical events in general, and E (in Mt) -~ m (in g)/10 I°
whether the Galaxy in turn controls the occurrence
of impacts.
I

~3
e'-
2. Episodic bombardment
.>,
__02
On the latter question, the discovery in recent
years that about half the mass of the interstellar
medium is in the form of giant molecular clouds is It.
crucial. These nebulae have masses 105-106 solar
masses, diameters 30-300 light years, and each
comprises 10-100 discrete cloudlets of diameters
1-5 light years and masses 102-104 solar masses.
0 5 10
In the neighbourhood of such nebulae consider-
able tidal forces are exerted, and passage through {t - ts)(lO 8 yr)
one leads to a temporary flooding of the inner Fig. l. Log F as a function of ( t - i s ) adapted from Wetherill
Solar System with small bodies. In this theory, [74]. Note that superposition of the many planetesimal capture
episodes throughout Galactic history will produce between
developed by Napier and Clube [15], the flooding spiral arm passages a background flux such that 2<log F<3.
is caused both by the capture of interstellar plane- The modulation on the average impact rate produced by
tesimals and the disturbance of indigenous ones bombardment episodes is thus ~ a few: 1.
 253

where 1 Mt -- 1 megaton of T N T equivalent = 4.2 150 Myr say), and such a concentration of as
× 1022 ergs. Equation (1) implie__s, however, that many similar events just after capture episodes
the interval 8t fluctuates about 8t both stochasti- ( t - t ~ ~<30 Myr) as to be possibly the primary
cally and with a quasi-periodic modulation. One cause of a permanently disturbed state of the
limitation of the theory at the time of writing is Earth's surface at these times. Thus, if we were to
that the proportion of genuinely interstellar plane- suppose in general that each bombardment by
tesimals to perturbed indigenous ones is uncertain, Apollo asteroids of mass m>~ 10~6g produces a
and the magnitude of the galactic modulation has significant geological signature, but only impacts
still to be assessed quantitatively. Semi-quantitative by bodies with rn~> 10J9g cause major extinction
estimates so far indicate that the impact rate may events, then an episodic pattern of discontinuities
vary by a factor of 3 or 4 to one due to episodic of the kind under discussion would in principle
capture, but brief bombardment episodes at - 10 emerge.
times the mean rate can probably not be excluded A theory of episodic bombardment, with
either dynamically or by current terrestrial and galactic control, thus exists. The question re-
lunar cratering data. Another consideration is that maining is to what extent an impact episode will
both the existence of interstellar comets and the manifest itself in geophysical phenomena. Even
assumption that Apollo asteroids are mainly de- without episodicity of course recent astronomical
gassed comets are controversial issues. The whole developments (e.g. [17,18]) indicate high impact
;luestion is reviewed by Clube and Napier [16]. In frequencies and require consideration of their geo-
the present paper, episodes of bombardment are physical consequences. The causes of geophysical
taken to be a fact of life: we seek to explore the phenomena are generally understood, if at all, only
terrestrial consequences. Weak bombardments at in a qualitative way, so the discussion here will be
50-Myr intervals are expected corresponding to limited to order of magnitude considerations: we
the tidal disturbance during Solar System passages pay particular attention to the changes in atmo-
through spiral arms; larger effects are expected at spheric, outer mantle and core boundary condi-
100- to 400-Myr intervals on average due to actual tions that arise as a consequence of impact. Now,
penetration of molecular clouds. in general, a collision imparts both energy and
The function F simply reflects the dynamical momentum to the Earth. The energy dissipated
processes affecting orbits of small bodies ran- beyond the crater perturbs the atmosphere and it
domly injected into the inner Solar System. Thus, turns out that energy considerations in particular
Apollo asteroids eventually suffer collisions or may be involved in glaciation and ocean regres-
close encounters resulting in their ejection from sion. Indeed an ice age appears to be caused by
the inner Solar System on a time scale - 3 0 Myr, many glaciations and regressions, both large and
though a substantial fraction may also be retained small, closely bunched in time. The momentum
in more stable orbits giving rise to a significant input to mantle and core on the other hand seems
number of impacts on longer time scales. It fol- to be playing the more fundamental role in geo-
lows that captured bodies whose average impact magnetic reversals and plate tectonics. Although
rates on the Earth are 1 in 50-100 Myr (corre- the magnitudes of all these effects increase mono-
sponding to m~> 1018g) will tend to be bunched tonically with the size of missile, each effect only
two or three at a time within periods of 30 Myr achieves qualitative significance in the geological
immediately following the passage of a dense and biological record at some particular size (in
nebula while a similar number of more isolated the range - 1 0 1 4 - 1 0 2 ° g - 0.4 40 km) peculiar to
events may occur between such passages. On the that effect. Below a particular level, therefore~ any
other hand, with bodies whose impacts are slightly one effect may be generally unobservable and
more numerous, 1 in 1- I0 Myr say (corresponding above some other level the effect may saturate or
tO m,~> 1016 g), there may be produced at a com- be swamped by other more dominant processes.
paratively regular rate - 2 0 separate identifiable Restriction solely to energy and momentum
events between capture episodes (i.e. 30 ~< t - t~ ~< related effects is clearly an oversimplification.
254

Thus, it is assumed in (1), (2) and (3) that missiles latter, for example, find that basaltic glass par-
have a mean density ~ 2.5 but at the low mass end ticles 0.1~m in diameter, distributed with optical
(m ~< 1014g) particularly, the impact effects will be depth ~ 0.2 through the stratosphere, would lower
different for cometary material (say ~< 2.5) and the mean global surface temperatures by 3°C.
meteoritic material (say ~>2.5). Since planetesi- Enhancement of this cooling by increased high-
mals captured from dense nebulae will be largely altitude snow and ice cover can be considered
undifferentiated proto-planetary and proto-stellar likely [25,26]. However, no volcanic outburst is
material, such m a t e r i a l - - duly volatilised-- known to have had energy ~> 100 Mt, whereas on
comprises the principal component of the larger timescales of orders 103-104 years, impacts of
Apollo asteroids (m~1016g). At the low mass energy ~>103-10 4 Mt (m~> 1013-1014g) are ex-
end, however, corresponding to missiles producing pected. Within the uncertainties the optical depth
most craters ~< 10 km diameter, the population of of the dust veil created might be ~ 0.1-1: there-
Apollos contains a greater proportion of collision- fore it is possible that dust veils created by impact
ally processed and differentiated primordial rather than volcano are a prime driving force in
material (i.e., meteoritic), much of it possibly com- climatic change over such timescales: surface cool-
ing from the asteroid belt. So far as terrestrial ing of order 1-10°C at intervals of 103-104 yr is
effects are concerned therefore, there is a discon- anticipated. According to Flohn [27], abrupt
tinuity in the mass function (equation (2)) around climatic coolings of - 5 ° C , with recovery times a
m ~ 1014g (say) below which impacting cometary few 100 years, have taken place at 103- to 104-year
material is strongly ablated by passage through the intervals throughout the past 0.7 Myr, even during
atmosphere and the population reaching the interglacials; and although he attributes these to
ground is increasingly dominated by meteorites. clustered volcanic activity, the observed magni-
The discussion here is largely concerned with im- tudes and time scales are broadly consistent with
pact energies ~> 104 Mt (m~> 1014g) and thus ap- impacts. It is possible that the recovery time of
plies mainly to Apollos originating from captured 100 years, which greatly exceeds the strato-
comets. spheric dust settling time, reflects the melting time
of the expected ice enhancements: Lamb [28] for
example found that high-latitude snow surfaces
3. Energy-related effects tended to be thermally isolated, protecting them-
selves from warm air transport through cold
According to Gault and Heitowit [20], a major troughs.
proportion of the energy of an impact reappears as In the next section it is shown that impacts may
kinetic energy of crater ejecta. Dust is injected into cause magnetic field reversals at - 0 . 1 - to 1-Myr
the stratosphere even at low energy levels and a intervals. A correlation between polarity reversals
planetesimal of mass m may excavate ~> 100 m of and climatic excursions is therefore expected. Such
material. Most of this forms the crater rim, but a a relationship is well established [29,30] and on the
proportion perhaps 10-3-+ J is injected as fine dust present theory each has impact as a common
into the stratosphere [21]; thus a dust veil of mass cause. On time scales -0.01 Myr, impacts are
10 - I-+1m is created. Discussions of the effect of a inadequate to reverse polarity but may neverthe-
dust or aerosol veil on climate have generally been less be associated with geomagnetic disturbance.
confined to the effects of volcanic eruptions pro- Whether such disturbances are seen in the
ducing optical depths in the stratosphere of a few palaeomagnetic record is a matter for some con-
percent. According to Hammer et al. [22], they troversy. Fairbridge [31] finds that coincident with
have had a cooling effect on the post-glacial the decreased field intensity of the alleged Gothen-
climate, temperature drops of order 0.2-0.5°C fol- burg Excursion of 13,750-12,350 years B.P. was a
lowing some months after large outbursts. Theo-- severe global cooling, a sea level fall of - 1 0 m,
retical modelling by Hunt [23] and Pollack et al. and an increase in glacier volume of - 4 × 106
[24] also indicates cooling of this order, and the km 3, all in a sense contrary to then prevailing
 255

trends. Over the period 16,000-8000 years B.P., perature differences it is likely that by the time the
Fairbridge demonstrates good correlations be- dust has "cleared from middle and high latitudes
tween climatic and sea-level fluctuations, and the ( - 1-3 years) a glaciation, all other factors being
volume of glacier ice melted. The indications then favourable, will be well under way. Gordon and
are that fluctuations in surface temperature (1- Davies [38] considered that ~>3% reduction in
10°C), glacier volume ( 1 - 1 0 × 106 km3), ocean insolation for - 1 year would lead to unstable
depth (5-30 m) and magnetic field intensity (0.1- growth of ice caps. Pollard et al. [39] found that
1 G) may be impact generated at intervals of 103- ice albedo feedback was not very important for
104 years. We note in passing that such minor small oscillations but suggested it might be im-
events are a major cosmic hazard on the time scale portant for a full glacial interglacial cycle. A glaci-
of human civilization. ation, therefore, may rapidly take hold consequent
The considerations so far relate to missiles ~< on a ~> 10~6-g land impact. Because of the extreme
10~g with energies ~< 105 Mt, but an Apollo instability of the system it might just as rapidly
asteroid of mass ~> 1016g and energy -~ 106 Mt decline in little more than the melting time of an
will impact on land at mean intervals At-- 1 Myr, ice sheet in a mild atmosphere (say t c - 103-104
ejecting ~> 1015+lg of dust into the stratosphere; years). However, given that appreciable cooling is
immediately following a capture episode At is ~ 0.1 initiated by impacts, a prolonged glaciation may
Myr. At such intervals a reflective blanket of large be maintained so long as the mean interval At
optical depth ( ~ > 1) is thrown around the Earth, between such impacts (of, say, 103-104 Mt en-
and although one may expect climatic cooling for ergy) is less than t c. This would correspond to a
lesser impacts to increase with the mass of dust somewhat marginal situation in general but follow-
veil produced, the effects will saturate at missile ing a capture episode such an impact rate would
masses ~> 1016g. Thus one expects a temperature be easily maintained. Thus if t c - 104 years and
drop in excess of that for average Flohn-type or A t - 103 years after an impact episode, then for a
Gothenburg-type events (10°C say) at 0.1- to 1- random rate of arrival, the mean duration of a
Myr intervals. The question arises whether the glaciation (assumed to be maintained until by
climatic perturbation created by such a dense dust chance there were no impacts over a 104-year
veil will decay or whether an instability of the type period) would be ~ 1 Myr in the absence of other
discussed by Brooks [32] and Budyko [25] may be factors. If t c = 2.5 × 104 years, At = 5 × 103 years,
induced, in which there is positive feedback be- the expected duration would be - 2 × 104 years.
tween enhanced snow cover, increased planetary It appears then that a medium-sized impact of
albedo and decreased solar heating. The growth the type currently expected at ~ 0.1-Myr intervals
time of large continental ice sheets has usually may induce rapid glaciation through an ice-albedo
been taken as 103-104 yr [25,26] but rapid precipi- instability and that frequent smaller impacts will
tation due to differential cooling of land and sea maintain it. The glaciation will terminate ran-
may greatly shorten this time scale [33,34,35]. The domly within a few 104 years to ~ 1 Myr de-
latter suggest that large areas of snow could form pending on the mean impact rate, though the
in - i0 years due to strong landward gales. Rud- whole may be modulated by the Milankovitch
diman et al. [36] have provided evidence for strong process [40-42]. The latter does not explain how
and virtually instantaneous ice growth in the the insolation variations produce the glaciations,
northern hemisphere - 1 1 5 , 0 0 0 and 75,000 years nor account for the observed brief episodes of
B.P. Modelling of climatic feedback systems by climatic cooling [27,31,39]; the Pleistocene glacial
Saltzmann and Moritz [37] indicates that for rea- variations may in any case be largely stochastic
sonable sets of parameters oscillatory, damped or [43]. We conclude that if there is extensive land
amplifying behaviour on many time scales is possi- precipitation of snow during the short period in
ble but that the ice-albedo feedback in particular which there is dense dust cover (z>~ 1), and if
may lead to strongly unstable behaviour. Since conditions under these circumstances are favorable
their model neglects the longitudinal land-sea tem- for triggering an ice-albedo instability, then be-
256

haviour corresponding quite closely to the significant poleward drift of suspended material
glacial/interglacial sequence of the Quaternary will before it is eventually removed after 3 years. Hoyle
occur following a capture episode, with in particu- proposes that a glaciation, once formed, is self-
lar a tendency for rapid switching between ex- sustaining, and to terminate it he postulates im-
treme states to occur at ~0.1-Myr intervals. pact by an iron meteorite raising metallic particles
Abrupt isolated sea-level drops of order 100m into the stratosphere that absorb rather than scatter
with a 10°C cooling are expected during periods of sunlight. However, an impact will eject mostly
lower impact rate and have apparently been de- ground material, meteoritic material being ~ 1%
tected by Vail et al. [12]; their onset seems to be of the total. And there is no episodicity in the
too rapid for ocean ridge spreading. They have the Hoyle scenario. It is thus one of short-lived sto-
appropriate frequency and magnitude throughout chastic glaciations rather than ice ages.
the Tertiary. Moreover there has been, during the
middle and late Tertiary, a global cooling [44], a
general growth of the Antarctic ice cap and a 4. Momentum-related effects
lowering of mean sea level, all coincident with the
passage of the Solar System through Gould's Belt The momentum implanted by an extrater-
[15], the most conspicuous spiral-arm feature in restrial missile is distributed throughout the man-
the solar neighbourhood. tle, because of the latter's rigidity, on a time scale
Geographic and tectonic factors may also be 1 hour. Thus a random rotational velocity Aw is
involved in creating the preconditions necessary suddenly added vectorially to that of the Earth,
for ice ages but it is not clear that one will always and is obtained from:
occur when these circumstances are favourable:
mvr = I A w / R e
glaciers vanished from Australia throughout much
of the Permian in spite of the continent's high where m, v represent, respectively, the mass and
latitude; there was a lack of glaciation from mid- velocity of the missile, and its impact parameter r
Cambrian to mid-Ordovician in spite of the pres- is typically - 0 . 7 times the radius of the Earth,
ence of North Africa at high latitudes; and so on. R e . The moment of inertia of the mantle I ~ 7 X
The placing of the continents may therefore be a 1044g cm 2, whence for v = 2 5 km/s, A w ~ 5 X
necessary consideration for ice ages but it is not 10-22m cm/s. If the radius of the core - 0 . 5 R e
sufficient. If a high rate of impacts plays the then an instantaneous velocity change Aw ~> 2.5 X
primary role in both triggering and maintaining a 10 -5 c m / s will be imparted to the inner edge of
global albedo feedback, then one expects ice ages the mantle by planetesimals of mass ~> 1017g col-
typically of 30-100 Myr duration--possibly mod- liding at mean intervals ~ 10 Myr, or ~ 1 Myr in
ulated by geographic factors-- following encoun- the period immediately following a capture epi-
ter of the Solar System with giant molecular clouds; sode.
separated by ice-free intervals of typically 100-200 We assume this motion is taken up at the outer
Myr duration. face of the core and transmitted inwards; a fluid
Mention should be made here of a similar pro- velocity Aw in the core gives rise to the usual
cess for initiating ice ages discussed by Hoyle and Coriolis force per unit mass, 2~o A Aw, where to is
his colleagues [35,45,46] involving first comet tails the Earth's angular rotation vector. Such electric
and then meteorites. Their mechanism requires a currents as may be associated with this will gener-
dense dust veil to exist for 10 years. In fact ob- ate a Lorentz force per unit mass ( I ~ / o ) ( j A A H ) ,
servations summarised by Lamb [28] indicate that where p is the fluid density, ff is the magnetic
the stratospheric residence time for fine dust is permeability, AH the magnetic field perturbation
1-3 years and that latitudinal inhomogeneities de- and j the electric current density. If we assume
velop even more rapidly: within - 0 . 2 years of that in a pre-existing fully magnetic turbulent flow
dust injection a uniform veil is created, but within (e.g. see Malkus [50]), a rough balance is struck
a further ~ 0.2-0.5 years there is likely to be a between the added Coriolis and Lorentz forces
 257

and these are of the same magnitude as the reac- which may be connected with mass extinctions [2].
tion force, then we may write: Causal mechanisms have been proposed [55] but
without much success [56]. McCrea [1] points out
12~A~w[~ ~(~AH)AAH that two great extinctions ~ 225 and 65 Myr ago
occurred in mixed magnetic intervals 230-204 Myr
resulting in characteristic magnetic fields induced and 70-50 Myr ago (the Cenozoic climate decline
by impact IAHI ~> 2 G. The vector character of the dates from the latter period). The time scale asso-
force requires that the magnetic field H be inter- ciated with high frequency of reversals is - 3 0
preted as that associated with the westward drift at Myr, and in accordance with theoretical expecta-
the surface and core circulation velocities - 10-2 tion, the reversal frequency declines after its highest
c m / s , while AH is characteristic of ( -+ ) changes in values are reached at the extinction boundaries.
the poloidal and toroidal components. H is not a There are notable coincidences in time between
directly observable quantity but the magnitude of these periods of high reversal frequency and cer-
the Earth's dipole field at the core-mantle tain tectonic phenomena also: see for example the
boundary is inferred from surface observations to discussion by Jacobs [57]. This leads to the ques-
be approximately 5 G. The velocity change ~> Aw tion of the role played by large impacts, or impact
at the core boundary, induced by an impacting episodes, in plate tectonics.
mass ~> 1017g at 1- to 10-Myr intervals is thus ~> About 10-2 of the impact energy of a missile
the characteristic flow speed of eddies within the will manifest itself as seismic waves, which for a
core and indeed, is capable of temporarily revers- 1017-g missile yields - 2 . 5 × 105 Mt of seismic
ing the direction of flow at the core boundary energy as against 100-500 Mt for a major earth-
relative to the mantle. Following such impact, on a quake: this will propagate globally and it is there-
time scale dictated by the circulation time of the fore very likely that weak crustal structures will be
core material (say 103-104 years), the magneto- violently disturbed, initiating volcanic and sec-
turbulent currents within the core will be strongly ondary earthquake activity on a worldwide scale.
perturbed with corresponding readjustment of the Because of the high rigidity of the mantle, how-
overall balance between the internal Lorentz and ever, flow will not immediately be induced in it by
other forces. The process simply reflects the sup- a large impact. Nevertheless, for a 1017-g missile,
posed instability of the geomagnetic dynamo the mantle spin induced yields at the core mantle
[47,48] and may even result in polarity reversal. interface, a relative speed ]Aw I-- 10 - 5_ ! 0 4 cm/s.
Any new configuration of the geomagnetic field As we have seen already, the magnitude of this
will of course require maintenance by driving forces velocity is at least at a level where it dominates
that we do not seek to specify in this discussion. existing motions at the core-mantle boundary and
At the present time, gravitational settling appears is sufficient to produce geomagnetic reversals as a
to be a favoured engine [49] but heat convection result of substantial reorganisations of the core's
and terrestrial precession have had their advocates convection pattern. With strong coupling of the
(e.g. see Malkus [50], Stacey [51]). Whichever turns core and mantle, it is to be expected that similar
out correct, polarity reversals can in principle be rearrangements will take place at the base of the
achieved by impacts at intervals of at least as mantle. Thus, during the period in which the core
small as 1-10 Myr (cf. 0.1- to 1-Myr intervals is accelerated to matching velocity, tangential
observed). The close coincidences of tektite ages stresses of similar magnitude are applied to the
and field reversals which have been noted [52,53] inner mantle surface. The time scale of the appli-
support this hypothesis. cation of these stresses is of the same order as the
Episodes of high bombardment rate will, by circulation time within the core, ~ 103-10 4 years.
this hypothesis, produce periods of high reversal Since the relaxation time of the mantle is - 3
frequency possibly coincident with other impact- years [58], the mantle response will be viscous and
induced phenomena. The geomagnetic data of the result will be to establish more or less synchro-
Creer [54] indicate the existence of such periods, nously with the core's reorganisation, a mantle
258

convection pattern, the magnitude of whose mo- turbations ~ 10 -1 cm,/s of duration ~ 103-104
tions for a 1017-g missile will be entirely compara- years will be introduced at the core-mantle
ble to those currently associated with currents of boundary. And if, as is generally supposed, the
10-4 cm/s at the core-mantle boundary. If it is convection pattern in core, mantle or aestheno-
assumed that large perturbations at the base of a sphere is strongly coupled throughout, then effects
convection pattern introduce large perturbations of the magnitude, duration and interval indicated
at its surface--and this may be model-dependent would now appear to be probable. With this the-
--then perturbations 10 -8 cm/s will be intro- ory, major impacts are the prime trigger of new
duced at the surface. For missiles -10t9-102°g patterns of continental drift.
on the other hand, and similar convection patterns
temporary surface motions of ~ 10-5_ 10-4 cm/s
are predicted with plate movements o v e r 1 0 3 - 1 0 4 5. The complexity of impact phenomena
years of ~ 30 km. It is clear that worldwide tectonic
activity for periods of ~> 103-104 years is antic- It is now clear, both from cratering data [17]
ipated for the largest and rarest impacts. Such and from observations of Apollo asteroids [18]
behaviour must involve the relatively rapid open- that impacts of the general magnitude and
ing of large lithospheric cracks whose presence, frequency given by equation (1) etc., must occur;
once established, can only serve thereafter to de- energy and angular momentum transfer as dis-
termine the subsequent pattern of motions in the cussed here is therefore unavoidable. Given only
aesthenosphere and hence, throughout the convec- the broad properties of mantle and core, we have
tive region. Over several thousand years therefore, demonstrated that the mutual slipping generated is
there is set up a new pattern of continental drift sufficient to greatly disturb the internal motions of
which may or may not simulate the existing one: both, at different levels of impact. We have there-
in general, we expect significant and sudden de- fore proposed that the large core disturbances may
flections of plate motion. In the ensuing period, cause polarity reversals, the large mantle dis-
plate tectonics are probably driven by differential turbances tectonic episodes, and we have dem-
heat loss, essentially controlled by the new onstrated that to order of magnitude the observed
boundary conditions established at the top of the time scales of each agree with impacts as a com-
mantle. Generally, such motions, if maintained by mon cause.
exceptional vulcanism over millions of years, will The energetic effects of modest impacts appear
involve significantly larger plate motions than to be largely climatic and here the discussion
those observed at present, but eventually, as the becomes very tentative because of the extreme
Earth's surface restores to its present state, the complexity of the atmosphere. But we have pointed
motions will be damped. Periods of rapid plate out that the dust veil created by modest impact
motion with eventual decline are thus required by does constitute a major, if short-lived, perturba-
the theory, in general accord with observation. tion on the insolation, and proposed that such a
A possible test of these ideas might be to relate perturbation might initiate an ice-albedo feedback
the "dog leg" in the Hawaii-Emperor and Austral- which is then maintained by frequent, small colli-
Gilbert-Marshall seamount chains to the geologi- sions until chance absence of such collisions brings
cal age of a documented impact event. The rela- it to an end. The characteristics of the Quaternary
tionship between hotspot tracks and the opening sequence of glaciations have been shown to be
of the Atlantic and Indian oceans (e.g. [80]) may broadly consistent with this scenario but model-
also be pertinent. ling, taking account of the longitudinal tempera-
Aspects of the deductions given here are evi- ture differences which develop between land and
dently of a tentative nature, the uncertainties lying ocean, is evidently called for.
mainly in those areas where the theory of plate These effects will operate whether or not the
tectonics is itself uncertain. However, it seems an Apollo population fluctuates due to spiral-arm
inevitable consequence of a large impact that per- effects. If it does, then episodicity of terrestrial
 259

phenomena, on characteristic time scales - 50 Myr seems unlikeiy to be a primary cause: dust veil
upwards, is predicted. The expected strength of optical deptlis probably saturate at missile masses
such episodicity is currently uncertain. Recurrence 1016g corresponding to impacts at - 1-Myr in-
of diverse geophysical phenomena on time scales tervals without catastrophic effects on the scale of
of this order has in fact frequently been claimed, the KT or Permo-Triassic extinctions, and the
along with synchronicities between, say, tectonic atmospheric data assembled by Lamb indicate that
activity and periods of high reversal frequency (see partial clearing due to latitudinal drift is expected
Seyfert and Sirkin [59] for a qualitative discussion). within - 0 . 5 years. Further, certain organisms
Known direct connections between specific im- which should have been extinguished by prolonged
pacts and geophysical changes are very sparse. darkness survived across the KT boundary [62].
Some well-established correlations between field Blast wave from a large impact is lethal, however,
reversals and the deposition of tektites have been over hemispheric dimensions, and this creates
mentioned. For larger impacts, an attempt was either a catastrophic or minor perturbation of the
made by the authors [15] to correlate termination biosphere depending on accident of geography
of geological periods with large known or probable (compare such an impact now on (0 °, 0 °) longi-
impact structures. While some suggestive connec- tude and latitude with one on (180 °, 0°)).
tions can be made the dating of periods is not yet Both the suddenness of the dinosaur extinctions
adequate for definitive statements. The geochemi- and their synchronicity with ocean extinctions im-
cal evidence at the KT boundary [7-10] seems to plicit in these models have been challenged and
point strongly to the occurrence of a large impact are the su~ect of controversy [63-65,76]. How-
65 Myr ago, but closely coincident in time with ever, if bombardment induces the geophysical dis-
this was the creation of the Deccan Traps 65-60 turbances discussed here the causal link between
Myr ago and the Brito-Arctic flood basalts, along impacts and extinction may become intractably
with a global synchronous discharge of hotspots complicated. For example, Pollack et al. [24] have
[60]. All of this occurred in the middle of the shown that doubling the CO~ content of the atmo-
mixed magnetic interval of polarity reversals 70-50 sphere will increase the mean global temperature
Myr ago: that is, the onset of the polarity reversals by 2.4°K neglecting feedback effects which might
preceded the "giant meteorite" by 5 Myr. That this increase this, in current circumstances, by a factor
is not mere coincidence is suggested by the fact, 4. Since a large volcanic explosion will increase
noted in the previous section, that the great Per- the atmospheric CO 2 by 10 5, several 105 volcanic
mo-Triassic extinction of 225 Myr also occurred explosions over a short space of time might be
within another mixed interval 230-204 Myr. The sufficient to exterminate at least the dinosaurs
implications are that, first, an impact of 65 Myr through the heat-infertility linkage envisaged by
ago caused the intense global vulcanism of the McLean [11]. Thus both fast catastrophe (extinc-
subsequent 5 Myr; and second, the impact oc- tion within 10 years) and slow catastrophe (within
curred not in isolation but in the midst of a 105 years) impact models are possible.
pattern of bombardment of - 2 0 Myr duration It might be hoped that an interstellar missile
corresponding to the mixed magnetic interval. would leave a chemical signature. One possibility
The occurrence of 3-6 impact episodes within would be to look for anomalous U isotope ratios
the Phanerozoic, with missiles of several l0 s Mt in significant strata (we are indebted to L. Alvarez
impact energy, was proposed as the cause of mass and F. Asaro for this suggestion). An interstellar
extinctions [15]: a variety of possible lethal mecha- comet may represent the aggregation of a wide
nisms was suggested (blast wave, ozone depletion, range of galactic material in time and space, and
collapse of food chains through obscuration of its chemical age will generally equal the cycling
sunlight). Subsequently Hs0 [61] has argued for a time of the interstellar medium. This is usually
KT comet impact and oceanic poisoning coupled taken as 3-5000 Myr and moderate U isotope
with blast, while Alvarez et al. [7] have preferred anomalies might be expected. However, a missile
the collapse of food chains hypothesis. The latter will excavate 102-103 times its own volume of
260

material and any such anomaly will be corre- 6. Discussion

spondingly diluted. Since U isotope ratios are gen-
erally only measured to - 1 % accuracy (O. van Despite the uncertainties, a theory of episodic
Breemen, personal communication) it seems bombardment is clearly not without support in the
doubtful that anomalies could be detected in strata geophysical field. In a recent review, Cloud [69]
deposited worldwide. However, for very large has remarked on the unprecedented synthesis of
structures such as the Sudbury or Popigai craters, geology and solid-state geophysics achieved
where much of the missile material may remain in through the unifying concept of plate tectonics. It
situ, this might provide a test. Large U isotope would be a considerable advance towards further
anomalies in several meteorites have been claimed unification if as a result of the present theory,
[66] and disputed [67]. If these w e r e real, U plate tectonics were seen to be actively controlled
anomalies in the ground would be evidence for an through episodic bombardments and hence ulti-
extra-terrestrial not just an interstellar connection. mately through galactic phenomena. And though
It has been suggested that the Solar System ratios hardly touched on in this paper the apparent
of Ir and Os detected at the KT boundary imply a correlation of great impact with the end Creta-
Solar System missile but this is not necessarily ceous at one extreme, through more recent
correct: the r-process environments which produce planktonic extinctions correlated with tektites, to
these elements, possibly in supernovae or red the coincidences between human evolutionary
giants, may individually produce variations - 2 : 1 crises and magnetic field excursions [72] at the
in the isotope ratios, but with - l 0 8 such stars other, suggest that similar considerations may
having appeared in the Galaxy within its chemical apply also to biological evolution. It has been
cycling time, the dispersion from one interstellar suggested for example that the palaeontological
object to another, assuming complete mixing, will record is better understood through Cuvier than
be ~< 2 × l0 -4 in the isotope ratios, well below the Darwin [70]. However, prospects are no substitute
level of detectability. A conceivable prediction is for proof, and for the p~esent theory to be fully
that the 12C/13C ratio of an interstellar planetesi- established, it will be necessary to identify particu-
mal will differ from the terrestrial value of 89 since lar impact events a n d demonstrate quantitative
this ratio is observed to vary appreciably from consistency with their consequences as recorded in
place to place in the molecular clouds within which or on the ground. Perhaps the most accessible tests
it grows [68]; a difference of several percent would are those relating to low mass projectiles at com-
be expected at significant strata. Even here, how- paratively recent epochs. Such tests are not without
ever, the situation is not clear cut for it is possible difficulty because the lower the energy and
the present theory of galactic enrichment is incor- momentum level, the greater the risk of confusion
rect. Recent observations of the denser regions of with purely terrestrial processes. However, if, at a
molecular clouds indicate a IEc/13C ratio surpris- certain level, perturbations of the terrestrial climate
ingly close to the Solar System value [77]. It is and geomagnetic field have impacts as a common
possible that material processed through normal trigger, the most conspicuous recent discontinui-
stars also cycles through some other physical en- ties in mean temperature and magnetic field may
vironment, for example supermassive stars, giving be as good candidates for study as any.
rise to primordial stellar material whose chemical If, for example, the northern limits of foresta-
mix does not vary with epoch. It is, therefore, tion during post glacial times are a good measure
surprisingly difficult to find critical isotope tests of global climate (e.g. [28, chapter 16 and refer-
for an interstellar origin: but it is to be hoped that ences therein]), there was an apparently significant
such tests will be forthcoming, for as McCrea [1] climatic deterioration around 5000 years B.P. as
has remarked, "the notion that in our journey well as others around 3600, 3000 and 700 years
round the Galaxy we may be able on Earth to B.P. (radiocarbon scale). The secular variation of
sample its material at every stage has potentialities the geomagnetic field intensity [71] also shows a
almost beyond belief". marked discontinuity around 5000 years B.P. In
 261

addition to these observations, the Greenland and 11 D.M. McLean, A test of terminal Mesozoic "catastrophe",
Antarctic ice cores are now known to provide a Earth Planet. Sc]. Lett. 53 (1981) 103.
12 P.R Vail, R.M. Mitchum and S. Thompson, in: Seismic
year-by-year record of stratospheric fallout; large
Stratigraphy-Applications to Hydrocarbon Explorations,
depositions of acidic material being frequently as- C.E Payton, ed. (American Association of Petroleum Geol-
sociated with known volcanic eruptions during the ogists, Tulsa, Okla., 1977) 83.
l a s t l 0 4 years. There are several significant fallouts 13 B.P. Glass and M.J. Zwart, in: Stratigraphic Micro-
in this period which are presently of unknown paleontology of Atlantic Basin and Borderlands, F.M.
Swain, ed. (Elsevier, Amsterdam, 1977) 553.
origin [22], and among the largest of these was one
14 J.A. O'Keefe, Tektites and their Origin (Elsevier, Amster-
3000 B.C. and another ~ 1250 A.D. The predic- dam, 1976) 180.
tion that the correlated effects around 3000 B.C. 15 W.M. Napier and S.V.M. Clube, A theory of terrestial
are of impact origin has been the subject of a catastrophism, Nature 282 (1979) 455.
separate astronomical investigation [73], but the 16 S.V.M. Chibe and W.M. Napier, Spiral arms, comets and
terrestial catastrophism, Q.J.R. Astron. Soc. (in press).
detection of composition anomalies consistent with
17 R.A.R. Grieve and M.R. Dence, The terrestial cratering
an extraterrestrial origin at either of these two record, II. The crater production rate, Icarus 38 (1979) 230.
levels would now clearly be of considerable inter- 18 E.M. Shoemaker, J.G. Williams, E.F. Helin and RF. Wolfe,
est. The acidity peaks generally relate to sulphuric Earth-crossing asteroids: orbital classes, collision rates with
acid and the products of volcanic eruption, but if Earth, and origin, in: Asteroids, T. Gehrels, ed. (University
of Arizona, 1979) 253.
an extraterrestrial missile interacts with the atmo-
19 W.M. Napier and M. Staniucha, Interstellar planetesimals,
sphere, concentrations of nitric acid are likely. I. Dissipation of a primordial cloud of comets by tidal
Typically 10~2g SO 2 are ejected during a large encounters with massive nebulae, Mort. Not. R. Astron.
volcanic outburst, as against 10~2g NO~ produced Soc. 198 (1982) 723.
by a 103-Mt fireball [21]. Favorable tests in this 20 D.E. Gault and E.D. Heitowit, in: Proc. 6th Hypervelocity
Impact Symposium Vol. 2 (Firestone Rubber Company,
respect at these levels might imply the ice cores
1963) 419.
carry a record of recent low-level impacts. 21 S. Glasstone and P.J. Dolan, The Effects of Nuclear
Weapons (U.S. Department of Defense, t977).
22 C.U. Hammer, H.B. Clausen and W. Dansgaard, Greenland
References ice sheet evidence of post glacial volcanism and its climatic
impact, Nature 288 0980) 230.
23 B.G. Hunt, A simulation of the possible consequences of a
1 W.H. McCrea, Long time-scale fluctuations in the evolution
volcanic eruption on the general circulation of the atmo-
of the Earth, Proc. R. Soc. London, Ser. A, 375 (1981) 1.
sphere, Mort. Weather Rev. 105 (t977) 247.
2 M.A. Whyte, Turning points in Phanerozoic history, Nature
24 J.B. Pollack, O.B. Toon, C. Sagan, A. Summers, B. Baldwin
267 (1977) 679.
and W. Van Camp, Volcanic explosions and climatic change:
3 D.J. McLaren, Time, life and boundaries, J. Palaeontol. 44
a theoretical assessment, J. Geophys. Res. 81 (1976) 1071.
(1970) 801.
25 M.J. Budyko, Climatic change, Sov. Geogr. 10 (1969) 429.
4 M.W. De Laubenfels, Dinosaur extinction: one more hy-
26 W.D. Sellers, A new global climatic model, J. Appl.
pothesis, J. Palaeontol. 30 (1956) 207.
Meteorol. 12 (1973) 241.
5 E.J. Opik, On the catastrophic effects of collisions with
27 H. Flohn, On time scales and causes of abrupt paleocli-
celestial bodies, Irish Astron. J. 5 (1958) 34.
matic events, Quaternary Res. 12 (1979) 135.
6 W. Whiston, A New Theory of the Earth (C.U.P. 1708).
28 H.H. Lamb, Climate, Vol. 2 (Methuen, London, 1977).
7 L.W. Alvarez, W. Alvarez, F. Asaro and H.V. Michel,
29 C.S.M. Doake, Climatic change and geomagnetic field re-
Extraterrestial cause for the Cretaceous-Tertiary extinction:
versals: a statistical correlation, Earth Planet. Sci. Lett. 38
experiment and theory, Science 208 (1980) 1095.
(1978) 313.
8 J. Smit and J. Hertogen, An extra-terrestial event at the
30 G. Wollin, G.J. Kukla, D.B. Ericson, W.B.F. Ryan and J.
Cretaceous-Tertiary boundary, Nature 285 (1980) 198; 287
Wollin, Magnetic intensity and climatic changes 1925-1970,
(1980) 760.
Nature 242 (1973) 34.
9 R. Ganapathy, A major meteorite impact on the Earth 65
31 R.W. Fairbridge, Global climate changes during the 13500
million years ago: evidence from the Cretaceous-Tertiary
B.P. Gothenburg magnetic excursion, Nature 265 (1977)
boundary clay, Science 209 (1980) 921.
430.
10 F.T. Kyte, Z. Zhou and J.T. Wasson, Siderophile-enriched
32 C.E.P. Brooks, Climate through the Ages (Ernest Benn Ltd.,
elements from the Cretaceous-Tertiary boundary, Nature
London, 1949).
288 (1980) 651.
262

33 C.J. Yapp and S. Epstein, Climatic implications of D / H 56 C. Sagan, Is the early evolution of life related to the
ratios of meteoric water over North America (9500-22000 development of the Earth's core?, Nature 206 (1965) 448.
B.P.) as inferred from ancient wood cellulose C-H hydro- 57 J.A. Jacobs, in: 75th Anniversary Volume (Ebro Observa-
gen, Earth Planet. Sci. Lett. 34 (1977) 333. tory, 1980).
34 W.R. Ruddiman and A. McIntyre, Warmth of the subpolar 58 C. Lomnitz, Global Tectonics and Earthquake Risk (Else-
north Atlantic ocean during northern hemisphere ice sheet vier, Amsterdam, 1974).
growth, Science 204 (1979) 173. 59 C.K. Seyfert and L.A. Sirkin, Earth History and Plate
35 F. Hoyle and N.C. Wickramasinghe, Comets, ice ages and Tectonics (Harper Row, New York, N.Y., 1979).
ecological catastrophes, Astrophys. Space Sci. 53 (1978) 60 P.R. Vogt, Evidence for global synchronism in mantle plume
523. convection and possible significance for geology, Nature
36 W.F. Ruddiman, A. McIntyre, V. Niebler-Hunt and J.J. 240 (1972) 338.
Durazzi, Oceanic evidence for the mechanism of rapid 61 K.J. Hsii, Terrestrial catastrophe caused by cometary im-
northern hemisphere glaciation, Quaternary Res. 13 (1980). pact at the end of Cretaceous, Nature 285 (1980) 201.
37 B. Saltzmann and R.E. Moritz, A time-dependent climatic 62 D.A. Russell, Proc. 2nd K-Tec Conf., Ottawa, Ont. (1981).
feedback system involving sea-ice extent ocean temperature 63 J.D. Archibald, The earliest known Palaeocene mammal
and CO 2, Tellus 32 (1980) 93. fauna and its implications for the Cretaceous-Tertiary tran-
38 H.B. Gordon and D.R. Davies, Mathematical prediction of sition, Nature 291 (1981) 650.
climatic change, Nature 253 (1975) 419. 64 R.F. Butler and E.H. Lindsay, Magnetostratigraphy, bio-
39 D. Pollard, A.P. Ingersoll and J.G. Lockwood, Response of stratigraphy and geochronology of Cretaceous-Tertiary
zonal climate-ice sheet model to the orbital perturbations boundary sediments, Red Deer Valley, Nature 284 (1980)
during the Quaternary ice ages, Tellus 32 (1980) 301. 375.
40 J. Imbrie and J.Z. Imbrie, Modelling the climatic response 65 D.A. Russell, The enigma of the extinction of the dinosaurs,
to orbital variations, Science 207 (1980) 943. Ann. Rev. Earth Planet. Sci. 7 (1979) 163.
41 J.D. Hays, J. Imbrie and N.J. Shackleton, Variations in the 66 M. Tatsumoto and T. Shimamura, Isotopic composition of
Earth's orbit: pacemaker of the ice ages, Science 194 (1976) uranium in Allende and other chordites, Meteoritics 14
ll21. (1979) 543.
42 T.M.L. Wigley, Climate and paleoclimate; what we can 67 J.H Chen and G.J. Wasserburg, The isotopic composition
learn about solar variations, Solar Phys. 74 (1981) 745. of uranium and lead in Allende inclusions and meteoritic
43 M.A. Kominz and N.G. Pisias, Pleistocene climate: de- phosphates, Earth Planet. Sci. Lett. 52 (1981) 1.
terministic or stochastic?, Science 204 (1979) 171. 68 F. Combes, E. Falgarone, J. Guibert and Nguyen-Q-Rieu,
44 G.H. Denton, R.K. Armstrong and M. Stuiver, in: The Late CO observations of interstellar clouds: isotopic ratios, As-
Cenozoic Glacial History of Antarctica, K.K. Turkekian, tron. Astrophys. 90 (1980) 88.
ed. (Yale University, 1971) 267. 69 P. Cloud, Beyond plate tectonics, Am. Sci. 68 (1980) 381.
45 E.J. Butler and F. Hoyle, On the effects of a sudden change 70 N.D. Newell, in: Uniformity and Simplicity, C.C. Albritton,
in the albedo of the Earth, Astrophys. Space Sci. 60 (1979) ed., Geol. Soc. Am. Spec. Paper 89 (1967).
505. 71 P.J. Smith, The intensity of the ancient geomagnetic field: a
46 F. Hoyle, Ice (Hutchison, 1981). summary of conclusions, in: Palaeogeophysics, S.K.
47 T. Rikitake, Oscillations of a system of disk dynamos, Proc. Runcorn, ed. (Academic Press, 1970) 79.
Cambridge Philos. Soc. 54 (1958) 89. 72 J.S. Kopper, Thesis, University of Columbia.
48 K. Ito, Chaos in the Rikitake two-disc dynamo system, 73 S.V.M. Clube and W.M. Napier, The Cosmic Serpent
Earth Planet. Sci. Lett. 51 (1980)451. (Faber, 1982).
49 D. Gubbins and T.G. Masters, Driving mechanisms for the 74 G.W. Wetherill, Evolution of the Earth's planetesimal swarm
Earth's dynamo, Adv. Geophys. 21 (I 979) 1. subsequent to the formation of the Earth, Proc. 8th Lunar
50 W.V.R. Malkus, Motions in the fluid core, Proc. Int. Sch. Sci. Conf. (1977) 1.
Phys. 50 (1971) 38. 75 A.R. Crawford, Tektites probably wholly terrestrial and
51 F.D. Stacey, Physics of the Earth (J. Wiley andSons, 1969). related to continental movement, Geol. Mag. 116 (1979)
52 B.P. Glass and B.C. Heezen, Tektites and geomagnetic 261.
reversals, Sci. Am. 217 (1967) 33. 76 D.V. Ager, Marine cycles in the Mesozoic, J. Geol. Soc.
53 P.V. Florenskii, V.P. Perelygin, M.L. Bazhenov, D.D. London 138 (1981) 159.
Lkhagvasuren and S.G. Stetsenko, A complex determina- 77 W.H. McCutcheon, R.L. Dickman, W.L.H. Shuter and R.S.
tion of the age of the Zhamanshin meteorite crater, Sol. Roger, The 13CO/CIsO ratio in interstellar dark clouds:
Syst. Res. 13 (1980) 137. evidence for isotopic fractionation, Astrophys. J. 237 (1980)
54 K.M. Creer, On a tentative correlation between changes in 9.
the geomagnetic polarity bias and reversal frequency and 78 D.M. McLean, Abstr., A.A.A.S. 147th Natl. Meet., Toronto,
the Earth's rotation through Phanerozoic time, in: Growth Ont. (1981).
Rhythms and the History of the Earth's Rotation, G.D. 79 D.M. McLean, A terminal Mesozoic "Greenhouse": lessons
Rosenbcrg and S.K. Runcorn, cds. (Wiley, London, 1975) from the Past, Science 201 (1978) 401.
293. 80 W.J. Morgan, Hotspot tracks and the opening of the Atlantic
55 R.J. Uffen, Influence of the Earth's core on the origin and and Indian Oceans, in: The Sea, Vol. 7, C. Emiliani, ed.
.evolution of life, Nature 198 (1963) 143. (1980).