(1989);L . E. Kay, D. Marion, A. Bax, J. Magn. Reson. 84, 72 (1989);E . R. 1'. 47. S. LT'. Frslk n al., ihid. 112, 886 (1990). Zuidenveg and S. W. Fesik, Biochemistry 28, 2387 (1989); B . A. Messerle. G. 48. hi. Ikura, L. E. Kay, A. Bax, Biuchernisrry 29, 4659 (1990);L . E. K J ~iM, . Ikura, Wider, G. Orring, C. Weber, K. Wuthrich, J. Mags. Reson. 85, 608 (1989);M . R. Tschudin, A. Ba,,, J. i h n . Rexor~. 89, 496 (1990); L. E. Kay, M. Ikura, A. Ikura, A. Bax, G. M. Clore, A. M. Gronenborn, J. Am. Chrm. Sor. 112, 9020 BAT, i6id. 91, 84 (1991); M. Ikura and A. Bax, J. Biomol. LVIMR,in press; R. (1990); T. Frenkiel, C. Bauer, IM. D. Carr, B. Birdsall, J. Feeney. J. M a p . Reson. Powers, A. M. Gronenborn, A. Bax, J. .$tap. Reson., in press. 90, 420 (1990). 49. G. M. Clore, P. C. Driscoll, P. T. Wingfield, A. M. Gronenborn, 1. !%l. Biol 214. :6. D. ,Marion rt al., Bioch.ernistry 29, 6150 (19891. 811 (1990). 57. G. M. Clore, A. Bax, A. M. Gronenborn, J. Biornol. IVMR, in press. 50. M. Karplus, J. A m . Chern. Soc. 85, 2870 (1963); A. DeMarco, M. Llina, K. 53. P. C. Driscoll, G. M. Clore, D. Marion, P. T . Wingfield, A. M. Gronmborn, Wiirhr~chBiopolymers, 17,617 (1978);A . Pardi, IM. Billeter, K. Wuthrich, J. Mol. Biochemktry 29, 3542 (1990); P. C. Driscoll, A. M. Gronenborn, P. T. Wingtield, Biol. 180, 741 (1984). G. M. Clore, ibid., p. 4468. 51. I.. E. Kay and A. Bax, J. Magn. Rrson. 86, 110 (1990); A. M. Gronenborn, A. 39. M. Ikura, L. E. Kay, R. Tschudin, A. Bay, J . Magn. Rrson. 86,204 (1990); E. R Bax, P.T. Win'eld, G. M. Clore, FEBS Lett. 243,93 (1989);J . D . Forman-Kay, P. Zuiderwcg, L. P. McIntosh, F. W. Dahlquist, S. W. Fesik, ibid., p. 210. A.M. Gronenborn, L. E. Kay, P. T. Wingfield, G. M. Clore,BiochemLtry 29, 1566 40. G. M. Clore, A. B a , P. T. Wing!ield, A. M Gronenborn, Biochemixtry 23, 5671 (1990). (1990). 52. L. LMueUer, J. :Magn. Rrjon. 72, 191 (1987); A. Bax and L. E . Lerner, ibid. 79, 41. A. Baxetal., J. Alagn. Rcson. 87,620 (1990);L . E. Kay and M. Ikura, A. Bax, J. 429 (1988); D. Marion and A. Bax, ibid. 80, 528 (1988). Am. Chem. Soc. 112, 888 (1990). 53. We thank our many colleagues, past and present, who have contributed to the work 42. A. Bax, G. M. Clorc, A. ,M. Gronenborn, J. ~Magrz.Reson. 88, 42.5 (1990). carried out in our laboratory. Above all others, we thank A. Baw, with whom we 43. G. M. Clore, A. Bax, P. C DriscoU, P. T. Wingkid, A. M. Gronenborn, have shared numerous stimulating discussions, fruitful experiments, and a contin- Biochemktry 29, 8172 (1990). uous and most enjoyable collaboration in the best of scientific spirits. In addition, 44. L. E. Kay, G. M. Clore, A. Bax, A. M. Gronenborn, Science 249, 411 (1990). we thank W. Eaton and A. Szabo for Inany constructive and helpful comments 45. G. M. Clore. L. E. Kay., . A. Bax. A. M. Gronenborn. Biochemistry 30, 12 (1991). during the preparation of this manuscript. The work in the authors' laboratory was 43. E. R. P. ~ u i d e n v eA.~ , M. Petros, S. W. Fesik, E. T. Olejniczak, J . A ~ . hem. supported in part by the AIDS Targeted Anti-Viral Program of the Office of the Sol. 113, 370 (1991). Director of the National Institutes of Hedth.

Asteroid 1986 DA: Radar Evidence for a Metallic Composition

fcrmztion, the collisional evolution of the asteroid belt, and the Schoes fiom the near-Earth object 1986 DA show it to delivery of small bodies into Earth-crossing orbits. As compositions kc significantly more reflective than other radar-detect- of asteroids bear directly on these issues, much astronomical effort is ed asteroids. This result supports the hypcthesis that devoted to obtaining and interpreting information about asteroid 1986 DA is a piece of NiFe metal derived fiom the mineralogy. Many NEAs and MBAs have been classified on the bsis kterior of a much larger object that melted, ditt'erenti- of their photometric colors and albedos. Visible-infrared (VIS-IR) ~ted,cooled, and subsequently was disrupted in a cata- reflectance spectra have established important mineralogical charac- s-rophic collision. This 2-kilometer asteroid, which ap- teristics that limit the set of plallsible meteorite analogs for the most Fears smooth at centimeter to meter scales but extremely populous classes (3-7). For example, C asteroids contain hydrated irregular at 10- to 100-meter scales, might be (or have silicates, carbon, organics, and opaques and are analogous to CI1 been a part of) the parent body of some iron meteorites. and CM2 carbonaceous chondrites. The S asteroids contain pyrox- ene, olivine, and NiFe metal, and probably correspond to stony iron meteorites or ordinary chondrites or both. The M asteroids contain NiFe nletal or assemblages of enstatite and NiFe metal or a EAR-EARTAHS TEROIDS, LIKE METEORITES, ARE THOUGHT combination of both, and probably correspond to iron meteorites or to come primarily from mainbelt asteroids (1, 2). The enstatite chondrites or both. relations among meteorites, near-Earth asteroids (NEAs), Iron and stony iron meteorites are igneous assemblages derived I mainbelt asteroids (MBAs), and comets are central to our under- from differentiated parent bodies, whereas chondrites are relatively srmding of conditions in the primitive solar nebula, planetary primitive, undifferentiated assemblages, so the different candidate / meteoritic analogs to the M and S classes have disparate cosrnogonic implications. Radar observations can probe asteroid-nletcorite rela- tions because metal concentration influences radar reflectivity dra- i' S. J. Ostro and K. D. Rosema are at the Jet Propulsion Laboratory, California Institute matically and iron and stony iron meteorites are much more metallic ofTechnology, Pasadena, CA 91109. D. B . Campbell is at the National Astronomy and Ionosphere Center, Cornell University, Ithaca, NY 14853'. J. F. Chandler and I. I. than chondrites. Only nvo M NEAs, 1986 DA and 3554 h u n Shapiro are at the Harvard-Smithsonian Center for Astrophysics, Cambridge, MA (1986 EB), have been identified (8, 9). We report radar observa- j' 02138. A. A. Hine is at the National Astronomy and Ionosphere Center, Box 995, Arecibo, PR 00613. R. S. Hudson is in the ~iectricaland Computer Engineering tions of 1986 DA that constrain this object's physical properties, i Dipamnent, Washington Sratc Unirenity, Pullman, WA 99164. and we argue that it is mineralogically similar to iron meteorites.

RESEARCH ARTICLE 1399 Table 1. The orbit of asteroid years. Over longer time scales, 1986 DA's orbit might intersect the 1986 DA. These osculating orbital Epoch: 1986 Mar. 31.0 TDT = orbit of Earthas a consequence of secular pert&bations by the JD 2446520.5 elemrnts are based on analys~so f the planets, and for this reason 1986 DA is referred to as an Earth- combined radar and opt~>alastro- metric data, and arc referred to the Semimajor axis, a crossing asteroid (17 ). mean ecliptic 2nd rqu;noa nf Eccentricity. P Shape and suriacc structure. T l ~ cb andwid~liB uf an asteroid's B1950.0. Our ertimate of standard Inclination, i instantaneous echo power spectnim is related to the breadth DSi, 120ngirudeo f error in each of these elements is (measured normal to the line of sight) of the asteroid's pole-on about I in the last decimal place ascending node, - = shown. based in part on a coipari- .IL silhouette by B ( 4 7 ~D,,, cos 6)/(U)w,h ere P is the apparent son of solutions using different sets of 126.7040" rotation period, 6 is the angle between the asteroid's equatorial of planetay ephemerides. The mean perihelion, plane and the radar line of sight, and h is the radar wavelength. Mean anomall',h f 358.41 17" motion and period are basrd on the Measurements of echo edge frequencies as functions of rotation orbit solution and averaged over the Mean 0.2085" Per phase can be used to estimate the convex envelope, or "hull," of the 50-year span starting in 1986. day Period 4.728 ye~r pole-on silhouette as well as the offset -f,, of the frequency of hypothetical echoes from the asteroid's center of mass (COLM)f rom the ephemeris prediction. (The hull can be thought of as the shape Observations. 1986 DA was discovered by M. Kizawa (10) at of a rubber band stretched around that silhouette.) The COM's Shizuoka, Japan, on 16 February 1986. Two months later, we locatior. in projection within the hull and the Fourier series expan- observed it with the Arecibo Observatoq's 2380-MHz (13-cm) sion of the hull's shape depend onf,,, which is treated as a free radar. Our experimental techniques were essentially the same as parameter in the least-squares estimation (18). Application of pro- those used in other asteroid radar investigations (1 1-14). During 18 cedures described in those references t o our echo spectra yielded the to 20 April we did "cw" runs, in which we transmitted an hull estimate in Fig. 1A. The apparently modest elongation of thc unmodulated, continuous wave and measured the distribution of hull, 1.4 ? 0.2, is consistent with the ratio of brightness extrema ir. echo power as a function of frequency. The resultant echo spectra optical light curves. However, as illustrated in Fig. lB, the data1> can be thought of as one-dimensional images, or brightness scans, noise level precludes high-precision estimation of the hull's detaile; across 1985 D,4 through a slit parallel to the asteroid's appzrt-nt s?in shape, thzr is, the curvature of the hull's approuching limb as : vector. We devoted 21 April to "ranging1 runs, in which a time- function nf roration phase. coded wave form was transmitted and the distribution of echo The C1Z and SC weighted sums of our cw echo spectra are show17 power in time delay and Doppler frequency was measured. D i l r i ~ g in Fig. ?.A. 1986 DA's SC/OC ratio (kc= 0.09 ? 0.02) indicates 18 to 21 April the asteroid's distance, right ascension, and dcclina that the cchoes are almost entirely due to single reflections fror- tion changed little, from (0.207 AU, 12.0 h, 18") to (0.210 Ai :, surfact- elements that are smooth at scales of centimeters to meters. 12.2 h, 16"). Each dafs observing window was 1.3 h to 4.1 h U T C However, pronounced variations in the O C spectral signature as ;- (coordinated universal time). function of rotation phase (Fig. 2B) indicate that the asteroid': Each run involved transmission for 3.5 min (the round-tr~pli ght surface possesses extreme structural complexity at 10- to 100-P ;imc to thc asteroid) fbllowcd by reception of' cchocs for a si~i-ilar scales. Fur example, bimodal spcctra occur within nvo phase intc: period. The transmission was circularly polariwd, and echoes were vals roughly one half of a rotation apart, which is the signature received simultaneously in che same circular polarization as trans- mitted (the SC sense) and in the opposite (OC) sense. 1985 DA's apparent radial motion intruduced a continuously changing Dopp- ler shift into tht- echoes, but we avoidcd spectral smear by tuning the reccivt-r according to an ephemeris based o n an orbit detcrmined from optical observations of the astcroid. Optical light clinres obtained within a fen. weeks of the radar obse~ationsi ndicatc a \raluc for thc apparent rotation peric)d. I'. hull , I \:.ithjn 0.1 hour oi' 3.S hours 1 15). C;onlparisc)ils :)t' ccho spc-ci-rz I I from 18 to 21 April support that value. Our obse~ationsw, hich spanned -2.5 hours on each date, yielded a total of 65 useful nins (57 cw plus 8 ranging), that individually span 6" of rotation phase and collectively provide thorough phase coverage (Fig. 1). Orbit determination. We derived an orbit for 1986 DA and searched for future close approaches using the radar delay-Doppler results in combination with all of the available optical astrometric data (Table 1). During the apparition in 1986, the asteroid ap- proached Eurth to within 0.21 AU, which is nearly as close as Fig. 1. Rotatio~p~ha ses of 1986 liA obsenl~tions~ I I Lth e estimate of :.:. possible in the present state of its orbital evolution because the asteroid's hull. (A) Phases of cw and raiging runs (light and dark radial i:.;. asteroid's perihelion roughly coincided with its opposition. A siniilar segments), based un 1' = 3.50 h and phase origi:~a t 1986 April 21.75 1-77,. approach vriU occur in 2038, but no other approaches closer than 0.6 The line segment for any given run givza the radar's orientatjor: with resp:: AU will occur until then. In that the asteroid's orbit crosses that of to the hull. Lengths of lines are proportioilal to backgl-ourd noise levcl 2-ni their variation rcflects the zenith-angle drpendrnce of the Arecibo tclescc~p:~. Mars, extremely close approaches are possible benvrcn those nbro sensitivit)..T hr convex hull can be thought of as thr shape of a mbbc; h;:!. bod~cs,b ut none closer than 0.2 "iU will occur in the next centu?. stretched around the asteroid's polr-on silhoucrte. The central dot is i''. Perhaps the most interesting fraturc o T 1986 1)A's present orbit is asteroid's projected center of mass. (B) Sin~ulationsd esignrd to cxp1or.c ::: that its period is {vithin 1 percent o i a 5:2 reronance with Jupiter. accuracy of cstilnates of 1986 D.4.s hu!l. Thc dotted profile is the truc ilu!' an ellipsoidal (convex) model asteroid anci the rolid curvcs arc csrlnx:: ~ ~ h i iisl ico rnpurahle to thc c1:)seness of Jupiter's period to th,n 512 dcrived from simulated echu spectra contaminated with noise. Llncc~-ra!:i:: a / . (;ti! th2; :csoi~~ilcc\~ .ithS aturn. Kahil cr ilc!nonrraccd :;:~c-h thc c\tir:iatc of' ! 986 0h.s hull 15 indicated graphically by thc d~rycr.J:. ~ zsrcri~ido rbits arc cha

RESEARCH ARTICLE 1401 expectations about such objects' shapes and surface-slope distribu- E-chondritic regolith o r a rather high-porosity, metallic regolith. tidns, G is probably within a few tens of percent o f unity (24),s o ,?c, F o r the smaller, more irregularly shaped NEAs, G might be a is a reasonable first approximation t o R. For dry, powdered mixtures strong function o f orientation and cause radar cross sections t o vary of meteoritic minerals, R depends almost linearly o n bulk density. much more dramatically than A,,,,, as the object rotates. Severe Hence R for any given mineral assemblage is a nearly linear function variation in a,, with rotation phase is in fact seen for 1986 D A (Fig. of porosity (Fig. 4C). For example, the highest M B A albedo, for the 2B). N E A albedos derived from observations with thorough rota- M asteroid 16 Psyche, is consistent with a metallic regolith having a tion-phase coverage might tend t o "average out" variations in G ,b ut porosity (-0.55) typical o f lunar soil, n,hereas the M asteroid possibly not enough t o justi$ treating kc,,a s an approximation t o R. 97 Klotho's albedo is compatible with either a rather low-porosity, In any event, we argue that 1 9 8 6 DA's large albedo relative t o those of S NEAs strongly favors its identification with irons rather than with E chondrites, as follon~s. If S NEAs are stony irons [a hypothesis defended by mineralogical Table 2. Asteroid radar cross sections u,,, diapeters D, radar albedos 6,, and SCiOC ratios. Albedos are calculated from the diameters and the interpretations of reflectance spectra for several large S MBAs (5)] radar cross sections listed here. All of the radar cross sections and SCiOC and 1986 D A is E chondritic, then w e would expect 1986 D A to be ratios were measured by the authors at Arecibo (24, 30) except for 1566 roughly 7 0 percent as bright as S NEAs with similar regolith Icarus (31). For lcarus and 97 Klotho, no SC data were talien. Relative properties. If S NEAs are ordinary chondritic and 1986 D A is uncertainties in Arecibo estimates of u,,, are -15 percent. MBA albedos chondritic, w e would expect D A t o be about 5 0 percent use Infrared Astronomy Satellite (IRAS! radiometric diameters (32). E 1986 Fractional standard errors in MBA diameter estimates are typically -5 brighter than "typical" S NEAs. However, 1986 1)A is about three percent. The relative uncertainty in MBA albedo estimates is probably times morc reflective than the average S N E A and about four to ten -20 percent. As discussed in (33, 34), radiometric diameters are normally times more reflective than the three radar-detected S NEAs that are based on either a "standard" or "nonstandard" thermal model, or some hybrid of these two lunds of models. All of the MRA diameters reported here are standard model values. For NEAs, we use the radiometric diameter (4, 34) corresponding to the thermal model providing the best match to the available V1S-1R ohsenrations 135). All of the NEA diameters are standard model results except for five objects. as follows. For 1685 Toro, neither thermal model provides an acceptable fit to the \TS-IR data, and we adopt the radar result (13). For 3757 (I982 );B), neither thermal model is judged preferable, and we adopt the mean of !he standard and nonstandard model diameters (0.5 and 0.7 km, respectively). Nonstandard model diameters are used for 1580 Bemiia, 210n Ra- Shalom, and 3199 Nefertiti. For the NEAs in this table, nonstandard diameters average -50% larger than standard model diameters.

M a i n belt asteroids (AnIBB4r) CEK C 33000 913 PAL C 21000 523 AST S 2400 125 HLB S 4300 192 I RI S 5900 203 FLO S 1500 141 MET S 3500 203 LT1C S 2100 117 1'SY M 1400 264 i- <, P, \ ~320ci Z Z 1 1)AI' ( 2900 182 SAP S 650 82 KLO M 1100 87 IUE C 1300 162 VIB C 1800 146 LIG C 1800 135 I'EK C 1600 99 EKA C 61 0 93 Fig. 3. Radar images of 1986 ]>A. The top two rows show images from tht. Near-Earth ncfe-ro -i ds (NEAs) ERO S / 3 22 eight ranging NnS taken at rotation phases indicated in the !abels here an*.: GAIN S 75 38.5 also in Fig- I . Time dela!.. c!r range, increases fi-om top to hortom anI, ; c L-<>;~ITLCS:::I,> .TC:~ 1982 YAK S 0.OiS 0.6 rh.lt drin in criiisrrucring :hi\ fi~urr Thi. i.irgc in.~yc.: r h,it!:~: i . s u p r r p c s i r ~ )(i~t' rh: e~gvt~ nd:\.iili:~tIr: :ncs. I.;hc.!s ,!I: :kc 4drs ot ti:.:.: 1986 !>A hl 2.4 2.3 .. . i>l;.ec. il;(licc;,, :):r ~-,.;lecIV .:., , , r . ~Crl-,(. ....i,;. ;, .,. . - , . - I - r ) : 7 x 1 : ! ! \ - S > I . ! , : :7:1:!?. less than one half its size. Therefore the hypothesis that 1986 DA is parent asteroids that melted, differentiated, cooled, and solidified ;ch ondritic seems untenable, independent of the S NEAs' neteor- during the first billion years of the solar system and that subsequent- ::ic kinship. ly suffered collisional disruption (26, 27). A meteorite's cosmic ray On the other hand, if 1986 DA is a relatively regolith-free, exposure (CRE) age marks the time when it was last shielded, by a ::on-meteorite analog, then we can easily understand its albedo meter or so of material, from energetic cosmic rays (26, 28). Most !:ring much higher than typical S NEA albedos as long as S N E h CRE ages reported for irons are between 200 and 1000 million Fossess a significant regolith or are ordinary chondritic, or bo.th. We years (My), comparable to the typical dynamical lifetime of an object rrject the possibility that most S NEAs are regolith-free stony irons in a Mars-crossing orbit, whereas only a few percent of those ages becaiise most S NEA albedos are comparable to or smaller than are as short as an Earth-crossing asteroid's typical lifetime of -30 ! those of so clan7 ob!ects that either are identified with intrinsically My (2). Any meteorites that originated as impact ejecta from unreflective 1ne:eorites /C NEAs), or are likely to possess substantial 1986 DA may or may not have been launched before the asteroid regoliths (S LMBA~)o,r both (C MBAs) (25). evolved into an Earth-crossing orbit. The set of plausible CRE ages Implications. The evidence that 1986 DA is metallic seems for such meteorites therefore includes virtually any value up to persuasive. How might this asteroid be "genetically" related to iron several billion years. Of course, 1986 DA might share the parentage i meteorites? Most irons are probably fragments of large cores of of some meteorites or part of their dynamical history without ever

Table 3. Mineralogical and meteoritical associations of asteroid classes in chondrites are an order of magnitude more abundant than very low iron T.lble 2. Meteorite analogs are from Lipschuu pt al. (6j. Typical densities (LL) and ensratite (E) chondrites. The extremely rare Q asteroids might be I ~ iapipro ximate abundances of metallics (native metals and metal suffides) analogous to ordinary chondritcs (4). Abbreviations: wt, weight; and vol, arc from tables 2.1 and 4.1 of Dodd (36), table 24 of Buchwald (37),t able volume. 2 ~f Keil (38), and table 4.6 of Glass (39). High-iron (H)a nd low-iron (L) -.-. Meteorite properties

Possible meteorite analogs Bulk Percent abundance of density metallics (g ~ m - ~ ) (wt) (~01) - M asteroids Irons 7.6 97 97 Enstatite chondrites 3.6 32 19

Stony irons 4.9 5 0 30 3:irtal + olivine + pyroxene Ordinary chondrites H 3.6 24 12 L 3.5 14 8 LL 3.5 2 1

Hydrated silicates Carbonaceous chondrites (CI1 and Ch42) 2.6 1I 6 + carbon-organics-opaques

C . 7 - v , -i 0.7 , STONY -S 0.6 - IRONS c 0 rn .d 0.5 - 0.8 - 19 CHONDRITES 0.4 -

NEF =i 0.3 - 3 2 0.2 -

0.1 - @@)0 " ~ ~ ~ ~ ) 0.1 - OUE GW t9a2.W , l / , I l 3.3 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.3 CIRCULAR POLARIZATION RATIO (P, - SClOc) u

0.0 I - 30% 40% 50% 60% Fig. 4. Asteroid radar albedo versus (A) eircular polarization ratio and (8) POROSITY diameter. The values plotted in (A) and (B) are listed in Table 2. In (A) small volume concentrations V of metal panicles in a silicate matrix can ''SMBA" and "CMBA" indicate mean values for seven S MBAs and nine C increase R above the value for V = 0 by an amount that depends on the M B L . (C) Reflection coefficient for meteorite types (Table 3). For dry, electrical properties of each phase, the metal panicles' dimensions and unconsolidated powders of meteoritic minerals, R depends primarily on bulk packing geometry, and V. Here the oval fields show ranges of R expected for d , and the curves show predictions of R(d) based on the average of several meteorite types on che basis of laboratory investigations (42) of three empirically determined fimctions (24, 40) over the range of lunar-soil "loaded dielectrics." Cross-hatched fields show measurements of R for six Porosities (41).T hese curves probably are reliable to 1 2 5p ercent. For solids, meteorite specimens (22).

RESEARCH ARTICLE 1403 having contributed any of its own ejecta to our meteorite sample Rosema, R. F. lurgens, ib~d.8 4, 334 (1990). Uncenaint~esq uoted for quantlriec related to the hull calculatton are estimated standard errors. ~alliim-germaniuma bundances in irons define groups that might 19. M. .4. Ramcci and M. Fulchignon~, In .rlc~rro~ds,C onlc~rs, Mrm~rs1 11, C.-I. share a common parent body, and -40 percent of iron meteorites Lagerkvist, R. Rtckman, B. A. Lindhlad, M. Lindgren, Eds. (Uppsala Univers~~.?, belong to two groups, IIIAB and IVA, each of whose CRE ages are Uppsala, 1990), pp. 101-105. 20. E. F. Tedesco, personal cornmunicatlon narrowly clustered around a specific age (650 and 400 My, respec- 21. D. S. McKay, T. D. Swindle, R. Greaiherg, in [3), pp. 6 1 7 4 4 2 ; K. R. Hoiisen, tively) thought to correspond to major disruptions of these two L. L. W~kening,C . R. Chapman, R. Greenberg. in .-lsrcro~ds.T Gehrcls. Ed. groups' parent bodies (26, 29). Does 1986 DA share one of these (Univ. of Arizona, Tucson, 1979). pp. h01-627. - - - 22. M. J. Camphell and J. Ulr~chs,J . C;rophp. Kc<.7 4, 5867 (1969). groups' parentage? O r is it associatcd with onc of the groups (such - - 23. R. A. Simpson, Pror. IEEE Trans. ilrlrerltla l'rilpag. AP-24. 17 ( 1976; as L&)- whosh CRE ages indicate involvement in at least four 24. S J. Ostro, D. R. Camphell, 1 I. Shapiro, Srlr,nrc 229. 442 119851 25 S 1. Ostro et a/., in preparation. collisions between 100 and 1200 Mv, aug o? The nlany diverse 26 E. R. D. Scorn, in,+rcroid.<, T. Gehrcls, Ed, (Unlv. of Arizona, Tucson. 19'9 . ; possibilities seem indistinguishable at present. Analysis of sanlples 892-925. returned from 1986 DA's surface by spacecraft would elucidate this 27. er al., in (3). pp. 701-739. asteroid's history and its mctcoritic kinship. 28. J . T. Pt'asson, Mcreorrtc,s: nterr Rrrord ojEarly Solar Systc2niIIlsrory (Freeman, I<:. York, 1985), pp. 58-61. 29. E. R. D. Scorn and j. T M'asson, Rw. Ccophys. S ~ ~ aPhys.r r 13, 527 (1975, REFERENCES AND NOTES 30. S J. Ostro, D B. Campbell, 1. 1. Shapiro, 51111.A m . ,+tron. Sor. 17, 729 (1QR: 31. R. M. Zaldstein. Srirrlcr 162, 903 (1968). 1. G. U'. Wetherill, Irorus 76, 1 (1988). 32. E. F. TcJssco. in ( z ) , pp. 1090-1 150. 2. -a nd C. R. Chapman, in Meteorires and the Early Solar Systern, J. F. Kerridge .; and hl. S. Manhews, Eds. (Univ. of Ar~wnaT, ucson, 1988). pp. 35-67. 33. L. A. Lch*?fsliy and 1. R. Spencer, in (3)- pp. 128-147; M. \' Sykes and R Walker, Scicnce 251. 777 ( 1991). 3. R. P. Birvrl, T. Gehrels, M. S. Marrhews, Eds , hrcroids 11 (Univ. of Arizona Prcss. Tucson, 1989). 34. G. J. Veedzr et al.. .+rmn. J. 97, 121 1 I 1989). Absolute uncenaint~esI n :\.: . 1. L. A. McFadden, D. J. Tholen, G. J. Veeder, ih~d.,p p. 442-467. radiometrtr diameters are probahly several tens of percent; relative uncenainric. . . . 5. M J. G d e y , J. F. Bell, D. P. Cmihhank. ihid.. pp. 98-137. probablv somewhat mailer. 6. M. E. L~pschutz,M . J. Gaffey, P. Pellas, ibid., p. 740-777. 35. G. 1.. Veeder, pcnonal cornmunicatlon. 7. E. A. Cloutis, M. J. Gaffey, D. G. W Smith. R. St 1. Lamben. J. C;cop/l)a. Rrs. 36. R T . Dodd, Mcrcoriles (Cambridge Univ Press, Cambridge. 198 1) 95, 281 (1990); D. F. Lupishko and I. N. Belskaya, Irarus 78, 395 (1989). 37. V. F. Buchadd. Handbook o f lror~A 4creorires (Univ. of Caltforn~aPrcss. Her,..:. 8. E. F. Tedesco and J. C. Gradie, h ~ r o ~Jl. .9 3. 738 11987). 1975). 1,. 86. 9. Among the several dozen classified NEAs, the S:C:M rauo 1s approximately 7:3: 1 4 ) . A tnong the hundreds of classified lMBAs, that ratio 1s approumatciy 7 :5 :1 , ~ec. (5) and D. J. Tholen, thesis, University of Arizona, Tucson (1984). Asteroid taxonomic methods are reviewed by D. J. Tholen and M. A. Bamccl, ir~(.?). pp 40. 1. B. Ga?,in, !. \V. Hcad, G. H . l'crtcngill, S. H. Ztsk, J. (;~~,nhysR.e x. 00. (I:.; 298-3 15. (1985); F. T l:l.~hy cf a/.. IEEE Pans. Gcosrl. Rrlnorc Sr~rrrrt2~8~. ~3 25 113J:

10. M. Kizawa. 1111. Act-rron. Union Clr. 4181 (, 1986,) . 41. W. D. C~rricr1 11, J. K. Mitchell. A. Mahmood, Pror. Lunar Sri. Conf 4. 1::': 11. S. J. Ostro, in (3). pp. 192-212, and rcfercnces therein. (1973). 12. el 01, S~ienrc2 48, 1523 (1990). 42. I. M. Keiiy, i.0 . Stcnoien, D. E Isbell, J. .4pp1. Phys. 24, 258 [1953), i. .. 13. S. J. Ostro, D. B. Campbell, I. I. Shapiro,.-t-tron. 1. 88, 565 (1983) Nielson, J. Phys. D:Appl. Phys. 7, 1549 (1974); G. H. Perneng~llP, . G For. .. i. 14. S. J. Ostro et al., ibid. 99, 2012 (1990) D. Chapmar., J. G~ophys.R cs. 93. 14881 (1988). 15. K. W Zeigler, M~tlorl'lar~cr Rull. 17, 1 (1990); W. 2. Wisniewski, Irarus 70, 566 43. We thank the staBof the Areciho Ohsenrator) for assistance wlth the ohscn-n:i- (1987). Pan of t h ~ sresearch was conducted at the Jet Propulsion Lahoraror), Cal~r!irr!:: 16. G. Hahn, C:I. Lagerhist. M. Lindgren. M. lhhlgren, In Ateroids, Cotnerc. Insuture cif Technolop. under contract with rhc N~ti(ln.~Ale ronaur~csa n<' :. ?vfcrmrs 111, C. -1. l.agcrk\,irr, h. K~ckmnn,H X Llndhlad. M Lindgrcn, I;<:, Admin~str,~:ion[N ASA). J.F.C. and I.1.S. ucrc upponcd in part hy NhS ': (Uppsala University. Llppiala, 19901, pp 95-98, Arecibc Ubservator) 1s part of the Xational Astronomy m d lcvnospiierc i:.:;.: 17. E. M. Shoemaker, R. F. Wolfe, C. S. Shoemaker, Geol. Sor. .4tt1. Spec. rdp2. 47, urh~chi s operated by Corncll Uni\ers~yu nder a coopcrattve qrecmcnr \r.tri- .:!,. 155 (1990); E. M. Shoemaker, J. G. Williams, E. F. Heltn, R. F. Wolfe, In National Science Foundation and with support from NASA. .rlcreroids, T . Gehrcls. Ed. (Univ. of Arizona, Tucson, 19791, pp. 253-282 18. S. J. Octro, R. Connelly, L. Belkora, 1rar1i.i 73, 15 11988); S J . Osrrj. K D. 14 MarLh 1991; accepted 26 April 1991