Abbildungen der Seite

low to give a short enough lifetime against collisional destruction; e.g., 107 yr. However, there is good reason to believe that the destruction lifetime is indeed on this order. Dohnanyi (1969) has reexamined the problem using an improved mass distribution function and cratering relations. He obtains mean destruction lifetimes of 3 to 10 million yr for objects 10 to 100 cm in diameter. Mazor et al. (1970) and Herzog (1970) have pointed out the curious fact that the radiation age distributions of meteorites show cutoffs related to crushing strength: ~15 million yr for the friable carbonaceous chondrites, ~60 million yr for all other stones, ~200 million yr for stony irons, and ~2 billion yr for irons. It appears that the age distribution is indeed governed by collisional destruction. Cutoffs of exactly the right order have been produced in Monte Carlo calculations, using a destruction lifetime of 10 million yr (Mellick and Anders, unpublished). Of course, if the majority of stony meteorites are destroyed by collisions, a correspondingly more intense source is needed to maintain the meteorite flux observed on Earth (10% to 10° kg/yr). The potential reservoir, from figure 1, is probably on the order of 101° to 1020 kg, so that even an extraction efficiency as low as 10-2 to 10−3 would suffice to maintain this flux for 10° to 1019 yr. However, if the correlations of a (the argument of perihelion) and e noted by Wetherill (1968a) prevent node and perihelion from coinciding even over periods of >107 yr, and even for ejecta, then there may indeed be a problem. Another observation to be explained is the predominance of p.m. falls among the chondrites. Wetherill (1968b, 1969) has pointed out that this requires a large orbit of low inclination. Moreover, the meteorite must be captured by Earth during the first few passes, otherwise a decreases, i increases, and the a.m./p.m. asymmetry is lost. Wetherill suggested that a special class of low-velocity, short-period, cometary orbits with aphelia near Jupiter would be suitable because objects in such orbits, if not quickly captured by Earth, are soon eliminated by Jupiter. However, it is difficult to envision circumstances where this type of orbit would dominate over more conventional short-period cometary orbits with higher geocentric velocities and/or smaller aphelia. It seems that the a.m./p.m. asymmetry can be equally well explained by the asteroidal model if collisional destruction is invoked to prevent “evolution” of the orbit by repeated close encounters with Earth. With a destruction lifetime of 3 to 10 million yr (Dohnanyi, 1969), meteorites will be captured by Earth in the first few passes, if at all. An unsolved problem still remaining is the relatively high frequency of meteorites with high geocentric velocities, U. = 0.5 to 0.7. Data are limited and of variable quality, but it appears from the available information on photographic and visual meteorite orbits (Millman, 1969), dense (“asteroidal”) meteors, and Apollo asteroids that perhaps one-third of all meteorites have Velocities in this range. Monte Carlo calculations for all Mars asteroid families give such velocities in much lower abundance (Anders and Mellick, 1969; Mellick and Anders, unpublished). In principle, the required acceleration could be achieved by an appulse to Jupiter, but such appulses lead to crossings and subsequent rapid elimination of the object. Perhaps commensurabilities or other factors stabilize some types of large orbit long enough for Earth capture to compete with Jupiter ejection. A theoretical investigation of this problem would be very desirable.


If most meteorites come from Mars asteroids, where then is the cometary debris? Three major clues are available: meteors, Apollo asteroids, and meteoritic material on the Moon.


It appears that the majority of photographic meteors, including the Prairie Network fireballs, are of cometary origin (McCrosky”). An asteroidal component seems to be present (Ceplecha, 1966; Kresák, 1969a,b) but is clearly subordinate in this mass range.

Apollo Asteroids

The Apollo asteroids seem to fall into two groups differing in geocentric velocity (table IV). Anders and Arnold (1965) suggested that the low-velocity group was asteroidal and the high-velocity group, cometary. Some support for this division has been obtained by Gehrels and his associates. The “cometary”

[merged small][ocr errors][merged small][ocr errors]

object 1566 Icarus shows only a minor variation (<0.24 mag) in its lightcurve, implying a nearly spherical shape, to <10 percent (Gehrels et al., 1970). According to Dohnanyi (1969), an object 0.54 km in radius, situated in the asteroid belt, has a lifetime of only 2.5 x 108 yr against collisional destruction. It seems inconceivable that an object as small as Icarus could have maintained its spherical shape for 4.5 aeons if it had resided in the asteroid belt during that time. Oort's comet belt, on the other hand, would be an acceptable, relatively collision-free place of storage. The asteroidal object 1620 Geographos, on the other hand, has a strongly elongated shape, with axial ratio 3.4:1 (Gehrels et al., 1970). This is entirely reasonable for an object that spent its entire life in the asteroid belt. If geocentric velocity is accepted as the criterion, three of the objects in table IV are cometary and seven asteroidal. (Comet Encke must be omitted because its discovery was aided by its light emission.) On a mass basis, the cometary contribution would seem to be smaller than 30 percent, but in view of the limited statistics, not much can be made of this trend.


A number of trace elements (Au, Ir, Bi, Te, etc.) are enriched in Apollo 11 lunar soils and breccias relative to igneous rocks, apparently reflecting addition of a meteoritic component (Ganapathy et al., 1970). From the abundance pattern it appears that this component consists largely of primitive, “carbonaceous-chondrite-like” material. The amount is about 1.9 percent, corresponding to an average influx rate of 4 X 10-9 g·cm−2-yr-". This agrees within a factor of 3 with a similar estimate for Earth, based on the Ir and Os content of Pacific and Indian Ocean sediments (Barker and Anders, 1968).

Apollo 12 soils collected some distance away from craters showed a very similar pattern, whereas those collected on crater rims gave a different pattern, resembling fractionated meteorites (irons, ordinary chondrites) in their low abundance of Bi, for example (Laul et al., 1971).

Six different impacts have thus far been characterized, and it seems that five of them were caused by fractionated meteorites (table V). On the basis of these

TABLE V.—Meteorite Impacts on Moon
[Laul et al., 1971)

Event Crater Projectile
diameter, km composition

12017 Glass o Primitive

Bench Crater .07 Fractionated
Head Crater .13 Fractionated
Surveyor Crater .18 Fractionated
10085 Anorthosite 2 Fractionated
Copernicus? 91 Fractionated

limited statistics, it appears that fractionated material dominates among the larger (>1 kg) bodies falling on both Earth and the Moon. Primitive material, on the other hand, seems to dominate among the subkilogram objects that are apparently responsible for the uniform blanket of C1-like material covering the Moon and for fireballs or meteors on Earth.

Probably the primitive component consists mainly of the debris of spontaneously disintegrating comets, with an unknown asteroidal contribution. Comets are rich in volatiles and hence almost certainly are of primitive composition. However, any material accreted at temperatures below ~400 K is likely to have this composition. An additional source of such material may thus be asteroids from the outer part of the belt and the surface layers of all asteroids.


Great efforts have been made to understand the chemical and thermal history of meteorites, starting with Urey’s (1952, 1954) classic papers. It appears that the observed chemical fractionations, involving some 55 elements, are due to only four basic processes that occurred in the solar nebula during cooling from high temperatures. I have reviewed the subject in a recent paper (Anders, 1971) and will therefore merely summarize the model that best accounts for the evidence (fig. 2). Degree of condensation is plotted on the ordinate; and degree of retention, on the abscissa.

(1) An early condensate, containing refractory elements (Ca, Al, Ti, U,
Th, lanthanides, Pt metals, etc.) was partially lost from ordinary and
enstatite chondrites.
(2) After condensation of the remaining material to grains of 107* to
10-9 cm, some of the nickel-iron was lost, at a temperature around
700 K. During this and the following stages, the enstatite chondrites
apparently found themselves in a more reducing environment,
perhaps a gas phase of C/O > 0.9.
(3) During or after the metal loss, about 30 to 80 percent of the
condensate was remelted to millimeter-sized droplets by local
heating events on a time scale of seconds to minutes (probably
electric discharges; Whipple, 1966). Volatiles were lost from the
remelted material.
(4) The unremelted, fine-grained material continued to take up volatiles
from the nebula (Pb, Bi, Tl, In, etc.) and accreted with the remelted
material. Accretion took place at P = 10-4” atm and falling
temperatures, as indicated in figure 2. The values for carbonaceous
and enstatite chondrites are only rough estimates.

The five principal chondrite classes were affected by these processes to a markedly different extent. Carbonaceous chondrites were generally affected least, and enstatite chondrites, most. Presumably this reflects differences in place and time of formation. If we only knew the original location of their

[merged small][merged small][ocr errors][ocr errors][ocr errors][merged small][ocr errors][merged small][ocr errors][ocr errors][ocr errors][merged small][merged small][merged small][ocr errors]

Figure 2.-Proposed outline of chondrite formation in a cooling solar nebula (Larimer and Anders, 1970). Shaded areas represent condensed phase at end of each stage. Fraction condensed is shown on ordinate; amount remaining after fractionation, on abscissa. Four partial fractionations seem to have taken place: (1) loss of initial condensate at ~1300 K, (2) loss of metal at ~700 K, (3) remelting at ~500 to 600 K, and (4) accretion at 350 to 550 K. The three main chondrite classes presumably represent three to five different regions of the inner solar nebula. Boxes at bottom of figure give estimated accretion temperatures of each petrologic type and percent remelted, volatile-poor materials. Numbers refer to petrologic types within each chondritic class.

parent bodies, we could correlate all this information with heliocentric distance, and thus reconstruct the chemical and thermal history of the inner solar nebula in considerable detail.


I am indebted to Michael S. Lancet and Rudy Banovich for the preparation of figure 1. This work was supported in part by the U.S. AEC Contract AT(11-1)-382 and NASA Grant NGL 14-001-010.


Anders, E. 1964, Origin, Age, and Composition of Meteorites. Space Sci. Rev. 3, 583-714.

Anders, E. 1965, Fragmentation History of Asteroids. Icarus 4,399–408.

Anders, E. 1971, Meteorites and the Early Solar System. Ann. Rev. Astron. Astrophys. 9, 1-34.

« ZurückWeiter »