Abbildungen der Seite


University of Chicago

Most meteorites come from a small number of parent bodies (6 to 11), with radii mainly between 100 and 300 km. The most likely sources of meteorites are seven asteroid families between 1.9 and 2.8 AU, whose eccentricities and inclinations are high enough to permit their collisional debris to cross the orbit of Mars. Comets are too small and too numerous to serve as a source of the major meteorite classes, but they may well be an important source of micrometeorites and carbonaceous chondrites.

Asteroids and comets are the two most plausible sources of meteorites. However, there is as yet no agreement on the relative importance of the two. Some authors believe that the great majority of meteorites come from comets (Öpik, 1965, 1968; Wetherill, 1968a). Others argue that they come mainly from asteroids (Anders, 1964; Wood, 1968). Let us review the principal clues to the nature of meteorite parent bodies, as obtained from the meteorites themselves.



Wasson has shown in an impressive series of papers (Wasson, 1969, 1970, and earlier publications cited therein) that most iron meteorites fall into 11 discrete groups, differing from each other in chemical composition and structure. In four-dimensional composition space (Ga, Ge, Ir, and Ni) these groups form very compact, well-defined clusters. Many of these clusters can also be recognized by other criteria, such as radiation age and shock effects (Jaeger and Lipschutz, 1967; Voshage, 1967). The observed infall rate of these meteorites requires a source at least 1 to 100 km3 in extent; and it therefore seems highly probable that each group represents either a separate parent body, or a sizable, compositionally distinct region within a parent body.

Clues to the size of these bodies have been obtained from the Widmanstätten pattern of iron meteorites. The formation of this pattern, long a subject of controversy, is now well understood, thanks to the work of Wood (1964), Goldstein and Doan (1971), and Goldstein and Ogilvie (1965). As a result of this understanding, it has been possible to estimate cooling rates of iron meteorites through the range in which the pattern formed, 700 to 300 C. The results for nearly 300 iron meteorites and pallasites range from 0.4 to 500 K/million yr, with the majority of values lying between 1 and 10 K/million yr (table I). These metallographically determined cooling rates are a direct clue to the size of the parent bodies, because the cooling rate of a planetary object is a sensitive function of size. Fricker et al. (1970) have shown that the above cooling rates correspond to radii between 10 and 500 km, with most values lying between 100 and 300 km. Some of Wasson's groups show little spread in cooling rates, which suggests that they come from a (nearly isothermal) core. Others show a nearly tenfold variation, which may imply that they are derived from a series of isolated iron pools extending from the center to the surface (Urey, 1966). At cooling rates less than 7 K/million yr, the radius depends strongly on whether the outer layers of the body are compositionally uniform or differentiated, with radioactive elements concentrated near the surface. Two different radii are therefore given in table I, for the uniform and differentiated CaSCS. The cooling rates have been confirmed by an independent method: fissiontrack measurements in the Toluca iron meteorite (Fleischer et al., 1968). By measuring tracks from extinct, 82 million yr 24*Pu in three minerals differing in track retention temperatures, Fleischer et al. obtained three points on a cooling curve for Toluca. The cooling rate found, 1.1+8% K/million yr, agrees well with the metallographically determined value, 1.6 + 0.6 K/million yr

TABLE I.—Parent Bodies of Iron Meteorites

Chemical Cooling Number Radius, km" Radiation
groupa rate, of - age, d
K/million yr members Uniform Differentiated million yr
body body
I . . . . . . . . . 1 to 3.5 87 150 + 30 230 +70 900
IIA . . . . . . . . . . . . . . . . . . . . 29 l . . . . . . . . . . . . . . . . . . . . . . . <100
IIB . . . . . . . . . . . . . . . . . . . . 11 | . . . . . . . . . . . . . . . . . . . . . . . . 1000
IIC . . . . . . 100 to 250 7 25 +5 25 +5 ! . . . . . . . . .
IID . . . . . . 1 to 2 6 165 + 15 260 +40 ~1000
IIIA . . . . . . 1.5 to 10 24 170 250 700
IIIB . . . . . . 1 to 2 14 165 + 15 260 +40 700
IIIC . . . . . . 2 to 5 5 130 + 10 180 +40 700
IIID . . . . . . 2 to 5 5 130 + 20 180 +40 200
IVA . . . . . . 7 to 90 16 100 110 400
IVB . . . . . . 2 to 20 7 150 220 Variable

aWasson (1969, 1970, and earlier publications cited therein.)
bColdstein and Short (1967a).

CFricker et al. (1970).

dVoshage (1967).

(Goldstein and Short, 1967b). Although a few uncertainties remain, the metallographic cooling rates for iron meteorites are probably reliable to within a factor of 2 to 3. The actual number of parent bodies involved may be as small as six. Though the subgroups IIIA to IIID are chemically distinct from each other, the differences are not drastic and in fact were not noticed until very precise analyses became available. Perhaps all came from the same body. All give essentially the same radius r, at least three of the four were produced in a single collision, judging from their common radiation age of ~700 million yr and the ubiquity of strong shock effects (Jaeger and Lipschutz, 1967). Similarly, IIA, B, and D may come from a single body. Assuming independent bodies for I, IIC, IVA, and IVB, we are thus left with only six bodies. Apparently more than 80 percent of all iron meteorites came from 11 bodies at most (possibly as few as six), of which all but 1 were larger than 100 km in radius.


According to chemical criteria, chondrites are divided into five groups (table II). The hiatuses separating these groups are not as wide as those for irons, and hence there is less reason to conclude, on chemical grounds, that only five parent bodies are involved. Here we must rely on other evidence.

Among the meteorites in table II, the L-chondrites stand out in having a preponderance of short K-Ar and U-He ages, discordant between l and 4 aeons, but becoming concordant at ~0.5 aeon. These short ages are correlated with shock and reheating symptoms. Detailed analysis of the data suggests that at least two-thirds of these meteorites were involved in a major collision 520 + 60 million yr ago, which caused partial or complete outgassing of “9Ar and “He (Anders, 1964; Heymann, 1967; Taylor and Heymann, 1969).

[merged small][ocr errors]

“Atom percentage of Fe?" in principal ferromagnesian silicate ((Mg, Fe)SiO3 in enstatite chondrites and (Mg, Fe)2SiO4 in all others).

bThe symbols H, L, and LL refer to total iron content, and stand for high, low, and low-low, respectively.

If this is true, then the L-chondrites come mainly from one or at most two bodies. Because L-chondrites are the most abundant class of meteorites now falling on Earth, it seems that the majority of chondrites, like the irons, come from a small number of bodies: probably no more than 10, perhaps as few as five. The total number of bodies need not be greater than perhaps six because some chondrites and irons may come from the same body. For example, the outgassing time of L-chondrites, 520 + 60 million yr, is rather close to the radiation age of group III irons, 600 to 700 million yr (Voshāge, 1967), and the systematic errors in the two dating methods are large enough to admit the possibility that both refer to the same event. Chemical resemblances between these two classes are sufficiently great to permit an origin in the same body. Wänke (1966) and Öpik (1968) have questioned the reality of the 520 million yr event, and the conclusion that most L-chondrites come from a single parent body. They suggest that the meteorite parent body was hot enough throughout its history to cause “He to partition between solid and pore space. On breakup, the “He in the pores would escape, leaving only a fraction of the total “He in the meteorite. If this fraction happened to be 8 percent, an apparent U-He age of 520 million yr would result. However, this explanation appears to be untenable. It does not account for the fact that U-He ages of 0.5 aeon are sometimes associated with concordant K-Ar ages of 0.5 aeon, and sometimes with discordant values as high as 3 aeons. It predicts a correlation between U-He age and porosity that is not observed, and yet fails to explain the correlation between U-He or K-Ar age and shock effects that is observed. (This correlation has been confirmed by several authors: Carter et al., 1968; Christophe, 1969; Taylor and Heymann, 1969; Van Schmus and Ribbe, 1968; Wood, 1967.) Finally, the reality of the 520 million yr outgassing event has been confirmed by the 49Ar/39Ar method (Turner, 1969). Stepwise heating of six L-chondrites with nominal K-Ar ages from 1.0 to 1.85 aeons showed that the least retentive *9Ar sites in each meteorite had been completely outgassed in a single event 500 + 30 million yr ago; the higher ages for the bulk meteorites represent incomplete outgassing of the more retentive sites. Two other L-chondrites gave shorter ages (305 + 30 million yr), implying partial or complete outgassing at a later time." The size of the chondrite parent bodies can be estimated again from cooling rates, as for the irons. Wood (1967) has shown that most ordinary chondrites have cooled through 500 K at rates between 2 and 10 K/million yr. These limits correspond to depths of 40 to 150 km in bodies of R P 150 km and 20 to 80 km in bodies of R = 90 km. Similar cooling rates, 5 to 9 K/million yr, have been estimated from 12°Xe diffusion (Manuel et al., 1968). They are also supported by various estimates of cooling times. (See Anders, 1971, for discussion and references.) It would seem that the parent bodies of ordinary chondrites were of about the same size as those of the irons. There are a few skeletons in this closet, however. Unequilibrated chondrites of three chemical groups and type III carbonaceous chondrites gave lower cooling rates, 0.2 to 1 K/million yr, corresponding to depths of 70 to 150 km in bodies >400 km in radius. This is about the size of Ceres; and though one cannot rule out the possibility that another Ceres-sized asteroid once existed but was destroyed, it does not seem very plausible that this one body should be the source of the least recrystallized, most primitive meteorites, and from four chemically distinct classes at that. Perhaps the metallographic method becomes unreliable in systems containing stony phases in addition to metal. Exceedingly low cooling rates were obtained for two other silicate-containing classes: pallasites, 0.4 K/million yr, and mesosiderites, 0.1 K/million yr. The latter corresponds to a temperature drop of only 450 K during the entire age of the solar system.



Comets are too numerous and too small to serve as the principal source of meteorites. Öpik (1965) has estimated the number of extinct, short-period comets with aphelion distance smaller than 4.94 AU as 2 X 10° to 10°. A significant fraction of these must have perihelia less than 1 AU. Even if one assumes that only a special subset of this population (e.g., low-velocity objects) can contribute meteorites, the resulting number far exceeds the apparent number of meteorite parent bodies, about 6 to 10. Available estimates of comet sizes (Roemer, 1971; Whipple, 1963) show them to be one to two orders of magnitude smaller than the meteorite parent bodies. Larger comets undoubtedly exist, but it is difficult to see how one giant comet, disrupted 520 million yr ago, could furnish half of Earth's meteorite influx. It is also hard to reconcile the fragility and high volatile content of comets with the prolonged high-temperature history of meteorite parent bodies and with the texture, chemistry, and mechanical strength of meteorites.


At first sight, the small number of meteorite parent bodies would seem to be incompatible with an asteroidal origin because asteroids, too, are very numerous. More than 4000 are known and at least 10 times as many undiscovered ones are thought to exist in the telescopically observable size range alone. However, two factors very greatly reduce this number. To change a typical asteroidal orbit into a meteoritic one, an acceleration of about 6 km/s is required. Cratering theory and experiments show that only a minute fraction of the ejecta in a hypervelocity impact can be accelerated to this velocity, and because half the total energy appears as heat, such material will be vaporized. The absence of lunar basalts from Earth's meteorite collections shows that acceleration of rocks to P2.4 km/s is indeed a very improbable process. Thus the majority of asteroids cannot contribute to Earth's meteorite influx.

« ZurückWeiter »