Monday, July 19, 1993 Mesosiderites, Irons, and Core Formation 8:15 a.m. Theater Chair(s): R. S. Clarke T. J. McCoy Drake M. J.* Hillgren V. J. DeAro J. A. Capobianco C. J. Metal/Silicate Partitioning, Melt Speciation, Accretion, and Core Formation in the Earth Core formation in terrestrial planets was concomitant with accretion. Siderophile and chalcophile element signatures in the mantles of planets are the result of these processes. For Earth, abundances of most siderophile and chalcophile elements are elevated relative to predictions from simple metal/silicate equilibria at low pressures [1]. This observation has led to three hypotheses for how these abundances were established: heterogeneous accretion [2], inefficient core formation [3], and metal/silicate equilibria at magma ocean pressures and temperatures [4]. Knowledge of speciation of siderophile elements in silicate melts in equilibrium with metal may help distinguish between these hypotheses. But there is some uncertainty regarding speciation. For example, Ni and Co have been reported to be present as 1+ or zero valence species in silicate melts at redox states appropriate to planetary accretion, rather than the expected 2+ state [5-7]. Independent metal/silicate partitioning experiments by three members of this group using two different experimental designs on both synthetic and natural compositions do not show evidence for Ni and Co in valence states other than 2+ over a wide range of redox states. For example, solid metal/silicate melt partition coefficients for Ni at 1260 degrees C obtained by VJH from experiments investigating the partitioning of Ni, Co, Mo, W, and P are indistinguishable from those obtained by JAD in similar experiments investigating the partitioning of Ni, Ge, and Sn. Both datasets define a line with the equation: log D(Ni) = - 0.54log fO2 - 3.14 with r^2 > 0.995. (Note that fO2 was calculated in both studies from thermodynamic data and phase compositions. A small, systematic offset from the true fO2 as measured by a solid electrolyte cell affects both equations similarly, but does not diminish their close agreement.) The valence of Ni in the silicate melt is obtained by multiplying the slope of the line by -4, indicating divalent Ni in both studies. Experiments by [8] between 1300 degrees C and 1550 degrees C and fO2 from air to just below iron-wustite in which Ni and Co are paritioned between Pt metal and CaO-Al2O3-SiO2 silicate melt also show evidence only for 2+ valence. Capobianco et al. [1] have noted that reliable extrapolation from current laboratory temperatures (1190 degrees C-1600 degrees C) to magma ocean temperatures is not possible. The hypothesis that siderophile and chalcophile element abundances in the mantle of Earth were established by metal/silicate equilibria at magma ocean pressures and temperatures needs to be tested using direct experimental measurements at magma ocean temperatures and pressures. Such experiments are currently being conducted. References: [1] Capobianco et al. (1993) J. Geophys. Res., 98, 5433. [2] Wanke (1981) Phil. Trans. R. Soc. London, A303, 287. [3] Jones and Drake (1986) Nature, 322, 221. [4] Murthy (1991) Science, 253, 303. [5] Schmitt et al. (1989) GCA, 53, 173. [6] Ehlers et al. (1993) GCA, 56, 3733. [7] Colson (1992) Nature, 357, 65. [8] Capobianco and Amelin (1993) GCA, 56 (in press). Ntaflos Th.* Kurat G. Koeberl C. Brandstatter F. Mincy Dunite F6241B: The Missing Ultramafic Component from Mesosiderites A dunitic clast from the Mincy mesosiderite was studied using microprobe and INAA techniques. This clast (1.5 x 1.2 cm) consists of olivine with minor amounts of enstatite, diopside, plagioclase, merrillite, troilite, taenite, kamacite, schreibersite, chromite, pentlandite, magnetite, and rust [1]. The texture is porphyroclastic with large olivine porphyroclasts (up to 0.3 cm) embedded in a fine-grained matrix consisting mainly of olivines (up to 60 micrometers) commonly forming triple point junctions. The opaque minerals show traces of a network penetrating the silicates. Olivine has a narrow range in composition (Fo(sub)90.9-Fo(sub)92.2) with both fine-grained and porphyroclastic olivines being slightly zoned with cores more forsteritic than rims. The FeO/MnO ratio decreases with increasing Fo, which is opposite from the trend reported for olivine clasts from mesosiderites [2]. Enstatite (En(sub)90.6Fs(sub)7.9Wo(sub)1.4) is associated with diopside (En(sub)51.9Fs(sub)3.1Wo(sub)45.0), chromite, and both types of olivine. Chromite occurs as euhedral and irregular grains. Compared to the wide variation of olivine compositions within and between other mesosiderites (Fo(sub)53-Fo(sub)92), the Mincy dunite contains only highly magnesian olivine. Triple points between fine-grained equigranular olivines clearly indicate recrystallization of coarse-grained porphyroclasts. Enstatite, diopside, and olivine appear to be in chemical equilibrium. The estimated equilibration temperature for enstatite-diopside pairs varies between 940 degrees and 1000 degrees C [3,4]. This temperature is consistent with those estimated by Nehru et al. [5] for the Emery mesosiderite. Chromite- olivine pairs reflect closure temperatures of about 550 degrees C [6]. The large differences in equilibration temperatures of pyroxenes and olivine- chromite pairs appear to be incompatible with a uniform steady-rate cooling history indicating drastically different cooling regimes for the pyroxenes and olivine-chromite pairs respectively. The dunitic clast has low REE abundances that display a V-shaped pattern (Fig. 1). This pattern indicates mixing of LREE-depleted and LREE-enriched components. The siderophile elements have high abundances (~0.2-0.5x CI) (Fig. 1) and an unfractionated Ni/Ir. Normalized abundances of Co and Fe are greater than those of Ni, indicating a substantial contribution from oxidized species. According to the planetesimal model, dunitic rocks should be common in mesosiderites [5,7,8], but are in fact rare. Besides the clast described here, only one other was reported from Morristown [9]. Trace elements indicate that the Mincy dunite is not of a simple magmatic origin. Lithophile elements indicate mixing of at least two reservoirs (high-LREE and low-LREE) similar to ureilites [10] and terrestrial metasomatic peridotites [11]. The high abundance of siderophile elements also excludes a simple magmatic origin. It rather suggests mixing of a metal with unfractionated Ir/NI ratio and an oxidized silicate of low siderophile contents. References: [1] Ntaflos Th. et al. (1991) Europ. J. Miner., 3, 200. [2] Delaney J. S. et al. (1980) Proc. LPSC 11th, 1073-1087. [3] Wells P. (1977) Contrib. Mineral. Petrol., 62, 129-139. [4] Bertrand P. and Mercier C. (1985) EPSL, 76, 109-122. [5] Nehru C. E. et al. (1980) GCA, 44, 1103-1118. [6] Fabries J. (1979) Contrib. Mineral. Petrol., 69, 329-336. [7] Mittlefehldt D. W. (1980) EPSL, 51, 29-40. [8] Hewins R. H. (1981) GCA, 45, 123-126. [9] Ruzicka A. and Boynton W. V. (1991) Meteoritics, 26, 391. [10] Janssens M. J. et al. (1987) GCA, 51, 2275-2283. [11] Frey F. A. (1984) REE Geochemistry (P. Henderson, ed.), 153-203. Fig. 1 appears here in the hard copy. Petaev M. I.* Clarke R. S. Jr. Jarosewich E. Lipschutz M. E. Wang M.-S. Davis A. M. Steele I. M. Olsen E. J. Wood J. A. Phosphate-Silicate Inclusions in Chaunskij: How Diverse are They? The Chaunskij meteorite was found in 1985 and was recently classified as the most highly metamorphosed, shock-modified, and metal-rich mesosiderite [1]. It contains ~10 vol% mono- and polymineralic troilite-phosphate-silicate inclusions, micrometers to centimeters in size. Metal in Chaunskij displays a mesosiderite structure and is described in some detail in an accompanying paper [2]. Here we present new data on polymineralic inclusions that shed additional light on their origin. Two dominant silicate lithologies have been found in the inclusions. One, making up the largest inclusion (2.2 x 1.7 cm), consists of a fine-grained (20-30 micrometers) aggregate of anhedral pyroxene, subhedral plagioclase laths, and silica, with larger poikilitic grains of the first two minerals. Whitlockite is minor. Textures vary from microophitic to xenoblastic. This lithology, called "igneous," also contains rare primary clasts enriched in pyroxene, whose boundaries are almost unresolvable from the ground mass in transmitted light. The second, "metamorphic" lithology occurs as separate small inclusions and as larger areas in intimate contact with the "igneous" lithology in the largest inclusion. This lithology is a fine-grained (typically 30-50 micrometers) xenoblastic intergrowth of low-Ca pyroxene, whitlockite, and cordierite, with rare larger porphyritic grains of the first two minerals. Porphyritic pyroxene grain edges are generally irregular, indicative of reaction with the ground mass. Plagioclase is present only as a rare accessory mineral. Minor minerals in both lithologies are silica, kamacite, taenite, troilite, chromite, ilmenite, and rutile. Rare grains of pyrophanite, zircon, alabandite, stanfieldite, and a graftonite-farringtonite mineral are also present in the inclusions. Mineral compositions of small inclusions are more diverse than those characteristic of the "igneous" and "metamorphic" lithologies. Many of them consist of cordierite, pyroxene, and whitlockite intergrowths, with or without silica and opaque minerals. However, some inclusions do not match the mineralogies of "igneous" or "metamorphic" lithologies. They consist of cordierite only (inclusion #1-10); cordierite and silica (#1-18); silica, whitlockite, and troilite with minor Al-rich chromite and rare pyroxene (#4- 5A); plagioclase and whitlockite (#4-5D); and silica and whitlockite (#4-6E). Mineral compositions vary considerably both between and within all inclusions. No systematic differences between separate inclusions were found. Compositions of pyroxene and plagioclase match those of mesosideritic minerals. The chromite and ilmenite display systematic variations in MgO, MnO, Al(sub)2O(sub)3 and V(sub)2O(sub)3 contents, suggestive of a precursor material consisting of a series of basaltic rocks. The bulk chemical composition of the largest silicate inclusion, recalculated to the silicate fraction only, is very close to that of eucrites and mesosiderites except for a large enrichment in P and volatile chalcophiles. Major-element and REE chemistry and bulk mineralogy point to cumulate eucrites as the precursor of the silicate inclusions. This precursor was apparently slightly fractionated during the remelting event inferred by the structure of the "igneous" lithology. The "metamorphic" lithology apparently was formed due to reaction between silicates and phosphorus dissolved in the metal: Px + An + P + O --> Cord + Q + Whit. This reaction took place under ~700 degrees C and ~4 kbar [3] in the interior of the Chaunskij parent body. Compositions and textures of small inclusions suggest that the metamorphic reaction took place before the incorporation of the inclusions into the piece of metal making up the Chaunskij main mass. References: [1] Petaev M. I. et al. (1993) LPS XXIV, 1131-1132. [2] Clarke R. S. Jr. et al., this volume. [3] Petaev M. I. et al. (1992) Meteoritics, 27, 276-277. Olsen E.* Steele I. New Alkali Phosphates and Their Associations in the IIIAB Iron Meteorites The IIIAB iron meteorites commonly contain inclusions of phosphate + oxide minerals within troilite nodules (occasionally directly within metal). The most frequently encountered associations are chromite with (1) graftonite, (Fe,Mn)(sub)3(PO(sub)4)(sub)2 (with Mn usually <4 mol%, but occasionally up to 50 mol%); (2) sarcopside (a polymorph of graftonite) with graftonite; or (3) sarcopside with beusite, (Fe,Mn)(sub)3(PO(sub)4)(sub)2 (Mn >= 50 mol%]. In addition, any of these three associations may contain one or two of three alkali-bearing phosphates, two of which appear to be new minerals: johnsomervilleite Na(sub)2Ca(Fe,Mn)(sub)7(PO(sub)4)(sub)6 (Fe >> Mn); new phase #1 (Na,K)(sub)2(Fe,Mn)(sub)8(PO(sub)4)(sub)6 (Fe >> Mn, Na > K); new phase #2 (K,Na)(sub)2(Fe,Mn)(sub)8(PO(sub)4)(sub)6 (Fe >> Mn, K > Na). Crystallographic work has been undertaken on these latter two phases in preparation for proposing them as new minerals. For the time being they are designated the 2:8:6 phosphates. Johnsomervilleite is a 2:1:7:6 phosphate; graftonite/sarcopside and beusite are 9:6 phosphates. Johnsomervilleite is the Fe end member of the new 2:1:7:6 alkali-Mg phosphate mineral, Na(sub)2 Ca(Mg,Fe,Mn)(sub)7(PO(sub)4)(sub)6, found in a silicate inclusion in the Carlton IIIC iron meteorite [1]. Five assemblages, all of which contain chromite, have been investigated thus far: Chupaderos (IIIB) with graftonite + johnsomervilleite; Grant (IIIB) with sarcopside + beusite + johnsomervilleite + trace of silica; Bella Roca (IIIB) with sarcopside + beusite + 2:8:6 [Na/(Na + K) = 0.97]; El Sampal (IIIA) with sarcopside + beusite + johnsomervilleite + 2:8:6 [Na/(Na + K) = 0.98]; Sandtown (IIIA) with graftonite (with up to 3 wt% CaO) + johnsomervilleite + K-rich 2:8:6 ([Na/(Na + K) = 0.40]. The 2:1:7:6 phosphates can be considered to have three end members: fillowite Na(sub)2CaMn(sub)7(PO(sub)4)(sub)6; johnsomervilleite Na(sub)2CaFe(sub)7 (PO(sub)4)(sub)6; new phase Na(sub)2CaMg(sub)7(PO(sub)4)(sub)6 [1]. The similarity of the X-ray diffraction patterns of all three suggests that their crystal structures are the same as that of fillowite [2]. No X-ray data have yet been obtained for the 2:8:6 phosphates. The lack of Ca, however, may indicate that their structures either differ from that of the 2:1:7:6 phosphates or that Fe can substitute in the Ca site in the fillowite structure. Because both the 2:8:6 and 2:1:7:6 phases coexist in at least two phosphate assemblages (Sandtown and El Sampal), we believe that the 2:8:6 phase has a different structure than that of the 2:1:7:6 phosphates. The Mg and Mn end members of the 2:8:6 trio have not been recognized, but based on the crystallographic similarity of Mg, Mn, and Fe, they are probable minerals given the right chemical environment. Tentatively, we consider the Na- and K- rich 2:8:6 phases may be an isostructural solid solution above some temperature, but forming separate K-rich and Na-rich phases at lower temperatures if the total K content is high enough. In the IIIAB irons we believe that these phosphates form from trace components (including oxygen) occluded from metal during its solidification and reaction with Fe and P components. References: [1] McCoy and Keil, personal communication. [2] Araki T. and Moore P. B. (1981) Am. Mineral., 66, 827-842. Yang C. W.* Williams D. B. Goldstein J. I. Metallographic Observation of the Cloudy Zone in Meteorites The cloudy zone in the retained taenite of meteoritic metal is composed of two phases, the high-Ni island phase and the low-Ni honeycomb phase [1,2]. There is a concentration gradient through the cloudy zone and the size of the island phase varies with local Ni concentration. We propose that this size variation can be used to estimate the low-temperature cooling rate of the meteorite since the island phase width is controlled by the cooling rate and Ni concentration. The purpose of this study is to develop a relationship between the size of the island phase in the microstructure, the composition of the cloudy zone in the retained taenite of iron, stony-iron, and stony meteorites, and the cooling rate of meteorites obtained by metallographic techniques. A JEOL 6300F high-resolution scanning electron microscope (HRSEM) and a JEOL 733 electron probe microanalyzer (EPMA) were employed to study the microstructure. The island phase size variation was measured using a Micro-Plan II image analysis system (DonSanto Co.). Five meteorites including one mesosiderite, Estherville (ES), one iron meteorite, Tazewell (TA), and three chondrites, Saint Severin (SS), Guarena (GU), and Kernouve (KE),were investigated. The island phase width was measured using high-magnification (50,000X) HRSEM images taken across the cloudy zone of the five meteorites and the local Ni concentration was measured for each image using EPMA. For all the meteorites, the size of the island phase increases with increasing Ni concentration. The Ni concentration of the cloudy zone, which abuts the clear taenite (tetrataenite) rim phase, has the highest Ni, and has the biggest island phases, is almost constant (~42 wt% Ni). The size of the biggest island phase in a meteorite can be used as a measure of the cooling rate. Figure 1 shows the variation of the island phase vs. cooling rate. The metallographic cooling rate data were taken from previous measurements (Estherville [3], Saint Severin [4,5], Guarena [5], Kernouve [4], and Tazewell [6]). A clear trend is observed: the size of the biggest island phase decreases from about 500 nm to 100 nm with increasing cooling rate of the five meteorites studied. Measurements for other meteorites are in progress. Acknowledgments: The suggestions of E. R. D. Scott, University of Hawaii, are greatly appreciated. References: [1] Reuter K. B. et al. (1988) GCA, 52, 617. [2] Yang C. W. et al. (1993) LPSC XXIV, 1557. [3] Wasson J. T. (1985) GCA, 33, 789. [4] Duffield C. E. et al. (1991) Meteoritics, 26, 97. [5] Willis J. and Goldstein J. I. (1983) LPSC, in JGR, 88, B287. [6] Saikumar V. and Goldstein J. I. (1988) GCA, 52, 715. Figure 1, which appears in the hard copy, shows the relationship between the size of the biggest island phase and the cooling rate of meteorites. Clarke R. S. Jr.* Jarosewich E. Petaev M. I. Metallography of Chaunskij, Russia, Mesosiderite Chaunskij is an unusually metal-rich meteorite with metallographic structures indicating a typical mesosiderite cooling history. Its metal is unusually Ni-rich, and it has been subjected to higher shock levels than other mesosiderites. Bulk metal composition in wt% is Ni 13.10, Co 0.54, and P 0.10. Troilite is abundant compared to other mesosiderites and schreibersite is rare. Silicates probably comprise less than 20 wt% of the meteorite. Equant to slightly elongated kamacite areas in the mm-size range are the dominant mineral on polished surfaces. Kamacite grains have been strongly shocked and display varying intensities of epsilon-structure, apparently depending on the orientation of the grain. At exterior edges of the specimen, a pronounced alpha(sub)2 heat-altered zone a mm or more in width is superimposed on the epsilon-structure. Recrystallized areas of kamacite 20 micrometers or smaller on an edge, some with a distinct alpha-2 structure, occur throughout the surface. They are particularly abundant at boundaries with taenite and in areas of kamacite where the epsilon-structure is not strongly developed. Cloudy taenite regions bordered by 10-20-micrometer wide rims of 47-53% Ni taenite are the second major feature of the structure. In areas of structure showing the least indications of shock effects, the taenite frequently displays the typical optical anisotropy of tetrataenite, indicating that the cloudy taenite was bordered by tetrataenite that has since been shock disordered. Traverses across two large cloudy taenite areas show a gradual decrease in Ni content to low points of 35 and 29% Ni. The cloudy taenite fields have a martensite-like texture that may well be a shock-induced feature. In a small number of cloudy taenite areas, Ni values appear to be low enough that some true martensite is present. Unevenly distributed, shock-twinned troilite is an important feature of polished surfaces, making up about 5% of the area. Globular masses are in the mm-size range with some few having surface areas as large as 25 mm^2. Schreibersite is distributed in minor amounts along kamacite grain boundaries and as very small crystals at taenite interfaces. Jones J. H.* Casanova I. Experimental Partitioning of As and Sb Among Metal, Troilite, Schreibersite, Barringerite, and Metallic Liquid We have performed a series of experiments to evaluate the behaviors of As and Sb in metallic systems. Because of the reputed chalcophile nature of these elements, we wrongly anticipated that they would follow S and that, compared to the Fe-X systems [1], (solid metal/liquid metal) partition coefficients would be considerably lower in S-bearing systems. Experimental and Analytical: Experiments were performed in sealed silica tubes as in [2]. Starting materials were high-purity metals, natural pyrite, and natural stibnite. Charges were doped either with As or Sb. Experiments were held at either 950 degrees C for six days or 1250 degrees C for three days. Typical experimental assemblages consisted either of taenite and coexisting Fe-Ni-S-X liquid (1250 degrees and 950 degrees C) or an assemblage of troilite, schreibersite, and Fe-Ni-S-P-X liquid (950 degrees C). The schreibersite-bearing, As-doped charge also contained barringerite (Fe,Ni)2P. Charges were mounted in epoxy, polished, and analyzed using a Cameca SX-50 electron microprobe and standard techniques. Results: Phases appeared homogeneous. Our results, along with partition coefficients inferred for the S-free system, are given in Table 1. Table 1 appears here in the hard copy. Discussion: Our results indicate that As behaves as a siderophile element at low temperatures, very analogous to Au. While the siderophility of Sb increases with decreasing temperature, it remains incompatible in solid metal. In this regard Sb is unique. Both As and Sb are very incompatible in troilite. Arsenic is weakly incompatible in schreibersite and strongly compatible in barringerite. Nickel shows no preference for either phosphide. Nickel partition coefficients for metal and schreibersite are similar to those measured previously [3]. On a lnD vs. ln(1-2 alpha X(S)) diagram [4], the data for Sb and As subparallel each other, indicating similar dependencies on S, despite their very different partition coefficients. Arsenic behaves similarly to P. The As and Sb partition coefficients for the S-free system, inferred for kamacite (alpha-iron) from the Fe-As and Fe-Sb phase diagrams [1], are probably not applicable to taenite (gamma-iron). Extrapolation of our data to zero S indicates that the taenite partition coefficients for As and Sb are likely to be much lower than for kamacite. In discussing the fractional crystallization of iron meteorites, Scott [5] originally grouped Au, As, Sb, and Co and assigned them a (solid metal/liquid metal) partition coefficient of about 0.4. This distinguished them from P, which was given a partition coefficient of 0.2. Given the strong decoupling of As and Sb in our experiments, the general coherence of As and Sb in iron meteorites [5] is surprising. To explore this further, we have derived a new equation for the slopes of LogEl vs. LogNi diagrams, which takes into account changes in D. References: [1] Moffatt W. G. (1986) Handbook of Binary Phase Diagrams, Genium. [2] Jones J. H. and Drake M. J. (1983) GCA, 47, 1199. [3] Jones J. H. et al. (1993) GCA, 57, 453-460. [4] Jones J. H. and Malvin D. J. (1990) Metall. Trans., 21B, 697-706. [5] Scott E. R. D. (1972) GCA, 36, 1205. Herpfer M. A.* Larimer J. W. Core Formation: An Experimental Study of Metallic Melt-Silicate Segregation To a large extent, the question of how metallic cores form reduces to the problem of understanding the surface tension between metallic melts and silicates [1]. This problem was addressed by perfoming experiments to determine the surface tensions between metallic melts with variable S contents and the silicate phases (olivine and orthopyroxene) expected in planetary mantles. The experiments were conducted in a piston-cylinder apparatus at P = 1GPa and T = 1250-1450 degrees C. Textural and chemical equilibration was confirmed in several ways: theoretical estimates were checked by conducting a series of experiments at progressively longer times (up to 72 hrs) until phase composition and dihedral angle ceased to change and the distribution of measured "apparent" angles matched the standard cumulative frequency curve. The dihedral "wetting" angles (theta) were measured from high resolution photomicrgraphs using a 10X optical protractor; 100-400 measurements were made for most experiments. The dihedral angle is related to the ratio of interfacial energies: gamma(sub)ss/gamma(sub)sl = 2 cos(theta/2), where gamma(sub)ss and gamma(sub)sl are the interfacial energies between solid-solid and liquid-solid. Since data exist for the pertinent solid-solid energies, the liquid-solid interfacial energies can be computed from measured theta values. However, the important relations are best expressed in terms of theta values. The extent to which a melt is interconnected along grain boundaries, and hence able to flow and segregate depends on the value of theta and the fraction of melt present. When theta < 60 degrees, the liquid can be interconnected at all melt fractions but when theta > 60 degrees, the melt fraction must be at least 1 vol% and increses as theta increases. Actually there is a predicted effect, analogous to a hysteresis effect, where for a given theta value the amount of melt that needs to be added for interconnection is greater than the amount left when the melt disconnects (pinches off). In our experiments, where dense metallic melt drained away, the disconnect theta values match the theoretical predictions. The composition of the metallic melt in the experiments was varied from stoiciometric FeS to Fe/S ratios near the the eutectic and on to more Fe rich compositons. The theta values vary in a systematic manner; for example, for melts in contact with olivine at 1300 degrees C the theta values range from 67 degrees for FeS to 55 degrees at the eutectic and back toward higher values at higher Fe contents. Theoretical considerations indicate that eutectic compositions are expected to have the lowest theta values, just as observed. The theta values indicate that melts with eutectic composition can interconnect and segregate at 1-2 vol% melt fraction at 1300 degrees C. Some previous estimates of the melt fraction required for interconnection are much higher [2,3], but the inferences were drawn from experiments that were not designed to test for textural equilibrium, fraction of melt present, etc. The present experiments clearly show that metallic melts can readily segregate from solid silicates. Simple extrapolations to other phases, compositions and PT conditions provide a rather complete picture of how the "plumbing" worked in the mantles of planetary objects during the initial stages of core segregation. References: [1] Stevenson D. J. (1990) In Origin of the Earth, 231-249. [2] Taylor G. J. (1989) LPSC XX, 1109. [3] Walker D. and Agee C. B. Meteor. 23, 81-91. Harper C. L. Jr.* 244Pu-Ru,Pd,Te Chronometers of Planetary Accretion and Core Formation Live ^99Tc in the early solar system is a necessary consequence of the well-established presence of live ^244Pu (T(sub)1/2 = 80 Ma), but at a rather low level: (^99TC/^99Ru)(sub)4.56Ga = (^244Pu/^99Ru)(sub)4.56Ga (lambda(sub)244/lambda(sub)99)(Y(sub)s.f.)(Y(sub)99) ~5 x 10-^11 in bulk solar system matter. (For the ^244Pu s.f. yield distribution, cf., [2].) Feeding from fission of the longer-lived ^244Pu parent, however, generates a substantially larger post- decay effect (^99*Ru/^99Ru ~2 x 10^-8), which is further amplified by early and large Pu/Ru fractionations in planetary reservoirs from which metal and sulfide have been removed (300 to >10^7). In fact, nature has provided three early solar system chronometers based upon refractory-siderophile (Ru, Pd) and volatile-chalcophile (Te) nuclide yields from ^244Pu spontaneous fission. In some cases, the maximum predicted isotopic effects are at the level of ~5-10 ppm, as in the bulk silicate Earth, but in others (e.g., in SNC meteorites, eucrites, and lunar samples) they range from >20 ppm to (possibly) factors. The development of high precision, high ion yield, negative thermal ionization mass spectrometry (NTIMS) techniques for mass analysis of these elements at low levels makes these chronometers attractive for solving a range of difficult unsolved problems in early planetary and planetesimal differentiation (e.g., the age of the Moon). A summary of relevant information for evaluating the potential for these chronometers is given in Table 1. As the ^244Pu spontaneous fission yields at all masses of interest are more than 30x the yields from s.f. of ^238U over the history of the solar system, the ^244Pu fission component dominates over the uranogenic component for the first few 100 Ma of solar system history in Pu/U-unfractionated reservoirs. (^244Pu and ^238U components also can be decomposed across the Pd mass region where the ^244Pu yields rise from mass 99 to 108 and the ^238U yields fall [3].) Spallation and neutron capture effects may complicate the isotopic systematics of extraterrestrial samples, but in contrast to Xe, where large spallation components from very high Ba/Xe and REE/Xe are common (and expected for Te), target element abundances are typically very low for yields at ^102.104Ru (n, gamma, and n,2n on ^98Mo and ^100Mo affect ^99,101Ru), and on all Pd masses. Experimental investigations are in progress. References: [1] Q. Yin et al. (1992) Meteoritics, 27, 310. [2] Allaert E. et al. (1982) Nucl. Phys., A380, 61-71. [3] Maeck W. J. et al. (1978) In Les Reacteurs De Fission Naturels, IAEA, Vienna, 521-533. [4] Anders E. and Grevesse N. (1989) GCA, 53, 197-214. [5] Hudson G. B. et al. (1979) LPS XIX, 547-557; Hagee B. et al. (1990) GCA, 54, 2847-2858. [6] Sun S. (1982) GCA, 46, 179192. [7] Treiman A. H. et al.(1985) GCA, 50, 1071-1091. [8] Morgan J. W. et al. (1978) GCA, 42, 27-38. [9] Newsom H. E. (1986) In Origin of the Moon (W. Hartmann et al., eds.), 203-209, LPI. Table 1 appears here in the hard copy. Allton J. H. Wentworth S. J. Gooding J. L.* Calorimetric Thermometry of Meteoritic Troilite: Early Reconnaissance Troilite (FeS) exhibits two solid-state phase transformations, which, according to conventional thermodynamic literature [1,2], occur at 411 +- 3 K (alpha/beta) and 598 +- 3 K (beta/gamma). In principal, the thermal history of a particular troilite sample might be expected to impart structural or strain characteristics that could be measured upon experimental inducement of the phase transformations. To investigate that possibility, we applied differential scanning calorimetry (DSC) to determine the temperatures and enthalpy changes for phase transitions of various troilite samples subjected to controlled heating [3]. Post-heating residues were prepared as polished grain mounts and petrologically characterized by scanning electron microscopy (SEM) and energy-dispersive X-ray spectrometry (EDS). Troilite grains separated from the Mundrabilla (octahedrite), PAT91501 (L7 chondrite), and EET83213 (L3 chondrite) meteorites were compared with terrestrial troilite from an ultramafic complex in Del Norte Co., California. Both the alpha/beta and beta/gamma transitions were easily measured by DSC applied to samples on the order of 5 milligrams. Extrapolated-onset temperatures (the DSC approximation of equilibrium transition temperatures) for the alpha/beta transition display a systematic progression as follows: Mundrabilla, PAT91501, EET83213, Del Norte (Fig. 1). At least for Mundrabilla, PAT91501, and Del Norte, there also exists an apparent correlation of transition enthalpy with onset temperature (Fig. 1). (Enthalpy measurements for EET83213 troilite were systematically low, by dilution effects from admixed Ni-Fe metal, and require correction before such comparison.) Although Mundrabilla troilite contains on the order of 1 wt% Cr, and some grains of PAT91501 troilite contain on the order of 1 wt% Ni, it is doubtful that compositional differences can fully account for the measured differences in thermodynamic properties. Instead, it is likely that first-order differences are the consequence of different thermal histories. Multiple samples of Del Norte troilite were used to determine the influence of laboratory-scale thermal histories on DSC signatures by heating and cooling each sample under different programmed conditions. In reconnaissance experiments, maximum temperature achieved during heat treatment appears to be more influential than does either the time maintained at temperature or the heating/cooling rate. The experimentally measured alpha/beta onset temperature shows a systematic decline with maximum temperature achieved during prior heating, suggesting that high onset temperatures are indicative of low maximum temperatures in the natural histories of the troilite samples. That trend is at least qualitatively consistent with the petrologic rankings of the meteorites in which troilite from the relatively unmetamorphosed L3 chondrite shows a higher onset temperature than does troilite from either the highly metamorphosed L7 chondrite or the octahedrite. Additional work should define the limits of a quantitative calibration that might ultimately permit derivation of meteorite thermal histories by calorimetric thermometry of troilite. Samples were kindly provided by E. R. D. Scott (Mundrabilla), C. B. Moore (Del Norte), and the Meteorite Working Group (PAT91501; EET83213 powder from E. Jarosewich). References: [1] Chase M. W. Jr. et al. (1985) JANAF Thermochemical Tables, 3rd ed., 1194. [2] Robie R. A. et al. (1979) Geol. Surv. Bull. 1452, 125. [3] Allton J. H. and Gooding J. L. (1993) LPS XXIV, 21-22. Fig. 1, which appears here in the hard copy, shows the thermodynamics of troilite alpha/beta phase transformations measured by DSC during first-heat cycles. Michel R.* Allegre C. J. Audouze J. Begemann F. Birck J. L. Cloth P. Dittrich-Hannen B. Filges D. Gilabert E. Herpers U. Lange H.-J. Lavielle B. Leya I. Meltzow B. Signer P. Simonoff G. N. Weber H. W. Wieler R. Zanda B. Simulation of the Interaction of Galactic Protons with Meteoroids: Isotropic Irradiation of an Artificial Iron Meteoroid with 1.6-GeV Protons Isotropic irradiations of artificial meteoroids have been succesfully used [1- 5] to simulate the interactions of galactic protons with extraterrestrial matter and to validate a physical model [6] for the calculation of cosmogenic nuclide production rates in stony meteoroids and lunar surface materials. Since the production of secondary particles by high-energy protons depends on the mean mass number of the target materials, significant differences in cosmogenic nuclide production rates are to be expected between meteoroid classes such as stones, stony irons, and irons. Experimental evidence for this effect of bulk chemical composition was found by Begemann and Schultz [7]. Model calculations [8,9] support this interpretation. In order to simulate the interactions of galactic protons with iron meteoroids, an artificial iron meteoroid with a diameter of 20 cm was isotropically irradiated with 1.6-GeV protons at the Saturne cyclotron at Laboratoire National Saturne/CEN Saclay. The artificial iron meteoroid was made of steel (99% Fe). There were 3 perpendicular bores that were filled with 9 cylindrical iron boxes containing more than 900 individual targets. These targets covered all elements investigated in the earlier simulation experiments [1-5]. During a 133-hr irradiation the artificial iron meteoroid received a (preliminary) proton dose of 2.17 X 10^14 cm^-2, which is equivalent to a 2.3- Ma exposure of a meteoroid in space. After irradiation the individual targets inside the artificial meteoroid were distributed to the collaborating laboratories, where the residual product nuclides are investigated by X-ray and gamma ray spectrometry as well as by conventional and accelerator mass spectrometry. Up to now, more than 200 depth profiles have been measured. Measurements of stable and long-lived nuclides will begin after the targets have cooled sufficiently. The experimental data are interpreted by model calculations based on spectra of primary and secondary particles derived by Monte Carlo techniques and experimental and theoretical thin-target cross sections of the underlying nuclear reactions. The calculations allow us to distinguish the influence of bulk chemical composition onto cosmogenic nuclide production rates and to improve the modeling of production rates in iron meteoroids. Acknowledgment: This work was supported by the Swiss National Science Foundation. References: [1] Michel R. et al. (1985) Nucl. Instr. Meth. Phys. Res., B16, 61-82; (1989) ibid., B42, 76-100; (1993) J. Radioanal. Nucl. Chem., 169, 13- 25. [2] Herpers U. et al. (1991) Meteoritics, 26, 344. [3] Wieler R. et al. (1992) Meteoritics, 27, 315-316. [4] Weber H. W. and Begemann F. (1992) Meteoritics, 27, 305. [5] Gilabert E. et al. (1992) Meteoritics, 27, 223-224. [6] Michel R. et al. (1991) Meteoritics, 26, 221-242. [7] Begemann F. and Schultz L. (1988) LPS XIX, 51-52. [8] Michel R. et al. (1990) Meteoritics, 25, 386-387. [9] Masarik J. et al. (1993) LPS XXIV, 937-938. Budka P. Z.* Viertl J. R. M. Mundrabilla: A Microgravity Casting The name "Mundrabilla" is applied to two nickel-iron meteorite masses (combined mass over 22,700 kg), which apparently were a single mass before atmospheric entry [1]. A medium octahedrite, Mundrabilla exhibits the microstructural features common to other nickel-iron meteorites such as Widmanstatten structure and troilite; however, its macrostructure is anything but common. Described by Buchwald as "anomalous" [1], Mundrabilla's macrostructural morphology is characterized by strikingly prominent, rounded Widmanstatten areas separated by regions of sulfur segregation (Fig. 1). While microstructural development of a metal can reflect both solidification and solid state reactions, macrostructural features are determined during solidification. Thus, a typical metallurgist, unfamiliar with microgravity solidification, might describe Mundrabilla's macrostructure as an "anomalous" casting. Those familiar with microgravity solidification might characterize Mundrabilla's macrostructural features as due to solidification of two immiscible liquids [2]--one rich in nickel-iron, the other rich in sulfur. Combining these observations, Mundrabilla's macrostructural features are consistent with that of a liquid mass solidified under microgravity conditions [3,4]. Since nickel-iron meteorite cooling rates often serve as the foundation for assumptions about the formation of solar system bodies, information on the solidification time for the Mundrabilla mass may give additional insights. How long did it take for Mundrabilla, with a minimum "as received" mass of approximately 22,700 kg to solidify? Because Mundrabilla's mass before atmospheric entry is unknown, we take as an upper boundary a mass of 4.1 x 10^15kg. These masses, assumed spherical, range in diameter between 1.8 meters and 10 kilometers, respectively. Mundrabilla can be idealized as a pure iron liquid mass cooling from the melting point of pure iron (1535C) by radiation into space at absolute zero. The latent heat of transformation for iron is used to calculate "excess temperature," i.e., the amount the mass temperature can be raised due to recalescence. Solidification is considered complete when the center of the mass is solid. Fig. 2, is a plot of the solidification times for an iron mass in the range 1.8 meters to 10 kilometers in diameter. At the lower bound, solidification time is about 1.6 hours; at the upper bound, solidification time is on the order of 3,400 years. References: [1] Buchwald V. F. (1975) Handbook of Iron Meteorites, University of California, Berkeley. [2] Carlberg T. and Fredriksson H. (1980) Metallurgical Transactions A, 11A, 1665-1676. [3] Budka P. Z. (1988) Metallurgical Transactions A, 19A, 1919-1923. [4] Budka P. Z. (1988) J. Metals, 40, 9, 6-9. Fig. 1, which appears here in the hard copy, shows Mundrabilla--a scale in inches. Figure 2, which appears here in the hard copy, shows solidification time vs. diameter. Takeda H.* Baba T. Saiki K. Otsuki M. Ebihara M. A Plagioclase-Augite Inclusion in Caddo County: Low-Temperature Melt of Primitive Achondrites Caddo County is an IAB-iron meteorite with silicate inclusions. Its silicate inclusion with olivine and clinopyroxene as major minerals was studied by INAA by Palme et al. [1]. We report mineralogy and INAA data of an inclusion with abundant augite and plagioclase, which will provide us with information on the missing materials removed from sourse regions of the primitive achondrites. The PTS of Caddo County, 7.5 x 5.5 mm in size has been studied by mineralogical techniques including electron probe microanalysis (EPMA) and scanning electron microscopy (SEM) with digital imaging utilities. A thick slice of the specimen next to the PTS was crushed and separated into metal- rich and a silicate-rich fractions for determination of lithophile and sidelophile element concentrations by INAA. The PTS shows aligned textures of silicate inclusions in metallic matrix, suggesting that some parts are intruded into this place. Augite (Aug) and plagioclase fragments are rounded and anhedral, and numerous metal veins fill interstices of the silicate grains. The silicate parts are mainly composed of augite and plagioclase with additional small amounts of minor orthopyroxene (Opx). Olivine is present only as small grains. Crystals of augite reach up to 0.8 x 0.4 mm in size and plagioclase crystals up to 0.6 x 0.3 mm in size. Many plagioclase grains in the metal matrix within one region shear a common crystallographic orientation as indicated by polysynthetic twin lamellae. The Opx distributes in a limited region and Opx and Aug are in contact with curved boundaries even within a grain. The pyroxene compositions are Aug (Ca44.5Mg52.5Fe3) and Opx (Ca2Mg91.5Fe6.5). The plagioclase compositions (Or3Ab80An17) are Na rich and show small ranges from grain to grain. The compositions of olivine (Fa97.5) are similar to that of winonaites [2]. The modal abundances are: Aug 20 vol%, plagioclase 28%, Opx 3%, metallic irons etc. 49% (excluding minor olivine and Opx). Na, Al and REE are enriched in the silicate phases in Caddo County: Na (5.79X CI), Al (6.17X CI) in contrast with Bild and Wasson's [3] results on Lodran, which showed depletion of Ca relative to Mg or Si. Mg-normalized abundance ratios of Ca in the silicates of Caddo County (11.8X CI) are one order of magnitude larger than Acapulco (0.9X CI). Ni content of metal phase of Caddo County is about 7%, and Ir is 2.5 ppm. The Ru content in the FeNi metal is 2 to 5 ppm. Abundances of Ni, Ir, and Ru are in the range of IAB iron meteorite in agreement with those by Palme et at. [1]. According to the formation model of lodranites [4], Ca-Al-rich melt and Fe-Ni- S eutectic melt are removed from source materials. Kracher and Wasson [5] suggested that the type IAB iron meteorites have been formed out of the Fe-Ni- S melt. It is interesting to note that Ca-Al-rich melts removed during formation of lodranitelike materials crystallize plagioclase and Aug. The Ca- Al-rich silicates like plagioclase and Aug are major minerals in the silicate inclusion of Caddo County. From bulk chemistry by INAA, Ca, Al, REE, typically absent in lodranites, were enriched in Caddo County. These facts indicate that this type of meteorite was formed from the Fe-Ni-S-rich and Ca-Al-rich melts. However, we have to admit that silicate inclusions in the IAB group have different oxygen isotopic abundances from those of lodranites. Lodranitelike meteorites in mineralogy and bulk chemistry are missing in the winonaite-IAB group and those of Caddo County have not been found in the Acapulcoite- lodranite group. References: [1] Palme H. et al. (1991) LPSC XXII, 1015-1016. [2] Kimura M. et al. (1992) Proc. NIPR Symp. Antarct. Meteorites, 5, 165-190. [3] Bild R. W. and Wasson J. T. (1976) Mineral. Mag., 40, 721-735. [4] Takeda H. et al. (1992) Proc. Japan Academy, 68, B, 115-120. [5] Kracher A. and Wasson J. T. (1982) GCA, 46, 2419. McCoy T. J.* Steele I. M. Keil K. Leonard B. F. Endress M. Chladniite: A New Mineral Honoring the Father of Meteoritics The IIICD irons are a small group of meteorites, three of which (Maltahohe, Carlton, and Dayton) contain silicate-bearing inclusions rich in troilite, graphite, schreibersite, and phosphates [1]. The Na,Ca,Mg-rich phosphates bnanite and panethite were first described in Dayton [2]. We have discovered a new mineral, Na(sub)2CaMg(sub)7(PO(sub)4)(sub)6, as a single grain within a silicate-bearing inclusion in the Carlton (IIICD) iron meteorite. The mineral and mineral name have been approved by the Commission on New Minerals and Mineral Names of the International Mineralogical Association. Chladniite occurs as a single grain near the edge of a silicate-bearing inclusion in polished section USNM 2707. This inclusion is dominated by chlorapatite and contains olivine, pyroxene, plagioclase, schreibersite, and troilite. Chladniite occurs as a single, massive grain (975 x 175 micrometers) and is cross-cut by hydrated iron oxides of terrestrial origin. In polished section, it is gray, dark, and weakly anisotropic. Cleavage is rhomboidal in plan and very likely rhombohedral in three dimension. The formula for chladniite (derived from five microprobe analyses) is Na(sub)1.77Si(sub)0.08 Ca(sub)0.98(Mg(sub)6.96Fe(sub)0.26Mn(sub)0.04)(sub)Sigma = 7.26(Po(sub)0.98 O(sub)4)(sub)6. The idealized formula is Na(sub)2CaMg(sub)7(PO(sub)4)(sub)6. Chladniite is related to two rare minerals, fillowite [3] and johnsomervilleite [4], where fillowite is the Mn-dominated and johnsomervilleite the Fe-dominated analog of chladniite. The unique occurrence of chladniite, the relatively small size of the grain, and the presence of terrestrial weathering veins all presented challenges for removing material for X-ray studies. A 30-micrometer-diameter spindle of material was removed after microdrilling a shallow trench and breaking the spindle with a surgical scalpel. Studies were performed using both a Gandolfi camera to obtain a powder pattern and a four-circle diffractometer to determine the unit cell. A total of 17 lines were observed in the powder pattern. Chladniite is hexagonal, R 3(bar), a = 14.967 angstroms, c = 42.595 angstroms, beta = 120 degrees. Attempts to determine the structure of chladniite are in progress. Chladniite is named for Ernst Florens Friedrich Chladni (1756-1827), who is widely regarded as the "Father of Meteoritics." After his initial training as a lawyer, Chladni turned his attention to science, particularly problems in acoustics. He was not, however, able to obtain a permanent position and embarked upon the life of a nomad, traveling among the great cities of Europe lecturing about acoustics. During these travels, he eventually gained an interest in meteoritics. It was Chladni's pioneering book of 1794 that, for the first time, presented strong evidence for an extraterrestrial origin of meteoritic stones and irons [5]. In addition, Chladni argued that meteorites must have been the building blocks of all planets and argued that a large iron core must exist inside the Earth. During his extensive travels, Chladni also established a meteorite collection that can still be seen at Humboldt University in Berlin. It is appropriate that a mineral be named in his honor as we approach the 200th anniversary of the publication of his monumental work. References: [1] McCoy et al. (1993) Meteoritics, in press. [2] Fuchs et al. (1967) GCA, 21, 1711-1719. [3] Araki and Moore (1981) Am. Mineral., 66, 827-842. [4] Livingstone (1980) Min. Mag., 43, 833-836. [5] Chladni (1794) Riga, J. F. Hartknoch (in German); reprinted (with introduction by G. Hoppe) by Akad. Verlagsgesellschaft Geest & Portig K.-G. (1982) (in German). Monday, July 19, 1993 Carbonaceous Chondrites, Alteration Processes, and Metamorphism 8:30 a.m. Cascade Ballroom Chair(s): W. Calvin G. Dreibus Palme H.* Spettel B. Ikeda Y. Origin of Chondrules and Matrix in Carbonaceous Chondrites We have recently shown that in Allende, chondrules and matrix are complementary in composition. Chondrules have low Fe/Cr ratios and matrix has high Fe/Cr ratios; the bulk meteorite has the CI ratio. Chondrules on the average have Ca/Al ratios below the solar system ratio of 1.10, which is also the bulk Allende ratio; matrix and related dark inclusions have ratios above 1.10. It was therefore concluded that chondrules and matrix formed from a single reservoir characteristic of the bulk Allende composition [1,2]. Recent reports on the CV meteorite Y-86751 [3,4] indicated the opposite relationship, with high Ca/Al ratios in chondrules and low ratios in matrix. Different compositions of individual components of Y-86751 and of Allende but similar bulk compositions of both meteorites may reveal important details on the preaccretionary evolution of the carbonaceous chondrites. We have therefore begun chemical analyses of the bulk of Y-86751 and its individual lithic components, chondrules, matrix, amoeboid olivine inclusions, Ca,Al-rich inclusions, etc. Results of the bulk analysis of Y-86751 show that this meteorite has (within the accuracy of the analysis) the same bulk composition as Allende. In particular, the Ca/Al ratio is within 3% of the average solar system ratio of 1.1 (by weight). The only statistically resolvable difference is the 50% higher content of Zn in Y-86751. Major-element INA analyses indicate a Ca/Al ratio above the chondritic ratio for several chondrules and opposite to that of average Allende chondrules [1]. Matrix samples have a tendency for low Ca/Al ratios, confirming EMP-matrix analyses [4]. The separates are presently analyzed for a large number of trace elements. After the analyses, separates will be investigated petrographically. The apparent disequilibrium among Allende components and the strong compositional zoning of olivine grains constrain the thermal history of the Allende parent body [e.g., 5]. Any metamorphic redistribution of Ca or Al between chondrules and matrix can be virtually excluded. Therefore, the complementary relationship in the distribution of Ca and Al between matrix and chondrules must be of nebular origin and must reflect conditions of formation of chondrules and matrix. Early condensation and separation of spinel may be responsible for the high Ca/Al in Y-86751 chondrules. The early separated spinel grains were later collected with the matrix and are responsible for the low Ca/Al ratio in Y-86751 matrix [3,4]. A different evolution, where chondrules incorporated less Ca than Al, must have occurred for Allende. Perhaps conditions were more oxidizing and some Ca remained as Ca(OH)(sub)2 in the gas before matrix was formed, as suggested by Hashimoto [6]. Allende dark inclusions representing matrix have excess Ca while all other refractory elements occur in CI proportions, suggesting a unique behavior of Ca [7]. A different nebular history for two members of the same group (CV) would indicate strong local differences in formation conditions for chondrules and matrix from two identical nebular reservoirs. In addition, if both meteorites come from the same parent body there could only be limited mixing on the parent body to retain the different signatures. References: [1] Palme H. et al. (1992) LPS XXIII, 1021-1022. [2] Palme H. (1992) 17th Symp. Antarc. Met., Tokyo, 71-1-3. [3] Murakami T. et al. (1992) 17th Symp. Antarc. Met., Tokyo, 11-1-2. [4] Murakami T. and Ikeda Y. (1993) in preparation. [5] Weinbruch S. et al. (1990) Meteoritics, 25, 115-125. [6] Hashimoto A. (1992) GCA, 56, 511-532. [7] Palme H. et al. (1989) Z. Naturforsch., 44a, 1005-1014. Dreibus G.* Palme H. Spettel B. Wanke H. Sulfur and Selenium in Chondritic Meteorites Selenium is the only truly chalcophile element in chondritic meteorites. It has no other host phases except sulfides. Since Se-volatility is similar to S-volatility one may expect constant S/Se ratios. To test this hypothesis chondritic meteorites were analyzed for Se and S. To avoid problems from inhomogeneous distribution of sulfides the same samples that had been analyzed for Se by INAA were analyzed for S (see Table 1) using a Leybold Heraeus Carbon and Sulfur Analyser (CSA 2002). Solar System Abundances of S and Se: The average S-content of CI- meteorites is with 5.41% in agreement with an earlier average of 5.25% for Orgueil [1], but not with higher S-contents for Ivuna, Alais, and Tonk. Inclusion of these data led to an average CI- content of 6.25% in the Anders and Grevesse compilation [2]. The essentially constant average S/Se ratio in all groups of carbonaceous chondrites of 2563 +- 190 suggests that our Orgueil S-content provides a reliable estimate for the average solar system. The new solar S/Se ratio and the CI-value of Se of 21.3 ppm [3] yield an atomic S/Se ratio of 6200 +- 170, 24% below that calculated from [2]. Weathering Effects: Some of the carbonaceous chondrite finds have similar S/Se ratios as falls (see Table 1). However the badly- weathered Arch (CVR) and Colony (CO) and the two C4-chondrites Mulga West and Maralinga have much lower S and somewhat lower Se contents compared to unweathered meteorites. Their S/Se ratios of 1000-230 indicate higher losses of S--probably by oxidation--as of Se. The low Na-contents in Arch and Colony rel. to CV3 and CO3 may also reflect weathering. Low S/Se ratios in the Sahara meteorites are also indicative of weathering processes. The depletion factors for the CV3- chondrite Acfer086 are, relative to average CV, 10 (S), 5 (Se), 6 (Na), and 4 (Ni). Lower absolute depletions, but the same depletion sequence are found for the CO-meteorite Acfer 202. In the CO/CM Acfer 094 only S and Na are depleted. The influence of weathering in the two CR-types Acfer 097 and Acfer 270 is less obvious. Although Se does not appear to be depleted in these meteorites [4] the lower S/Se ratios of 1660 res. 1970 rel. to CI and the low Na-contents indicate weathering related losses of S and Na. Losses of Ni by weathering are more pronounced in meteorites containing Ni-rich sulfides, whereas metallic Ni is apparently less affected (CR-meteorites). A high depletion of S and Ni but none for Se and Na is found in the Carlisle Lake-type, Acfer 217. In summary, weathering effects in the carbonaceous chondrites result in losses of S, Se, Na, and Ni. Sulfur is in all cases significantly more affected by weathering than Se resulting in low S/Se ratio rel. to CI. References: [1] Mason B. (1962) Space Sci. Rev., 1, 621-646. [2] Anders E. and Grevesse N. (1989) GCA, 53, 197-214. [3] Spettel B. et al. (1993) this volume. [4] Bischoff A. et al. (1993) GCA, 57, in press. Nakamura N.* Kimura M. Shimoda H. Nohda S. Correlated Study of Rb-Sr Systematics and Petrologic Properties of Chondrules from Allende (CV): Evidence for Secondary Alteration In order to clarify when and where the distributions of alkalis in chondrules were established [1-3] we have undertaken a correlated study of Rb-Sr systematics and petrologic properties of chondrules from the Allende (CV) meteorite. We report here the results of Rb-Sr isotopic analyses combined with petrologic examinations for 16 chondrules (19 specimens) from Allende. The whole-rock analyses indicate a model age of 4.5 +/- 0.05 Ga (using ALL [4]), suggesting that the Rb-Sr isotopic system has been closed in bulk Allende for 4.5 yr. For the chondrules, the earlier works [4,5] include a few barred-olivine and many unknown petrographic types (probably mostly porphyritic) chondrules. In this work, we present results for 1 radial pyroxene, 4 porphyritic (7 specimens), and 11 barred olivine chondrules, covering major petrographic types. The results are shown in an 87Rb-87Sr evolution diagram (Fig. 1). Including earlier results [4,5], no systematic differences are found for different textural types, but data points mostly deviate from the 4.5-Ga line to the right, forming a rough linear array. This means that the Rb-Sr isotopic system in chondrules have been disturbed by a late event(s). The slope of the dotted line corresponds to an age of ~4.0 Ga. The meanlng of the linear trend is not clear because of the lack of isotopic equilibrium. Ground masses of chondrules analyzed for Rb-Sr isotopes were examined by optical microscope and EPMA. Some chondrules have primary glassy ground masses enriched in the anorthite component, especially in their central parts. However, as demonstrated previously [6], most of the chondrules analyzed here for Rb-Sr isotopes were also altered. They have abundant nepheline and sodalite components that replace primary glassy ground masses. Sodium was secondarily introduced and Ca was lost from the Allende chondrules. Using abundances of dodalite and nepheline components, the degree of alteration was tentatively classified into three categories: A (least altered), B (middle), and C (most altered). In Fig. 1, the chondrules with higher 87Rb/86Sr are mostly in categories of B or C. On the other hand, the chondrules of category A have the lower 87Rb/86Sr ratio, relatively close to that of bulk Allende. The 87Rb/86Sr ratios are related to the alteration categories but are not directly related to the isotopic deviation. It is thus considered that the general trend of high and low elemental ratios of Rb/Sr had been basically established during the early nebular processes and were then modified significantly more recently. The diffusivities of alkalis in glassy materials [7] at low temperatures (~400 degrees C) suggest a possible migration of Rb as well as Na from matrix to ground masses of chondrules. The old model age for Allende "matrix" [5] is consistent with such a possibility. We therefore strongly suggest that the Allende chondrules were subjected to a low- temperature alteration reaction after consolidation of the chondrules. References: [1] Grossman J. N. and Wasson J. T. (1983) In Chondrules and Their Origins (E. A. King, ed.), 88-121, LPI. [2] Hewins R. H. (1991) GCA, 55, 935-942. [3] Matsuda H. et al. (1990) Meteoritics, 25, 137-143. [4] Gray C. M. et al. (1973) Icarus, 20, 213-239. [5] Tatsumoto M. et al. (1976) GCA, 40, 617-634. [6] Kimura M. and Ikeda Y. (1992) Papers 17th Symp. Antarct. Meteor., 31-33. [7] Jambon A. and Carron J. P. (1976) GCA, 40, 897-903. Fig. 1, which appears here in the hard copy, shows an 87Rb-87Sr evolution diagram for chondrules from the Allende meteorite. Endress M.* Bischoff A. Mineralogy, Degree of Brecciation, and Aqueous Alteration of CI Chondrites Orgueil, Ivuna, and Alais The textural and mineralogical properties of the CI-chondrites Orgueil, Ivuna, and Alais were studied by electron microscopy in order to obtain new insights into the history and evolution of the CI parent body. Degree of Brecciation. The electron optical investigation of Orgueil, Ivuna, and Alais revealed that the three samples experienced different degrees of brecciation: (1) Orgueil is highly brecciated. It consists of abundant clastic matrix (clasts usually <10 micrometers) surrounding fragments of variable sizes (10-700 micrometers). (2) Alais is much less brecciated than Orgueil and in this manner more closely related to Ivuna (see below); fragments (100-500 micrometers in size) are embedded in clastic matrix. (3) Ivuna shows the lowest degree of brecciation and contains only small areas of clastic matrix. The major part of the sections studied consists of huge fragments (up to 1 x 3 mm in their largest dimension) representing at least three distinct lithologies: The first lithology can be characterized by a high modal abundance of carbonates (up to 10 vol%) and a lack of sulfates; the second is enriched in sulfates, but contains no carbonates, and the third lithology can be characterized by its high abundance of coarse-grained phyllosilicate fragments. Mineralogy. As phases >5 micrometers in size we observed magnetite, sulfides, carbonates, Ca-phosphate, olivine (Fa<5; 14 grains), pyroxene (Fs<1; 6 grains), sulfates, phyllosilicates, and as accessory phases Mn-rich chromite (15 wt% MnO), Mn-rich ilmenite (11 wt% MnO), Fe,Ni-metal grains in olivine relics and a yet unidentified Na,K-rich silicate (7 wt% Na2O; 6 wt% K2O; only in Ivuna). Magnetites (pure Fe3O4) occur as platelets, framboids, and spherules. As sulfides pyrrhotite, pentlandite (partly exsolved in pyrrhotite) and cubanite were observed. Carbonates (dolomite, breunnerite, calcite) either occur as single grains (usually <40 micrometers) or as big polycrystalline chunks (up to 300 micrometers in size; especially in Ivuna). The Ca-phosphates contain small amounts of FeO (1-2 wt%), MnO (1 wt%), and Na2O (up to 3 wt%). Sulfates mostly occur as vein fillings in Alais (mainly gypsum) and Orgueil (mainly epsomite), but such veins are lacking in our Ivuna sample. Here, single sulfate grains (<10 micrometers) are rarely dispersed throughout the matrix. Aqueous alteration. Based on a study of Orgueil and Y-82162, [1,2] proposed a model, that during late stage aqueous alteration Fe-rich solutions readily decompose coarse-grained phyllosilicate clusters to fine-grained phyllosilicates and ferrihydrite and that contemporaneously sulfates were formed. Since Y-82162 contains abundant phyllosilicate clusters, but no ferrihydrite and no sulfate, Y-82162 should be less altered than Orgueil [1]. Considering this model we can distinguish between Orgueil, Ivuna, and Alais based on the degree of alteration. The abundance of phyllosilicate fragments is low in Orgueil, but high in Ivuna and Alais. In contrast to Orgueil, Ivuna contains no ferrihydrite [3,4] and very little sulfate. Therefore, the degree of aqueous alteration appears to be increasing in the order: Y-82162 << Ivuna < Alais << Orgueil. Furthermore, preliminary studies may indicate a link between the degree of aqueous alteration and the degree of brecciation. TEM studies are in progress in order to resolve this question. References: [1] Tomeoka K. and Buseck P. R. (1988) GCA, 52, 1627- 1640. [2] Tomeoka K. et al. (1989) Proc. NIPR Symp. Antarct. Met., 2, 36-54. [3] Zolensky M. et al., GCA, submitted. [4] Brearley A. J. (1992) LPSC XXIII, 153-154. Tomeoka K.* Microtextures of Phyllosilicates in the Mokoia CV Chondrite: Evidence for Aqueous Alteration Prior to Consolidation The CV and CO chondrites were less affected by aqueous alteration than the CI and CM chondrites; thus they may serve as important indicators of the earliest stages of carbonaceous chondrite formation. Mokoia is one of the rare CV chondrites that contains considerable amounts of phyllosilicates [1,2]. They occur in CAIs, some chondrules, and matrix; they also occur abundantly in some chondrule rims. Tomeoka and Buseck [2] showed that the phyllosilicates in Mokoia are saponite and Na-rich phlogopite and were formed by aqueous alteration of olivine, pyroxene, anorthite, and glass. However, the mineralogy and occurrence of phyllosilicates in Mokoia are different from those in CI and CM chondrites, suggesting that the alteration of Mokoia occurred in a condition distinct from those for CI and CM chondrites. Most phyllosilicates in CI and CM chondrites probably resulted from alteration of olivine and pyroxene by the activity of liquid water on the meteorite parent bodies (e.g., [3]). Major questions are (1) What were the conditions responsible for the alteration in Mokoia? (2) Were these conditions different from those experienced by the CI and CM chondrites? (3) Where did the aqueous alteration in Mokoia occur? To address these questions, I performed extensive observations of the Mokoia CV chondrite by using a scanning electron microscope equipped with an EDS analysis system. The Mokoia chondrite is a breccia composed of submillimeter to millimeter clasts. Its matrix consists largely of fine grains (<0.1 to 10 micrometers in diameter) of Fe-rich olivine. Phyllosilicates occur in narrow interstices (<1 micrometer) between the olivine grains. The abundance of phyllosilicates in matrix differs within a clast and among clasts. Some clasts that are rich in phyllosilicates are adjacent to phyllosilicate-poor or -free clasts; the evidence suggests that aqueous alteration of the matrix preceded some of the brecciation events that produced the clasts from earlier matrix. Phyllosilicates are much more abundant in some chondrules and CAIs than in matrix in Mokoia; in some chondrules, phyllosilicates extend over an area of 100 x 100 micrometers. Chondrules and inclusions that were altered in various degrees occur together with those that remain unaltered within the same clasts Because of small sizes, olivines in the Mokoia matrix are more permeable to fluids and thus more reactive to aqueous alteration than the coarser olivines and pyroxenes in chondrules. Therefore, if alteration.of chondrules occurred in situ in the parent body (after consolidation), the matrix olivines should have been altered preferentially to those in chondrules, which is just the opposite of what I observe. These observations suggest that some chondrules and inclusions experienced a distinct environment in terms of hydration from that of the matrix materials. Mineralogy itself does not provide precise constraints on whether the alteration occurred via a solar nebular gas or aqueous solutions. However, based on the present observations, it is likely that some chondrules, inclusions, and portions of matrix in Mokoia have experienced aqueous alteration prior to consolidation into the present configuration. References: [1] Cohen R. E. et al. (1983) GCA, 47, 1739-1757. [2] Tomeoka K. and Buseck P. R. (1990) GCA, 54, 1745-1754. [3] Tomeoka K. (1990) Nature, 345, 138-140. Brearley A. J.* Geiger T. Fine-grained Chondrule Rims in the Murchison CM2 Chondrite: Compositional and Mineralogical Systematics The CM2 carbonaceous chondrites contain numerous chondrules, silicate grains, CAIs, etc, which are mantled by rims of fine- grained material [1]. These rims often consist of two or more layers that can be distinguished on chemical and textural criteria. The formation mechanisms of rims, and the timing and location (i.e., nebular vs planetary) of aqueous alteration of rim materials is the subject of some controversy. We are examining, in detail, the relationships between the mineralogy and bulk composition of individual rims in the Murchison CM2 chondrite in order to place constraints on their mechanisms of formation and alteration histories. We have carried out detailed SEM, electron microprobe and TEM studies of a number of rim sequences in Murchison. Our data provide further confirmation of some of the observations of [1]. For example, we have found that on any given chondrule, inner rims are almost invariably more Mg-rich than outer rims. However, when the entire population of rims is considered it is evident that the compositional field for inner rims overlaps that for outer rims in terms of Mg/Fe ratio. Na and K are also consistently enriched in inner rims, but all the other analyzed elements show variable behavior. Elemental ratio diagrams for rims show some variation in their shape, but in most cases are relatively flat. The elements that consistently show exceptions are Ca and S are frequently depleted relative to CI values in both inner and outer rims. We have also examined the interelement variations in inner and outer rims. One of the surprising results of this study is that some elements may be correlated in inner rims, but not in outer rims and vice versa. Fe and S show a strong positive correlation in outer rims, but have no correlation in inner rims. The reverse is true of Na and S. Our TEM studies of the fine-grained mineralogy of 5 rims have, so far, revealed consistent relationships between rim composition and mineralogy. All the inner rims studied consist dominantly of microcrystalline Mg-rich serpentine, rare platy cronstedtite crystals and poorly crystalline pentlandite and pyrrhotite. The sulfides are disseminated throughout the regions of microcrystalline serpentine. Tochilinite has not been found. The compositions of serpentine and cronstedtite in the different rims studied are very similar and define distinct, tightly clustered compositional groups on Si-Fe-Mg ternary diagrams. Outer rims are mineralogically distinct. For intermediate Mg/Fe ratios, outer rims are dominated by relatively coarse-grained platy cronstedtite, a minor amorphous component and sulfides, whereas the most Fe-rich outer rims contain tochilinite and minor cronstedtite. Our present data indicate clear relationships between bulk rim composition and mineralogy, which would appear to support a parent body location for aqueous alteration, rather than nebular. In addition, the evidence that some elements show variable correlations in inner and outer rims, indicates that there must be mineralogical controls on the major and minor element chemistry of rims. This may reflect variations in the mineralogy of the precursor components of rims, or the mineralogical constraints imposed on elemental mobility between rims and chondrules during alteration. The depletions in Ca and S in rims may be a reflection of the high mobility of these elements during alteration, as is certainly the case in CI chondrites. Finally the textural characteristics of inner rim materials appear to be inconsistent with alteration of a crystalline precursor, because there is no evidence of pseudomorphic replacement of phases. Many of the textures are similar to those produced during low temperature alteration of basaltic glass [2]. The possibility that the precursor was an amorphous material, perhaps of the type observed in ALH A77307 [3] and several of the least equilibrated ordinary chondrites [4], should be considered. Funded by NASA grant NAGW-3347 to J. J. Papike (P.I.). References: [1] Metzler K. et al. (1992) GCA, 65, 2873-2897. [2] Tazaki K. et al. (1989) Clays Clay Minerals, 37, 348-354. [3] Brearley A. J. (1993) GCA, 57, 1521-1550. [4] Alexander C. M. et al. (1989) EPSL, 95, 187-207. Kojima T.* Tomeoka K. Takeda H. Unusual Dark Clasts in the Vigarano CV3 Chondrite: Record of Parent Body Process A variety of dark lithic clasts have been reported from CV3 chondrites and are commonly called "dark inclusions" (DIs). The DIs widely range in texture from chondritic with chondrules and Ca-Al-rich inclusions (CAIs) embedded in a matrix (similar to host meteorites), to fine-grained aggregates of Fe-rich olivine free of coarse-grained components [1,2]. The DIs have been interpreted to represent (1) primary aggregates of materials in the solar nebula [3-5] and (2) materials that were affected by thermal metamorphism on their parent bodies [6]. We present the results of petrographic and scanning electron microscope (SEM) studies of two unusual clasts found in the Vigarano CV3 chondrite. The two unusual clasts, which we will call CL1 and CL2, are approximately 1.2 x 1.0 mm^2 (CLl) and 0.8 x 0.6 mm^2 (CL2) and occur within one of the large clasts (2.8 x 1.0 mm^2). CL1 and CL2 have very similar mineralogies and textures; they contain irregular to oval-shaped inclusions consisting mostly of fine grains of Fe-rich olivine embedded in the matrix of the clasts, and are free of distinct chondrules, CAIs, and coarse mineral fragments. Thus, they resemble the fine-grained variety of DIs. Under the optical microscope, most inclusions resemble chondrules or chondrule fragments in shape and size. However, they are brownish-translucent in transmitted light and are clearly distinct from chondrules in the Vigarano host. The inclusions are characteristically flattened in direction, exhibiting apparent foliation. Our SEM observations reveal the following unusual characteristics: (1) the inclusions are not mere random aggregates of olivine grains but have peculiar internal textures, i.e., assemblies of round or oval-shaped outlines, which are suggestive of pseudomorphs after porphyritic or granular olivine chondrules; (2) one of the thick inclusion rims contains a network of veinlike strings of elongated olivine grains, (3) an Fe-Ni metal aggregate in CL1 has an Fe-, Ni-, S-rich halo, suggesting a reaction between its precursor and the surrounding matrix; and (4) olivine in the clasts commonly shows a swirly, fibrous texture similar to that of phyllosilicate. These characteristics are not reconciled with a primary (unprocessed) origin for the clasts but suggest that they were involved in a secondary process such as aqueous alteration. The fine grains of olivine in these clasts were presumably produced by thermal transformation of phyllosilicate. Serpentine is known to start transforming to olivine at >300 degrees C [7]. The inclusions in CL1 and CL2 resemble chondrules: Some of them are surrounded by distinctive rims similar to chondrule rims commonly seen in the carbonaceous chondrites. Therefore, we believe that the precursor material of the clasts would have been related to some types of chondrites, possibly the Vigarano host itself, and that aqueous alteration and thermal metamorphism occurred in the meteorite parent body. We suggest that some of the dark inclusions and clasts previously reported from CV3 chondrites and other types of meteorites may have origins common to these clasts in Vigarano. References: [1] Fruland R. M. et al. (1978) Proc. LPSC 9th, 1305-1329. [2] Johnson C. A. et al. (1990) GCA, 54, 819-830. [3] Bischoff A. et al. (1988) LPSC XIX, 88-89. [4] Kurat G. (1989) Z. Naturforsch., 44a, 988-1004. [5] Palme H. (1989) Z. Naturforsch., 44a, 1005-1014. [6] Bunch T. E. and Chang S. (1983) LPSC XIV, 75-76. [7] Akai J. (1992) Proc. NIPR Symp. Antarct. Meteorites, 5, 120-135. Symes S. J. K.* Guimon R. K. Benoit P. H. Sears D. W. G. Thermoluminescence and Metamorphism in CV Chondrites One of the effects of metamorphism in meteorites is the production of feldspar, a thermoluminescence (TL) phosphor, through the devitrification of primary chondrule glass [1]. The 105-fold variation in TL sensitivity among the ordinary chondrites reflects this process and has been used successfully to subdivide the petrographic type 3 meteorites into types 3.0-3.9 [2]. Although less pronounced, the variability exhibited by the CO chondrites has also allowed petrographic subdivision of these meteorites [3]. It is possible that the CV chondrites have also experienced a range of metamorphic intensities, although McSween has warned that their petrography does not indicate a simple sequence [4]. On the other hand, Scott et al. show that the homogeneity of matrix olivine increases along the series Kaba, Mokoia, Vigarano, Grosnaja, Allende, which may indicate progressive thermal metamorphism [5]. Here we report TL sensitivity measurements for 12 whole-rock samples of CV chondrites and we suggest petrographic type assignments and discuss their metamorphic history. Samples of bulk powder were ground, the magnetic fraction removed, and the TL of 4-mg aliquots was measured three times for duplicate splits. Averages are given in Table 1, which appears in the hard copy. The CV chondrites, like the CO chondrites, generally display three peaks in their glow curves; one at 130 degrees C, which is sensitive to metamorphism at temperatures below 650 degrees C [3], one at 250 degrees C, which is metamorphism independent, and one at 350 degrees C, which might be associated with refractory minerals in CAI [6]. The TL sensitivities of these samples show a >100-fold range, the lowest being below detection limits (<0.01) and the highest being greater than the Dhajala H3.8 chondrite, which we use as a standard. Six of the 12 samples have TL sensitivities corresponding to type 3.0 if we apply the criteria proposed by Sears et al. to subdivide the CO chondrites (which are similar to those used for the ordinary chondrites) [3]. All but one of the remainder have sensitivities corresponding to minimal metamorphism (type 3.2-3.3). The exception is Coolidge, whose TL, like many other properties [4], indicates that it is petrographic type 4. Our TL data therefore indicate that, with the exception of Coolidge, the CV chondrites have experienced minimal metamorphism, although evidence for some slight variation is present. Even though Coolidge has been metamorphosed to type-4 levels, like CO chondrites and unlike ordinary chondrites, the temperatures involved were <650 degrees C since the TL peak temperature is still 130 degrees C. Apparently, the most metamorphosed CV and CO chondrites were metamorphosed for longer times than ordinary chondrites, but at lower temperatures. References: [1] Guimon et al. (1985) GCA, 49, 1515-1523. [2] Sears et al. (1980) Nature, 287, 791-795. [3] Sears et al. (1991) Proc. NIPR Symp. Ant. Met., 4, 319-343. [4] McSween H. Y. (1977) GCA, 41, 1777-1790. [5] Scott et al. (1988) in Meteorites and the Early Solar System (Kerridge and Matthews, eds.), 718-745. [6] Guimon R. K. and Sears D. W. G. (1986) Meteoritics, 21, 381-382. Lerner N. R.* Influence of Murchison Minerals on Hydrogen-Deuterium Exchange of Amino Acids The amino acids found on the Murchison meteorite are deuterium enriched. For the glycine-alanine fraction, delta D = +2448 per mil, and for the alpha-amino isobutyric acid fraction, delta D = +149 per mil [1]. In order to retain such levels of deuterium enrichment, the amino acids found in Murchison must have not only retained the deuterium enrichment of their interstellar precursors (delta D > +1500 per mil [2]) during synthesis, as has been recently shown [3], but they must have also retained their deuterium label during the aqueous alteration phase [4]. By measuring the rates of deuterium exchange of amino acids with D(sub)2O, limits can be set on the length of time and the conditions under which the Murchison parent body experienced an aqueous environment. The rates of hydrogen-deuterium exchange of nondeuterated glycine, alanine, alpha-amino isobutyric acid, and imino diacetic acid have been measured in D(sub)2O as a function of temperature, pH, and the presence of Murchison minerals. In addition to the amino and carboxylic hydrogens, only the alpha- hydrogens of glycine, alanine, and imino diacetic acid are found to exchange. Even for solutions maintained for weeks at temperatures as high as 120 degrees C, no exchange was observed with the hydrogens of the methyl groups of alanine or alpha-amino isobutyric acid. The rate of exchange for alpha-hydrogens of amino acids is first-order with respect to the amino acid concentration. Increasing the pH of the solution markedly increases the rate of exchange. For example, at 115 degrees C and pH 4.0, 7.0, and 10 the rates are 14, 30, and 125 yr^-1 respectively for glycine and 2.0, 3.5, and 14 yr^-1 respectively for alanine. In a pH-6.0 D(sub)2O solution of amino acids containing Murchison dust the rates are 135 yr^-1 for glycine and 32 yr^-1 for alanine, rates close to those for the pH 10 solution. Activation energies for exchange were obtained from Arrhenius plots constructed from measurements made between 70 degrees C and 155 degrees C in solutions containing Murchison dust. For both glycine and alanine the activation energy is -25 kcal/mole. Using this value, we have calculated the half-lives for complete exchange of the alpha-hydrogens of glycine and alanine for the temperature range thought to have existed on the parent body during aqueous alteration [5]. The half-lives at 0 degrees C and 20 degrees C are 7500 yr and 300 yr respectively for glycine and 55,000 yr and 2100 yr respectively for alanine. Murchison amino acid fraction IV [1] was known to contain impurities and hence the measured delta D value represents a lower limit for alpha-amino isobutyric acid. Assuming that all the deuterium recovered from fraction IV came from alpha-amino isobutryric acid, and that one atom of nitrogen is recovered for each molecule of alpha-amino isobutyric acid, a maximum delta D value of +2600 per mil can be calculated for this amino acid. This is comparable to delta D for the glycine-alanine fraction, which is mainly glycine [6]. In an aqueous environment glycine loses deuterium relatively rapidly while alpha-amino isobutyric acid does not undergo exchange. Hence the similarity in the delta D values of both fractions indicates that the period of aqueous alteration is less than the half-life for hydrogen-deuterium exchange of glycine. References: [1] Pizzarello S. et al. (1991) GCA, 55, 905-910. [2] Zinner E. (1988) In Meteorites and the Early Solar System (J. R. Kerridge and M. S. Matthews, eds.), 956-983, Univ. of Arizona. [3] Lerner N. R. et al. (1993) GCA, in press. [4] Bunch T. E. and Chang S. (1980) GCA, 44, 1543-1577. [5] Clayton R. N. and Mayeda T. K. (1984) EPSL, 67, 151-161. [6] Shock E. L. and Shulte M. D. (1990) GCA, 54, 3159-3173. Romanek C.* Gibson E. Socki R. Burkett P. J. Microscale Variations in the 13C Content of the Murchison Meteorite Heretofore unresolved micrometer-scale carbon isotopic zonation in the Murchison meteorite (CM3) is documented using a laser microprobe mass spectrometer. High-resolution isotopic gradients and heterogeneities between high- and low-temperature textural components help to constrain the processes that have shaped the physiochemical character of this carbonaceous chondrite. Previous bulk samples of Murchison yield an average delta ^13C value of - 5.7 +/- 4.3 per mil [1] while individual components such as micrometer-sized mineral separates (e.g., C(sub)graphite , C(sub)diamond, and SiC), acid- soluble extracts (e.g., CaCO3 and polar hydrocarbons), and insoluble residues (e.g., polycyclic aromatic hydrocarbons) are isotopically diverse (delta ^13C of -1000 to 29,000 per mil). While these studies shed light on the origin and occurrence of C-bearing phases, they fail to constrain intrinsic spatial isotopic heterogeneities. The power of the laser microprobe lies in the fact that in situ chemical and isotopic compositions are measured simultaneously for volatiles extracted from extremely small sample volumes (i.e., 0.025 mm^3 for 5 wt% C). Nd-YAG laser irradiation (1.06 micrometers) is directed onto texturally defined targets (>=50 micrometers wide) from which solid material is volatilized. Condensible gaseous phases are collected in a variable-temperature cold trap while the more volatile species (CH4 and CO) are quantified using an ion trap mass spectrometer. All gases are then converted to CO2 in a CuO furnace (containing Pt) held at 600 degrees C and analyzed for carbon and oxygen isotope ratios. The concentration and isotopic composition of condensed species are determined by stepped sublimation of unstable components and conversion to CO2. Preliminary isotopic analyses of the total volatile C content (i.e., bulk microanalysis) from distinct textural components at least 0.05 mm^3 in volume are described below. The most ^13C-depleted components within Murchison reside within the cores of chondrules and/or aggregates. Three typical cores were analyzed, with an average bulk composition of -21.0 +/- 0.5 per mil (n = 7). The bulk ^13C content of C-bearing phases increases monotonically outward in all directions within 100 to 200 micrometers of each core (i.e., within dust mantles) to a constant matrix value of -12.5 +/- 0.5 per mil (n = 40). The most isotopically enriched textural component found in Murchison is a regolith breccia clast without chondrules that has an average bulk delta ^13C value of -10 +/-0.5 per mil (n = 5). The clast was originally detectable only under cathodoluminescence, but with the aid of the laser microprobe it is now characterized by an unusually low volatile content and enriched ^13C composition. In general, the most isotopically enriched components also produce the lowest yield of gas (normalized to sampling volume). This trend of isotopic enrichment from chondrule to matrix has been documented previously for oxygen isotopes in carbonaceous chondrites [2]. Carbon isotopic gradients and heterogeneities within Murchison reflect fundamental changes in the chemical speciation and/or isotopic content of the main C-bearing components (i.e., acid-soluble and insoluble hydrocarbon fractions) within the meteorite. Perhaps core interiors and dust mantles are responding to environmental changes reflected in the speciation of C-bearing species distributed within the solar nebula or the parent body. Inverse correlations between hydrocarbon atomic mass number and ^13C abundance in the acid-soluble [3] and insoluble residues [4] of Murchison have been documented. Alternatively, micrometer-scale isotopic gradients may reflect fundamental changes in the isotopic composition of individual C-bearing species through time. Enrichments may represent kinetically controlled processes related to hydrocarbon formation. In contrast, assuming an equilibrium fractionation mechanism, isotopic enrichments may record a temperature-dependent component to hydrocarbon delta ^13C values. These opposing alternatives will be discussed in light of the isotopic composition of individual C-bearing components volatilized from tightly constrained sample volumes within Murchison. References: [1] Kerridge J. F. (1985) GCA, 49, 1707-1714. [2] Clayton R. N. and Mayeda T. K. (1984) EPSL, 67, 151-161. [3] Yuen G. et al. (1984) Nature, 307, 254. [4] Gilmour I. et al. (1991) Meteoritics, 26, 337-338. McSween H. Y. Jr.* Riciputi L. R. Johnson C. A. Prinz M. Trace-Element Concentrations in Calcite and Dolomite from Carbonaceous Chondrites Determined by Ion Microprobe, and Coexisting Fluid Compositions Carbonates in CI and CM chondrites precipitated from aqueous fluids, probably derived from melted ices originally accreted into the parent asteroids [1]. Previous attempts to constrain the proportions and properties (temperature, pH, Eh, and dissolved C) of these fluids have been based on oxygen isotope fractionations [2] and on computer simulations of the alteration process [3]. The concentrations of minor and trace elements in carbonates can also provide insights into the composition of the alteration fluids. However, the carbonate grains are usually very small, and it is difficult to obtain reliable results for element concentrations <500 ppm using the electron microprobe. Utilizing a Cameca 4f ion microprobe at ORNL, we have analyzed calcites and dolomites in four CM chondrites (Boriskino, Murchison, Nogoya, and ALH 83100) and dolomites in one CI chondrite (Orgueil) for Fe, Mg, Mn, Sr, Na, Ba, and B, using Ca concentrations determined by electron microprobe as a reference element. Electron microprobe analyses for Fe, Mg, and Mn [4] (when detectable) are in good agreement with our results. Mean element concentrations (ppm) for 25 calcites in CM chondrites are Fe 2272, Mg 876, Mn 446, Sr 104, Na 685, Ba 8, B 13. Corresponding data (ppm unless otherwise noted) for 5 dolomites in CM (and 10 dolomites in CI) chondrites are Fe 2.55 wt% (1.65%), Mg 10.02% (10.73%), Mn 1.98% (2.69%), Sr 135 (142), Na 410 (286), Ba 10 (7), B 10 (6). Calcites are highly variable in composition, even within the same CM chondrite, and there are no apparent interelement correlations. ALH 83100 is visibly brecciated, but the range of calcite compositions in this sample is no greater than in other meteorites, and calcites within one clast show appreciable compositional scatter. Dolomites in both CM and CI chondrites appear to be more restricted in composition compared to calcites, and CI dolomites are consistently less Fe- rich and usually more Mn-rich than those in CM chondrites. We have estimated element ratios (relative to Ca) for fluids in equilibrium with these carbonates using element/fluid partition coefficients for calcite [5] and dolomite [6]. Such Ds have large uncertainties due to possible influences of temperature, pH, and other factors, but some general properties of the fluids can be ascertained. For CM chondrites, dolomites indicate fluids having consistently higher Fe/Ca, Mg/Ca, Mn/Ca, and Sr/Ca ratios than do co- existing calcites. All these fluids have high Na/Ca, and appear similar in many respects to terrestrial brines. The nearly stoichiometric compositions of dolomites also suggest high salinity. Dolomites in CI chondrites suggest fluids indistinguishable from those in equilibrium with CM dolomites. Fluids in equilibrium with calcite in highly altered meteorites like Nogoya have higher Fe/Ca, Mg/Ca, and Mn/Ca than those calculated for less-altered meteorites like Murchison. Moreover, the fluids that produced calcite in CM chondrites were apparently distinct from those that produced dolomites in CM and CI chondrites. These observations reinforce previous conclusions [4,7,8] that fluid compositions in the parent asteroids were locally and temporally variable. The elevated trace-element contents of the carbonates and the brinelike compositions of the assocated fluids further indicate that dissolution of soluble components may have been more pervasive than suggested by the preservation of bulk compositions thought to be cosmic. References: [1] Grimm R. E. and McSween H. Y. (1989) Icarus, 82, 244-280. [2] Clayton R. N. and Mayeda T. K. (1984) EPSL, 67, 151-161. [3] Zolensky M. E. et al. (1989) Icarus, 78, 411-425. [4] Johnson C. A. and Prinz M. (1993) GCA, in press. [5] Veizer J. (1983) SEPM Short Course No. 10, 3-100. [6] Kretz R. (1982) GCA, 46, 1979-1981. [7] Richardson S. M. (1978) Meteoritics, 13, 141- 159. [8] Fredriksson K. and Kerridge J. F. (1988) Meteoritics, 23, 35-44. Nakamura T.* Tomeoka K. Sekine T. Takeda H. Shock Metamorphism of Carbonaceous Chondrites: Textural Diversity of Experimentally Shocked Allende in Various Conditions Introduction: It has been found that shock effects are recorded in many carbonaceous chondrites [e.g., 1,2], suggesting that impact events are a common process in the early evolution of planetesimals. In order to understand textural and mineralogical characteristics of carbonaceous chondrites affected by multiple impacts and also by impacts under high temperature, we carried out high-temperature, multiple shock experiments using the Allende CV3 chondrite. One of our goals is to reproduce the texture of the Leoville CV3 chondrite, which shows a strong preferred orientation of the flattened chondrules with high aspect ratios (average of 1.9) [3]. Leoville shows much evidence of deformation by moderate shock pressures [1]. However, it remains uncertain whether such moderate shock pressures can flatten round chondrules to such high aspect ratios. There was an unsuccessful attempt to reproduce the texture of Leoville by a single shock [4]. Results: Shocked once (21.7 GPa) at room temperature. Chondrules are flattened with an average axial ratio (long axis/short axis) of 1.4 and show a moderate preferred orientation nearly perpendicular to the compacting axis. Almost all olivine and low-Ca pyroxene in chondrules exhibit numerous fractures and show wavy extinction. The matrix is strongly compacted, but each olivine grain is distinguishable in SEM images. Metal and sulfide grains in the matrix are considerably deformed, but do not appear to have melted. Shocked twice (21.3 and 21.7 GPa) at room temperature. Chondrules are clearly more flattened with an average axial ratio of 1.9 and show a more pronounced preferred orientation. Olivine and low-Ca pyroxene in chondrules show extremely fine fractures and exhibit wavy extinction. The matrix is so strongly compacted that each olivine grain can hardly be distinguished in SEM images. Metal and sulfide grains in the matrix are strongly elongated along surfaces of chondrules. Shocked once (21.3 GPa) at high temperature (607 degrees C). Chondrules are flattened with an average axial ratio of 1.8 and show a strong preferred orientation. Olivine grains in the matrix look very smooth in SEM images. Many Si- and Ca-rich glassy grains ranging 10-50 micrometers in diameter occur in the matrix. Metal and sulfide are apparently melted to form networklike veins in the matrix. In places, the Fe-Ni-S melt is segregated in areas ~400 micrometers across. Conclusions: The doubly shocked Allende is most similar to Leoville in chondrule and matrix texture. The singly shocked Allende at room temperature and 607 degrees C also shows chondrule flattening and preferred orientation, but differs considerably in matrix texture from Leoville. This appears to indicate that the deformation texture of Leoville resulted from repeated impacts under low temperature. The doubly shocked experiment reveals that repeated impacts by relatively mild shock pressures apparently facilitate mechanical compaction of chondrite without intensive heat generation. The effects of shock at high temperature (607 degrees C) are quite different from those at room temperature. The flattening of chondrules and the compaction of matrix are clearly more pronounced at 607 degrees C than at room temperature. This can be explained by an increase of the plasticity of components constituting chondrules and matrix at the high temperature. Melting of metal, sulfides, and silicates are also significantly facilitated at the high temperature. References: [1] Nakamura T. et al. (1992) EPSL, 114, 159-170. [2] Scott E. R. D. et al. (1991) LPS XXII, 1205-1206. [3] Cain P. M. et al. (1986) EPSL, 77, 165-175. [4] King E. A. et al. (1978) Meteoritics, 13, 549. Calvin W. M.* King T. V. V. Reflectance Spectra of Fe-bearing Phyllosilicates: Applications to CM Chondrites The composition of the carbonaceous chondrites is dominated by a fine-grained opaque mineral mixture called matrix. In the lowest petrologic type C-chondrites significant alteration of matrix minerals has occured, resulting in compositions dominated by aqueous alteration products such as phyllosilicates, sulfates, oxides, hydroxides, and carbonates. The phyllosilicates top the list of abundant phases, and in the CM chondrites in particular, Fe-rich serpentines are the most important phases [e.g., 1]. King and Clark [2] have characterized the Mg-serpentines and chlorites, noting certain spectral similarities between chlorites and CI1 and CM2 chondrites. However, they found no exact spectral matches. We present here the results of an examination of the reflectance spectra of Fe-serpentines and two varieties of chamosite (chlorite group). We find these minerals can provide a reasonable spectral match to features seen in certain CM chondrites and by extension, the dark asteroids. We have measured the reflectance spectra of several different high-iron phyllosilicates. Samples were primarily obtained from the National Museum of Natural History (NMNH) with one extremely high iron chamosite provided by the University of Munster. Samples were hand picked and ground and measured in bidirectional reflectance from 0.3 to 25 micrometers. Samples were also characterized by x-ray. In the serpentine group we have measured samples of greenalite (Fe^2+,Fe^3+)(sub)2- 3(Si)2O(sub)5(OH)(sub)4, berthierine (Fe^2+,Fe^3+,Mg)(sub)2- 3(Si,Al)2O(sub)5(OH)4, and cronstedtite Fe^2+2Fe^3+(Si,Fe^3+)O5(OH)4. In the chlorite group we have measured two different samples of chamosite (Fe^2+,Mg,Fe^3+)(sub)5Al(Si(sub)3,Al)O(sub)10(OH,O)(sub)s. We have measured an additional Mg-serpentine amesite Mg(sub)2Al(Si,Al)O(sub)5(OH)(su)4, not examined by [2]. (Chemical formulas cited reflect the ideal given in [3].) For comparison to spectra of CM-type chondrites, samples of Murchison and Murray were available to us and these were also measured in reflectance from 0.3 to 25 micrometers. There are a number of spectral differences between the Fe- and Mg- serpentines, most notably that the Fe-bearing minerals lack the strong, narrow feature at 1.4 micrometers. They also lack the strong Mg-OH features between 2.2 and 2.4 micrometers. In addition several of the samples exhibit absorptions near 0.7 and 0.9 micrometers. The absence of the near-infrared features coupled with the presence of absorptions at the long end of the visible allows the Fe endmembers to provide a much better spectral match to near-infrared characterisitics of CM chondrites like Murchison and Nogoya. Additionally, the general slope characteristics below 0.58 micrometers in CM2 chondrites are also well matched by those observed in the Fe-serpentines, particularly the berthierine that we measured. Vilas and Gaffey [4] compared the absorptions near 0.7 and 0.9 micrometers in several CM chondrites with those observed in main- and outer-belt asteroids. They argued for a similar origin for the spectral features so Fe-phyllosilicates may contribute to the observed spectral characteristics of certain asteroids as well. In the infrared the spectra of CM chondrites Murray and Murchison are quite similar with broad absorptions at 3 micrometers, and from 8-12 micrometers, with a narrower feature centered on 6.2 micrometers. The Mg-serpentine, amesite, has abundant spectral features beyond 13 micrometers, which are not seen in the CM chondrite spectra. The Fe-serpentines have absorptions that can contribute to those seen in the CM chondrites, but lacks the large absorptions beyond 13 micrometers, again providing a better spectral match than the Mg-serpentines. In the future we hope to compare the spectra of these Fe- serpentines with a wider variety of CM chondrites. Additionally a theoretical modeling study is planned, which will attempt to match meteorite spectra using their mineralogy and grain size distribution as the initial input to the models. Acknowledgements: This work was begun while W. M. Calvin was a Humboldt Research Fellow at the Inst. fur Planetologie at the University of Muenster. References: [1] Zolensky M. and McSween H. Y. (1988) Meteorites and the Early Solar System (Kerridge and Matthews, eds.), 114- 143. [2] King T. V. V. and Clark R. N. (1989) JGR, 94, 13997- 14008. [3] Fleischer M. and Mandarino J. A. (1991) Glossary of Mineral Species. [4] Vilas F. and Gaffey M. J. (1989) Science, 246, 790-792. Browning L. B.* McSween H. Y. Jr. Zolensky M. An Alteration Scale for CM Chondrites and Implications for Planetary Noble Gas Abundances Three progressive alteration parameters have been identified from the mineralogical and textural analyses of 7 CM chondritic falls. These indices predict the following order of progressive alteration [3]: Murchison (MC)10^4 yr to equilibrate a pyroxene 1 mm in radius), suggesting that fractional partial melting may have been common. Thus, calculations and experiments that assume equilibrium partial melting may be only approximations to the conditions that existed in asteroids. The rapid migration time also suggests, along with the lack of strong density traps, that melts will not readily form intrusions; thus, it may be difficult to explain the formation of cumulates such as diogenites by conventional intrusive processes. Perhaps they formed in a magma ocean. References: [1] Taylor G. J. et al. (1993) Meteoritics, 28, 34-52. [2] Scott E. R. D. et al. (1993) GRL, 20, 415-418. [3] Wilson L. and Keil K. (1991) EPSL, 104, 505-512. [4] Keil K. and Wilson L. (1993) EPSL, in press. [5] Muenow D. W. et al. (1992) GCA, 56, 4267-4280. [6] McKenzie D. (1984) J. Petrol., 25, 713-765. [7] Snow D. T. (1969) Water Resour. Res., 5, 1273-1289. Monday, July 19, 1993 Chondrules 2:00 p.m. Cascade Ballroom Chair(s): T. Bunch H. Nagahara Huang S.* Benoit P. H. Sears D. W. G. Metal and Sulfide in Semarkona Chondrules and Rims: Evidence for Reduction, Evaporation, and Recondensation During Chondrule Formation The fact that many chondrules in UOCs contain metal associated with sulfide has been attributed to either low temperature of formation (<680 K) and lack of subsequent heating sufficient to cause evaporation [1] or metamorphism after chondrule formation [2]. We have examined the metal and sulfide in group A and B chondrule interiors and rims in the most primitive ordinary chondrite, Semarkona, in order to further explore these options. Most group A1 chondrules contain abundant metal(3-4 wt%), which is mainly as rounded grains of kamacite (< 1 micrometer~60 micrometers) usually situated in the mesostasis near chondrule edge. For the 37 group A1 chondrules investigated, only five contain sulfide and in only one was it abundant. Usually the sulfides were found associated with metal near the chondrule surfaces, and in a few cases, the metal grains were enclosed in sulfides. The "dusty metal" [3] is common in group A1 chondrules, but is not found in group B chondrules, and the host olivine is often embayed by metal-free pyroxene, which has lower Fe/Fe+Mg ratio than the coexisting olivine. In contrast, metal in group B1 chondrules is much less abundant (generally less than 1 wt%) and occurs as both kamacite and taenite. It is often associated with sulfide, with the sulfide being more abundant than metal. Metal in group A1 chondrules is generally poorer in Ni than the metal in group B1 chondrules (Fig. 1). A similar observation was made for type IA and II chondrules [4,5], which are subsets of group A and B respectively. Additionally, metal in chondrules with Fe-poor olivine contains lower abundance of Ni and Co than metal in chondrules with Fe-rich olivine (Fig. 1) [6]. Group A1 chondrules are more frequently rimmed than group B1 chondrules (~70% by number, compared with ~30% ) and seem to have higher ratios of rim thickness to chondrule diameter (Fig. 2). Most group A1 chondrule rims contain ultra-fine-grained metal- and sulfide-rich materials, which are not observed in chondrites of higher petrographic grades. In contrast, group B1 chondrule rims, when present, contain fine-grained matrix-like materials with dispersed or massive sulfide and metal, which, in contrast to the ultra-fine sulfides/metal-rich rims in group A chondrules, are also observed in higher petrographic types [7]. These results can best be explained by reduction of ferrous olivine and loss of FeS by evaporation during group A1 chondrule formation with the recondensation of FeS and/or reactions between recondensed metal and H2S in the nebular gas at lower temperatures. Thermoluminescence, cathodoluminescence, and compositional zoning in several Semarkona group A1 chondrules has also been interpreted in terms of recondensation of major volatile elements like Na and Mn [8,9]. References: [1] Grossman J. N and Wasson J. T. (1983) In Chondrules and their Origins (E. K. King, ed.), 88-121. [2] Wood J. A. (1993) personal communication; see also Grossman J. N. (1988) In Meteorites and the Early Solar System (J. F. Kerridge and M. S. Matthews, eds.), 680-696. [3] Rambaldi E. R. and Wasson J. T. (1982) GCA, 46, 929-939. [4] Jones R. H. and Scott E. R. D. (1989) LPS XIX, 523-536. [5] Jones R. H. (1990) GCA, 54, 1785- 1802. [6] Snellenburg J. (1978) Ph.D. Thesis, State University of New York at Stony Brook. [7] Allen J. S. et al. (1980) GCA, 44, 1161-1176. [8] DeHart J. M. (1989) Ph.D. Thesis, University of Arkansas. [9] Matsunami S. et al (1992) GCA (in press). Fig. 1, which appears here in the hard copy, shows chondrule melt compositions (data from [4,6,8]). Fig. 2, which appears here in the hard copy, shows rim thickness against chondrule diameter with regression lines. Bunch T.* Paque J. M. Reynolds R. Podolak M. Prialnik D. Meteorite Ablation Rinds as Analogs for the Origin of Rims on Chondrules Conventional wisdom holds that UOC chondrule rims were formed in the nebula by dust accretion. Following the accretion stage, some investigators suggest that these porous rims were subjected to thermal alteration that ranged from sintering to melting [e. g., 1-3]. To understand the evolutionary history of chondrules we need to ask: (1) What nebular mechanism(s) concentrated the dust for rapid accretion? (Addressed in a companion paper at this meeting [4]). (2) What thermal event(s) welded or melted the dust? (3) Is this dust solely responsible for the rim composition, or are some rims composed, in part, of the parent chondrule? Production and/or modification of rims during atmospheric entry onto a parent body is a scenario that is testable by examination of ablation rinds produced on meteorites during entry into Earth's atmosphere. Comparison of ablation rind features with opaque rims on UOC chondrules will indicate whether this is a viable method for the production of chondrule rims. Terrestrial ablation rinds on UOCs and carbonaceous chondrites have been examined both texturally and chemically. Ablation rinds have these distinct characteristics: (1) The bulk composition of the rind is a reflection of the bulk chemistry of the host object, including Na, K, and P, but with the exception of much lower S. (2) Boundaries between unmelted bulk meteorite and rind silicates are physically sharp over distances of microns, similar to boundaries between rims and their chondrules. However, compositional transition zones extend inward from the boundaries for 10s of microns. (3) Melted meteorite matrix in the rind is compositionally similar to unmelted matrix and is texturally and chemically similar to rims. (4) Mineral texture and chemistry at chondrule/rim and meteorite/rind interfaces indicate significant thermal processing has occurred. For example, sulfides show high concentrations of included, more refractory phases at the melt interface with a corresponding loss of S. Overall, the comparison of ablation rinds with rims strongly suggests that opaque rims formed by melting of dusty accretion mantles. This melting event may have continued into the outer margins of host chondrules, or may be restricted to the accreted dust. SEM examination of the boundaries between chondrules and rims indicate that both cases probably occur. The major and minor element composition of opaque rims is similar to "accretionary" rims on objects in CM meteorites [5]. We suggest that both types of rims were formed from the same basic anhydrous dust, although CM rims acquired more O^16-bearing component than UOCs. From here, their evolutionary paths diverged: Opaque rims were thermally processed and CM rims were aqueously altered. Calculations of rim melting due to entry into a transient atmosphere of low scale height [6] indicate that encounter velocities in the range 2-4 km/sec are sufficient to melt the outer parts of chondrules. If the thermal conductivity of porous accretionary rims is as low as that of powdered chondrite [7], gas dynamic deceleration can produce totally or partially melted rims on chondrules without melting the chondrule itself. References: [1] Rubin A. and Wasson J. (1987) GCA, 51, 1923-1937. [2] Podolak et al. (1990) Icarus, 84, 254-260. [3] Bunch T. et al (1991) Meteoritics, 26, 326. [4] Cuzzi J. and Dobrovolskis A. (1993) this meeting. [5] Metzler et al. (1992) GCA, 56, 2873- 2898. [6] Podolak et al. (1993) Icarus, in press. [7] Wechsler A. E. and Glaser P. E. (1965) Icarus, 4, 335. Wasson J. T.* Krot A. N. Rubin A. E. Sibling and Independent Compound Chondrules SIBLING AND INDEPENDENT COMPOUND CHONDRULES. J. T. Wasson, A. N. Krot, and A. E. Rubin, Institute of Geophysics and Planetary Physics, University of California, Los Angeles CA 90024-1567, USA. We studied compound chondrules in 79 cm2 of ordinary chondrite (OC) thin sections. Compound chondrules consist of a primary that solidified first and one or more secondaries attached to the primary. Sibling compound chondrules have very similar textures and compositions; most, perhaps all, seem to consist of chondrules melted in the same heating event. About 1.4% of all chondrules are the primaries of sibling compound chondrules. A smaller fraction, 1.0%, of all chondrules are the primaries of independent chondrules, the members of which were melted in separate heating events. Independent chondrules show appreciable differences in texture and/or composition. We propose that sibling chondrules originated when numerous chondrules were created from one large, more-or-less homogeneous, precursor assemblage that was flash-melted to produce a large set (perhaps 100-1000) of chondrules; some of these collided while molten, probably within several centimeters of the production site. We envision that small radial velocities were imparted to the members of the set, with small differences in velocity causing collisions among those few in intersecting trajectories. If all chondrules were produced this way, the collision efficiency was 1.4%; if only 10% were produced in this fashion, the efficiency rises to 14%. The original Gooding-Keil model of independent compound chondrule formation calls for random collisions to occur while the secondaries were molten. This appears improbable because the mean period between collisions in the dusty midplane of the nebula is estimated to be hours (or days), orders of magnitude longer than the period during which chondrules could have retained low viscosities following a flash-heating event in a cool (<700 K) nebula. We suggest that most independent compound chondrules formed by the mechanism that accounts for chondrules with relict grains and for chondrules with coarse- grained rims: the primary chondrule was embedded in a porous dust assemblage at the time of the second heating event; it experienced minimal melting because melting efficiency increases with increasing surface/volume ratio. There is a minor tendency for the FeO/(FeO+MgO) ratio in independent secondaries to be higher than in primaries, as expected if this ratio increased with time in the nebular dust. However, Monte Carlo calculations confirm that the compositions of independent secondaries are not randomly distributed, but related to those of primaries. Some exchange probably occurred during the fusion of the two chondrules, but this mechanism seems unable to account for the general similarity of independent primary/secondary compositions. This suggests that, in the environment where, at any one time, chondrules were forming (perhaps the interface between the gaseous nebula and the dusty midplane), the dust composition was more uniform than it was in the central midplane at a later time when agglomeration occurred. Krot A. N.* Wasson J. T. Silica-Fayalite-bearing Chondrules in Ordinary Chondrites: Evidence of Oxidation in the Solar Nebula Most ordinary chondrite (OC) chondrules have compositions similar to those of bulk OC in terms of lithophile-element abundances. There are only a few rare chondrule classes that deviate significantly from OC-like compositions; these include Al-rich chondrules, chromitic and chromite-bearing silicate chondrules, and silica-rich chondrules. We studied 41 thin sections of unequilibrated OC and found 82 silica-bearing chondrules that can be divided into two major categories: silica-pyroxene chondrules and silica-fayalite- pyroxene chondrules. These chondrules are more common in H (>3/cm^2) than in L and LL chondrites (<1/cm^2). Silica-pyroxene chondrules consist mainly of low-Ca pyroxene and silica and have radial and porphyritic textures. Silica-bearing radial pyroxene (RP) chondrules contain 5-10 vol% silica grains; the low-Ca pyroxene is uniform in individual chondrules but varies from one chondrule to another (Fs(sub)10.2- Fs(sub)31.5). Silica-bearing porphyritic pyroxene (PP) chondrules contain 15- 40 vol% silica; the low-Ca pyroxene varies in composition within individual PP chondrules and tends to be more magnesian than in the silica-bearing RP chondrules (Fs(sub)5.0-Fs(sub)21.1). Petrographic observations suggest that some PP chondrules were not completely molten; they appear to have cooled more slowly than the silica-bearing RP chondrules. Silica-fayalite-pyroxene chondrules consist of silica, fayalite, and low-Ca pyroxene; accessory high-Ca pyroxene, plagioclase mesostasis, troilite, and metallic Fe-Ni are also present. Based on texture and the modal abundances of pyroxene and silica these chondrules can be divided into two types: (1) radial or porphyritic silica-fayalite-pyroxene chondrules containing 5-40 vol% silica and (2) granular silica-fayalite-pyroxene chondrules consisting almost entirely (90-95 vol%) of silica. Silica-fayalite-bearing pyroxene chondrules are texturally and compositionally similar to the silica-bearing pyroxene chondrules described above; the principal difference between them is the presence of fayalite-forming veins within or rims around the silica grains. The continuum between these chondrule categories implies that they are genetically related: We infer that the fayalite veins and rims formed by nebular alteration of the silica grains. Fayalite forms veins along the silica grain boundaries in granular silica-fayalite-bearing chondrules. Fragments of granular silica chondrules occur as relict clasts within two pyroxene chondrules in Sharps. These fragments were altered after chondrule solidification. Conclusions: (1) Silica-bearing chondrules have similar textures to common mafic silicate chondrules and were formed by melting silica-rich precursor material that possibly formed by nonequilibrium condensation. (2) The higher abundance of silica-bearing chondrules in H than in L and LL chondrites may indicate a greater degree of silica condensation in the H-formation region. (3) Silica-fayalite-bearing chondrules formed by alteration of silica-bearing chondrules. The common occurrence of both categories within the same chondrite suggests that oxidation and fayalite formation by nebular gas was an inefficient process. Jones R. H.* Layne G. D. Partitioning of Trace Elements Between Pyroxene and Liquid in a Porphyritic Pyroxene Chondrule in Semarkona The unequilibrated chondrite Semarkona (LL3.0) enables us to investigate primary properties of chondrules that have not been overprinted by secondary processes. Electron microprobe studies of the compositions and zoning properties of silicate phases in these chondrules have helped to interpret crystallization behavior and, hence, offer important insights into formation conditions [e.g., 1,2]. However, the behavior of trace elements in these systems has not been investigated, largely because of the difficulties encountered in analyzing such elements in chondrule silicates. Here we report preliminary ion microprobe data obtained on coexisting pyroxene and glass phases from a pyroxene-rich chondrule in Semarkona. Trace elements analyzed are REE (La, Ce, Nd, Sm, Eu, Dy, Er, Yb), Sr, Y, and Zr. The chondrule studied is a typical example of textural type IAB [2]. It contains phenocrysts of olivine (Fa(sub)3) and clinoenstatite and a glassy mesostasis occupying approximately 15 vol% of the chondrule. Augite (Fs(sub)3, Wo(sub)44) occurs as narrow (10-micrometer) rims on clinoenstatite phenocrysts. Clinoenstatite is FeO-poor (Fs(sub)3, Wo(sub)0.4) and shows little zoning in major and minor elements. Trace-element analyses have been carried out on clinoenstatite, augite, and glass in this chondrule. REE contents in clinoenstatite are extremely low, lying in the range 0.01-0.1 x CI, and show a smooth increase in abundance from La to Yb. REE abundances are enriched in both augite and glass at levels approximately 4-10 x CI, with a small negative Eu anomaly in augite and a small positive Eu anomaly in glass. Olivine is likely to contain REE abundances similar to low-Ca pyroxene [3]. These relative abundances are consistent with closed-system crystallization of the chondrule, assuming that its bulk composition has chondritic abundances of REE [4]. Trace-element partition coefficients (Ds) for the two pyroxene phases are shown in Fig. 1. Clinoenstatite Ds vary smoothly, increasing from 0.0006 (La) to 0.02 (Yb). These data are broadly consistent with equilibrium D values obtained experimentally [3]. However, chondrule Ds for the LREE and Sr are consistently higher than equilibrium experimental values. This could be attributed to the effect of rapid cooling in chondrules [3]. Values for D(sub)Y and D(sub)Zr are also consistent with the experimental data. For augite, the Ds we determined are approximately flat, at values around 1, with a decrease in the LREE and a negative Eu anomaly. The chondrule data are consistently higher than equilibrium experimental data for pyroxenes of composition Wo(sub)40 [5]. This may also be attributable to the effect of rapid cooling rate. However, the partitioning behavior of REE in Ca-rich pyroxene as a function of melt composition is not fully understood. D(sub)Sr, D(sub)Y, and D(sub)Zr are comparable with REE values, consistent with the data of [6]. In summary, trace-element partitioning among chondrule silicate phases appears to be entirely consistent with closed-system crystallization of the chondrule. Data such as these will be valuable in assessing the origins of, and relationships between, various chondritic components. They may also provide a valuable tool for studying metamorphism in ordinary chondrites. This work is supported by NASA grant NAGW-3347 (J. J. Papike). SIMS analyses were performed at the UNM/SNL Ion Microprobe Facility, a joint operation of the Institute of Meteoritics, UNM, and Sandia National Laboratories. References: [1] Jones R. H. (1990) GCA, 54, 1785-1802. [2] Jones R. H. (1992) LPS XXIII, 631-632. [3] Kennedy A. K. et al. (1993) EPSL, 115, 177-195. [4] Grossman J. N. et al. (1988) In Meteorites and the Early Solar System, 619- 659. [5] McKay G. et al. (1986) GCA, 50, 927-937. [6] Hart S. R. and Dunn T. (1993) CMP, 113, 1-8. Fig. 1 appears here in the hard copy. Lofgren G. E.* Jurewicz A. J. G. DeHart J. M. CaO Partitioning Between Olivine and Melt in Highly Magnesian Systems Minor-element contents of olivine are used to make inferences about petrogenetic history. For example, anomalously high CaO contents (>0.5 wt%) are attributed to rapid crystal growth, while low CaO contents (<0.1 wt%) are attributed to the effect of metamorphism or high pressure. In complicated chondrule systems, we often do not consider that the CaO contents of olivine are a function of the partition coefficient, ^s/l^D(sub)CaO, between olivine and liquid. Accordingly, before being considered anomalous, ideal CaO contents in olivine should be calculated from a ^s/l^D(sub)CaO taken from a model. Unfortunately, models for CaO partitioning currently in the literature are largely empirical and derived primarily using terrestrial basalts [1-3]. Our results show that these models vary in how well they can be extended to the more magnesian systems encountered in chondrules. The model of [3] appears to describe the data best, while those of [1,2] have considerably more scatter. We present a new model easier to use than that of [3]; this model is perhaps not quite as accurate, but it underscores that ^s/l^D(sub)CaO is a function of melt structure as well as liquid composition. The dataset we used to derive our model and to compare with [1-3] was compiled from both equilibrium experiments and dynamic crystallization experiments for which it was demonstrated that there was no effect on CaO partitioning with cooling rate [2, 4-9]. Our new model uses the "corrected" calcium partition coefficient of Jurewicz and Watson [2], Kd(sub)90, and the "normative" plagioclase and olivine contents of the coexisting melts, as calculated for an Ol-Si-Pl projection from Di [10]. The empirical parameterization, Kd(sub)90/(PL/OL), was chosen because (1) the Kd(sub)90 is an empirical representation of the effect of relative forsterite/fayalite activity on calcium partitioning and (2) the normative components of the melt generally reflect the melt structure, with the Pl component especially relevant to the complexation of Ca with Al in the melt. When plotted against 1/T, this new parameter describes a variety of iron-free, chondritic, and basaltic compositions. A similar plot of -ln(calculated ^s/l^D(sub)CaO) vs. 1/T using our data in the model of [3] also describes the data well. We note that the arguments of [3] are thermodynamic, whereas our approach concentrates on melt structure, especially the complexation of calcium with plagioclase structures in the melt. Since we can now reasonably estimate CaO partitioning in magnesian systems at least two ways, the results of this study constrain the interpretation of the CaO contents of olivine in chondrules. Both models and experiments indicate that the ^s/l^D(sub)CaO is lower in the high-temperature, high-magnesian, iron-containing melts than in typical basaltic melts. Thus, low CaO contents in olivine may be equilibrium values and should not automatically be interpreted as a metamorphic or condensation effect. Moreover, for bulk melts with less than 8 wt% CaO, rapid growth (cooling >100 degrees C/hr) does not result in higher CaO contents [6,7]. Therefore, high CaO content is not necessarily due to rapid growth, but more likely controlled by bulk composition (greater than 8 wt% CaO in the melt) combined with rapid growth. It is important for interpretation of olivine composition to determine the composition of the liquid from which the olivine crystallized. References: [1] Ford et al. (1983) J. Petrol., 24, 256-265. [2] Jurewicz and Watson (1988) Contrib. Mineral. Petrol., 99, 176-185. [3] Snyder and Carmichael (1992) GCA, 56, 303-318. [4] Watson (1979) Am. Mineral., 64, 824- 829. [5] Lofgren and Lanier (1990) GCA, 54, 3537-3551. [6] Lofgren and DeHart (1991) LPS. XXII, 823-824. [7] Lofgren et al. (1991) Geol. Soc. Am. Abstr. with Progr., 23, A271. [8] Radomsky and Hewins (1990) GCA, 54, 3475-3490. [9] Connolly and Hewins (1990) LPS XXI, 222-223. [10] Walker et. al. (1979) Contrib. Mineral. Petrol., 70, 111-125. Yu Y.* Hewins R. H. Clayton R. N. Mayeda T. K. High-Temperature Oxygen Isotope Exchange Between Meteorite Sample and Water Vapor: Preliminary Experimental Results Chondrules in carbonaceous and ordinary chondrites show slope-1 mixing lines on the oxygen three-isotope diagram, suggestive of a gas-melt exchange process during chondrule formation. In order to test this conjecture and to extend our existing knowledge of chondrule thermal history and the kinetics of reaction of interstellar dust with solar nebula gas, an experiment involving high- temperature oxygen isotope exchange between a 16O-rich sample (meteorite) and water vapor (terrestrial) has been designed. The experiment was conducted with a DELTECH vertical tube furnace with ceramic parts shielded with metal foil. The starting meteorite powder (one of two C3 carbonaceous chondrites--bulk Allende and Ornans) was pressed into a pellet and suspended at the hot spot inside the furnace. The furnace gas was a mixture of H2O vapor and H2 (1 atm total pressure, fO2 = IW-0.5) [1]. The preliminary experiments were performed at 1400 degrees C for durations from 5 minutes to 36 hours, and were terminated by quenching the samples into liquid nitrogen. The meteorite charges and the water samples collected were later analyzed for their oxygen isotope compositions. The experimental results (Fig.1) show that the exchange process has greatly modified delta-18O and delta-17O for both meteorites, which move towards the projected equilibrium point as the heating time increases. For Allende samples, the exchange proceeds quickly in the first 5 minutes, which accounts for most of the isotope exchange (~84% of total change in delta-18O(sub)A-W, and ~57% of total change in delta-17O). Then the exchange is dramatically slowed down, and takes at least 12 hours to finally reach equilibrium with the ambient water vapor. The approach to equilibrium is not a straight line on the three-isotope graph, possibly due to the presence of residual 16O-rich solids in the molten sample. A similar exchange profile is observed for Ornans samples. However, it takes longer for the Ornans sample to reach equilibrium after the initial fast exchange. The 15-hour run for Ornans is still away from the TF line, and it is moving toward the TF line about 1 permil lighter than the expected equilibrium value. Microscopic and electron microprobe studies on the heated Allende and Ornans samples (parallel runs) show that the quenched charges are composed of olivine relics and glass. The initial fast exchange observed is probably due to the rapid exchange between the ambient gas and the molten part of the meteorite sample, and the existence of olivine crystals eventually slows down the exchange process because of its much lower rate of oxygen diffusion [2]. The concentration of exchangeable gas molecules in our experiments is much greater than that in the solar nebula. The next step of this study will be experiments at higher temperatures, under conditions more similar to the chondrule-forming environment, such as flash heating and with gas diluted with Ar to obtain fewer oxygen molecules. Acknowledgments: We thank T. Grove for technical guidance, E. Jarosewich (NMNH, Washington) and B. Zanda (MNHN, Paris) for meteorite samples, and NASA (OSS) and NSF for financial support. References: [1] Baker M. B. and Grove T. L. (1985) Am. Mineral., 70, 279-287. [2] Jaoul O. et al. (1983) JGR, 88, 613-624. Nagahara H.* Young E. D. Hoering T. C. Mysen B. O. Oxygen Isotopic Fractionation During Evaporation of SiO2 in Vacuum and in H Gas Chondritic components, chondrules, CAIs, and some parts of the matrix are believed to have formed and/or thermally processed in the solar nebula. If this scenario is the case, they should be fractionated for major and minor elements and isotopes according to the formation temperature. This is true for major and trace elements, but is not the case for isotopes. Differences in oxygen isotopic composition among meteorite groups are interpreted to be the results of mixing of gas and dust from different oxygen reservoirs, and the effect of isotopic fractionation is negligible for most meteorites except for rare CAIs. Davis et al. [1] studied the isotopic fractionation of SiO2, MgO, and forsterite and showed that oxygen isotopic fractionation from solid materials is very small, but that from liquid is significant. Evaporation in the solar nebula should, however, be in hydrogen gas, which is reactive with silicates. Therefore, the effect of hydrogen gas on the evaporation behaviors of silicates, including mode of evaporation, evaporation rate, and compositional and isotopic fractionation, should be studied. Nagahara [2] studied the evaporation rate of SiO2 in equilibrium, in constant evacuation (free evaporation), and in hydrogen, and showed that the rate in hydrogen gas is orders of magnitude larger than that in vacuum; the mode of evaporation also differs from that in vacuum. Oxygen isotopic fractionation during evaporation of SiO2 in constant evacuation and in hydrogen gas at two different total pressures are studied in the present study. The starting material is a single crystal of natural quartz, which should transform into high cristobalite at experimental conditions. The powdered starting material was kept in a graphite capsule without a cap and set in a vacuum chamber with and without hydrogen gas flow. Experimental temperature was 1600 degrees C. Oxygen isotopic compositions (^18O/^16O) were measured with the CO2laser heating fluorination technique. Oxygen isotope measurements, including ^17O and silicon isotope measurements, are now in progress, and some of the results are shown in this paper. Oxygen isotopic compositions of residues in vacuum and in hydrogen gas of total pressure of 2.6 x 10^-5 bar, which approximates the pressure of the solar nebula at the midplane at 2-3 AU, are shown in comparison with evaporation rate (Figs. 1 and 2). Oxygen isotopic fractionation is remarkable in a constant evacuation, but is negligible in hydrogen gas of 2.6 x 10^-5 bar total pressure. In vacuum, delta ^18O of solid residue increases with increasing degree of evaporation. The curve is best fit to delta ^18O = 0.00094x^2 + 0.00173x + 19.606 (r = 0.997), where x is the degree of evaporation in weight percent. The curve is fit to the Rayleigh fractionation curve with a constant fractionation factor (alpha(sub)vap-sol) of 0.9970. Figures 1 and 2 show that evaporation is significant but oxygen isotopic fractionation is insignificant in hydrogen gas in the approximate solar nebular condition. The high evaporation rate in hydrogen gas is due to the fact that evaporation is a decomposition reaction of an oxide, which should be accelerated in reducing condition. The rate, however, can be explained by an unknown diffusion process that is possible when hydrogen is reactive with silica [2]. In a fairly high hydrogen pressure, isotopic fractionation is suppressed. On the other hand, in vacuum, the evaporation rate is small but the degree of isotopic fractionation is significant. The results suggest that chondrules and CAIs without isotopic mass fractionation could have been formed in the solar nebula, but that mass loss during heating should have been significant. The CAIs with significant mass fractionation such as HAL could have been formed in vacuum. References: [1] Davis A. et al. (1990) Nature, 347, 655-658. [2] Nagahara H. (1993) LPS XXIV, 1045-1046. Fig. 1, which appears here in the hard copy, shows the evaporation rate of SiO2 heated at 1600 degrees C in vacuum and in hydrogen gas of 2.6 x 10^-5 bar as a function of time. Fig. 2, which appears here in the hard copy, shows oxygen isotopic composition (delta ^18O) of evaporation residue of SiO2. Maharaj S. V.* Hewins R. H. Vesicles in Experimental Chondrules as Clues to Chondrule Precursors The processing of chondrule precursors during melting is so extensive that there are few unambiguous indicators of their mineralogical composition. The specific combination of peak temperature and heating time, i.e., the heating mechanism, is also unknown. The general absence of vesicles in chondrules is a potential constraint on both questions. Meteor ablation spherules, whose origins are well understood, differ from chondrules in having abundant vesicles [1]. Chondrules simulated experimentally in a variety of ways have vesicles in many cases, but it has been suggested that the presence of vesicles rules out flash heating [2]. We therefore examine in detail the formation of vesicles in synthetic chondrules. Vesicles have been produced in experiments with long heating times [3] as well as short [2]. They are most prominent in charges that experienced low degrees of melting, probably because of surface tension effects that trap bubbles between relict grains, aided by high melt viscosity. The gas could be derived from air trapped when the powdered sample is prepared, binding agents (acetone, water), or volatiles in the starting minerals (Na, H2O). We have conducted experiments to determine the source of vesicles in synthetic chondrules initially heated slightly below the liquidus and cooled at 500 degrees C/hr. Runs made in pairs included charges with and without acetone binder and charges baked out at 200 degrees C for different lengths of time. Charges with acetone produced more vesicles, which could be avoided to some extent by preliminary baking. Charges with no binder had very few vesicles if baked for 1/2 hour. Vesicles are more prominent when using a well-sorted fine-grained powder than with an unsorted more uniform size distribution. Pulling a vacuum on pellets had no effect on subsequent vesicle development. Vesicles are unlikely to be due to loss of Na from the charge, because vesicles are equally prevalent in flash-heated charges, which retain most of their Na, and earlier experiments that spent longer times at temperature. Experiments with serpentine in the starting materials resulted in a popcorn vesicle texture with voids as large as 3 mm, like some ablation spherules [1]. Trapped air and binding agents cause most vesicles in experimental charges. Chondrule precursors must have consisted of olivine, etc., with no hydrous minerals, assembled at low pressure, or they would have generated vesicles. The absence of vesicles in chondrules does not rule out flash heating mechanisms. References: [1] Brownlee D. E. et al. (1983) In Chondrules and Their Origin (E. A. King, ed.), 10-25, LPI, Houston. [2] Wdowiak T. J. (1983) In Chondrules and Their Origin (E. A. King, ed.), 279-283, LPI, Houston. [3] Radomsky P. M. and Hewins R. H. (1990) GCA, 54, 3475-3490. Connolly Jr. H. C.* Hewins R. H. Lofgren G. E. Possible Clues to the Physical Nature of Chondrule Precursors: An Experimental Study Using Flash Melting Conditions In a continuing series of experiments to assess how flash melting conditions can reproduce chondrule textures we have investigated how the initial grain size of the starting material can affect chondrule textures. A previous series of experiments [1] has been carried out using an Fa analog composition (T(sub)L 1211 degrees C) with initial grain size ranges from 23-45 micrometers, 45-63 micrometers, 63-125 micrometers, and 125-250 micrometers. New experiments ([2], this research) use an average Type IIA chondrule composition (TL 1550 degrees C) with initial grain size ranges of 23-45 micrometers and 125-250 micrometers. All experiments utilized the flash melting techniques of [1,2]. Experiments performed with the finer grain fractions (<125 micrometers) of these composition for flash melting conditions less than 50 degrees C above their respective TL yield microporphyritic textures typical of Type IA chondrules [3]. Higher initial flash melting temperatures generate textures typical of Type IIA chondrules (i.e., PO,BO). However, the same experimental conditions that produced Type IA textures produced porphyritic textures typical of Type IIA chondrules [4] with starting grain size of 125-250 micrometers. Experiments with 125- 250-micrometer grain size produced only the typical Type IIA chondrule textures from initial flash melting temperatures ranging to 125 degrees C above TL. These charges also had a high number of large (100-200 micrometers) relict grains indicating that melting was not very extensive. Fewer nucleation sites survived the melting process allowing larger crystals to grow in the Type IIA textured charges than charges that produced Type LA textures. From our experiments it is clear that for the flash melting conditions used we cannot produce the microporphyritic textures typical of Type IA chondrules from precursors with an initial grain size of 125-250 micrometers with our compositions. If chondrules were produced by flash melting conditions similar to our experimental conditions it is clear that Type IA chondrules had precursors with a relatively homogeneous grain size that is less than 125 micrometers. Based on our experiments, in order to obtain the high nucleation density needed after melting to produce textures truly analogous to Type IA chondrules a grain size of 45 micrometers or less is favored. Therefore, Type IA chondrules could not have experienced high degrees of melting, either due to a relatively short, high-temperature melting event (i.e. flash melting) or a longer, lower-temperature (subTL) melting event(s). The presence of BO textures and large phenocrysts within PO textures of Type IIA group indicates that melting was more extensive, thus fewer nucleation sites survived, for Type IIA than for Type IA chondrules. Relict grains produced from flash melting with 125250-micrometer precursor material in the lab may be analogous to relict grains found in Type IIA chondrules. However, these experiments cannot rule out the possibility that Type IIA chondrules were totally melted and their texture is a function of the number of nucleation sites caused by collisions with dust and the temperatures at which those collisions occur. Type IA chondrules were made by partial melting of fine grained precursor dust. Type IIA chondrules could either have been made by partial melting of coarse grained precursor dust or they could have been totally melted precursor dust (with any size characteristics) provided dust of any size range collided with molten chondrules during their formation to act as nucleation sites. References: [1] Connolly H. C. Jr. et al. (1991) Meteoritics, 26, 329. [2] Connolly H. C. Jr. et al. (1993) LPSC XXIV, 329-330. [3] Jones R. H. and Scott E. R. D. (1989) Proc. LPSC 14th, 559-566. [4] Jones R. H. (1990) GCA, 54, 1784-1802. Eisenhour D. D.* Buseck P. R. Constraints on Nubular Electromagnetic Pulses Chondritic meteorites contain an abundance of silicate minerals with opaque inclusions of oxides, sulfides, and metals. These host silicates interact differently from their enclosed opaques to electromagnetic (EM) radiation; specifically, silicates are inefficient at absorbing EM energy in the visible and near infrared while metals, sulfides, and Fe oxides absorb strongly in this frequency range. In the presence of a strong electromagnetic pulse (EMP), this preferential absorption leads to the selective heating of the opaque inclusions and can produce unique textures ("dirty snowballs": intimate, ~spherical intergrowths of silicate and opaque minerals with radii of < 1 to 10 micrometers) that record the passage of the EMP. Many chondrules, CAIs, and isolated silicate grains within chondritic meteorites exhibit these unique features, suggesting that strong EMPs were common in the early solar nebula [1]. Here we discuss new constraints on nebular EMPs obtained from both experimental simulations and calculations of radiative heat transport. To test the feasibility of producing "dirty snowball" textures by EMP heating, olivines and pyroxenes containing metal and sulfide inclusions were heated with a 10 watt, argon-ion, CW laser operated at 514 nm. Comparisons between meteoritic "dirty snowball" textures and experimentally produced textures confirm the ability to produce the meteoritic textures by EMP heating and suggest heating times and fluxes of 0.25 to 10 seconds and 10^9 to 10^10 ergs cm^-2 sec^-1. Fluxes less than 10^9 ergs cm^-2 sec^-1 were insufficient to melt metal and sulfide inclusions, while fluxes greater than 10^10 ergs cm^-2 sec^-1 resulted in complete melting of metal, sulfide, and silicates. The experimentally determined heating time scales suggest that radiative equilibrium was reached in the "dirty snowball" formation process, indicating that the range of observed textures is controlled by cooling rates. Calculations of radiative absorption and emission allow further constraints to be placed on the EMPs responsible for "dirty snowball" formation. The absorption and emission efficiencies of grains in a blackbody radiation field were determined by calculating Planck mean cross sections for olivine, pyroxene, and iron as a function of grain size [2,3]. This information was combined with conductive heat flow calculations to determine the behavior of olivine and pyroxene grains with small inclusions of metal. Results indicate that "dirty snowball" formation results only over a narrow flux range for a given multiphase assemblage, with higher fluxes required for smaller, more transparent, or more refractory grains. For a 100-mm olivine chondrule containing a 10-micrometer "dirty snowball," the required flux is ~9 +- 1 x 10^8 ergs cm^-2 sec^-1, with a minimum pulse duration of 4 seconds (assuming an initial grain temperature of 500 K prior to heating). These values are in good agreement with experimentally determined values. The results show that pulses energetic enough to create "dirty snowballs" are also capable of producing the total melting required for chondrule formation with only slight increases in flux, or with only marginally different grain properties (e.g., more opaque inclusions, lower melting points, higher absorption cross sections). Because of the temperature and grain size dependence of the Planck mean cross sections of silicates, an EMP of the type described above will selectively melt larger aggregates and individual grains (>100 micrometer) while leaving smaller aggregates and grains unmelted. Therefore, natural products of EMP heating are: 1) the formation of chondrules in a sustained dusty environment, 2) a paucity of small chondrules, and 3) residual grains relatively unaffected by the EMPs. References: [1] Eisenhour D. D. and Buseck P. R. (1993) LPSC XXIV, 435-436. [2] Falk S. W. and Scalo J. M. (1975) Ap. J., 202, 690-695. [3] Gilman R. C. (1974) Ap. J. Supp., 268, 28, 397-403. Sanders I. S.* Reappraisal of Planetary Collision as a Mechanism for Chondrule Formation Planetary models for chondrule formation are not widely favored [1]. During the past decade, however, the importance of planetary impacts in the early solar system has featured prominantly in the fashionable "single large impactor" hypothesis for the origin of the Moon. If the Moon was, indeed, born out of the rapidly re-accreted debris of a planetary collision, then it may be supposed that other, smaller bodies (such as the chondrite parent bodies) could have been formed in a similar way, albeit on a more modest scale. The recent discovery in Bovedy (L3) of an immiscible glassy chondrule that formed from a silica pyroxenite precursor (a fractionated planetary rock) argues strongly in favor of this concept [2]. When the Bovedy evidence is placed alongside the considerations listed below, a general case for chondrule formation by early planetary collision seems attractive. 1. Radiogenic 26Mg began to accumulate in Ste. Marguerite about 5 Ma after it was first trapped in Allende CAIs [3]. Assuming uniform distribution of 26Al, planetesimals of 100 km radius or more formed during the first 2 Ma of this 5- Ma interval would have overheated and melted [4]. Collisions between such bodies would have discharged showers of molten silicate and metal into space, some of it escaping from the solar system altogether, some of it re-accreting sooner or later into new planetesimals (or onto existing ones) where it would again become ingested as melt. Through time, planetesimals would become fewer and larger, and their mutual collisions more energetic. Between about 4 and 5 Ma, with 26Al radioactivity now mostly spent, the bodies would have begun to cool and fractionate into a carapace of volcanic igneous rock (mixed with heterogeneous ejecta) overlying residual silicate melt, crystal cumulates, and molten metal/sulphide. Collision at this stage would yield droplets and rock fragments broadly comparable to those seen in chondritic meteorites. The accreted debris would be insufficiently radioactive to melt. 2. Alternative models involving remelting of dust in the nebula [1] require the fortuitous aggregation of precursor material into discrete (e.g., olivine-rich, pyroxene-rich, and metal-rich) compositional clusters before melting. 3. Collisions would produce enormous numbers of incandescent droplets simultaneously, sufficient to retard radiative heat loss and yield the relatively slow cooling rates implied by chondrule textures [5]. Also, volatile elements like Na may have remained close at hand as vapor, and later recondensed onto the cooling chondrule surfaces [6]. 4. Prior to collision, each planetesimal would have been surrounded by an orbiting torus of infalling debris, including primary interstellar dust. Mixing of the incandescent spray with the orbiting dust could explain a number of chondritic features, including accretionary chondrule rims, a source of nuclei to initiate crystallization, and the presence of unheated interstellar dust in the matrix of some meteorites. Also, reintegration of the disrupted planetesimal materials and intermingling of interstellar dust and other debris would help to maintain the primitive chemistry of chondrites and to provide an appropriate variety of chondrules and clasts. 5. The presence of more than one example of a very unusual and distinctive kind of chondrule in a particular meteorite (e.g., silica pyroxenite in Bovedy) suggests that accretion occurred near to the chondrule source, and was probably therefore rapid. In this case, hot accretion rather than metamorphic reheating is a plausible scenario. Besides, heating by 26Al decay is only possible during a short and specific time interval [4]. References: [1] Grossman J. N. (1988) in Meteorites and the Early Solar Sytem (J. F. Kerridge and M. S. Matthews, eds.), 680-696, Univ. of Arizona. [2] Hill H. G. M., this volume. [3] Zinner E. and Gopel C. (1992) Meteoritics, 27, 311. [4] Morden S. J. (1992) Meteoritics, 27, 263-264. [4] Radomsky P. M. and Hewins R. H. (1990) GCA, 54, 3475-3490. [6] Matsunami S. et al. (1992) Meteoritics, 27, 256-257. Monday, July 19, 1993 Poster Presentations 5:30 - 7:30 p.m. CENTENNIAL A, B, C Baba T. Takeda H. Saiki K. Mineralogy of Three EUROMET Ureilites Including an Orthopyroxene-Augite Achondrite Three ureilite specimens in the EUROMET collection have been studied by mineralogical techniques as part of the consortium studies. Grady and Pillinger [1] reported a preliminary result on carbon and nitrogen geochemistry. Their carbon and nitrogen data for FRO90054 serve to reinforce the belief that this meteorite, if it is a ureilite, is most unusual. They gave the bulk carbon content of 0.24 wt%. Our mineralogical data also show intermediate nature between augite-bearing lodranites [2] and magnesian orthopyroxene-augite ureilites [3]. A PTS (13 X 8 mm in size) of FRO90054,007 shows a coarse-grained equigranular texture. FRO90054 contains considerable orthopyroxene (Opx) and augite (Aug). Opx crystals reach up to 2 X 1.5 mm in size, olivine up to 2 X 1.5 mm, and Aug up to 3 X 2 mm. Olivines have parallel fractures, indicating minerals were shocked. The Opx crystals poikilitically enclose rounded olivine and Aug and show good cleavage. The silicates contain very fine- grained dusty inclusions. The modal abundances (olivine 23 vol.%, Aug 22 %, Opx 48 %, troilite and opaque vein 7 %) show that this PTS is rich in Opx more than Aug. Carbonaceous matrices at grain boundaries are smaller than other ureilites. The compositions of olivine (Fa(sub)12) is more Mg-rich than common ureilites. The Aug (Ca(sub)38Mg(sub)56Fe(sub)7) and Opx (Ca(sub)4.8Mg(sub)84Fe(sub)11.2) compositions give an equilibration temperature of 1245 degrees C [4]. Some Opx at the rim or small grains show higher Fe concentrations. A PTS (7 X 6 mm in size) of FRO90036,008 shows heavily shocked textures of olivines and pigeonites. The olivine crystals were converted into mosaics of tiny grains (Ca 0.03 mm in diameter). The mineral chemistry (Fa(sub)12 to Fa(sub)23) and textures are similar to those observed in LEW86216 [5]. The PTS (8 X 6.5 mm in size) of ACF277 shows a coarse-grained texture with elongated crystals. The modal abundances of silicates are: olivine 45 vol% and pigeonite 55%. The compositions of olivine (Fa(sub)21) and pigeonite (Ca(sub)9.7Mg(sub)73.2Fe(sub)17.1) are uniform. This is a typical ordinary ureilite. Ordinary ureilites have been said to be olivine-pigeonite achondrites such as ACF277. However, recently Opx-bearing ureilites [3] and Aug-bearing one [6] have been found. According to the mineral compositions, Opx-bearing ones are magnesian ureilites [3]. Coexisting Opx and Aug are known for a magnesian ureilite (LEW85440). FRO90054 can be classified as Opx-Aug- bearing ureilite of the above group if it is a ureilite, because the FRO90054 Opx compositions are located at the Fe-rich extension of the ureilite Opx trend and at a little above the lodranite Opx. The fact that some of the magnesian ureilites and lodranites contain Opx as a major phase implies that these two groups of meteorites are more closely related in their classification. The mineral assemblage of FRO90054 is rather similar to Aug- bearing lodranite, MAC88177 than other ureilites. If the oxygen isotope ratios indicate otherwise, FRO90054 would be classified as a Aug-bearing lodranite such as MAC88177 with a little metal, because it consists of Opx, Aug and olivine with similar Mg/Fe ratios [2]. Only evidence indicating affinity to ureilites is that the CaO concentration of olivine (0.30 wt%) is within the range of ureilites and that the chromite has not been found. Although they have different oxygen isotope trends [7] and interstitial grain boundary materials, ureilites and lodranites may have formed from different source materials, but the formation mechanism may be similar. We thank EUROMET for the meteorite samples, and Drs. T. Ishii and H. Yoshida for their technical assistance. References: [1] Grady M. M. and Pillinger C. T. (1993) LPS XXIV 551-552. [2] Takeda H. et al. (1992) Proc. Japan Academy 68B, 115-120. [3] Takeda H. (1989) EPSL, 93, 181-194. [4] Ishii T. et al. (1979) Min. J., 9, 460-481. [5] Saito J. and Takeda H. (1990) Proc. NIPR Symp. Antarct. Meteorites, 3, 132-146. [6] Takeda H. et al. (1989) Meteoritics, 24, 73-81. [7] Clayton R. N. et al. (1992) LPS XXIII, 231-232. Brearley A. J. Spilde M. N. Papike J. J. Pyroxene Microstructures as Recorders of the Thermal Histories of Unequilibrated and Equilibrated Eucrites It has been widely recognized that the non-cumulate eucrites consist of two groups, which reflect different degrees of post- crystallization equilibration. The unequilibrated eucrites, exemplified by Pasamonte, retain clear evidence of primary igneous zoning in pyroxenes, while equilibrated eucrites such as Stannern and Juvinas have lost their primary magmatic mineral compositions. Although some petrographic observations have been made, the details of the thermal histories of unequilibrated and equilibrated eucrites remain unclear [e.g., 1,2]. We have begun a systematic study of pyroxene mineral chemistry and microstructures in Pasamonte, Stannern, and Juvinas to examine this problem in detail. In Pasamonte pigeonites show strong Mg-Fe zoning from core to rim and contain very thin exsolution lamellae. TEM observations show that lamellae of augite have exsolved primarily parallel to (001), but rare, extremely thin (100) lamellae are also present. The average width of the lamellae is 120 nm with a wavelength of 270 nm in the rims and 70 nm with a wavelength of 200 nm in the Mg-rich cores. The compositions of the existing lamellae determined by AEM show that as the bulk pigeonite composition becomes more Ca and Fe-rich (i.e. towards the rims), the exsolved augite lamellae decrease in Ca content. The relationship is consistent with experimental data for the subsolidus phase relations of augite and pigeonite, which show that equilibration of the lamellae occurred between 800-850 degrees C [3], indicating rapid cooling. The microstructures present in Stannern pigeonites are significantly different from Pasamonte. Exsolution has occurred exclusively on (001) pigeonite and the lamellae have extremely variable widths, ranging from 70 nm up to a maximum of 1 micrometer. In any one area [30 x 30 micrometers) the lamellae widths are all similar, but there is considerable variation in the mean lamella width from region to region. The compositions of coexisting pigeonite and augite lamellae show that extreme unmixing of the two phases has occurred and that equilibration of coexisting lamellae occurred at temperatures between 500 and 600 degrees C [3]. As reported by [1] we have found chromites exsolved within the Stannern pyroxenes, but no ilmenite or metal particles have been observed. Chromite has exsolved exclusively within augite lamellae and appears to have nucleated coherently at pigeonite-augite interfaces. AEM of augite lamellae in Stannern shows that they are depleted in Cr and Al relative to augite in Pasamonte consistent with exsolution of chromite. Cr contents in both pigeonite and augite in Stannern are essentially zero, showing that the Cr systematics measured by electron microprobe are the result of the presence of fine-grained exsolved chromite. For Pasamonte an initial rapid cooling stage is indicated by the small wavelength of the exsolution lamellae in pigeonite cores. Cooling in the subsolidus region slowed after exsolution in the Mg-rich cores had occurred. The thermal history of Stannern appears to be more complex. The variability in the thickness and wavelength for exsolution lamellae in Stannern strongly indicates that there were compositional gradients in the crystal while exsolution was occurring, i.e. in the subsolidus region. Therefore equilibration could not have taken place above the solvus. The magma may have been emplaced into a hot environment such as the base of a thick flow and cooled slowly producing thick exsolution lamellae. After significant exsolution had occurred cooling slowed, perhaps as a result of burial under additional flows. During this period compositional equilibration of pyroxene occurred, but the spatial distribution of exsolution lamellae was preserved. Alternatively a second separate reheating event could be invoked, perhaps as a result of contact metamorphism by other, later intrusive rocks. In the latter scenario the exsolution of opaque phases is probably associated with this reheating event. Funded by NASA grant NAGW-3347 to J. J. Papike (P.I.). References: [1] Duke M. B. and Silver L. T. (1967) GCA, 31, 1637. [2] Harlow G. E. and Klimentidis R. (1980) LPS XI, 1131. [3] Lindsley D.H. (1983) Am. Mineral., 68, 477-493. Briscoe J. F. Moore C. B. Determination of Formic and Acetic Acid in Chondritic Meteorites The concentrations of formic and acetic acid have been determined using ion exclusion chromatography after water extraction from several chondritic meteorite samples. Monocarboxylic acids are of great importance because of their high concentration in meteorites and for their role as precursor molecules in organic synthesis [1]. The concentration of acetic acid has been determined previously using gas chromatography [2,3]. Prior gas chromatographic analyses failed to resolve formic acid and so the results were limited to carboxylic acids having two or more carbons. Alternatively, wet chemical methods for the determination of formic acid, although precise, are lengthy and difficult to reproduce [4]. Ion exclusion chromatography (ICE) is an excellent technique for the simultaneous determination of formic and acetic acids. Using ICE the carboxylic acids can be determined in less time and with minimal sample handling. In most cases the amount of formic acid present is found to be lower than the amount of acetic acid present. This contradicts the accepted synthesis scheme of higher homologs being made from lower members, where the formic acid would be expected to have a higher concentration than acetic acid. Other monocarboxylic acids in the homologous series (C(sub)2-C(sub)7) have been shown to decrease with increasing carbon number as expected [2,3]. This data suggests that either the formic acid may have been preferentially depleted or it may have a different synthesis mechanism as compared with the other monocarboxylic acids present in meteorites. Additionally, there is a relationship between the amount of formic and acetic acid present and the oxidation state of the iron in the chondrites. As the matrix environment becomes more oxidizing, the amount of the two monocarboxylic acids increases comparatively. Furthermore, the ratio of formic to acetic acid starts to increase as the metal phase is more oxidized, suggesting that a more oxidized matrix environment in some way makes the production of higher homologs from lower members more favorable. References: [1] Cronin J. R. et al. (1988) In Meteorites and the Early Solar System (J. F. Kerridge and M. S. Matthews, eds.), 819-857. Univ. of Arizona. [2] Yuen G. U. and Kvenvolden K. A. (1974) Nature, 246, 301-303. [3] Yuen G. et al. (1984) Nature, 307, 252-254. [4] Kimball B. (1988) M.S. thesis, Arizona State Univ. [5] Urey H. C. and Craig H. (1953) GCA, 4, 36-82. [6] Sears D. W. and Dodd R. T. (1988) In Meteorites and the Early Solar System (J. F. Kerridge and M. S. Matthews, eds.), 3-31. Univ. of Arizona. Table 1, which appears here in the hard copy, shows a representative concentration of formic and acetic acid (in ppm) for select chondrites as measured by ion exclusion chromatography. Britt D. T. Bell J. F. The REAACT Multiple Asteroid Rendezvous Mission The REAACT (Rendezvous with Earth Approaching Asteroids) mission was a proposal to last year's Discovery Program Workshop. This mission is designed to address the major question in asteroid science, the link between the spectral diversity of the asteroids and their geochemistry. Spectral diversity is perhaps the overriding feature of the asteroids. There are currently 17 different asteroid spectral types with at least 5 spectral types having no established meteoritical analogs and several more types having dubious or poorly defined analogs. The best way to link the large body of meteorite geochemical data with the groundbased remote sensing spectral data is asteroid rendezvous that yield imaging, IR-spectroscopy, and high-precision geochemical data. However, part of the scientific landscape are the severe programmatic and budgetary constraints that are imposed by limited NASA resources. The Discovery program was conceived as a way of maximizing those limited resources by emphasizing small, short missions with highly focused scientific objectives. Even with the lowcost Discovery-type missions, we cannot expect more than one or two asteroid rendezvous/sample return missions per decade. But missions that only target a single asteroids for rendezvous cannot provide a statistically meaningful sample of the wide diversity within the asteroids. It will take decades under the current budget constraints to build up an accurate picture of asteroidal geochemistry from single missions. The only way to address the diversity of the asteroids is with multiple rendezvous and REAACT proposes to do exactly that by buying at least four high-quality rendezvous spacecraft under a single Discovery cost cap. This is accomplished by adherence to four cost-saving principals: Maximum Use of Existing Commercial Equipment: REAACT uses the TRW Eagle-class lightsats for the spacecraft bus and a commercial space-qualified CCD camera for the imager. Scientific Instruments of Simple Deisign and Extensive Heritage: The primary geochemical instrument is an Alpha X-Ray spectrometer that first flew on the Surveyor missions in the 1960s. Target Near-Earth Asteroids: The program targets small asteroids with low- i, low- e and semi-major axis near 1.0 AU using single launches that place the spacecraft on a direct rendezvous trajectory. This keeps cruise times short and makes data return much easier. Built-in Redundancy: Risk is reduced by having multiple spacecraft rather than expensive redundant systems on each spacecraft. This greatly reduces spacecraft cost, complexity, and mass; and with multiple spacecraft the loss of one or even two spacecraft will not cause the failure of the program. In addition the mission plans for all the rendezvous are essentially identical to reduce mission planning and operations costs to a minimum. REAACT offers a cost-effective way to address the fundamental questions of asteroid science. Its advantage lies in its simplicity and redundancy. REAACT can deliver at least four identical commercial-grade spacecraft for quick rendezvous with four asteroids within the budget constraints of one Discovery mission. Chochula P. Masarik J. Povinec P. On the Production of Cosmogenic Nuclides in Extraterrestrial Bodies by Galactic Cosmic Rays In the interplanetary space, meteoroids are continuously irradiated by high-energy ray particles. These particles have enough energy to penetrate into and interact with matter in the solar system. Some of these interactions leave reaction products that persist for long period of time. These reaction products can be used both to determine the nature and behavior of cosmic rays in the past and to study the history of the target. The accurate modeling of the reaction products is necessary for the interpretation of measured values. In this paper we present the results of the simulation of production rates of cosmogenic nuclides in lunar and meteoritic samples. The presented calculations use the system of coupled Monte Carlo codes KASKADA [1]. These treat different physical phenomena that have to be considered in the accurate computer simulation of radiation transport and interaction. The simulation of the interaction of an incoming high-energy particle is started by a choice of primary particle coordinates and direction relative to the target. The incident particle is followed to its first collision with a target nucleus, in which the production of secondaries is performed using the intranuclear cascade- evaporation model. Then the histories of individual secondary particles are followed one after the other until the predefined cut-off energies are reached or the geometry is left. Production rate P(sub)j of cosmogenic nuclide j at depth d in an irradiated body was calculated with the equation, which appears below in the hard copy. Where N is the number of atoms for target element i per kg material in the sample, sigma is the cross section for the production of nuclide j from target element i by particle k, and J is flux of primary and secondary particles of type k with energy E(sub)k. In the case of meteoroids the irradiated body was divided into concentric shells. The lunar surface was simulated with a large cylinder, which was also divided into smaller cylindrical sublayers. The corresponding fluxes were calculated within each sublayer for both geometries. In recent years a series of studies of different aspects of production and transport processes have been carried out for a variety of targets and shielding conditions [2]. All these calculations showed the importance of detailed simulation of neutron and proton fluxes for accurate calculation of production rates of cosmogenic nuclides. In our present calculations, we have used spectrum of the shape given by [3], normalized to unit integral flux above 10 MeV. One of goals of our calculations is the determination of the average GCR spectrum. Our earlier calculations [1] and results of work [4] allowed us to use linear dependence of production rates on the integral flux. The average fluxes of GCR particles were determined by fitting the measured depth profiles with calculated ones. In the lunar case, on the base of simulation of ^26Al and ^53Mn production in Apollo 15 drill core samples we obtained best fit for integral flux 4.85 part.cm^-2s^-1 that corresponds to the primary GCR proton spectrum with modulation parameter 467 MeV. Simulation of the production of the same nuclides in Knyahinya and St.Severin led to the initial integral particle flux 5.29 part.cm^-2s^-1 that is equivalent to the spectrum with modulation parameter 419 MeV. Our results are in fairly good agreement with [4], however the increase of primary GCR particle flux with distance from the Sun is a few percent higher than measured by Pioneer 10 and 11 [5]. References: [1] Masarik J. et al. (1991) J. Phys. G. Nucl. Part. Phys., 17, S493-S504. [2] Masarik J. et al. (1992) Meteoritics, 27, 209-210. [3] Castagnoli G. and Lal D. (1980) Radiocarbon, 22, 133-159. [3] Michel R. et al. (1991) Meteoritics, 26, 221-242. [5] McKibben R. B. (1987) Rev. Geophys., 25, 711-722. Delaney J. S. Redox Controls of Fe-Mn-Mg in Ordinary Chondrites If the Fe/Mn ratio of CI chondrites (~95) is an average for the solar system, then the variation of Fe/Mn ratios (50-150) among other chondritic meteorites reflects processing of the prechondritic material. The known, moderate volatility of Mn is often invoked to explain these variations [1], but CI normalized abundance data for O and C chondrites [2] reveals that although carbonaceous chondrites are indeed depleted in Mn, as reflected in their high Fe/Mn ratios, the ordinary chondrites are essentially undepleted. Relative to CI, however, the ordinary chondrites have variable Fe abundances and this is reflected in their bulk Fe/Mn (atom) ratios. The low Fe/Mn ratios of ordinary chondrites are, therefore, not a reflection of Mn enrichment relative to CI but of Fe depletion. In detail, ordinary chondrites also reveal evidence of processing prior to their assembly as O-chondrites. Figure 1 shows the variation of Fe, Mn, and Mg in ordinary chondrites relative to CI. The line shows the trace of iron loss or addition relative to Orgueil [2]. Both bulk and silicate fractions of OCs fall on or close to this line, confirming that they are related to the CI average by Fe loss. The order of increasing Fe loss (H-L-LL) reflects the general classification of these meteorites. However, the silicate-only fraction of the OC [3] shows the opposite trend (LL-L-H) with Fe loss, reflecting the metal/silicate fractionation present in the meteorites. The variation of bulk OC Fe/Mn reflects the removal of Fe metal from the OC precursors by a reduction reaction presumably of the type FeO(silicate) + CO(nebula) = Fe(metal) + C02 (gas). Clearly H-chondrites show little of this metal removal in bulk. L and LL chondrite precursors, however, have been progressively depleted in Fe resulting in lower bulk Fe/Mn ratios. In contrast, the metal rich H-chondrites have the most reduced silicates (lowest silicate Fe/Mn) although much of the metal fraction is still present in the meteorites. The LL chondrites contain the least evidence of in situ reduction, as their silicates have Fe/Mn ratios close to their bulk Fe/Mn. The reaction that reduced the FeO to metal was probably controlled by in situ carbon rather than nebular CO. Since the lowering of Fe/Mn in the bulk OCs reflects nebular reduction by CO, other evidence of that reduction should be sought. The progressive shift of the oxygen isotope ratios in sequence H-L-LL toward heavier oxygen probably reflects this reduction reaction with a gas reservoir enriched in heavy oxygen as suggested by [4]. The removal of Fe metal prior to the assembly of OC meteorites, therefore provides a link to the undifferentiated iron meteorites. The fractionation of the silicates in OCs relative to the bulk meteorites may reflect either in situ reaction of FeO with C in the chondritic host, or incomplete separation of metal formed by nebular reaction. Correlation of the Fe-Mn-Mg results for chondrules from UOCs with their oxygen isotope signature should provide a measure of these competing effects, although the effects of Mn volatility may obscure the data at the scale of a chondrule. References: [1] Ganapathy and Anders (1977) Proc. LPSC 8th. [2] Wasson and Kallemeyn (1988) Phil. Trans. R. Soc. A235. [3] Jarosewich (1990) Meteoritics, 25. [4] Clayton and Mayeda (1992) Ann. Rev. Earth Planet. Sci. Acknowlegment: NAG9-304 Flynn G. J. Catastrophic Disruptions or Slow Erosion as the Dominant Mechanism for IDP Production Evidence from the degree of entry heating [1,2] and solar flare track densities [3] suggests a large fraction of the silicate IDPs recovered from the stratosphere are derived from main-belt asteroidal parent bodies. The two dominant mechanisms by which main-belt asteroids contribute to the interplanetary dust are slow erosion and catastrophic disruption with subsequent comminution of the debris. These mechanisms produce profoundly different IDP populations. If slow erosion dominates, the IDPs sample the diversity of the asteroid population in rough proportion to the surface areas of the individual asteroids [4], although probably modified by fragmentation effects [5]. If catastrophic collisions dominate, the IDPs principlly sample the debris of a few recent disruptions. Comparison of the compositional diversity of the IDP population with that of the main-belt asteroids and with the asteroid families associated with recent disruptions should allow a choice between the two mechanisms. Diversity in the Main-Belt: Reflection spectroscopy indicates that the main-belt asteroids include primitive, metamorphic, and igneous objects showing a great range of compositional diversity within each group [6]. Likely parent bodies for most types of meteorites have been identified in the main-belt, and several types of asteroids remain without analog meteorites [6]. Many of these asteroids have relatively high albedos. Diversity of the IDPs: Both the anhydrous and the hydrated silicate IDPs have high contents of carbon [4,7] and volatiles [8]. Only 4 of 30 silicate IDPs analyzed by [4] and 2 of 11 analyzed by Thomas et al. [7] had C/Si lower than CM meteorites, suggesting that most silicate IDPs are carbonaceous chondrites [4]. The parent bodies of these carbon-rich IDPs are likely to be dark objects. Unless many higher albedo interplanetary particles are hidden among the "terrestrial" dust on the collectors, the silicate IDPs do not sample the higher albedo asteroids in proportion to their surface areas in the main belt. The lack of compositional diversity suggests the silicate IDPs principly sample only a few source types, mostly carbonaceous chondrites. Catastrophic Collision Origin: The three major dust bands, identified in the main-belt by IRAS measurements, are believed to have resulted from the catastrophic disruption of larger asteroids [9]. These dust bands are associated with the Themis, Eos, and Koronis families of asteroids (see Table 1). Two of these families are dominated by asteroids similar to carbonaceous chondrites: Themis, associated with hydrated carbonaceous chondrite material, and Eos, associated with anhydrous carbonaceous chondrite material. The Koronis family is classified as S-type, generally thought to be differentiated stony-iron bodies [6], but analog meteorites have not yet been identified. If most of the collected IDPs were derived from comminution of the debris from the catastrophic collisions responsible for the IRAS dust bands we would expect IDPs to consist principlly of carbonaceous chondrite material of both the anhydrous and hydrated types as well as S-type material of unknown chemical and mineralogical composition. The low abundance of non-carbonaceous, silicate IDPs is consistent with catastrophic collisions being the dominant dust producing mechanism in the main belt. References: [1] Flynn G. J. (1989) Icarus, 77, 287-310. [2] Brownlee D. E. et al. (1993) LPS XXIV. [3] Sandford S. A. (1986) Icarus, 68, 377-394. [4] Schramm L. S. et al. (1989) Meteoritics. [5] Flynn G. J. (1990) ACM III, 59-62. [6] Bell J. F. et al. (1989) In Asteroids II, 921-945. [7] Thomas K. L. et al. (1993) GCA, in press. [8] Flynn G. J. et al. (1993) LPS XXIV. [9] Sykes M. V. (1989) In Asteroids II, 336-367. Table 1, which appears here in the hard copy, shows dust band associations with asteroid families. Gilmour J. D. Johnston W. A. Lyon I. C. Turner G. RELAX Update: Recent Developments in the Resonance Ionization Mass Spectrometry of Xenon Refrigerator Enhanced Laser Analyser for Xenon (RELAX) [1] is an ultrasensitive mass spectrometer designed for the analysis of xenon from meteorites. It combines a selective laser resonance ionization ion source using a Nd:YAG based laser system for the generation of UV light pulses at 10Hz with a low volume (400cc) time-of-flight mass spectrometer. A cryogenic sample concentrator is used to increase sensitivity to the point where one count per second is produced from a sample of 1000 atoms, and this, combined with the continuous measurement of all isotopes and baselines implicit in the time-of-flight technique gives an effective sensitivity 2 orders of magnitude in excess of conventional, single-collector, magnetic sector instruments. In the past year the data acquisition system has been radically altered by the addition of discrimination against electronic noise and the development of a pulse counting system for small quantities of gas (<5,000 atoms of a single isotope). Peak fitting is still used to determine the heights of the peaks in the mass spectrum corresponding to individual isotopes. However, the fitted peak height can now be corrected if the isotopic abundance is sufficiently low for the count rate to be below the saturation level (approximately 5cps). This has resulted in an increase in precision in the measurement of the low abundance light isotopes ^124Xe and ^126Xe. Precisions of close to 10% are achievable in under 10 minutes in calibration aliquots that contain approximately 600 atoms of ^124Xe. Gas can be released from samples using either an argon ion laser microprobe or a tantalum filament microfurnace while being observed through the laser port via a video monitoring system. The blank from the sample extraction chamber in which either the filament furnace or the laser microprobe sample mount can be fitted is not measurably higher than the dynamic blank of the spectrometer (10^-15ccSTP xenon total) however, at high filament temperatures (~1000 degrees C) the filament furnace produces a larger blank (up to 5 x 10^-15ccSTP) over the 5 minute duration of a typical sample extraction. No corresponding increase in blank has been noted when the laser microprobe is in use. High resolution stepped pyrolysis analyses of acid residues from the Murchison meteorite have been performed using the filament furnace in preference to the laser probe because of the greater stability and reproducibility of its temperature control. It is hoped to obtain an optical pyrometer in the near future to allow temperature measurements to be made during sample release. References: [1] Gilmour J. D. et al. (submitted) G Rev. Sci. Inst. Goswami J. N. Sahijpal S. Swindle T. D. Musselwhite D. S. Grossman J. N. Ion Microprobe Studies of Iodine Contents in Silicate Glasses and in Semarkona Chondrules Isotopic studies of electronegative elements (e.g. H, C, O, S, I, etc.) by the ion microprobe is best done in the negative secondary mode as the negative ion yields for these elements are much higher compared to their positive ion yields. However, analysis of non-conducting solids (e.g., silicates) in the negative secondary mode is beset with the problem of sample charging. In addition, for heavy elements like iodine, the problem of molecular interferences is also difficult to resolve. We have used a normal incidence electron gun, that generates a cloud of low energy electron near the sample surface, to overcome the problem of sample charging. The problem of molecular interferences was effectively removed by the energy filtering technique, commonly used for trace element studies [1]. Since the normal energy filtering procedure that involves introduction of appropriate offset to the sample high voltage cannot be followed in the negative secondary mode, we have introduced offset to the electrostatic analyzer (ESA) voltage, to achieve the required energy filtering. A silicon sample was analyzed to calibrate ESA voltage offset with sample voltage offset. We have initially analyzed silicon samples and a set of silicate glasses doped with iodine (0.1 to 1.5% by weight) to check for optimum conditions for measurement of low iodine concentration (= 50V should be sufficient to suppress most of the molecular interferences excluding the hydrides. Analysis of iodine doped silicate glasses showed that even for an energy filter of 40V, the ion signals at masses 111 and 126 are sufficiently small for any oxide or hydride to contribute effectively to the iodine signal at mass 127. The count rate at mass 127 for the silicate glasses showed a linear relation with the iodine content for both 40V and 50V energy filter. Following the optimization of the instrument parameters and calibration with silicate glasses, we have analyzed iodine content in several well defined phases (FeS, glass, silicates) in several individual chondrules from the Semarkona(LL3) meteorite. These chondrules have been analyzed earlier for their trace- element content and I-Xe systematics. Because of the small size of the analyzed phases we have used a primary beam of 2nA and an energy filter of 40V for analyzing the silicate phases and both 40 and 50V energy filter for the sulphide. The results show that the iodine signal at mass 127 is maximum for the sulphide followed by glass and silicates. If we assume that the iodine ion yield for all the phases (FeS, glass, and silicates) to be similar to that for the iodine doped silicate glasses, the estimated iodine concentration varies from ~2.5 ppm in FeS to <20 ppb in olivines. References: [1] Zinner E. and Crozaz G.(1986) Int. J. Mass Spec. Ion Proc., 69, 17-38. Greenwood R. C. Hutchison R. Jones C. G. The Structure and Evolution of a CM2 Regolith: A Three-dimensional Study of Cold Bokkeveld The matrices of CM2 chondrites are a complex assemblage of high- and low-temperature components, some of which may have formed in a nebular environment, others by reprocessing in an asteroidal regolith. A necessary first step in identifying the primitive components is to understand the processes by which they were modified following incorporation into their parent bodies. Here we report the results of a textural investigation of Cold Bokkeveld. This work follows an earlier study [1] that had identified a planar fabric within Cold Bokkeveld, defined by the alignment of the long axes of various macroscopic objects. However, on sectioning the meteorite it was realized that it is composed of a more diverse range of lithic material than had been previously recognized. The nature and origin of these lithic fragments have therefore been examined in some detail. Method: To study the structure and fabric of Cold Bokkeveld a single fusion-crusted stone (maximum diameter 8cm) was cut along three directions at right angles and a series of slices removed. The stone was photographed before and after cutting to record the relationships between the slices and to document the major structural features. A polished section from each of the orthogonal cuts was prepared (total area 9 cm^2) and these were photographed using a Hitachi S2500 SEM. Montages of back- scattered electron images (x30 magnification), covering the full area of each section, were assembled. Results: Cold Bokkeveld is an inhomogeneous breccia comprising lithic fragments enclosed in a matrix of comminuted clastic material. Two end-member lithic fragment-types are present, fine- grained dark clasts and lighter-colored, coarse-grained fragments. Dark clasts are up to 1.2 cm diameter and consist predominantly of fine-grained Mg-phyllosilicate-rich material with a variable Fe-Ni sulphide content; coarser-grained, anhedral olivine grains (Fo(sub)98.1-99.5) are sometimes present. Raster- beam analysis of the four largest dark clasts examined indicates that they have a major element composition similar to dust mantles [2]. Light-colored, coarse-grained lithic fragments are up to 1.3 cm diameter, consist of abundant high-temperature objects (chondrules, etc.) enclosed by dust mantles. Features present on cut surfaces and on back-scattered montages demonstrate clearly that Cold Bokkeveld possesses a weakly- developed planar fabric defined by the alignment of the long axes of most components. Dark clasts are generally more deformed than light-colored fragments, a feature that presumably reflects the higher phyllosilicate content of dark clasts. In general the fabric within individual lithic fragments is parallel to that in the meteorite as a whole, however, in a few cases foliations are present, which show a marked discordance to that in the host. Discussion: The results of this and previous studies [2] indicate that clastic matrix in CM2 chondrites is produced within a parent body regolith by disaggregation of lithic fragments. Since it has been shown that clastic matrix in Cold Bokkeveld and Murchison is the host to interstellar silicon carbide [3] it is clearly important to identify the full range of lithic material that contributed to its formation. It remains a possibility that presolar grains may be present in one lithic component and not others. It has been proposed by [2] that clastic matrix in CM2s was formed from only a single lithic component termed by them 'primary accretionary rock' and equivalent to the light-colored lithic fragments described here. However, our evidence suggests that at least two lithic components are required to produce clastic matrix, namely i) fine-grained phyllosilicate dark clasts and ii) coarse-grained light colored fragments. References: [1] Greenwood R. C. et al. (1991) Meteoritics, 26, 340. [2] Metzler K. et al. (1992) GCA, 56, 2873-2897. [3] Alexander C. M. O'D. et al. (1990) Nature, 348, 715-717. Hashizume K. Sugiura N. A Nitrogen-concentrated Phase in IA Iron Meteorite Acid Residue Introduction: Iron meteorites are considered to have experienced a complex history, which is indicated by the variations in trace element chemistry (e.g., [1]). Among iron meteorite groups, the so called nonmagmatic groups, such as IAB, IIE, and IIICD, may have passed through different formation paths compared to others. Nitrogen isotopes can be a useful tool to understand the origin and formation processes of iron meteorites. Nikogen isotopes in a number of iron meteorites are measured [2,3], although trapping sites of nitrogen in iron meteorites are not yet clear. This is an important issue because nitrogen, a typical mobile element, may well reflect thermal history of their parent bodies (c.f., [4]). Generally, a major portion of nitrogen in iron meteorites is expected to be in a solid solution in Fe-Ni, especially in f.c.c. Fe-Ni (taenite). Franchi et al. [3] report that at least 25 to 35% of nitrogen in magmatic iron meteorites is in acid insoluble phases, however, not in those of non-magmatic meteorites. This result contradicts with the result [5] who report that a significant portion of nitrogen seems to be trapped in acid residues not only of magmatic meteorites but also of non- magmatic meteorites. To resolve the contradiction described above, and to identify the trapping site, we started measuring nitrogen isotopes in acid residues of iron metcorites. We report here preliminary results on acid residues of Canyon Diablo (IA). Procedures: Acid residues were prepared by Dr. J.-I. Matsuda and his colleagues. Different blocks of Canyon Diablo, "Can-1" and "Can-2" were treated by 14M HCl, 10M-HF + 1M-HCl, 1M-HCl, and by aqua regia, which destroyed Fe-Ni, sulfides, silicates, and shreibersite. Acid residues of these two blocks, "Can-1bn" and "Can-2b," yielded 0.102 wt% and 0.299 wt% of their original masses, respectively These residues seem to consist mostly of graphite No diamond was detected by powder X-ray analysis [6]. Preliminary Results: A predominant portion of nitrogen is released at 500 degrees C and 600 degrees C temperature fractions. Total nitrogen amounts and average delta^l5N values of the two acid residues are described in Table 1. Discussion and Summary: Sample "Can-1bn" is 3-4 times concentrated in nitrogen than "Call-2b," although its delta^15N value is within terrestrial range (0 < delta^15N < +20 per mil). Presently, we cannot deny the possibility that nitrogen in "Can-1bn" is dominated by terrestrial nitrogen, which may have been acquired during the acid treatment. Nevertheless, nitrogen isotope data of "Can-2b" suggests that indigenous nitrogen is indeed concentrated in the acid residue of Canyon Diablo. Bulk nitrogen isotope data of Canyon Diablo is reported to be delta^15N= -61.8 +- 10.4 per mil, N= 38.0 +- 155 ppm [2]. Therefore, delta^15N values of "Can-2b" can be resulted by a mixing of indigenous nitrogen and contaminating nitrogen. However, distinct delta^15N values of these two samples may indicate, in turn, that nitrogen isotopes in inclusions of Canyon Diablo are truly heterogeneous because carbon isotopes of graphite inclusions in IA iron meteorites seem to be heterogeneous [7]. Acknowledgments: We thank Dr. J.-I. Matsuda of Osaka University for providing samples and information on these samples. References: [1] Scott E. R. D. and Wasson J. T. (1975) Rev. Geophys. Space Sci., 13, 527-546. [2] Prombo C. A. and Clayton R. N. (1983) Meteoritics, 18, 377-379. [3] Franchi I. A. et al. (1988) Meteoritics, 22, 379-380. [4] Hashizume K. (1993) Doctor Thesis. [5] Murty S. V. S. et al. (1983) GCA, 47, 1061-1068. [6] Ogata Y. et al. (1990) In Abstract of the 1990 Annual Meeting of the Geochemical Society of Japan, 57. [7] Deines P. and Wickman F. E. (1973) GCA, 37, 1295-1319. Table 1 appears here in the hard copy. Hill H. G. M. Three Unusual Chondrules in the Bovedy (L3) Chondrite The Bovedy (L3) chondrite [1] has recently been studied petrographically using SEM and EMPA as part of a general review of the Irish meteorites. The following chondrules are notable: Chondrule 1. A covered thin-section of the Bovedy (Sprucefield) meteorite contains a very highly-strained, ellipsoidal, radiating pyroxene chondrule with a semi-major axis of 2mm. The elongation ratio, 2.6 x, is higher than values published elsewhere [2]. Chondrule 2. A slab of Bovedy (~48 cm^2) contains an exceptionally large, ellipsoidal, porphyritic olivine chondrule (semi-major axis = 1.4 cm, minor axis = 0.8 cm). This is among the largest droplet chondrule on record [2]. The chondrule is texturally identical to other PO chondrules in the meteorite. Chondrule 3. A polished thin-section, prepared from the above slab, contains an ellipsoidal-to-irregular shaped glassy chondrule (Fig. 1). SEM and EMPA confirm a composition of pyroxenitic glass (brown) with globular and elongate inclusions of silica glass (colorless). Representative EMPA of the brown glass (in wt%) is: SiO2 57.49, Al2O3 0.93, Cr2O3 0.38, FeO 14.22, MnO 0.63, MgO 23.32, CaO 2.69, Na2O 1.03 (no other elements detected). This can be recast as a pyroxene with formula Ca(sub)0.10 Na(sub)0.07 Fe(sub)0.43Mg(sub)1.26Al(sub)0.04 Cr(sub)0.01Mn(sub)0.02Si(sub)2.08O(sub)6. The composition corresponds closely with that reported by [3] for a silica pyroxenite clast from the same meteorite. It suggests that the chondrule was derived by rapid melting of the material represented by the clast, which has been interpreted as an igneous fractionate formed in a planetary environment. References: [1] Graham A. L. et al. (1976) GCA, 40, 529-535. [2] Grossman J. N. et al. (1988) In Meteorites and the Early Solar System (J. F. Kerridge and M. S. Matthews, eds.), 619-659, Univ. Arizona. [3] Ruzicka A. and Boynton W. V. (1992) Meteoritics, 27, 283. Fig. 1, which appears here in the hard copy, shows a photograph of chondrule 3 photographed in plane polarized light. Darker areas within chondrule boundary are pyroxenitic glass. The (white) globular and elongate inclusions are silica glass. The width of the image is 1.7 mm across. Hough R. M. Sigurdsson H. Franchi I. A. Wright I. P. Pillinger C. T. Gilmour I. Carbon and Oxygen Isotopic Measurements of K/T Boundary Spherules from Haiti Glass spherules thought to be tektites from Haiti have previously been analyzed for their mineralogy and chemical composition to identify their origin and mode of formation [1]. They contain bubbles and occur in various colors dependent upon the original target rock. To investigate these spherules and the nature of any gas phase, several dark brown glasses have been analyzed for their carbon content and isotope composition, using stepped combustion analysis and static mass spectrometry. Both brown and yellow spherules were analyzed for oxygen isotope composition using laser fluorination and conventional dynamic gas-source mass spectrometry. Some spherules were analyzed whole for carbon but one was broken into fragments for the purpose of replication. Individual fragments were initially analyzed and found to yield a total of 0.2 wt% carbon in two components of different isotopic composition. The first, released between 350-400 degrees C had a delta^13C of -22 per mil whereas the second, between 500-600 degrees C had a delta^13C of -6.3 per mil. As the lower temperature release was presumed to be contaminated, other spherule fragments were pre- treated with 0.1M chromic acid to remove organic and carbonate components. Analyses of cleaned fragments indicated a variable carbon content from 0.005 to 2.6 wt% carbon but still with two isotopically different components. The first with a delta^13C of -0.8 per mil and the second, a delta^13C of -19.0 per mil. The spherules are both variable and heterogeneous. The -19 per mil component is apparently present in most of the spherules and released by 600 degrees C. A component with a similar combustion temperature and delta^13C has been encountered in K/T residues containing nanodiamonds [2]. There is currently no information available confirming its identity, but it does not appear to be surficial or an oxidizable organic. Identification of these carbon components by future work may reveal a possible source and mode of formation for the spherules and will also clarify the effect of the internal bubbles upon the compositions. Dark brown spherules selected for oxygen isotope measurements were broken into fragments to allow repeat analyses on the same spherule. Due to the smaller size of the yellow spherules they were analyzed whole. The dark brown spherules yield a delta^17O of 4.97 to 3.65 per mil and a delta^18O between 9.47 to 7.15 per mil. The yellow spherules yield a delta^17O of 6.77 per mil and a delta^18O of 13.02 per mil. Both closely follow the terrestrial fractionation line for Delta^17O with only a slight deviation and the delta^18O values agree with those previously measured by [3]. Heterogeneity's seen in the carbon data for the dark brown spherules seem to be reflected in the oxygen data with variations between fragments of the same spherule and between whole spherules. The yellow spherules appear to be homogeneous in terms of their oxygen isotopic compositions. References: [1] Koeberl C. and Sigurdsson H. (1992) GCA, 56, 2113-2119. [2] Gilmour I. et al. (1992) Science, 258, 1624-1625. [3] Sigurdsson H. et al. (1991) Nature, 349, 482-486. Howard W. M. Arnould M. Rayet M. Production of Intermediate Mass Elements in the High-Entropy Supernova Bubble Recently much interest has been generated by the study of the neutron-rich alpha-rich freeze out that is expected to occur in the neutrino-energized wind that emerges from the nascent neutron star in a Type II supernova [4]. Such a wind is expected to have high entropy and relatively low electron-to-baryon ratio, Y(sub)e, due to the interaction of the electron neutrinos and anti-neutrinos with the protons and neutron escaping in the wind. It was recognized by [4] that such conditions provide an attractive site for the astrophysical r-process, as well as contributing to the Galactic production of some intermediate mass elements that are usually ascribed to the astrophysical s- and p- processes. Meyer et al. [3] and [1] demonstrated that a solar-system distribution of r-process nuclei might be produced for an interesting range of entropy and Y(sub)e. In this paper we study the production of intermediate mass elements in this high-entropy wind and discuss the implications for isotopic anomalies found in meteorites. The calculations are performed with the help of a single reaction network that is able to describe the establishment of nuclear statistical equilibrium, the charged-particle reaction freeze-out, and the subsequent neutron captures and Beta-decays that produce the r-nuclides. We find that for an entropy per baryon (in units of kappa) of less than s = 200, an r-process does not occur. However, there may be a significant production of intermediate mass elements (especially near the neutron closed shell N = 50). In particular, for entropies near s = 100, there is signficant production of the nuclides ^90Zr, ^89Y, ^88Sr, ^87Rb, ^86Kr, ^82Kr, ^72Ge, and ^70Ge. For entropies near s = 20, the neutron-rich isotopes ^70Zn, ^64Ni, ^60Fe, ^54Cr, and ^50Ti are produced. Of course, these results depend on the Y(sub)e of the emerging wind. We investigate the production of these nuclides within the context of [2] supernova model and discuss the implications of this type of nucleosynthesis for some of the isotopic anomalies found in meteorites. This work was performed under the auspices of the U. S. Department of Energy by Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48. References: [1] Howard W. M. et al. (1993) Ap. J., in press. [2] Mayle R. W. and Wilson J. R. (1993) in preparation. [3] Meyer B. S. et al. (1992) Ap. J., 399, 656. [4] Woosley S. E. and Hoffman R. D. (1992) Ap. J., 395, 202. Howell E. S. Britt D. T. Bell J. F. Binzel R. P. Lebofsky L. A. 0.4-3.5-micrometer Observations of 4179 Toutatis We obtained nearly simultaneous observations of 4179 Toutatis over a 0.3-3.5 micrometer wavelength range on 4 January 1993 UT. Howell obtained a 1.2-2.5 micrometer spectrophotometry using the Multiple Mirror Telescope in Arizona. Britt and Bell obtained narrowband photometry in the 3-micrometer region as well as broadband JHK photometry from the Infrared Telescope Facility in Hawaii. Binzel measured the visible spectrum using a CCD spectrograph at the McGraw-Hill Observatory in Arizona. Using V photometry reported by Pravec in the Czech Republic on adjacent nights [1], we were able to combine all these spectral regions. The rotation period of this object is approximately 10 days, so the time differences between the measurements of different spectral regions are negligible. Tholen has classified 4179 Toutatis as an S-type asteroid based on visible photometry. We measure a pyroxene absorption band near 2 micrometers, present in most S-type asteroid spectra. Unfortunately, a gap in spectral coverage prevents us from determining the characteristics of the 1-micrometer absorption band accurately. The spectral slope as measured from 1.25 to 2.2 micrometers is 6-10%, which is modest compared to other S-type asteroids. The spectrum of this asteroid is similar to other near-Earth S-type asteroids that have been observed in the near-infrared wavelength region. On 4 January 1993, 4179 Toutatis was 0.182 AU from the Earth, and 1.158 AU from the Sun. At this solar distance, the thermal emission contributes substantially to the flux at 3 micrometers. The determination of thermal emission is complicated by the slow rotation rate and the irregular shape of this object that was revealed by radar observations [2]. Preliminary results suggest that no 3-micrometer absorption feature is present, indicating that this object is anhydrous. Using these spectral data, we will compare 4179 Toutatis to other S-type asteroids, both in the main belt and the near-Earth environment. References: [1] Pravec P.(1993) Personal communication. [2] Ostro S. J. (1993) paper presented at Hazards Due to Comets and Asteroids Conference. Jenniskens P. Betlam H. Betlam J. Barifaijo E. Schlueter T. Hampton C. Heusser G. Laubenstein M. Mbale Meteorite Fell On August 14, 1993, an ordinary chondrite of type L6 entered the atmosphere of Mbale, Uganda, broke up, and caused a strewn field 1 x 5 km2 in size. Shortly after the fall, an expedition was organized in order to gather eye witness accounts and locate the position of impacts. The impact points of 48 specimens were located. For 267 specimens mass estimates were made. Two specimens were measured for short-lived radio nuclides, one of these only 12 days after the fall. The meteorite had an initial mass between 400 and 2200 kg (most likely some 600 kg) and approached Mbale from Az = 170-210, H = 65-85, V = 13- 20 (15) km/s. Catastrophic fragmentation occurred at an altitude of about 10-14 km, although breakup started above an altitude of 25 km. The surface of the fragments was melted. Some 210+/-40 kg reached the Earths surface several minutes later. 134 kg could be recovered. The impact caused minor damage. One person was hit on the head by a small specimen that fortunately was slowed down by the leaves of a tree. Ivanova M. A. Assonov S. S. Kononkova N. N. Shukolyukov Yu. A. The Dengli (H3.8) Complex Breccia: Petrological and Isotopic Studies The Dengli meteorite was found in the Karakum desert in July 1976. Previous studies [1] have shown that the meteorite is a complex chondritic breccia containing unusual components of achondritic affinity. In order to understand the origin of the breccia, we studied its mineralogy, petrology, and isotopic composition of noble gases. In addition, in cooperation with Dr. J. N. Goswami, we are now studying minerals in achondritic clasts using an ion microprobe. The Dengli meteorite has a distinct chondritic texture and consists of chondrules and chondrule fragments embedded into a fine-grained matrix. Equilibrated and unequilibrated populations were found among the chondrules, the latter of which are characterized by normal or inverse Mg-Fe zoning. Olivine (Fa19.6, N = 52, C.V. = 19.3) and pyroxene (Fs18.2, N = 27, C.V. = 17.0) displays large compositional variations. Based on textural and chemical characteristics, Dengli can be classified as an H3.8 chondrite. Two objects with unique compositions were identified. One has a round shape and consists of silica, orthopyroxene (Fs18.1 Wo6.2), clinopyroxene (Fs17.6 Wo27.6), and feldspar (An76.2 Ab23.2). The other object is an angular clast, which is composed of olivine (Fa17.2), Al-rich clinopyroxene (Fs6.3 Wo46.8), and feldspar (An44.7 Ab54.7). The presence of clinopyroxene and Ca-rich feldspar in these objects indicates their similarity with an achondritic material. The Ar, Kr, and Xe contents measured in Dengli are very close to those in other ordinary chondrites of petrological types 3 through 5 [2]. The Xe and Kr isotopic compositions at any temperature step differ from those of solar Xe and Kr. High ^38Ar concentrations were found. The exposure age was estimated to be 7.6 Ma, which is close to the average (6.3 +/- 0.2 Ma) exposure age of 350 H chondrites. The Dengli K/Ar age, 3.73 +/- 0.10 Ga, is close to K/Ar ages of the majority of H chondrites. Thus, Dengli is depleted in noble gases as compared to regolith breccias, and its noble gas isotopic compositions differ from those of the solar wind. This means that Dengli cannot be a regolith breccia. Preliminary ion-microprobe measurements of Mg isotopes in feldspar grains in the clasts described above did not reveal a clear excess of ^26Mg, which could result from an ^26Al decay. This suggests that the clasts were formed when ^26Al became extinct. In summary, we have to conclude that Dengli is a complex breccia that was not formed on a surface of a parent body. It should be of a nebular (accretional) origin. Meteorites with a possibly similar origin are Severnyi Kolchim [3], Study Butte, Beddgelert [4], and Barwell [5]. These meteorites also contain clasts of achondritic affinity as well as silica-bearing and other unusual objects that, in turn, should be considered as products of nebular processes. References: [1] Ivanova M. A. et al. (1992) Meteoritics, 27, 463-465. [2] Ivanova M. A. et al. (1993) LPSC XXIV, 695-696. [3] Nazarov M. A. et al. (1993) LPSC XXIV, 1055-1056. [4] Fredriksson K. et al. (1989) Z. Naturforsch., 44a, 945-962. [5] Hutchison R. et al. (1988) EPSL, 90, 105-118. Johnston W. A. Gilmour J. D. Arden J. Lyon I. C. Turner G. High-resolution Stepped Heating Analyses of Murchison Residues Using Resonance Ionization of Xenon RELAX [1] (Refrigerator-Enhanced Laser Analyzer for Xenon) is an ultrasensitive mass spectrometer designed for the analysis of xenon from meteorite samples. It combines a selective, resonance ionization ion source that has a cryogenic sample concentrator with a low-volume (400 cc) time-of- flight mass analyzer. Acid residues have been prepared from the Murchison meteorite by methods similar to those of Tang et al. [2]. Sample MM was produced from an interior sample of the Murchison C2 chondrite by HF/HCl treatment. A portion of this was oxidized using Cr2O7- followed by HClO4 to produce MM1. A proportion of C delta was removed from some of this into an ammonia suspension, leaving sample MM2 as a residue. Samples of MM1 and MM2 were analyzed by direct loading onto previously degassed filaments constructed by spot welding strips of Ta ribbon onto vacuum feedthroughs and mounting them in the sample extraction chamber. The samples were mounted in well-localized spots to minimize variations in temperature. Gas was released in a series of temperature steps, each of which were 5 min in duration; typically the filament required an initial 1 min to achieve equilibrium. Repeat extractions were made at each temperature stage in an attempt to degas the sample at that temperature before increasing the filament current. Typically 10 separate temperature steps, each consisting of at least 4 extractions, contained evidence of the presence of Xe-HL, the exotic xenon component associated with C-delta microdiamonds. No trace of Xe-s was detected in any filament furnace analysis, although it has been released from smaller quantities (25 micrograms) of residue MM2 by laser heating. Indeed, small amounts of Xe-HL were usually still evolving when the analysis was terminated as the filament broke. Figures 1 and 2 show three isotope plots appropriate to the examination of the heavy and light components of Xe-HL from stepped heating of sample MM1b (215 micrograms). The data points are compiled from the sums of the gas released in each analysis at a given temperature step, which consisted of up to nine separate heatings. The data are consistent with mixing between a Xe-HL end member from a previous study of the Murchison meteorite [3] and a planetary component. The ability of RELAX to make reliable measurements on gas samples smaller than previously possible underlines its potential contribution to the understanding of presolar xenon components. References: [1] Gilmour J. D. et al., Rev. Sci. Instr., submitted. [2] Tang M. et al. (1988) GCA, 52, 1221-1234 [3] Alaerts L. et al. (1980) GCA, 44, 189- 209. Figure 1 appears here in the hard copy. Jull A. J. T. Cielaszyk E. Brown S. T. Donahue D. J. 14C Terrestrial Age Distribution of Meteorites from the Allan Hills Region, Antarctica The terrestrial age or residence time on the Earth's surface is an important parameter in determining the history of a meteorite. Terrestrial age can be determined using several radioisotopes, predominantly ^14C [1,2] and ^36Cl [3]. Terrestrial ^36Cl [3] and ^81Kr ages [4] established that the meteorites collected from the main ice field at Allan Hills were generally over 100,000 yr, beyond the range of ^14C dating. The old ages are generally thought to be due to burial and transport of meteorites in the ice. The observed age distribution gives some indication of the processes that have occurred. We have therefore undertaken a study of ^14C ages of meteorites from the other Allan Hills ice fields, where younger meteorites were expected from the few data previously available [3]. In this report, we concentrate on the Far Western ice field. Groups of meteorites from the other Allan Hills ice fields (Middle and Near Western) have also been studied. A suite of over 50 samples from the Far Western ice field was used in this study. The ^14C age distribution for meteorites at this site is shown in Fig. 1. Some achondrites from the Far Western ice field have also been studied and generally appear to be young in terrestrial age. The number of meteorites decreases with increasing age, with a possible small age offset of the maximum, indicating there might be some transport of meteorites through the ice. However, this transport cannot be for more than 5000-10,000 yr, and this indicates that the Far Western meteorites came from within a few kilometers of the site. Detailed AMLAMP sample location maps allowed us to plot ages as a function of location to indicate any dynamic effects. This map indicates little evidence for transport outside the area of several kilometers indicated from the total ^14C age distribution. These results will be compared to other locations in the vicinity and the effects of weathering will also be discussed. References: [1] Jull A. J. T. et al. (1989) GCA, 53, 2095. [2] Jull A. J. T. et al. (1993) Meteoritics, in press. [3] Nishiizumi K. et al. (1989) EPSL, 93, 299. [4] Freundel et al. (1986) GCA, 50, 2663. Fig. 1, which appears here in the hard copy, shows terrestrial ^14C ages for the Allan Hills Far Western ice field. Meisel T. Lange J.-M. Langenauer M. Fehr K. T. The Lusatian Moldavites and Muong Nong-type Moldavites The Lusatian subfield within the moldavite strewnfield was proposed by [1], based on petrographical criteria. During the course of this study we analyzed 18 moldavites for their major- and trace-element composition (2 from the Radomilice, 2 from the Bohemian, 4 from the Moravian, and 10 from the Lusatian subfield). It was one aim of this study to elucidate geochemical relations of the Lusatian tektites in respect to tektites from the other subfields. Including data from [2], tektites from the Moravian and Bohemian subfield are quite distinct on the basis of their major-element composition. The variation of the elements can be explained by varying mixtures of clay and dolomite components besides quartz. Moravian tektites are dominated by a clay component and the Bohemian by a dolomite component respectively. Lusatian glasses are intermediate in their major- and trace-element abundances and in petrographical features as well. A mixing of glasses from the Bohemian and Moravian subfield by ground transportation can be excluded, because fluviatile transportation of Lusatian tektites was short [1]. In six moldavites the halogen abundances (F, Cl, Br, and I) were analyzed. The average abundance is lower in comparison to most of the other tektite strewnfields. The MN-7 tektite reported by [3] and reanalyzed in this work has halogen abundances 5X average, but shows no unusual content of other trace elements. One moldavite (JKV 12) shows halogen contents 30X to 50X higher than average. The Cl and Br abundances are similar to Muong Nong-type tektites from the Australasian strewnfield [4], and JKV 12 also exhibits very unusual major- and trace-element composition. According to its composition, JKV 12 belongs to the Muong Nong-type tektites and is one of the few known Muong Nong-type moldavites [3-7]. Optical and microchemical inhomogeneity in the micrometer scale within JKV 12 can be explained by the formation of two distinct type of microtektites, like the Muong Nong-type moldavite 2205 from Lhenice [7]. One phase shows a pronounced Moravian signature and the other a Bohemian respectively. We suggest that all Muong Nong-type tektites from the moldavite strewnfield are composed of two distinct glass phases besides lechatelierite. References: [1] Storr M. and Lange J.-M. (1992) GCA, 56, 2937- 2940. [2] Lange J.-M. (1993) Ph.D. thesis, in preparation. [3] Meisel T. et al. (1989) Meteoritics, 24, 303. [4] Meisel T. et al. (1992) Meteoritics, 27, 576-579. [5] Rost R. (1966) Acta Univ. Carolinae-Geologica, 4, 235-242. [6] Glass B. P. et al. (1989) LPS XX, 341-342. [7] Fehr K. T. and Preuss E. (1990) Ber. DMG, Beih. Eur. J. Min., 2, 56. Fig. 1, which appears here in the hard copy, shows analyses of moldavites by [2] in a plot of oxide weight percent ratios after [8]. There is a clear separation of the Bohemian and Moravian subfields, whereas the Lusatian samples plot in both fields. Miura Y. Takayama K. Iancu O. G. Material Evidence for Ocean Impact from Shock-Metamorphic Experiments Continental impact reveals an excavated crater that has few fresh fine ejecta showing major high shock metamorphism due to weathering [1]. A giant ocean impact rarely remains as an excavated crater mainly due to crushing by dynamic platetectonic movements on the crust [2]. However, all impact materials, including finegrained ejecta, can be obtained with artificial impact experiments [3]. The purpose of this study is to discuss material evidence for ocean impact based on shock-metamorphic experiments. Artificial impact experiments indicate that fine shocked quartz (SQ) aggregates can be formed on several target rocks (Table 1) [1]. It is found in Table 1 that (1) the largest-density deviation of SQ grain is found not at the wall-rock or the impact crater but at fine-grained ejecta, and (2) silica-poor rocks of basalt, gabbro, and anorthosite can also make fine SQ aggregates by impact. Table 1, which appears here in the hard copy, shows formations of fine shocked quartz aggregates from ocean-floor rocks of basalt, gabbroic anorthosite, and granite [3]. An asteroid (about 10 km across) hits the Earth ~65 m.y. ago [4] to result in global catastrophe by titanic explosion and climate change. But shocked quartz grains found in the K/T boundary layer were considered to come from crystalline continental rocks [5]. The present result as listed in Table 1 indicates that fine SQ aggregates can also be formed at sea-floor basaltic and gabbroic rocks [3]. The present result of formation of the SQ grains from sea- floor target rocks is nearly consistent with the finding of a sea-impact crater at the K/T boundary near the Caribbean [6]. Impact-induced volcanism at the K/T boundary can explained by the penetration from thin ocean crust to upper mantle reservoirs, if giant impact of a 10-km- diameter asteroid hit the ocean [2,7]. The present result can explain "phreatomagmatic (magmatic vapor) explosion," which is created by abrupt boiling between high-temperature magma and cold sea water to produce a titanic explosion of the asteroid disintegrated in a mass of exploding steam and vaporizing soil, including the SQ aggregates, and to create the Atlantic Ocean floor by the continental drift [8]. References: [1] Miura Y. (1991) Shock Waves, 1, 35-41. [2] Miura Y. and Takayama K. (1993) Symp. Shock Waves (Japan), 2, 193-196. [3] Miura Y. et al. (1992) Proc. Shock Waves, 18, 403-408, Springer-Verlag. [4] Alvarez L. W. et al. (1980) Science, 208, 1095-1107. [5] Bohor B. F. et al. (1984) Science, 224, 867-869. [6] Hildebrand A. R. et al. (1991) Geology, 19, 867-871. [7] Barlow N. G. (1990) Geol. Soc. Am. Spec. Pap. 247, 181-187. [8] Hartmann W. K. and Miller R. (1991) The History of Earth, 165, Workman. Miyamoto M. Takeda H. Rapid Cooling of Pallasite: Comparison of Chemical Zoning with Primitive Achondrites We estimated cooling rates of the Yamato (Y) 8451 pallasite by applying the diffusion calculation similar to that used in our previous studies [1] to measure chemical zoning of its olivines with an electron microprobe. Because of textural similarity [2], these results were compared with those of the Y 74357 and MAC 88177 primitive achondrites to obtain some constraints on the formation and structure of their primitive crusts. The Y 8451 pallasite is described by Yanai and Kojima [3]. Olivine (Ol) in Y 8451 shows the reverse zoning of Mg-Fe toward the rim from Fa(sub)10.4 to Fa(sub)9.8 within a few hundred micrometers (Fig. 1). The MnO content in Ol of Y 8451 gradually increases toward the rim from about 0.35 to 0.45 wt%. The CaO content in Ol of Y 8451 gradually decreases toward the rim within a few hundred micrometers from about 0.1 to 0.01 wt%. The Cr(sub)2O(sub)3 content in Ol also decreases toward the rim from about 0.1 to 0.04 wt% within a few hundred micrometers. The Al(sub)2O(sub)3 content slightly decreases toward the rim, although it is very low (0.01 wt%). The preservation of chemical zonings in Ol suggests rapid cooling of the Y 8451 pallasite. Diffusion calculations show that a cooling rate of about 500 degrees C/yr from 1100 degrees C [2] to 600 degrees C gives the best fit for the observed Mg-Fe profile at the rim of Ol in the Y 8451 pallasite (Fig. 1). This cooling rate corresponds to a burial depth of about 5 m under solid rock. Y 74357 orthopyroxene (Opx) shows the reverse zoning of Mg-Fe at the rim within a few tens of micrometers, probably due to reduction [4]. In contrast to this, Fa(sub)8 of Ol in Y 74357, which is a similar value to the outermost rim composition of Opx, is almost constant throughout the grain. This result has been interpreted as Y 74357 cooling slow enough to reduce the whole grain of Ol, but too fast to reduce the entire Opx. The MnO content in Opx slightly increases at the rim, probably due to reduction. The CaO content in Opx gradually decreases toward the rim from about 1.3 to 0.7 wt%. The Cr(sub)2O(sub)3, Al(sub)2O(sub)3, TiO(sub)2, and Na(sub)2O contents in Opx of Y 74357 and MAC 88177 and in augite of Y 74357 gradually decrease at the rim within a few tens of micrometers. This result suggests that liquid rich in these elements was present and was extracted at the late stage of formation of this meteorite, because these elements tend to concentrate in Opx compared with liquid. These results may be attributed to the growth process and not to subsolidus cooling episodes. The CaO content in Y 74357 Ol gradually decreases from core to rim from 0.05 to 0.02 wt%. The other elements in Ol show no clear zoning profiles. In contrast to the Y 8451 pallasite, a cooling rate of 1.5 degrees C/yr (70 m in depth) gives the best fit for the observed Mg-Fe profile at the rim of Opx and the CaO profile at the rim of Ol in the Y 74357 primitive achondrite [4]. These results suggest that the Y 8451 pallasite records relatively rapid cooling compared with the primitive achondrites. The cooling rate of the Y 78451 pallasite is slower than that (a few degrees C/hr) calculated for ureilites [5]. References: [1] Miyamoto M. et al. (1986) JGR, 91, 12804-12816. [2] Hiroi T. et al. (1992) 17th Symp. Antarct. Meteorites, 53-56. [3] Yanai K. and Kojima H. (1987) Photographic Catalog of Antarctic Meteorites. [4] Miyamoto M. and Takeda H. (1991) Meteoritics, 26, 374. [5] Miyamoto M. et al. (1985) Proc. LPSC 16th, in JGR, 90, D116-D122. Fig. 1, which appears here in the hard copy, shows the chemical zoning profile of an olivine in the Y 8451 pallasite. Open circles indicate the Fa [=100x Fe/(Mg + Fe)] component. Curves show calculated diffusion profiles. Numbers on curves show cooling rates in degrees C/yr. Data are at 5-micrometer intervals. Niedermann S. Eugster O. Frei R. Krahenbuhl U. Kramers J. D. Thalmann Ch. Formation of Alpine Au ~30 Ma Ago: Further Results of the Development of a Dating Method for Native Au We have continued the development of a direct dating method for natural gold deposits based on U/Th and noble gas analyses. In our previous work [1-3], we have demonstrated that U/Th-^4He dating as well as U-^136Xe dating should be feasible for gold because of its high retentivity and the large U concentrations in certain samples. For vein-type gold from Brusson, Northern Italy, our U/Th-^4He age agreed with the K-^40Ar age of associated muscovite [1,4]. Improved U data yield consistent ages of 30 +/- 3 and 39 +/- 13 Ma for two splits. For placer gold from the Elvo river, 30 km south of Brusson, we obtain a somewhat lower age (22 +/- 6 Ma). These results confirm an age of approximately 30 Ma for gold from that region of the Southern Alps. On the other hand, for Archean gold from the South African Lily Gold Mine (Eastern Transvaal), we calculate an age of 1200 +/- 400 Ma. Since He concentrations vary in different splits of a sample [2], these ages can be improved by analyzing U, Th, and noble gases in the same split. In a double-walled quartz extraction line we heated gold samples, wrapped in Pt or Ni foil, above the melting point (1064 degrees C). After gas extraction the molten samples were retrieved for U and Th analysis. So far we have obtained results for two placer gold samples from Tertiary sediments in Central Switzerland extracted by this method: Both Krumpelgraben and Grosse Fontanne yield U/Th-^4He ages around 500 Ma. If the Tertiary origin is correct, an explanation for the large He excesses is required. Atmospheric He (with ^3He/^4He = 1.4 x 10^-6) is negligible: In two samples of Grosse Fontanne, ^3He/^4He is <2 x 10^-8 [5]. Several potential He sources have been excluded previously [2]. The following possibilities remain as explanations for the high He concentrations. Crustal noble gases containing radiogenic ^4He may have been trapped by the gold during formation. This is probably true for ^40Ar, which is overabundant by up to 2 orders of magnitude relative to the production from ^40K decay in ~30 Ma [1,3] . Correction for "inherited" components is possible by plotting isochrons for several samples having the same age but different U or trapped gas concentrations. Our current database allows us to estimate the trapped ^4He/^40Ar ratio from the Krumpelgraben and Grosse Fontanne analyses. We obtain ^4He/^40Ar <= 3 in the trapped component, i.e., the high ages cannot be explained by inherited He. The ^4He that is not accounted for by the U and Th content and the estimated age may have been produced in inclusions of U-rich minerals, such as garnets or zircons, that were worked into the placer gold grains during residence in the river detritus. They either have a higher age than the surrounding gold or were not dissolved in the chemical processing preceding the U analysis. The latter possibility will be tested thoroughly in future analyses. However, we cannot exclude the possibility that placer gold from Central Switzerland is actually older than believed. Prior to solving this problem we shall establish the dating method, concentrating on Tertiary gold from the Brusson/Elvo region and on Archean gold from Eastern Transvaal. Acknowledgment: This work was supported by the Swiss National Science Foundation. References: [1] Eugster O. et al. (1992) Meteoritics, 27, 219-220. [2l Eugster O. et al. (1993) LPS XXIV, 455-456. [3] Niedermann S. et al. (1993) LPS XXlV, 1073-1074. [4] Diamond L. W. (1990) Am. J. Sci., 290, 912-958. [5] Kamensky I. L., personal communication. Reed S. J. B. Chinner G. A. The Bawku LL5 Chondrite Introduction: A 59-g piece of stony meteorite with black fusion crust and medium grey interior was recovered in 1992 from Northern Ghana by Dr. S. Abudulai of the Centre for African Studies, University of Cambridge. This meteorite was reported to be part of a fall that occurred in the vicinity of Bawku (11 degrees 05'N, 0 degrees 11'W) at 1630 hours (local time) on December 29, 1989; the meteorite was found between the villages of Naarango-Anisi and Kpukparigu, east of Bawku. A further mass of 1.5 kg was recovered by Dr. Abudulai during a visit in January 1993. The name "Bawku" is proposed for this meteorite. Description: The Bawku chondrite is a monomict breccia. Some well-preserved chondrules with a variety of textures are present. Olivine exhibits undulose extinction and extensive irregular cracking. Other shock features include cracks infilled with metal/troilite. Electron microprobe analyses of olivines gave a mean Fe/Fe + Mg ratio of 26.8%, with a standard deviation of 1.0% (relative) and an average CaO content of 0.03% (by weight). The mean Fe/Fe + Mg content of orthopyroxene was found to be 22.6% (standard deviation 1.2% relative). Small (10-micrometer) grains of plagioclase (75-85% Ab) were found. Automated point counting (700 points) gave volume fractions of 4.6% troilite and 1.3% metal. Compositions of 200 randomly selected points on metal grains fell into three groups, corresponding to kamacite (4-7% Ni, 1.4% Co), taenite (34-38% Ni, 0.35% Co), and tetrataenite (48-56% Ni, 0.2% Co). The inferred volume fractions of these phases were 52%, 45%, and 3% respectively. The average Ni content of the metal was 21%, with 0.9% Co. On etching, Neumann lamellae became visible in kamacite, while the taenite took on a cloudy appearance. Tetrataenite was observed as narrow rims adjacent to cloudy taenite, but also occurred sometimes in larger areas. Minor phases noted included chromite, ilmenite, apatite, and whitlockite. Classification: The Fe/Fe + Mg ratios of olivine and pyroxene place this chondrite at the lower limit of the range for the LL group, adjacent to the hiatus between LL and L [1]. The low metal content is consistent with this classification. The relatively high proportion of kamacite compared to other LL chondrites is in accordance with Prior's law, which applies within this group, unlike others [2]. The homogeneity of olivine and orthopyroxene, the existence of moderately abundant preserved chondrules, and the presence of feldspar grains comparable in size to those observed in LL5 chondrites [3] point to Bawku being a member of this class. The undulose extinction and irregular fracturing of olivine are indicative of shock class S2 [4]. References: [1] Fredriksson K. et al. (1978) In Origin and Distribution of the Elements (L. Ahrens, ed.), 457-466, Pergamon, Oxford. [2] Sears D. W. and Axon H. J. (1976) Meteoritics, 11, 97-100. [3] Heyse J. V. (1978) EPSL, 40, 365-381. [4] Stoffler D. et al. (1991) GCA, 55, 3845-3867. Rider P. E. Noble Gases and Nitrogen Released from a Lunar Soil Pyroxene Separate by Acid Etching We report initial results from a series of experiments designed to measure recently implanted solar wind (SW) ions in lunar soil mineral grains [1]. An acid-etching technique similar to the CSSE method developed at ETH Zurich was used to make abundance and isotope measurements of the SW noble gas and nitrogen compositions. Among the samples examined was a pyroxene separate from soil 75081. It was first washed with H2O to remove contamination from the sample finger walls and grain surfaces. H2O also acted as a weak acid, releasing gases from near-surface sites. Treatment with H2SO3 followed the water washes. Acid pH (~1.8 to ~1.0) and temperature (~23 degrees C to ~90 degrees C) and duration of acid attack (several minutes to several days) were varied from step to step. Finally, the sample was pyrolyzed in several steps to remove the remaining gases, culminating with a high-temperature pyrolysis at 1200 degrees C. Measurements of the light noble gases were mostly consistent with those from previous CSSE experiments performed on pyroxene [2,3]. It should be noted, however, that the Zurich SEP component was not easily distinguishable in the steps where it was expected to be observed. We suspect our experimental protocol masked the SEP reservoir, preventing us from seeing its distinctive signature. The most interesting results from this sample are its Kr and Xe isotopic and elemental compositions. Pyroxene apparently retains heavy noble gases as well as ilmenite (and plagioclase [4]). The heavy noble gas element ratios from this sample along with those previously reported [5,6] are, however, considerably heavier than the theoretically determined "solar system" values [7,8]. Explanations for the difference include the possibility that the derivations are incorrect, that there is another component of lunar origin mixing with the solar component, or that some type of loss mechanism is altering the noble gas reservoirs of the grains. The Kr and Xe isotopic compositions for the two acid-etch steps most likely to have released SW gases were identical to the "solar" values reported by the Zurich group [5]. The krypton from both steps appeared to be mixtures of "solar" krypton, an isotopically heavier component (perhaps the Zurich SEP component [5]), and a spallation component. There was, however, no evidence for such a mixture in the xenon. The compositions of the two acid-etch steps were clearly combinations of a solar Xe component and a Xe spallation component. They were also identical to that of U-Xe [9] for isotopes up to 132Xe, with the exception of an ~300 per mil enhancement of the 126Xe/132Xe ratio. This anomaly does not appear to be an artifact of spallation correction. These measurements constitute the first experimental verification of the U-Xe composition for isotopes lighter than 134Xe. Persistent contamination problems and the possibility of nitrogen being held back in the acid residue during the etching process make interpretation of the nitrogen data uncertain. However, results from the steps not obviously affected by contamination show an enhancement of N over Ar by 2x to 12x the "solar" value (from [8]). References: [1] Rider P. E. and Pepin R. O. (1993) GCA, submitted. [2] Wieler R. et al. (1986) GCA, 50, 1997-2017. [3] Benkert J.-P. (1989) Ph.D. thesis, ETH Zurich (No. 8812). [4] Wieler R. (1993) personal communication. [5] Wieler R. et al. (1992) LPS XXIII, 1525-1526. [6] Wieler R. et al. (1993) LPS XXIV, 1519-1520. [7] Anders E. and Grevesse N. (1989) GCA, 53, 197-214. [8] Cameron A. G. W. (1982) In Essays in Nuclear Physics (Barnes et al., eds.), 23-43. [9] Pepin R. O.(1991) Icarus, 92, 2-79. Ruzicka A. Kring D. A. Hill D. H. Boynton W. V. The Trace-Element Composition of a Silica-rich Clast in the Bovedy (L3/4) Chondrite The discovery of a ~4 X 4.5 X 7 mm^3, igneous-textured, silica-rich clast in the Bovedy chondrite [1] may have important implications regarding igneous processes that occurred on chondritic parent bodies [1,2]. This clast, designated Bo-1, is comprised of orthopyroxene, a silica polymorph, two feldspars, pigeonite, and minor chromite and trace metal and sulfide [1]. Bulk SEM/EMPA analyses of the clast indicated superchondritic Si/Mg and Si/Fe ratios, which Ruzicka and Boynton [1] proposed was produced by extensive olivine fractionation from a melted L-chondrite precursor. The low Fe/Mn ratio and low metal and sulfide abundances also suggest that the clast is largely missing a chondritic complement of metal and sulfide. To test these hypotheses, we measured the bulk composition of the clast using INAA techniques and found that the siderophile elements were lost in a two-step process and that partial melting also depleted incompatible lithophile elements. Lithophile Elements: Two splits (2.94 and 2.39 mg) of Bo-1 were analyzed. The concentrations of major elements (Ca, Fe, Cr, K, Na) bracket those previously determined by SEM/EMPA [1], suggesting that the two splits are reasonably representative of the bulk clast. If olivine and metal had been removed from an ordinary chondrite melt to produce the clast, then incompatible lithophile trace elements should have been enriched. Contrary to this expectation, however, the REE, Zr, Hf, Th, Sr, Rb, Cs and Br are consistently depleted to a level of 0.5-1.0 X CI abundances, while all of them (except the highly volatile Cs and Br) have concentrations of ~1.0-2.0 X CI abundances in ordinary chondrites. If the clast had been derived from melted ordinary chondrite material, then an additional step that removed incompatible elements, such as the loss of a partial melt, must have occurred. Siderophile Elements: Unlike lithophile trace elements, which are relatively unfractionated, the siderophiles Ni, Co, and Au are dramatically fractionated from Re, Os, Ir, and Ru. Nickel, Co, Au, and the chalcophile element Se are present at approximately 0.004-0.015 X CI abundances, compared to Re, Os, Ir, and Ru at 0.1-0.3 X CI. All the former elements have high affinities for S- rich metallic liquids, while the latter elements prefer solid metal [3], implying that the petrogenesis of Bo-1 included the loss of a S-rich metallic liquid. An equilibrium batch melting model of these trace siderophile elements, using the partition coefficients of [3], was constructed assuming an ordinary chondrite precursor. In these models, the proportion of liquid to solid silicate or of silicate to metal is unimportant, because the siderophile elements partition almost entirely into the metallic phases. The model results suggest that the siderophile trace elements can be adequately accounted for by a two-step process: (1) loss of a S-rich metallic liquid at high degrees of melting; and (2) subsequent loss of much of the remaining solid metal fraction. For an L-chondrite precursor, an optimal model involves the complete removal of metallic liquid generated by 90% partial melting of the metal + sulfide system, followed by the loss of 80% of the remaining 10% of the solids. Together, these two steps remove all but 2% of the initial metal + sulfide complement, consistent with the presence of only trace metal and sulfide in Bo-1. The large fraction of metallic melt involved in the first step implies that metallic liquid segregated from the remainder of the system at relatively high temperatures (~1325 degrees C for an L-chondrite precursor, based on the Fe-S phase diagram of [4]). References: [1] Ruzicka A. and Boynton W. V. (1992) Meteoritics, 27, 283. [2] Ruzicka A. and Boynton W. V. (1992) Meteoritics, 27, 284. [3] Jones J. H. and Drake M. J. (1986) Nature, 322, 221-228. [4] Kellerud G. and Yoder H. (1959) Econ. Geol., 54, 533-572. Ryerson F. J. McKeegan K. D. Determination of Oxygen Self-Diffusion in Akermanite, Anorthite, Diopside, and Spinel: Implications for Oxygen Isotopic Anomalies and the Thermal Histories of Ca-Al-rich Inclusions Oxygen self-diffusion coefficients have been measured for three natural clinopyroxenes (diopside end member), a natural anorthite, a synthetic magnesium aluminate spinel, and a synthetic akermanite over oxygen fugacities ranging from the NNO to IW buffers. The experiments employed a gas-solid isotopic exchange technique utilizing 99% ^18O-enriched COCO2 gas mixtures to control both the oxygen fugacity and the isotopic composition of the exchange reservoir. Diffusion profiles of the ^18O tracer were obtained by in-depth analysis with an ion microprobe. The experimental results yield Arrhenius relations that appear here in the hard copy. At a given temperature, oxygen diffuses about 100 times more slowly in diopside than indicated by previous bulk-exchange experiments [1]. Our data for anorthite, spinel, and akermanite agree well with prior results obtained by gas-solid isotopic exchange and depth profiling methods [2-4]. Since these other experiments were conducted at different oxygen fugacities, this agreement indicates that diffusion of oxygen in these nominally iron-free minerals is not greatly affected by fO2 in the range between pure oxygen and the iron-wustite buffer. The oxygen diffusion data are used to evaluate the effects of three different types of therrnal histories upon the oxygen isotopic compositions of minerals found in Type B calciumaluminum-rich inclusions (CAIBs): (1) gas-solid exchange during isothermal heating, (2) gassolid exchange due to instantaneous heating followed by cooling at different rates, and (3) isotopic exchange with a gaseous reservoir during partial melting and recrystallization. With the assumptions that the mineral compositions within a CAIB were uniformly enriched in ^16O prior to any thermal processing, that effective diffusion dimensions may be estimated from observed grain sizes, and that diffusion in diopside is similar to that in fassaite, all the above scenarios fail to reproduce either the relative oxygen isotopic anomalies observed in CAIBs and/or yield improbably long or unrealistically intense thermal histories relative to both current theoretical models of nebular evolution and inferences from other isotopic systems. The failure of these simple models, coupled with recent observations of "disturbed" Mg isotopic abundances and petrographic features in anorthite and melilite indicative of alteration and recrystallization [5,6], suggests that the oxygen isotopic compositions of these phases may have also been modified by alteration and recrystallization during multiple melting events. Because the modal abundance of spinel remains relatively constant for plausible melting scenarios and its relatively sluggish diffusion kinetics prevent substantial equilibration, Mg-Al spinel is a reliable indicator of the oxygen isotopic composition of precursor material that formed CAIBs. References: [1] Connolly C. and Muehlenbachs K. (1988) GCA, 52, 1585-1592. [2] Elphick S. C. et al. (1988) Contrib. Mineral. Petrol., 100, 490-495. [3] Reddy K. P. and Cooper A. R. (1981) J. Am. Ceram. Soc., 64, 368-371. [4] Yunmoto H. et al. (1989) GCA, 53, 2387-2394. [5] Podosek F. A. et al. (1991) GCA, 55, 1083-1110. [6] MacPherson G. J. and Davis A. M. (1993) GCA, 57, 231-243. Sahijpal S. Ivanova M. A. Goswami J. N. Isotopic Studies of Cr-rich Objects in the Raguli (H3.8) Chondrite Cr-rich objects (chromite-rich inclusions and chondrules) have been documented in several ordinary chondrites in recent years [1]. They are most abundant in H chondrites and the dominant phases in these objects are chromite and sodic plagioclase with ilmenite, pyroxene, and phosphate occurring as accessory phases. The genesis of these objects and their interrelationship are not clearly understood as yet. Condensation from nebular gas of nonsolar composition, gas-phase metasomatism, and oxidation of metal phases are some of the proposed mechanisms for the formation of the chromite phases found in these objects [2]. We have carried out ion microprobe studies of isotopic compositions of magnesium, chromium, and iron in a set of Cr-rich objects in the Raguli (H3.8) chondrite to further address these questions. The measurements were carried out at appropriate mass resolution to resolve hydride and other isobaric interferences. The analyzed phases include two chromite-rich chondrules, one of which contains two large euhedral grains of chromite, two chromite-rich inclusions, and isolated chromite grains in matrix. The magnesium isotopic compositions of plagioclase phases in these objects were measured to look for the possible presence of excess 26Mg. The contribution from chromite toward the magnesium signal made these measurements difficult and a relatively clean signal could be seen only for the plagioclase phase in one of the chondrules (measured 27Al+/24Mg+ = 52). No evidence for excess 26Mg was found. Magnesium (24,25,26), chromium (52,53), and iron (56,57) isotopic compositions of the chromite phases in all the objects were measured to determine isotopic mass fractionation. Terrestrial chromite (USNM 117075) was used as a standard. The measured magnesium isotopic mass fractionation for the chromite phases in the inclusions and in the chondrules are similar and are also close to the measured value (-21.0 +- 0.72 permil/amu) for the terrestrial chromite. The data for the isolated chromite grain in matrix are suggestive of a small intrinsic fractionation (a few per mil per amu) favoring the lighter isotope. The measured chromium isotope mass fractionation for all the chromite phases, including the matrix grain, are similar and again these values are also close to the measured value for the terrestrial standard (-10.3 +- 1.43 permil/amu) There is no hint in the data for an intrinsic mass fractionation favoring either the lighter or heavier isotopes of magnesium and chromium for the chromite phases in both the Cr-rich chondrules and inclusions. The above results suggest that the precursor material from which the Cr-rich objects were formed had nearly unfractionated magnesium and chromium isotopic compositions and also the process(es) leading to the formation of these objects did not result in any detectable isotopic fractionation in the chromite phases. This would argue against the suggestion that the chromite phases in these objects could be of condensation origin. On the other hand, the isotopic data are not incompatible with the suggestion that incomplete melting of chromite-rich inclusions followed by rapid crystallization led to the formation of the chromite-rich chondrules. References: [1] Krot A. N. and Ivanova M. A. (1992) LPSC XXIII, 729-730. [2] Krot A. N. et al. (1992) LPSC XXIII, 731-732. Schutt J. Fessler B. Moore A. Benton M. The Antarctic Meteorite Location and Mapping Project (AMLAMP) Geographic Information System (GIS) The Antarctic Meteorite Location and Mapping Project (AMLAMP) was developed to document and provide location data on meteorites recovered from Antarctica by the Antarctic Search for Meteorites project (ANSMET). Since 1976 approximately 6000 specimens have been recovered from numerous localities along the Transantarctic Mountains by ANSMET expeditions. The geographic locations of most of these specimens have been documented. The AMLAMP maintains databases of this information at the Lunar and Planetary Institute and produces meteorite location maps of significant meteorite standing sites. Recent computer hardware and software advances allow the AMLAMP data to be easily accessed and used in many different ways. The increasing use and availability of geographic information systems (GIS) has led us to begin development of a GIS application of the meteorite location data. GIS is ultimately the ideal environment for the AMLAMP data because meteorite locations are spatial data that have attributes (i.e., weight, classification, etc.). These data can then be analyzed in a spatial context. AMLAMP data have been ported to the ARC/INFO GIS platform and a demonstration application developed. This poster session is designed to exhibit the AMLAMP GIS application and other AMLAMP products and services, and provide users an opportunity to make suggestions and recommendations. The ARC/INFO GIS platform is made available through the Office of Polar Programs of the National Science Foundation . Smith T. R. Hodge P. Microscopic Meteoritic Material Surrounding Meteorite Craters Meteoritic impact-related particles around meteorite craters can have several forms: (1) ablation spherules formed from the melt layer during atmospheric entry; (2) fragments of meteoritic metal formed by the shattering of the meteorite on impact; (3) fragments of metal oxide with meteoritic Fe/Ni ratios; (4) glassy spherules made up of a mixture of target rock and meteoritic material, formed by condensation of impact vapor; and (5) fragments of vesicular material formed from the impact melt. We are investigating the nature of the particles collected from soil surrounding the following craters: Odessa (Texas), Kaalijarvi (Estonia), Boxhole, Dalgaranga, Henbury, Snelling, Veevers, and Wolfe Creek (all Australia). No impact-related particles have been identified in the Veevers or Snelling samples. The Odessa samples include both meteoritic fragments (type 3) and Fe/Ni spherules (type 1). The Henbury samples include particles of type 4 [1] and type 2. The Boxhole samples include particles of types 1 and 4 [2]. The Kaalijarvi particles, being studied cooperatively with Reet Tiimaa of the Institute of Gelogy of the Estonian Academy of Sciences, include particles of type 3 and 5. The type 3 particles from Kaalijarvi are primarily kamacite, with small amounts of taenite. They have oxidized, Ni-free surface layers, probably formed by weathering. The vesicular particles are primarily made of glass that has a bulk composition that indicates that they are about half meteorite and half target rock material. The glass suggests partial recrystallization, with dendritic patterns of slightly different composition. Inclusions of quartz grains also occur and the outer layer in some cases is pure iron oxide. Many of the bubbles have their inner walls laced with patterns of iron condensate, often dendritic and in some cases in the form of stars. References: [1] Hodge P. W. and Wright F. W. (1971) JGR, 76, 3880-3895. [2] Hodge P. W. and Wright F. W. (1973) Meteoritics, 8, 315-320. Socki R. A. Romanek C. S. Gibson E. K. Jr. D/H Exchange Reactions in Salts Extracted from LEW 85320 Understanding the effects of terrestrial weathering on meteorites has been shown to be critical in distinguishing primary chemical and isotopic features from secondary alterations [1]. To further constrain weathering effects we report here the D/H composition of water thermally extracted from three distinct generations of efflorescence (,98, ,99, and ,102) occurring on the Antarctic H-5 chondrite LEW85320. To better understand the hydrogen isotope exchange systematics of these precipitates, an experiment was performed to characterize the rate of isotope exchange between a synthetic analog to the predominant weathering product, nesquehonite (Mg(HCO3)(OH).2H2O), found on the exterior of LEW85320 [2], and water. Synthetic nesquehonite, produced following the procedure of Ming and Franklin [3], a dehydrated CaSO4 standard, and deuterium-spiked water (deltaD = +701 permil SMOW) were placed together in a closed box and allowed to exchange hydrogen isotopes at constant temperature and humidity (30 degrees +- 2 degrees C and 75% +- 5%). Samples of each solid phase were taken initially and at 1, 3, 20, and 30 days. These samples along with three generations of efflorescence on LEW85320 (,98, ,99, and ,102) were weighed and loaded into separate high-purity, prebaked, 9-mm (O.D) quartz tubes. After degassing for two hours under high vacuum, samples were heated to 625 degrees C for 4 hr while all condensable gases were collected in a trap immersed in liquid nitrogen. CO2 was separated from water by exchanging the LN2 trap with a dry ice/alcohol mixture. All evolved water was frozen into a tube containing Zn turnings, which was then heated to 450 degrees C for 30 min, producing hydrogen gas for isotopic analysis. Results of our exchange experiment show that the CaSO4 standard quickly assumes the deltaD composition of the water (from -29 permil to +581 permil in 30 days). On the other hand, nesquehonite becomes only slightly enriched in deltaD (from -29 permil to +51 permil). Mass balance calculations reveal that absorption of the spiked water is stoichiometric with respect to the formation of CaSO4.2H2O, while within limits of sampling error no net change of weight was observed for the nesquehonite. Assuming that the change in deltaDnesq. is due entirely to exchange (i.e., no absorption), mass balance constraints dictate that less than 5 wt% of water exchanged. These data suggest that nesquehonite retains its original deltaD composition even under conditions of relatively high temperature and humidity. Hydrogen isotope data of water extracted from three generations of nesquehonite on LEW85320 are plotted as a function of the theoretical delta18O composition of water in equilibrium with the carbonate at 0 degrees C (where delta18Onesq. is derived by phosphoric acid digestion of the carbonate, assuming a calcite-CO2 fractionation factor of 1.01012). Our data plot very near the meteoric water line indicating formation from slightly enriched Antarctic meltwater. Water extracted from generations II (,99), salts consisting mostly of hydromagnesite (Mg5(CO3)4(OH)2.4H2O) (Gooding, 1993, personal communication), and III (,102), with mineralogy as yet unknown, is enriched in D (deltaD = -55 and -75 permil, respectively) and plot above the meteoric water line. Both generations precipitated in the Houston curatorial facility. Data suggest either that hydrogen isotopes have exchanged at least partially with local (i.e., Houston) water, or that the exchange reactions differ between strucural sites within or among the various generations of efflorescent salts. Hydrogen isotopes extracted from hydrous weathering products can reveal information about the environment of crystal growth. However, hydrogen isotope exchange systematics could be complicated if water within the crystal structure of the mineral is located in multiple sites. Furthermore, these results could have profound implications for curation and long-term storage strategies in curatorial facilities. References: [1] Socki R. A. et al., (1991) Meteoritics, 26, 396-397. [2] Gooding J. L. et al., (1988) LPSC XIX, 397-398. [3] Ming D. W. and Franklin W. T. (1985) Soil Sci. Soc. Am. J., 49, 1303-1308. Spettel B. Palme H. Dreibus G. Wanke H. New Analyses of CI Chondrites: Refinement of Solar System Abundances We have compiled our instrumental neutron activation data of CI chondrites and performed several new analyses. We find a surprisingly good agreement of our results with the CI data of Kallemeyn and Wasson [1]. In Table 1, which appears in the hard copy, averages of CI analyses done in Mainz and at UCLA [1] are listed. The standard deviations are, in most cases, calculated from variations of analytical results of individual analyses. Since both sets of data are completely independent we believe that the agreement justifies a high degree of confidence in the accuracy of the data. We therefore suggest that the mean of the two datasets represents the best available estimate for solar system abundances of the elements listed in Table 1, except for the REE. For the elements Mg, Al, and Ca there are more accurate analytical procedures than INAA. However, the INAA-results are in good agreement with wet chemical data and XRF analyses. A very small internal spread in both datasets is obvious for Sc, Cr, Mn, Fe, and Co. In addition, the agreement between the two datasets is better than +- 3% (except Mn, with -3.1%). The average solar system ratios among these five elements are known to within at least 2%. The Anders and Grevesse [2] compilation has within a 3% limit the same abundances for these elements except for Fe, which is 4.4% higher for average CI chondrites, but comes closer to the INAA data when only Orgueil is considered. More serious differences between INAA data and the A&G compilation are found for Se (12.7%), Au (7.9%), and Ir (4.7%). The internal spread in Se data from Mainz and UCLA is very small and the agreement is better than 1%. Recent data on S and Se on the same samples in chondrites have shown a constant S/Se weight ratio of 2518 (+-6%) in CI and CM chondrites [3] and an average CI content of 5.41% S in Orgeuil and Ivuna, in agreement with older Orgueil data of 5.25%. The very constant S/Se ratio in all chondrites and the CI content of 21.3 ppm Se lead to CI sulfur content of 5.36%, which is considerably below the A&G CI average of 6.25%. A very special case is Hg, which can be easily determined by INAA at the expected abundance level in CI chondrites. All our Orgueil samples appear to be more or less contaminated with Hg. A single Ivuna analysis gave a value of 0.31 ppm. This is basically in agreement with s-process calculations, which predict for 198Hg an abundance of 0.38 ppm [4]. The good agreement of solar and meteoritic abundances justifies a more sophisticated treatment of CI abundances. Small differences in element ratios between CI and other chondrite groups may require different nebular processing of these components and thus provide information on early solar nebula conditions. The INAA data on CI also demonstrate the basic homogeneity of CI chondrites for compatible elements. References: [1] Kallemeyn G. W. and Wasson J. T. (1981) GCA, 45, 1217-1230. [2] Anders E. and Grevesse N. (1989) GCA, 53, 197-214. [3] Dreibus G. et al. (1993), this volume. [4] Palme H. and Beer H. (1993) Landolt Brnstein, in press. Spilde M. N. Papike J. J. Use of Major- and Minor-Element Mapping to Measure Chemical Variability in Diogenite Pyroxenes Diogenite orthopyroxene grains have been shown to exhibit chemical variability within individual meteorite samples, e.g., the population groups reported by Hewins [1]. Our previous work [2] has shown a great deal of inter- and intragrain variability in OPX. The Garland diogenite, for example, appears to have two distinct populations of OPX, based on Cr/Al ratios. However, within individual crystals, excursions of Cr/Al ratios are present that may span a wide range within each population group. We are presently conducting SIMS analysis of OPX in diogenites and, therefore, the chemical variability of analyzed pyroxenes must be completely determined in order to find that portion of the crystal that most accurately records the igneous, rather than the metamorphic history. Optical analysis alone is not sufficient to categorize the pyroxene crystals. For example, numerous grains in EET 83246 appear to be zoned, with changes in interference colors toward the rims. However, EDS mapping indicates that the rims are chemically similar to the cores; the interference colors are interpreted to be due to some sort of grain abrasion process that occurred during brecciation. Backscattered image mapping at low magnification can identify intergrain variations and gross intragrain chemical variations, but this must be followed up with more detailed elemental mapping. Our microprobe employs a Link (Oxford) eXL II analyzer with full-stage automation, so that we can combine EDS maps of major elements with WDS maps of minor elements (Al, Ti, Mn, Cr). Large area maps (>1 mm) are produced using stage rastering to avoid defocusing of the WDS spectrometers. In a final step, quantitative elemental maps of target grains are produced, whereby the characteristic X-ray intensity collected at each map pixel is background suppressed and fitted against a standard intensity to yield an apparent concentration. Appropriate standards, close to the composition of the mapped grain, are used such that the apparent concentration is close to the real (ZAF-corrected) concentration. A number of grains have been mapped in Roda, Garland, EET 83246 and LEW 88008. Except for Garland, OPX grains are relatively homogeneous. Some large Roda grains show very fine (100) augite lamellae along with small augite blebs and scattered spinel along relict grain boundaries. Orthopyroxenes in Garland exhibit rare zoning and thin augite lamellae. Both symmetrical and asymmetrical zoning are evident in elemental maps; an example of nearly symmetrical zoning is shown in the Fe and Mg quantitative maps in Fig. 1, which appears in the hard copy (note a companion abstract [3] that illustrates a microprobe traverse across this grain). In addition, some grains show evidence of partial relict rims, indicating that the grains are brecciated fragments of larger, zoned grains. These brecciated pieces may explain some, but not all, of the intragrain variability seen in Garland. Because Cr decreases toward the rim while Al remains relatively constant, those fragments from the grain core will have higher Cr/Al ratios than fragments from the rim. Acknowledgment: This research was supported by NASA Grant NAGW-3347 and the Institute of Meteoritics. References: [1] Hewins R. (1980) LPSC XI, 441-443. [2] Papike J. J. et al. (1993) LPSC XXIV, 1109-110. [3] Papike J. J. and Spilde M. N. (1993), this volume. Wasson J. T. The Role of Body Crystallization in Asteroidal Cores Large fractionations (factors of 2000-6000) in Ir/Ni and other ratios demonstrate that the magmatic groups of iron meteorites formed by fractional crystallization, and thus that the residual liquid remained well stirred during core crystallization. Past models have relied on solidification at the base or the top of the core, but body crystallization offers an attractive alternative. The simplest of the earlier models involved convective maxing induced by the liberation of heat and light elements (especially S) during upward crystallization from the center of the core. Other models involving downward crystallization from the core-mantle interface are based on the fact that temperatures at this location are slightly lower than those at the center; no whole-core stirring mechanism is provided by these models. Haack and Scott recently published a variant of the downward crystallization model involving the growth of giant (kilometer-scale) dendrites. Because crystallization creates a boundary layer enriched in S that does not participate in the convection, these models require several K of supercooling to induce crystallization (this undercooling is much greater than the temperature difference between the center of the core and the core-mantle interface). Buoyant forces will occasionally remove droplets of the basal boundary fluid; thus it was thinner and its degree of undercooling less than in that at the ceiling of the magma chamber. Homogeneous nucleation of metals is difficult to achieve; generally 200-300 K of undercooling is required, much more than could possibly occur in an asteroidal core. Crystals could, however, nucleate in the magma body on chromite, probably the first liquidus phase (A. Kracher, personal communication, notes that this is required to explain why Cr behaved like a compatible element despite having a solid/liquid D < 1). In addition, some tiny, submillimeter dendrites that formed at the top of the core must have pinched off and fallen into the magma. Such seeds settle as a result of buoyant forces (thus stirring the magma) and, as a result, achieve very thin boundary layers and require low degrees of undercooling in order to crystallize. The rate of core crystallization is limited by the rate of heat transport across the core-mantle interface. If sufficient nuclei are available at different sites, the bulk of the crystallization occurs where undercooling is least. It is possible that a larger fraction of the total crystallization occurred in the body of the magma than at its base or ceiling. Weber D. Ross C. R. II Bischoff A. X-Ray Data on Extraterrestrial Ca Dialuminate (CaAl4O7) After the first discovery of Ca-dialuminate (CaAl4O7) in Allende [1], in recent years this phase has been found in several carbonaceous chondrites. Ca- dialuminate is a major phase in Ca,Al-rich inclusions from ALH85085 (e.g., [2]) and a dominating phase in CAIs from Acfer 182 ([3,4]). X-ray data on Ca-dialuminate are known from synthetic (e.g., [5-8]; cell constants) and terrestrial CaAl4O7 ([9]; only d-spacings), but are not available from extraterrestrial Ca-dialuminate. We report here the results of the first X-ray study of extraterrestrial Ca- dialuminate. The data (Table 1) were obtained by microdiffraction using a Rigaku PSPC microdiffractometer at the Bayerisches Geoinstitut. Ni-filtered Cr radiation was used with a direct beam diameter of about 50 micrometers. This powder diffraction method allows in situ measurement of polycrystalline Ca- dialuminate in a thin section. The CaAl4O7-rich inclusion 022/9 described in [4], consisting of a ~200-micrometer-sized core of Ca-dialuminate surrounded by layers of melilite and Ca-pyroxene, was chosen for analysis. The polycrystalline core contains only a small number of tiny inclusions (especially perovskite) and is therefore an excellent candidate for an X-ray study. For determination of the d-spacings of Ca-dialuminate an external standard (Ag6Ge10P12) was used for detector calibration. A large number of reflections could be indexed based upon comparison with the X-ray pattern of synthetic CaAl4O7 available in the JCPDS compilation [7]. The comparison was simplified because of the high purity of CaAl4O7 in inclusion 022/9 [4], and suggests the same structure for synthetic and extraterrestrial Ca-dialuminate. For determination of lattice parameters (cell constants, cell volume) refinement calculations were made based on 14 reflections (Table 1). The data for extraterrestrial CaAl4O7 shown in Table 1 indicate a close similarity to those obtained for synthetic CaAl4O7. The cell constants a, b, and therefore the cell volume, are slightly higher in Ca-dialuminate from Acfer 182 than from synthetic CaAl4O7. This may be due to the incorporation of traces of refractory elements (REE) with large ionic radii, which were analyzed within inclusion 022/9 [10]. With the determination of the cell constants of natural Ca-dialuminate combined with data on synthetic CaAl4O7, sufficient X-ray data should be available required to nominate this mineral. References: [1] Christophe Michel-Levy M. et al. (1982) EPSL, 61, 13-22. [2] Kimura M. et al. (1993) GCA, in press. [3] Weber D. and Bischoff A. (1992) Meteoritics, 27, 304-305. [4] Weber D. and Bischoff A. (1993) GCA, submitted. [5] Boyko E. R. and Wisnyi L. G. (1958) Acta Cryst., 11, 444-445. [6] Goodwin D. W. and Lindop A. J. (1970) Acta Cryst., B26, 1230-1235. [7] Baldock P. J. et al. (1970) J. Appl. Cryst., 3, 188-191. [8] Geiger C. A. et al. (1988) GCA, 52, 1729-1736. [9] Gross S. (1977) Geol. Surv. Israel Bull. 70, 1-80. [10] Bischoff A. et al. (1992) Meteoritics, 27, 204. Table 1, which appears in the hard copy, shows unit-cell constants of Ca- dialuminate (monocline; space group C2/c) and X-ray powder diffraction data (CrK-alpha (Ni-beta), 45 kV, 30 mA) on extraterrestrial CaAl4O7 in comparison to JCPDS data [7]. Numbers in parentheses are uncertainties in last significant figures. Wlotzka F. A Weathering Scale for the Ordinary Chondrites Weathering categories A, B, and C are used by the Meteorite Working Group at the NASA Johnson Space Center in Houston for Antarctic meteorite finds, denoting minor, moderate, and severe rustiness of hand specimens. A different scale can be set up from the weathering effects seen in polished sections with the microscope. These weathering effects finally lead to the disintegration of the meteorite; they are important in connection with its terrestrial age and an estimate of the true fall rate of meteorites. In order to avoid confusion with the hand specimen classification A, B, C, the weathering grades determined on polished sections were named W1 to W6. Weathering affects first the metal grains, later troilite, and finally the silicates. The following progressive stages can be distinguished: W0: No visible oxidation of metal or sulfide. A limonitic staining may already be noticeable in transmitted light. Fresh falls are usually of this grade, although some are already W1. W1: Minor oxide rims around metal and troilite, minor oxide veins. W2: Moderate oxidation of metal, about 20-60% being affected. W3: Heavy oxidation of metal and troilite, 60-95% being replaced. W4: Complete (>95%) oxidation of metal and troilite, but no alteration of silicates. W5: Beginning alteration of mafic silicates, mainly along cracks. W6: Massive replacement of silicates by clay minerals and oxides. More or less massive veining with iron oxides can already be found in stage W2. These veins develop independently from the weathering grade, apparently in cracks that form through mechanical forces. Broad cracks are often filled with carbonates. Grades W5 and W6 are rare. The silicate alteration affects first the olivines; it starts inside the grains, not from the rim. In stage W6 intact chondrules were found, where olivines were completely replaced by a mixture of clay minerals and iron oxides, the feldspathic mesostasis being unaffected. A correlation between these weathering grades and the terrestrial ages was shown for meteorite finds from Roosevelt County, New Mexico [1]. In these climatic conditions the weathering grades W2 to W6 develop in the following times: W2, 5000 to 15,000 yr; W3, 15,000 to 30,000 yr; W4, 20,000 to 35,000 yr; W5 and W6, 30,000 to >45,000 yr. Similar terrestrial ages were found for chondrites of these weathering grades from the Lybian and Algerian Sahara [2,3]. Antarctic meteorite finds weather much more slowly. A check of 53 Antarctic ordinary chondrites (of hand specimen weathering categories A to C) showed only 9 of grade W2, the rest being W1. Among the W1s is ALHA77278 (category A) with a terrestrial age of 320,000 yr [4]. References: [1] Jull A. J. T. et al. (1991) LPSC XXII, 665. [2] Jull A. J. T. et al. (1990) GCA, 54, 2895. [3] Jull A. J. T. et al. (1993), this volume. [4] Nishiizumi K. et al. (1989) EPSL, 93, 299. Yang S. V. Zolensky M. Golden D. C. Ming D. W. Ivanov A. Phyllosilicates in the Carbonaceous Chondrite Breccia Kaidun Kaidun appears to predominantly be a CR chondrite, containing other diverse components, including enstatite chondrites. Previous observations indicate that the dominant phyllosilicates in Kaidun are serpentine and saponite, suggesting that the Kaidun parent body has undergone aqueous alteration [1]. Phyllosilicates in the smectite group are important in that they act as ion exchangers, which can retain alkalimetal, alkaline earth, or ammonium ions in their interlayers in exchangeable form while their structure may contain hydroxyl groups derived from the aqueous alteration process. The purpose of this investigation was to study the layer charge of these smectites and to make an attempt to understand the interlayer chemistry, which was the result of rock-water interaction in its parent body. An alkylammonium method coupled with high-resolution transmission electron microscope (HRTEM) was used to study layer charge and the electron microprobe was used to study the composition and the interlayer chemistry of phyllosilicates. Ultramicrotomed Kaidun matrix samples (on TEEM grids) were pretreated with C12-alkylammonium solutions [2]. This procedure was performed to expand and stabilize the smectite (e.g., saponite) for HRTEM study and permit characterization of the relative charge density of the interlayer sites. This latter feature is a potentially important indicator of the environment of the Kaidun parent body (probably a hydrous asteroid). Recent work by Ming et al. [2] shows that the basal lattice fringes of C12-alkylammonium treated saponites in Kaidun meteorite are typically 1.3-1.4 nm, which indicates low-charged interlayer sites. However, in this study saponites with much larger layer spacing (1.3-2.6 nm, mostly 2 nm) were observed. suggesting the presence of high-charge interlayer sites. Another distinct feature observed in this study is that saponite is clearly the dominant phyllosilicate phase in some Kaidun matrix lithologies, with serpentine being rarely observed. In contrast, most reported Kaidun and CR lithologies have approximately subequal amounts of saponite and serpentine in matrix. Phyllosilicates in Kaidun are commonly associated with sulfides; no phyllosilicates have been observed as direct overgrowths on olivine or pyroxene. Microprobe analyses of coarse-grained Kaidun saponites indicate that the majority of the exchangeable cations in the saponites studied are Mg2+ and Ca2+, with mior Na+. However, since the results of this study suggest that the saponite in Kaidun has a highly charged interlayer environment, one might speculate that any ammonium (NH4+) if present in the original parent body atmosphere or the reacting solution might be fixed in the interlayers. High- charge smectites are known to fix ammonium ions from solution [3]. There is spectroscopic evidence for ammonium-bearing phases on asteroid Ceres 1 [4]. Most carbonaceous chondrites are known to contain relatively high amounts of nitrogen (up to 3000 ppm) [5]. In order to detect if any of this N is in NH4+ form in the interlayers, we set up our Cameca electron microprobe to detect the nitrogen K-alpha X-ray peak using an ODPB crystal of a wavelength dispersive spectrometer. No nitrogen peak was positively identified on the carbonaceous matrix, nor on any saponites, although it is possible that the electron beam neutralized and evaporated any NH4+ cations before detection. In conclusion, the phyllosilicates in Kaidun are heterogeneously distributed from clast to clast, with highly charged saponite predominating in some clasts; serpentine and saponite are more nearly equally abundant in other clasts. No nitrogen was positively detected in the matrix or in any components in Kaidun by the electron microprobe in this study, although further studies of Kaidun phyllosilicates are in progress. References: [1] Zolensky M. and McSween H. Y. Jr. (1988) in Meteorites and the Early Solar System, Univ. of Arizona, 114-143. [2] Ming D. W. et. al. (1992) LPSC XXIII, 913-914. [3] Krishnamoorthy C. and Overstreet R. (1950) Soil Sci., 69, 41-53. [4] King T. V. V. et al. (1992) Science, 255, 1551-1553. [5] Handbook of Geochemistry, 7-C-5. Xue S. Herzog G. F. Hall G. S. Stable Nickel Isotopes in Fusion Crusts from Iron Meteorites and from Metallic Particles in a Black Wabar Impact Glass Iron and nickel isotopes may undergo mass fractionation in systems subjected to high-temperature vaporization [1-3]. We report here a search for nickel fractionation in fusion crusts from iron meteorites and in metal-rich material separated from Wabar impact glasses. Fusion-crust bearing samples of Bogou (IA), N'Goureyma (I-an), and Pitts (IB) were potted in epoxy and were "shaved" with a milling machine. Microscopic examination of the shavings showed the presence of some material from the interior of the meteorites as well as from the fusion crust. A fourth meteorite, Cape of Good Hope (IVB), was prepared for use as a reference standard. About 1.4 mg of magnetic material was collected from a 2-g sample of black Wabar impact glass ground in a Spex mill; microscopic examination indicated that adhering silicates comprised ~5% of the sample. These (terrestrial) silicates contain relatively little Ni [4] so their presence does not interfere with the nickel analysis. Nickel was separated from all samples and its isotopic composition determined as in [2]. Results and Discussion: Nickel isotopic abundances are given in Table 1 both as delta values and as an average fractionation, PHI, where PHI is the slope of a plot of delta vs. mass for each sample. Within the precision of our measurements (from 0.3 to 1.5%, depending on the isotope) all the samples had normal (i.e., terrestrial) isotopic abundances of Ni. Clayton et al. [5] reported that delta-18O in fusion crust is lower than in the atmosphere, probably as a result of a kinetic isotope effect, while in metallic deep-sea spheres, heavy oxygen isotopes are enriched. They inferred that the metallic spheres are not the ablation products of larger meteorites. Similarly, the Ni isotopic abundances in fusion crust are normal, while those in deep-sea metallic spheres are enriched in the heavier isotopes [1]. We note, however, that material ablated from the surface of an iron could have undergone fractionation after separation from the incoming meteorite (see [4]). Horz et al. [6] found variable Fe/Ni ratios (from 0.1 to 222) in black melt glasses associated with the Wabar impact. The Fe/Ni ratio in our metal sample is 2, which is considerably lower than that in the bulk meteorite (~12.4). Several lines of evidence suggest that vapor fractionation is to be expected in samples that have Fe/Ni ratios greater than those in the bulk impactor [2-6]. Thus it is not surprising that our first results for Wabar impactites show no Ni isotopic fractionation. Isotopic analyses of Wabar impactites with high Fe/Ni ratios should be made to test the importance of vapor fractionation. References: [1] Herzog G. F. et al. (1992) LPSC XXIII, 527-528. [2] Xue S. et al. (1993) LPSC XXIV, 1547-1548. [3] Davis A. et al. (1993) LPSC XXIV, 373- 374. [4] Mittlefehldt D. W. et al. (1992) Meteoritics, 27, 361-370. [5] Clayton R. N. et al. (1986) EPSL, 79, 235-240. [6] Horz F. et al. (1989) Proc. LPSC 19th, 697-710. Table 1, which appears in the hard copy, shows delta (permil) and average isotope fractionation PHI (%/amu) for Ni isotopes in iron meteorites and black Wabar impact glass. Tuesday, July 20, 1993 Cosmogenic Nuclides 8:15 a.m. Theater Chair(s): O. Eugster T. Swindle Masarik J.* Reedy R. C. Production Profiles of Nuclides by Galactic-Cosmic-Ray Particles in the Tops of Lunar Rocks The determination of the flux and spectral shape of solar-cosmic-ray (SCR) particles from measurements of cosmogenic nuclides in lunar samples requires many pieces of information. An important part of the analysis is the correction of the measured nuclide concentration for the contribution by galactic-cosmic-ray (GCR) particles in the surface layers, where SCR production is important. Usually the GCR contribution is inferred from the concentration of the nuclide measured at a depth where SCR production is negligible and using a GCR production profile to extrapolate back to the surface. The shapes of these GCR profiles were found to be important for neon and argon [1] and ^10Be [2] in lunar rock 68815. Better determination of this near-surface GCR production profile would improve this important correction. As almost all nuclides have some production by SCR particles, it is hard to determine experimentally this GCR profile. The GCR production profile in the top of lunar rock 68815 was calculated using the Los Alamos Monte Carlo LAHET Code System (LCS). LCS has yielded calculated production rates that almost always are in good agreement with cosmogenic- nuclide measurements in meteorites [3,4]. The fluxes of protons and neutrons in rock 68815 was calculated for 1 g/cm^2 layers down to a depth of 25 g/cm^2 and with a coarser depth mesh to a depth of 500 g/cm^2 with LCS. These fluxes for each layer were then multiplied by the relevant cross sections and integrated over energy for eight nuclides: ^10Be, ^14C, ^21Ne, ^22Ne, ^26Al, ^36Cl, ^38Ar, and ^53Mn. The calculated production profiles below about 30 g/cm^2 agree well with lunar core measurements. All the cosmogenic nuclides studied have GCR production profiles that increase from the surface of the Moon to a maximum at depths of ~20-50 g/cm^2 and then decrease with increasing depth. The amount of this increase varies considerably, with products made by higher-energy particles and by protons having less increase in production rate with depth. The amount of the increase from the surface to depths of 10 or 20 g/cm^2 correlates with the depth of the peak production rate. The highest-energy product in this study, ^10Be, has a peak production rate near ~20 g/cm^2 and has ratios of the production rate at 0-1 g/cm^2 to those at 9-10 and 19-20 g/cm^2 of 0.965 and 0.940 respectively. With typical lunar saturation activities of ~12 dpm/kg, the surface has a ^10Be activity due only to GCR particles that is ~1 dpm/kg less than that near the production peak. The flat ^10Be activities measured in lunar samples [2] would thus imply a SCR contribution about equal to this small GCR difference, restricting the spectral shapes of solar protons that made ^10Be to those with few high-energy particles. For the composition of lunar rock 68815, the nuclides with the steepest GCR profiles were ^26Al and ^38Ar, with ratios of production rates at 0-1 g/cm^2 to those at 9-10 and 19-20 g/cm^2 of about 0.79 and 0.71 respectively, and with peak production rates near ~50 g/cm^2. The production ratios for these depths were, respectively, about 0.82 and 0.74 for ^53Mn and ^14C, about 0.84 and 0.76 for ^36Cl, and 0.88 and 0.84 for ^21Ne and ^22Ne. These calculations for the production of nuclides in the top layers of lunar rocks by GCR particles show that there are increases with depth down to ~20-50 g/cm^2 and that the amount of the increase varies with the nuclear reactions making the nuclide. Experimental confirmation of these calculated GCR production profiles should be made, possibly with 35-day ^37Ar in lunar rocks from a mission like Apollo 16 that had few solar particle events prior to it. This work was supported by NASA and done under the auspices of the U.S. Department of Energy. References: [1] Garrison D. G. et al. (1993) LPS XXIV, 521. [2] Nishiizumi K. et al. (1988) Proc. LSPC 18th, 79. [3] Masarik J. and Reedy R. C. (1993) LPS XXIV, 937. [4] Reedy R. C. et al. (1993) LPS XXIV, 1195. Gilabert E.* Lavielle B. Simonoff G. N. Rosel R. Herpers U. Schnatz-Buttgen M. Lupke M. Michel R. Production of Krypton and Xenon Isotopes by Galactic Protons The production of krypton from target elements Rb (Rb(sub)2SO(sub)4), Sr (SrF(sub)2), Y, Zr, and of xenon in Ba (Ba glass), La (LaF(sub)3) is studied in a simulation experiment of the galactic cosmic-ray proton bombardment of stony meteoroids in space [1,2]. This investigation is part of the experiment LNS 172 by which a 50-cm-diameter artificial meteoroid (gabbro) was isotropically irradiated at Saturne with 1.6 GeV protons. Measurements of krypton production vs. depth are now complete in the four investigated target elements. In the ^81Kr-^83Kr dating method, the production ratio P(sub)81/P(sub)83 can be evaluated from the cosmogenic spectrum of krypton in the meteorite according to the formula: P(sub)81/P(sub)83=0.95[(^80Kr/^83Kr)(sub)c+(^82Kr/^83Kr)(sub)c]/2 [3] where (^80Kr/^83Kr)(sub)c and (^82Kr/^83Kr)c represent the measured cosmogenic ratios assuming no contribution from (n,gamma) nuclear reactions on Br. Applying this formula to this experiment, a good agreement with the measured production ratio is obtained for Zr and Y targets. On the other hand, this formula overestimates the measured production ratio by 6% for Sr and 15% for Rb. Taking a mean composition of ordinary chondrites [4], the production ratio ^81Kr/^83Kr decreases from the surface to the center by 4% but the value calculated with the formula still exceeds the measured ratio by 7%. The ratio ^78Kr/^83Kr also shows a decrease by 10% from the surface to the center. Variation by 20% of the concentration of target elements can change this ratio by 10%, but, for the same variation, dependence on the target chemistry is less than 4% for ^81Kr/^83Kr. For Xe, depth profiles of production in Ba and La are reported. Production of ^126Xe shows a steep increase from the surface to center by a factor of 1.5 for Ba and of 2 for La. All the production ratios also increase from the surface to the center except ^124Xe/^126Xe, which is decreasing and ^136Xe/^126Xe, which is almost constant. This work was partially supported by C.N.R.S., by IN2P3, and by INSU (Programme National de Planetologie). References: [1] Michel R. et al (1991) Meteoritics, 26, 372a. [2] Gilabert et al. (1992) Meteoritics, 27, 223. [3] Marti K. (1967) Phys. Rev. Lett., 18, 264-266. [4] Wasson J. T. and Kallemeyn G. W. (1988) Phil. Trans. R. Soc., A325, 535-544. Sisterson J. M.* Jull A. J. T. Beverding A. Koehler A. M. Castaneda C. Vincent J. Donahue D. J. Englert P. A. J. Gans C. Young J. Reedy R. C. Cross-Section Measurements from 40 to 450 MeV for the Production of 14C from Silicon and Oxygen: Better Estimates for Cosmogenic Production Rates Cosmogenic nuclides in extraterrestrial materials allow studies to be made of the solar cosmic ray (SCR) flux over time periods in the past [1], the constancy of the galactic cosmic ray flux, and even of the sample's recent history. To interpret such measurements, especially for SCR-produced nuclides, it is essential that the cross sections for the reactions of all cosmic ray particles with each constituent of the sample be very well known. Approximately 98% of SCR particles are protons and their interactions are the major source of cosmogenic nuclides in the surface layers of extraterrestrial materials. Until the development of accelerator mass spectrometry (AMS), few of the needed cross sections were known well enough to be used with confidence. Now using thin target irradiations and the improved sensitivity of AMS, good cross-section measurements for the production of these cosmogenic radioisotopes can be made. Preliminary cross-section measurements for 16O(p,3p)14C and natSi(p,x)14C made at the Harvard Cyclotron Laboratory (HCL) using SiO2 and Si targets have been reported for the proton energy range 65-160 MeV [2]. They confirm earlier data for the 16O(p,3p)14C cross section and are the only measurements for the natSi(p,x)14C cross section. New measurements made at the cyclotron at the University of California at Davis for proton energies from 40 MeV to 67.5 MeV and at TRIUMF for proton energies from 200 to 450 MeV have extended the energy range over which these cross sections are well known, including the important region near the threshold of the excitation function. In all cases, targets of SiO2 and Si were irradiated in thin target conditions that kept the energy loss in a single target to <2 MeV. The total energy lost in the entire target stack for the Davis irradiations was <8 MeV; for the TRIUMF irradiations <1 MeV; and at HCL it ranged from <5 MeV at 160 MeV to <10 MeV at 65 MeV. Thus both the secondary neutron production within the target stack and loss of protons due to scattering were minimized. The proton fluence was determined using Faraday cups and the 27Al(p,3p3n)22Na reaction measured in aluminum monitor foils. All the samples were analyzed for 14C at Arizona using well known methods [3,4]. Details of these measurements and new values of the 14C production cross sections will be presented in the context of their importance to lunar sample and meteoritic studies. These measurements represent the first data available from Si, SiO2, Al, Mg, and C targets that have already been irradiated at some proton energies. These targets will be analyzed for 14C, 10Be, 26Al, 7Be, 22Na, and the noble gases, while additional relevant target materials will be irradiated in this ongoing systematic study. References: [1]) R. C. Reedy and K. Marti (1991) in The Sun in Time, 206, Univ. of Arizona. [2] J. M. Sisterson et al. (1992) LPSC XXIII, 1305. [3] A. J. T. Jull et al. (1989) GCA, 53, 2095. [4] T. W. Linick et al. (1986) Radiocarbon, 28, 522. Schiekel Th.* Rosel R. Herpers U. Bodemann R. Leya I. Gloris M. Michel R. Dittrich B. Kubik P. Suter M. Cross Sections for the Production of Cosmogenic Nuclides with Protons up to 400 MeV for the Interpretation of Cosmic-Ray-produced Nuclides Integral excitation functions of the cosmogenic nuclides are the basic requirement for the interpretation of interactions between cosmic ray particles and extraterrestrial and terrestrial matter. Together with the knowledge of primary and secondary particle fields inside an irradiated body, model calculations can be developed to interpret abundances of cosmogenic nuclides in dependencies of the irradiation history of the irradiated body and of the cosmic particle ray itself. The quality of those model calculations depends on the quality of the available cross-section database, which is neither comprehensive nor reliable for the most important nuclides like the long-lived radionuclides (i.e., 10Be, 26Al, 36Cl, 41Ca) and the stable rare gas isotopes. For a systematic investigation in this field of science we carried out several irradiation experiments with protons in the energy region between 45 MeV and 400 MeV at the Paul Scherrer Institut (Villigen, Switzerland) and the Laboratoire Nationale Saturne (Saclay, France) using the stacked foil technique. We included 21 different target elements with Z between 6 and 79 (C, N as Si3N4, O as SiO2, Mg, Al, Si, Ca as CaC2H2O4, Ti, V, Mn as Mn/Ni alloy, Fe, Co, Ni, Cu, Sr as SrF2, Y, Zr, Nb, Rh, Ba as Ba containing glass and Au) in our experiments. The proton fluxes were monitored via the reaction 27Al(p,3p3n)22Na using the evaluated data of [1]. Residual nuclides were measured by X-, gamma-, and after a chemical separation by accelerator mass spectrometry. In order to check the quality of our experimental procedures we included some target elements in our new experiments for which consistent excitation functions have already been determined [2,3,4]. Our new data show excellent agreement with the earlier measurements. We measured cross sections for more than 120 different reactions. Here we report on the results for target elements with Z up to 28. The exsisting database of experimental excitation functions for the production of some radionuclides relevant for cosmic ray interactions with extraterrestrial and terrestrial matter, i.e., 7Be, 10Be, 22Na, 26Al, and 36Cl is discussed in detail. By a comparison of our new cross-section database with theoretical calculations, the ability to predict unknown excitation functions on the basis of equilibrium and preequilibrium reactions (code ALICE) and an internuclear cascade evaporation model (code HETC) is analyzed. The new data are applied to model calculations for the production of long-lived radionuclides by solar cosmic rays in lunar surface materials and in meteorites. Acknowledgment: This work was supported by the Deutsche Forschungsgemeinschaft and by the Swiss National Sience Foundation. References: [1] Tobailem J. and de Lassus St. Genies C. H. (1981) CEA-N-1466. [2] Michel R. et al. (1984) JGR, 89, B673-B684. [3] Michel R. et al. (1985) Nucl. Phys., A441, 617-639. [4] Luepke M. et al. (1992) in Nuclear Data for Science and Technology (S. M. Qaim, ed.), 702, Springer Verlag. Toe S. Lavielle B.* Gilabert E. Simonoff G. N. Depth Profiles of Cosmogenic Noble Gases in the Chondrite Knyahinya Concentrations and isotopic ratios of Ne, Ar, Kr, and Xe have been analyzed in 5-g size samples from different positions within the L5 chondrite Knyahinya. A previous work [1] has shown that Knyahinya experienced a single-stage exposure history (duration 40.5 Ma) as a meteoroid of approximately spherical shape (radius 45 cm). For these reasons, this meteorite represents a very interesting object to study depth profiles of cosmogenic nuclide concentrations and to test and improve model calculations of production rates. The procedure of extraction of noble gases adopted for this work, includes two pyrolyses respectively at about 450 degrees C and 650 degrees C, followed by a combustion step in pure O2 (15-25 torr pressure) at 650 degrees C before the complete melting of the sample [2]. This procedure allows a low-temperature extraction of a significant fraction of the Kr and Xe trapped noble gas component, leading to an enrichment of the cosmogenic component during the last temperature step. Concentration of trapped Ar, Kr, and Xe is 2-3 times lower than expected for a type 5 chondrite. The isotopic composition of the trapped Xe component analyzed in the combustion step is identical with the OC- Xe composition measured in Forest Vale [3]. Preliminary results show that concentration of cosmogenic 83Kr increases by 16% from the surface to the center when the ratio of cosmogenic 78Kr to 83Kr decreases from 0.157 to 0.136. The concentration of 81Kr has been measured in each sample. It increases from 0.0220 10^-12 cm^3 STP/g near the surface to 0.0255 10^-12 cm^3 STP/g at the center, in excellent agreement with the variations measured by Eugster [4] in other ordinary chondrites. Acknowledgments: This work was supported by C.N.R.S., by IN2P3 and by INSU (Programme National de Planetologie). References: [1] Graf Th. et al. (1990) GCA, 54, 2511-2520. [2] Gilabert E. and Lavielle B. (1991) Meteoritics, 26, 337. [3] Lavielle B. and Marti K. (1992) JGR, 97, 20875-20881. [4] Eugster O. (1988) GCA, 52, 1649-1662. Graf Th.* Nishiizumi K. Finkel R. C. Caffee M. W. Southon J. Toe S. Lavielle B. Gilabert E. Simonoff G. N. New Model Parameters for the Production of Cosmic-Ray-produced Nuclides Derived from Measurements in Knyahinya A model for the production of cosmogenic nuclides in chondrites based on a production rate equation with two free parameters (A(sub)i, B(sub)i) for each nuclide i has been presented earlier [1]. The model parameters were determined by fitting measured depth profiles in the large Knyahinya chondrite [2]. New measurements of ^26Al, ^10Be, and ^36Cl in metal separates and Kr in bulk samples of Knyahinya have been carried out [3,4]. Here, we present model calculations for these newly measured nuclides. The A(sub)i and B(sub)i parameters for the production of ^10Be, ^36Cl from metal and ^78,81,83Kr in bulk samples were fitted to the measured depth profiles in Knyahinya and are given in Table 1. The model predictions for ^36Cl are in good agreement with the nearly constant activities observed for St. Severin [5], if corrections for the high Ni content in the metal of this meteorite are applied. However, ^78,83Kr profiles measured in St. Severin [6] show larger variations with shielding depth than predicted by the model. This may partly be due to the complex preatmospheric shape of St. Severin and to uncertainties arising from corrections for the low Rb content of this meteorite. The model predicts that any two production rate ratios are correlated in a way independent of the size and shape of a particular meteoroid. The calculated correlation line in a ^81Kr/^83Kr vs. ^22Ne/^21Ne diagram agrees well (+-3%) with the line obtained from data of many different chondrites [7]. On the other hand, the calculated correlation of ^78Kr/^83Kr vs. ^22Ne/^21Ne yields ^78Kr/^83Kr ratios that are 5-10% lower compared to those observed in samples from near surface locations in small meteorites (^22Ne/^21Ne>1.15) [7]. A(sub)i and B(sub)i parameters for the production of ^36Ar in metal (Table 1) can be derived from those of ^36Cl because >80% of ^36Ar in metal is produced indirectly via decay of ^36Cl and the ratio of the cross-sections for the production of ^36Cl and ^36Ar from Fe is quite constant. Parameters for ^38Ar in metal are obtained from those of ^36Ar by using the known shielding dependence of the ^36Ar/^38Ar ratios. A comparison of the calculated production rates of ^38Ar in bulk samples and metal separates indicates that the production rate ratio P(sub)38(Ca,K)/P(sub)38(Fe,Ni) increases by ~40% with increasing shielding (^22Ne/^21Ne ratios of 1.2 and 1.06, respectively). ^10Be activities in metal show a small decrease with increasing shielding depth except for one unusually low value in a near surface sample. The resulting value for B(sub)10 is 0.77. Model calculations based on ^10Be measurements in the Grant iron meteorite [8] yielded B(sub)10=0, indicating that ^10Be is produced solely by high energy particles. Therefore, the Grant model should be applicable to the Knyahinya data if the interaction mean free path for high energy particles is adjusted properly. The activities predicted in this way show a much larger decrease with increasing shielding (~50%) than observed. A possible solution to the problem may be that the preatmospheric radius of Grant was larger than the 40 cm derived in [9]. References: [1] Graf Th. et al. (1990) GCA, 54, 2521-2534. [2] Graf et al. (1990) GCA, 54, 2511-2520. [3] Reedy et al. (1993) LPSC XXIV, 1195-1196. [4] Toe et al. (1993) Meteoritics, this issue. [5] Nishiizumi et al. (1989) LPSC XIX, 305-312. [6] Lavielle B. and Marti K. (1988) LPSC XVIII, 565-572. [7] Eugster O. (1988) GCA, 52, 1649-1662. [8] Graf et al. (1987) Nucl. Instr. and Methods, B29, 262-265. [9] Signer P. and Nier A. O. (1960) JGR, 65, 2947-2964. Table 1, which appears here in the hard copy, shows model parameters. Reedy R. C.* Masarik J. Jull A. J. T. Donahue D. J. Wasson J. T. Studies of Cosmic-Ray-produced Carbon-14 in the Vaca Muerta Mesosiderite We have studied the production of 5730-yr ^14C in samples of the VM10 mass of the mesosiderite Vaca Muerta. This ~165-kg mass was buried with only one surface exposed at ground level. Because of heavy corrosion, it could only be excavated in pieces ranging in size up to 4 kg. Our samples ranged from the exposed surface to about 25 cm depth. These samples were analyzed for ^14C content at the University of Arizona by accelerator mass spectrometry, and their elemental compositions were measured. The results are compared to calculations of ^14C in Vaca Muerta produced by cosmic rays, and constraints on the recent exposure record of Vaca Muerta, such as terrestrial age and preatmospheric size, are obtained. Five of the samples were fairly pure silicate, and three other samples were about 40-50% silicate. The major target for ^14C, oxygen, is ~43% by weight in the silicate. The results of surface silicate inclusions give ^14C of 37-43 dpm/kg. Samples of mesosiderite, ~45% metal, are about 23 dpm/kg ^14C at the same surface. Production rates of ^14C were calculated using the Los Alamos Monte Carlo LAHET Code System (LCS). LCS has yielded calculated production rates for ^14C and other nuclides that are in good agreement with cosmogenic-nuclide measurements in meteorites [1,2]. Earlier calculations showed that the bulk composition of a meteorite affects production rates of cosmogenic nuclides [1]. In the calculations done for Vaca Muerta, spherical objects with radii of 50 and 70 cm, a density of 5 g/cm^3, and typical bulk mesosiderite compositions were irradiated with galactic-cosmic-ray protons. Layers 2.5 cm thick were used. The calculated fluxes for each layer were then multiplied by the relevant cross sections and integrated over energy to get ^14C production rates for pure silicate and for the other samples. The calculated production profiles increased from the surface to depths of ~20-30 cm. In the 50-cm object, the profile was fairly flat to the center, but the profile dropped slightly for depths >30 cm in the 70-cm object. Production rates for ^14C were about 60 atoms/(min kg) in pure silicate but only a few atoms/(min kg) (not well determined because of the lack of cross sections) in metal. The rates for a 50-cm radius were ~10% higher than for the 70-cm radius. The ^14C measurements and the calculated production rates are only consistent with a meteoroid of short terrestrial age, as practically none of the ^14C can have decayed. The ^14C must have been produced in a body not substantially larger than a radius of 70 cm and probably not more than ~100 cm. This indicates a lack of significant erosion of the meteoroid in the last 20,000 years. If the current sampled surface were actually at some depth in the preatmospheric meteoroid, then some terrestrial age of up to about 2500 years is possible using these data. However, data from deeper samples argue against any significant terrestrial age. With some additional studies of samples at additional depths and other cosmic-ray-produced isotopes, particularly ^36Cl and Ne isotopes to constrain shielding, we may be able to derive a better model for the preatmospheric irradiation history. Acknowledgments: This work was supported by NASA, and the work at Los Alamos was done under the auspices of the U.S. Department of Energy. References: [1] Masarik J. and Reedy R. C. (1993) LPS XXIV, 937. [2] Reedy R. C. et al. (1993) LPS XXIV, 1195. Herpers U.* Bremer K. Klas W. Michel R. Metzler K. Stoffler D. Dittrich-Hannen B. Kubik P. Suter M. 10Be and 26Al Concentrations in Bulk Material and Mineral Separates of Antarctic and Non-Antarctic Achondrites The investigation of radionuclide concentrations in bulk material and mineral separates of achondrites allow the study of the irradiation and collision history of meteoroids and their parent body as well as the determination of elemental production rates. Numerous Antarctic and non-Antarctic achondrites of types howardite, monomict, and polymict eucrite, and diogenite were mineralogically investigated. Bulk material and mineral separates, e.g., plagioclase, pyroxene, basalt, and impact- fusion-breccia were taken and the chemical composition of each sample was analyzed. In these specimen ^10Be- and ^26Al- concentrations were determined by means of radiochemical separation and accelerator mass spectrometry. The experimental data are interpreted by model calculations [1] based on spectra of primary and secondary particles derived by Monte Carlo techniques and on experimental and theoretical thin-target cross sections of the underlying nuclear reactions. The results will be presented and discussed. Acknowledgement: This work was partially supported by the Bundesminister fur Forschung und Technologie. Reference: [1] Michel R. et al. (1991) Meteoritics, 26, 221-242. Eugster O.* Michel Th. Cosmogenic Noble Gases and Their Production Rates in Eucrites, Diogenites, and Howardites: Common Asteroid Break-up Events 38 Ma, 21 Ma, and 6 Ma Ago It is likely that the eucrites and their associates, the howardites and diogenites, sample the surface and shallow interior of a single parent body, possibly 4 Vesta (cf. [1] and [2]). A break-up event that reaches deep enough may, thus, eject asteroidal fragments representing meteorites from all three classes. In this work we present a comprehensive investigation of the exposure age clusters for howardites, eucrites, and diogenites (HEDs). Cosmic-ray exposure ages critically depend on the production rates for cosmic-ray produced nuclei. For eucrites shielding independent production rates for ^21Ne and ^38Ar have been determined previously [3,4]. We now present production rates of ^3He, ^21Ne, ^33Ar, ^78Kr, ^83Kr, and ^126Xe for eucrites, howardites, and diogenites as a function of shielding, where appropriate, and of target element abundances as derived on the basis of ^81Kr-Kr ages. E.g., for ^21Ne we obtain: P(sub)21 (EUC) = 8.43 P^1(sub)21 [16.1 (^22Ne/^21Ne)(sub)c - 10.3]^-1, P(sub)21 (HOW) = 6.16 P^1(sub)21 [18.1 (^22Ne/^21Ne)(sub)c - 14.1]^-1, P(sub)21 (DIO) = 4.81 P^1(sub)21 [25.7 (^22Ne/^21Ne)(sub)c - 23.7]^-1, where P^1(sub)21 = 1.63 [Mg] + 0.6 [Al] + 0.32 [Si] + 0.22 [S] + 0.07 [Ca] + 0.021 [Fe + Ni] as given by [3]. (Elemental abundance [x] in weight %, P(sub)21 in 10^10 cm^3 STP/g, Ma). Average cosmic-ray exposure ages were derived from as many nuclei as possible for 14 HEDs analyzed by us (see also [5,6]) and for those compiled by [7]. Two major exposure age clusters at 21 and 38 Ma are represented in all three meteorite classes (Fig. 1). In the cluster at 21 +- 4 Ma are 12 out of 39 eucrites, 6 out of 14 howardites, and 7out of 12 diogenites. In the cluster at 38 +- 8 Ma are 6 eucrites, 5 howardites, and 4 diogenites. A third common break-up event at 5 +- 1 Ma is indicated by the remaining diogenite, three eucrites, and one howardite. Schultz [8] found major clusters for eucrites at 13, 21, 26, and 40 Ma for howardites around 10 and 24 Ma, and for diogenites at 17 Ma, whereas the two clusters for diogenites calculated by [9] are in excellent agreement with our results. The cluster distribution observed in our work is consistent with the stratigraphy of 4 Vesta [10] and strengthens the view that this asteroid is the HED parent body: large collisional events liberate material representing all three classes, whereas smaller ones do not reach down to depths where diogenites can be ejected. Therefore, eucrites that are suggested to originate from the outer layer are more frequently expelled than diogenites and show additional exposure age clusters. Howardites are in all respects (mineralogy, chemistry, exposure age distribution) an intermediate case between eucrites and diogenites. Acknowledgement: This work was supported by the Swiss NSF. References: [1] McCord T. B. et al. (1970) Science, 168, 1445-1447. [2] Consolmagno G. J. and Drake M. (1977) GCA, 41, 1271-1282. [3] Schultz L. and Freundel M. (1985) In Isotopic Ratios in the Solar System, (CNES ed.), CEPAD, 27-33. [4] Freundel M. et al. (1986) GCA, 50, 2663-2673. [5] Michel Th. and Eugster O. (1989) Meteoritics, 24, 304. [6] Michel Th. et al. (1991) Meteoritics, 26, 372. [7] Schultz L. and Kruse H. (1992) A Data Compilation, MPI, Mainz. [8] Schultz L. (1987) LPS XVIII, 884-885. [9] Welten K. C. et al. (1991) Meteoritics, 26, 408. [10] Dreibus G. et al. (1977) Proc. LPSC 8th, 211-227. Swindle T. D.* Burkland M. K. Kring D. A. Noble Gases in the Brachinites Eagles Nest and LEW 88763 Although ultramafic regions should be plentiful on differentiated chondritic bodies, our collection of meteorites only contains a few meteorites that are dominated by olivine, the bulk of which are from the ureilite parent body. Another olivine-rich specimen, Chassigny, is believed to be a piece of Mars, while two others, Brachina and ALH84025, appear to represent a third parent body. We have analyzed two more finds, Eagles Nest and LEW 88763, which have been classified as brachinites [1,2]. We analyzed Ne, Ar, Kr, and Xe in unirradiated samples of each, primarily to study their exposure histories, and Ar and Xe in an irradiated sample of Eagles Nest, to use radiogenic gases to constrain its parent body history. It turns out that the noble gases in brachinites do not testify to such simple relationships as are seen among the SNC meteorites. Eagles Nest and LEW 88763 contain 17 and 22 x 10^-8 ccSTP/g, respectively, of cosmogenic 21Ne, considerably more than Brachina [3,4] or ALH 84025 [5]. However, if standard production rates without shielding corrections [6,7] are used, the exposure ages based on 21Ne (50-70 Ma) and 38Ar (15-20 Ma) are not concordant. LEW 88763 has a K-Ar age of about 4500 Ma, implying early crystallization. On the other hand, it has a puzzling lack of radiogenic 129Xe (129Xe/132Xe < 1.04 for all four temperature extractions). For Eagles Nest, the best chronological information comes from the irradiated sample. Most (94%) of the K-derived 39Ar came out in a single step, giving an apparent age of 955 +- 8 Ma. The next three steps (88% of remaining gas), give apparent ages of 1300-1500 Ma, suggesting that this sample's K-Ar system has been partially reset no more than 955 Ma ago. This may have been induced by the shock event that produced irregular and planar fractures in olivine and mechanical twinning of pyroxene in Eagles Nest. While these features only indicate very weak to weak shock (S2 or S3 levels [8]), this is sufficiently high to have degassed radiogenic 40Ar from many ordinary chondrites [8]. Eagles Nest has very high 129Xe/132Xe ratios (>10) in high-temperature extractions, like Brachina and ALH 84025 [3-5], but not LEW 88763. The "radiogenic" 129Xe does not correlate with iodine, again consistent with a shock disturbance. By comparison, 129Xe and iodine do correlate in Brachina, and the K-Ar system also appears to be less disturbed than that of Eagles Nest, although a 40Ar-39Ar study gave a mean age of 4100 Ma and evidence for more recent resetting [3]. The Rb-Sr system in Brachina is disturbed [3], and Brachina has a U-Th-He age of 400 Ma [4], suggesting a disturbance at least that recent. We can compare the noble gases in the four brachinites to those in the nine SNC meteorites. The brachinites have at least three different exposure ages, while all but one of the SNCs fall into one of two groups. All SNCs have crystallization ages of <1300 Ma, while most of the brachinites have some indication of much older crystallization ages. The Xe isotopes and Ar/Kr/Xe elemental systematics of SNCs appear to be relatively simple mixtures of a Chassigny-like component (129Xe/132Xe =1.0) and a component like the martian atmosphere (129Xe/132Xe = 2.4) [9,10], while there is no systematic relationship among elemental abundances or Xe isotopes of brachinites. Many of these differences probably reflect the difference between a planet-sized shergottite parent body (probably Mars) and the smaller body (or bodies) on which the brachinites evolved. References: [1] Kring D. A. and Boynton W. V. (1992) LPSC XXIII, 727-728. [2] Nehru C. E. et al. (1992) Meteoritics, 27, 267. [3] Bogard D. D. et al. (1983) Meteoritics, 18, 269-270. [4] Ott U. et al. (1985) Meteoritics, 20, 69-78. [5] Ott U. et al. (1987) Meteoritics, 22, 476-477. [6] Schultz L. and Freundel M. (1985) Isotopic Ratios in the Solar System, 27-33. [7] Freundel M. et al. (1986) GCA, 50, 2663-2673. [8] Stoffler D. et al. (1991) GCA, 55, 3845-3867. [9] Ott U. (1988) GCA, 52, 1937-1948. [10] Swindle T. D. et al. (1989) LPSC XX, 1097-1098. Nishiizumi K.* Arnold J. R. Sharma P. Two-Stage Exposure of the Fayetteville Meteorite Based on 129I Cosmogenic nuclides and cosmic ray track densities in the solar noble gas-rich brecciated H-chondrite Fayetteville have been reported [1] . The authors found a discrepancy between exposure ages based on noble gases (27-32 m.y.) and on cosmic ray tracks (4-7 m.y.). We obtained nearby samples from Dr. P. Pellas and measured ^129I (half-life = 1.57 x 10^7 year) in two locations, "S" and "UV" of Fayetteville. The steep track density gradient indicates that sample "S" was a few centimeters from the preatmospheric surface [1]. The distance between "S" and "UV" is 3-4 cm. The ^129I AMS measurements were made at the University of Rochester and the results are shown in Table 1. The blank correction was (49 +/- 27) x 10^-15 ^129I/^127I. Measurements of ^129I in the two samples are in good agreement. Since the main target element for production of cosmogenic ^129I in meteorites is Te, the ^129I concentrations were normalized to Te content and are shown in Table 1. We adopted 0.50 ppm as the Te concentration in Fayetteville [2]. The measured ^129I concentration is higher than the saturation values of Abee or Dhajala and similar to that of a sample of Allende (USNM 3529) that was located 25-30 cm from the surface of a preatmospheric body with a 55-65-cm radius [3]. Iodine-129 in Fayetteville was produced by ^128Te(n,gamma) reactions in addition to ^130Te(n,2n) reactions. Based on comparison with other (n,gamma) products such as ^60Co, ^41Ca, and ^36Cl in Allende and in the Apollo 15 core, we conclude that the ^129I in Fayetteville was produced at a depth below 20 cm within a 60-150-cm-radius object. The observed ^129I activity cannot be explained by either a shallow depth within an object of the recovered size or by 2-pi bombardment in a large parent body. Our ^129I results indicate that Fayetteville experienced a two- stage irradiation with 4-pi production dominating in both bombardments. In this scenario most of the ^129I and cosmogenic noble gases were produced at or below a depth of 20 cm in a large body during a 25-40-m.y. first-stage bombardment. The meteoroid (or parent body) was then broken up and sample "S" was exposed in a near-surface location. The cosmic ray tracks and short half- life nuclides were produced during the 4-7 m.y. of this second-stage bombardment. Measurement of ^36Cl in the stony phase and measurement of ^53Mn will constrain the duration of the second-stage exposure and the size of the object during this stage. The slightly high ^22Ne/^21Ne ratio observed in the meteorite [1] may suggest a complex irradiation prior to the first stage described here. The measurement of ^129I extends the study of exposure histories of extraterrestrial materials to tens of million years. References: [1] Wieler R. et al. (1989) GCA, 53, 1449-1459. [2] Xiao X. and Lipschutz M. E. (1991) GCA, 55, 3407-3415. [3] Nishiizumi K. et al. (1983) Nature, 305, 611-612. Table 1, which appears here in the hard copy, shows ^129I results in the Fayetteville meteorite. Shima M. Honda M.* Yabuki S. Takahashi K. Cosmogenic Radionuclides in Recently Fallen Chondrites Mihonoseki and Tahara Introduction: The chondrite Mihonoseki, L6, 6.38 kg, fell on December 10, 1992 [1]. The other chondrite, Tahara, fell on March 26, 1991, on the deck of car- carrier ship, M.S. Century-Highway No.1 of Kawasaki Kisen Kaisha Ltd., anchored at T-3 berth of Toyota Pier, at Toyohashi harbor, in Tahara-Center, Toyota Motor Corp., Tahara-machi, Atsumi-gun, Aichi-ken, Japan. Although the total mass is estimated to be more than 5 kg, only several fragments were recovered by crews. In fact, this was recognized by the event of Mihonoseki. Tahara was classified as H5 [2]. Gamma-Ray Counting: With whole mass of Mihonoseki, nondestructive gamma-ray countings started on December 15, 1992, using a pure Ge detector (ORTEC), 45 mm x 39 mm, horizontal type. Data collections were performed every day in the beginning and later about every week through February 3, 1993. A sample chamber was shielded with 15-cm-thick lead, 6-cm-thick iron, and 0.5-cm-thick plastic plates. For Tahara, another set (Canberra), 44 mm x 42 mm, coaxial type, was used. The 420-g fragment was mounted in the sample chamber shielded with 15-cm-thick lead, 2-cm-thick iron, 2-cm-thick copper, and 2-cm-thick plastic plates. The counting started in January 1993. The counting efficiencies for gamma rays as a function of energy, ranging between 122 keV (57Co) and 1809 keV (26Al), have been determined using three different standards. A mixed standard solution of nine-species gamma-ray emitters, QCY-44, reference time 12:00 GMT on February 1, 1993, was supplied from Amersham, England. The solution was dropped onto (1) chips of Al-foil, (2) chips of filter paper, or (3) olivine sand. Those standards were mixed thoroughly with mock materials, fine and coarse olivine sand and iron powder, and reagent KCl, standard for 40K, then filled into mock shells of Mihonoseki and Tahara, which were made of hard plastic and aluminum foil with epoxy resin, respectively. For Tahara, mocks with all three types of standards were examined for comparison, while for Mihonoseki only (3) was used. The difficulty was to prepare a suitable mock sample having the same density as chondrites 3.5. Especially for a large sample like Mihonoseki, even when we use about equal fractions of olivine sand and metallic iron, the weight of the mock was about 80%, and when we intend to obtain heavier than 90%, we have to use a larger portion of metallic iron, which causes some reductions in the efficiencies of 20-30%, depending on energies. Results: The contents of 14 gamma emitters were studied as shown in the Table (which appears in the hard copy). Errors quoted are only from counting statistics. The most striking may be to learn that Mihonoseki contains a very low level of 60Co; the content is lower than 1 dpm/kg, which could not be determined accurately by a current-direct gamma counting. This reflects the smaller preatmospheric size of the body, and consistent with other observations such as 22Ne/21Ne = 1.180 [1] and lower activity levels of general products such as 46 dpm 26Al/kg, which is about three-fourths of a common level among L chondrites. Besides, relatively high 56Co in respect to 58Co is also noticed in Mihonoseki. References: [1] Shima M. et al.(1993) LPSC XXIV 1297-1298. [2] Shima M. et al.(1993) Meteoritical Bull., in press. Romstedt J.* Pedroni A. Irradiation History of Acfer 111, Inferred from Nuclear Tracks and Rare Gases Acfer 111 is a regolith breccia consisting of H4-H6 chondritic and igneous clasts embedded in a fine-grained unequilibrated clastic matrix. The matrix has a high concentration of solar noble gases of virtually unfractionated composition [1]. To investigate the irradiation history of Acfer 111, we analyzed the cosmic ray tracks and noble gases at different locations in the meteorite. Noble gases were measured by conventional mass spectrometry. For cosmic ray track analyses, 18-200-mg fragments of nine clasts and five matrix locations were crushed and sieved. Ten to fifty grains of transparent olivine were picked from the 60-200-micrometer fractions, mounted in epoxy, polished, and etched (about 4 hr in a boiling WN solution [2]). Clasts: Out of two clasts having a small ^21Ne(sub)c excess (5-10%), one had a few grains with clearly higher track densities, the other a small amount of solar ^4He. Since the track-rich grains were not identified in situ (i.e., in an etched section), matrix contamination cannot be ruled out. The presence of preirradiated clasts, detected in many other gas-rich meteorites [see references in 4], remains an open question for Acfer 111. Clastic Matrix: Track densities at each location in the matrix show a main peak distribution that can be attributed to the galactic-cosmic-ray (GCR) irradiation of the meteoroid. Ten percent to fifty percent of the grains, however, have a track density higher than the main peak, and thus are preirradiated. Fourteen percent of these grains exhibit a steep track density gradient, indicating solar-flare irradiation at the surface of the parent body. Comparable abundances of preirradiated grains are found in the gas-rich meteorites Kapoeta and Fayetteville [3,4]. Modal and INAA analyses revealed 25-30% more metallic Fe-Ni in the matrix than in the clasts. ^21Ne(sub)c- deficits observed for matrix samples are attributed to these differing target- element chemistries. Gcr Exposure Age and Preatmospheric Size: As shown in Fig. 1, the depth- dependent main-peak track densities (owing to the meteoroid irradiation) correlate with the shielding-sensitive (^22Ne/^21Ne)(sub)c ratio. Furthermore, there is a good correlation between these two parameters and the position of the sample within the meteorite. The preatmospheric radius inferred from noble gases is at least 12 cm. Assuming production rates given by Eugster [6], the cosmic ray ages of all clasts cluster at 37.6 +/- 2 Ma. Comparing this noble gas exposure age and the measured main-peak track densities with track production rate profiles modeled for spherical meteoroids [5] and taking into account the differing track recording effiencies of olivines and pyroxenes [7], a preatmospheric meteoroid radius of 13-14 cm is inferred. The excellent agreement in size, age, and geometry found for track and noble-gas data strongly supports a single-stage exposure for Acfer 111. References: [1] Pedroni A. and Begemann F. (1992) Meteoritics, 27, 273-274. [2] Krishnaswami S. et al. (1971) Science, 174, 287-291. [3] Price P. B. et al. (1975) Proc. LSC 6th, 3449-3469. [4] Wieler R. et al. (1989) GCA, 53, 1441-1448. [5] Bhattacharya S. K. et al. (1973) JGR, 78, 8356-8363. [6] Eugster O. (1988) GCA, 52, 1649-1662. [7] Pellas P. et al. (1973) Meteoritics, 8, 418-419. Fig. 1, which appears here in the hard copy, shows the correlation between the 22Ne/21Ne ratio and track densities. Michlovich E.* Vogt S. Wolf S. F. Elmore D. Lipschutz M. E. Cosmogenic Radionuclides in Antarctic Meteorites: Preliminary Results on Terrestrial Ages and Temporal Phenomena Since 1969, more than 15,000 meteorites have been recovered from various sites in Antarctica. Differences have been reported between the Antarctic populations and the population of non-Antarctic meteorites in volatile trace- element content, thermoluminescence properties, physical size, and relative distribution of meteorite type [1]. Lipschutz and Samuels [2] developed a method based upon multivariate linear and logistic regression that they applied to interpret trace-element content in Antarctic and non-Antarctic meteorites, showing that the two populations can be chemically distinguished. Since Antarctic meteorites have, on the whole, much longer terrestrial ages than non-Antarctic falls, such differences have been used to support the notion that the flux of meteorites sampled by the Earth has changed in the recent past. A subsequent study [3] showed a statistically significant difference in trace-element content between meteorites from Victoria Land and those found in Queen Maud Land, two groups that seem to have different terrestrial age distributions. Changes in meteorite flux patterns on the order of 60 yr are indicated from a study of Cluster 1 vs. non-Cluster 1 falls [4]. Rapid fluctuations would almost certainly require the existence of co-orbital meteoroid streams, an idea that has been criticized by some [5] on dynamical grounds. To quantify the discussion of a temporal dependence of meteorite flux patterns, and to continue systematic study of Antarctic meteorites, we have measured the contents of the cosmogenic radionuclides ^10Be and ^26Al in the bulk phase, and ^36Cl in the metal phase, of 40 Antarctic specimens that are from the same suite of samples analyzed in the trace-element studies and that were chosen to minimize any chances of paired meteorites. The means and standard deviations of ^10Be and ^26Al activities are 16.4 +/- 3.5 and 48 +/- 8 dpm/kg respectively. Correction for cosmic ray exposure [6,7] and terrestrial ages allows us to estimate the production rates for these radionuclides in this group of meteorites to be 18.2 +/- 2.3 and 58 +/- 13 dpm/kg respectively, consistent with production rates cited for falls [8]. Cosmic ray exposure ages using the ^10Be/^21Ne method outlined by Graf et al. [9] substantially agree with ages calculated from noble gases alone. Similar agreements are obtained between cosmic ray exposure ages based solely on noble gases and those calculated using ^26Al/^21Ne [9]. We calculated terrestrial ages using the secular equilibrium distribution for ^36Cl of 22.8 +/- 3.1 dpm/kg [10]. Our results are similar to those seen by Nishiizumi et al. [10], with a few ages ranging up to several hundred thousand years. It is worth noting that the Yamato meteorites measured in the present study, all of which happen to have been collected in the 1979 recovery effort ("Y79"), have a much older terrestrial age distribution (median age of 140 ka) than the Yamato distribution shown in [10]. We find it interesting that our Yamato age distribution is, however, consistent with the distribution of Y79 ages (median age, 110 ka) listed in [10], and that non-Y79 Yamato meteorites (median age in [10], 22 ka) seem to be responsible for a disproportionate number of the youngest Yamato meteorites. This possible collection area phenomenon is under investigation. Preliminary statistical analysis of the results using the preliminary terrestrial ages calculated here, trace-element data [3,4,11], and the methods elucidated in [2] is consistent with the notion that the meteorite flux sampled by the Earth has changed as a function of time. The latest results will be presented in Vail. References: [1] Koeberl C. and Cassidy W. A. (1991) GCA, 55, 3-18. [2] Lipschutz M. E. and Samuels S. M. (1991) GCA, 55, 19-34. [3] Wolf S. F. and Lipschutz M. E. (1992) LPS XXIII, 1545-1546. [4] Dodd R. T. et al. (1993) JGR, submitted. [5] Wetherill G. W. (1986) Nature, 319, 357-358. [6] Schultz L., personal communication. [7] Schultz L. et al. (1991) GCA, 55, 59-66. [8] Vogt S. et al. (1990) Rev. Geophys., 28, 253-275. [9] Graf Th. et al. (1990) GCA, 54, 2521-2534. [10] Nishiizumi K. et al. (1989) EPSL, 93, 299-313. [11] Lingner D. W. et al. (1987) GCA, 51, 727-739. Jull A. J. T.* Wlotzka F. Bevan A. W. R. Brown S. T. Donahue D. J. 14C Terrestrial Ages of Meteorites from Desert Regions: Algeria and Australia The terrestrial age or residence time on the Earth's surface is important in determining the history of a meteorite. Carbon-14 has been used for a terrestrial-age indicator since 1962 [1,2]. Since 1984, small samples of meteorites of 0.1 to 0.5 g have been dated using accelerator mass spectrometry [3-5]. The precision of terrestrial age estimates is limited by the accuracy to which the saturated activity of ^14C in the meteorite is known. Jull et al. [4,5] used Bruderheim and some other chondrites to establish a saturated activity reference level. It is important to be aware that ^14C can vary with the depth and size of the object, and ^14C as a function of accurate depth has so far been measured only for one object, Knyahinya [7]. Carbon-14 is of particular interest in warmer climatic regions, where the storage time before a meteorite weathers away is expected to be much less than other locations, for example, Antarctica. This view was originally based on the work of Boeckl [7], who determined a "weathering half life" of some 3500 yr for chondrites from the southwestern U.S. This work was reinvestigated [5] and it was determined that the ^14C age distribution of the meteorites was longer than the earlier report. We have studied ^14C ages of meteorites from Roosevelt County, New Mexico [8], and from the western Libyan desert [9]. In both these areas meteorites of ages as old as 35,000 yr are observed, and the mean survival time at both locations is well over 10,000 yr. We have studied the ^14C age distribution of a large number of meteorites from Acfer, Algeria, and the Nullarbor Plain, Australia. Figure 1 presents the ^14C age distribution of Acfer samples compared to some other locations where a substantial number of ^14C ages have been obtained. The Algerian site shows a simple exponential dependence of terrestrial age vs. time, and no meteorites of >25 K.y. age. This is in contrast to the results from the southwestern U.S. [7] and from Roosevelt County [8]. One might expect that meteorites would be more well preserved in a very arid, hot climate, and some meteorites of longer age would be present, but this appears not to be the case. This interpretation is strengthened by the results from Nullarbor Plain, although the Australian collection does show some older samples. However, these two regions do show the expected exponential drop-off in number of meteorites of a given terrestrial age with time, which indicates the collections have been undisturbed over at least the last 20,000 yr. This is not seen in the U.S. meteorites. The less arid and colder high plains of Texas and New Mexico may be more conducive to storage of meteorites for long periods of time than these areas, but we believe some selection processes must be at work here and there is a deficit of "young" meteorites. References: [1] Suess H. and Wanke H. (1962) GCA, 26, 475. [2] Fireman E. L. (1978) Proc. LPSC 9th, 1647. [3] Beukens R. P. et al. (1988) Proc. NIPR Symp. Antarc. Met., 1, 224. [4] Jull A. J. T. et al. (1989) GCA, 53, 2095. [5] Jull et al. (1993) Meteoritics, in press. [6] Reedy R. C. (1993) LPS XXIV. [7] Boeckl R. P. (1972) Nature, 236, 25. [8] Jull A. J. T. et al. (1991) LPS XXII, 665. [9] Jull A. J. T. et al. (1990) GCA, 54, 2895. Fig. 1, which appears here in the hard copy, shows terrestrial ^14C ages from desert regions. Tuesday, July 20, 1993 Chondrites and Antarctic Weathering 8:15 a.m. Cascade Ballroom Chair(s): J. C. Bridges B. Zanda Jochum K. P.* Palme H. Fractionation of Volatile Elements by Heating of Solid Allende: Implications for the Source Material of Earth, Moon, and the Eucrite Parent Body CI-chondrites have average solar-system abundances of moderately volatile (Na, K, Rb, Sn, etc.) and highly volatile (Cs, Pb, etc.) elements. In most other types of chondrites and in samples from differentiated planetary bodies, these elements are more or less depleted relative to CI chondrites. Volatile-element fractionation occurred either by evaporation or incomplete condensation [1]. Recent data on the isotopic composition of K indicate that depletion of volatiles did not occur by evaporation from a melt of CI-chondritic composition [2]. Evaporative loss from a solid, however, would not necessarily lead to isotopic fractionation of K in the residue [e.g., 3]. In order to study loss of volatile elements from solids, we performed a series of heating experiments under variable oxygen fugacities at temperatures of 1050 degrees C to 1300 degrees C. Residues were analyzed by INAA [4]. We report here additional analyses (K, Rb, Cs, Sn, Pb) of these residues by isotope dilution-SSMS. Results (including Na data from INAA) are shown in Fig. 1. Results at other oxygen fugacities are similar, i.e., there is no strong dependence on fO2, contrary to the results for Au, As, and Zn [4]. Elements are arranged in the order of decreasing condensation temperatures. Depletions increase with increasing temperature and, at least for the 1050 degrees C experiment, with decreasing condensation temperature. The CI- normalized Allende pattern has no strong depletions of Cs and Pb, unlike the experimental results, indicating that evaporation from a solid cannot produce patterns observed in volatile-element-depleted meteorites. Even heating at temperatures as low as 1050 degrees C, affecting alkali elements only slightly, leads to large losses of lead, which are an order of magnitude greater than required for producing CV chondrite patterns. Depletions of these elements apparently occurred in the solar nebula before accretion by incomplete condensation or removal of gas during condensation. Nearly-CI-chondritic Sn/Pb ratios are observed in Allende and other carbonaceous chondrites. Evaporation from a solid leads to a severe increase in this ratio. Similarly, Rb/Cs ratios (about 12) are approximately CI-like in all groups of carbonaceous chondrites, perhaps reflecting the inability of nebular processes to fractionate these ratios. In contrast, terrestrial, lunar, and eucritic rocks have much higher Rb/Cs ratios [5]. As volatile loss from molten magmas is excluded [2], their low Cs contents must be characteristic of the parent material. This may exclude carbonaceous chondrites as source materials of eucrites, the Earth, and the Moon. The low Cs in planetary precursor materials may have been produced by secondary heating of small fragments of solid matter at subsolidus temperatures before final accretion. Equilibrated chondrites also show high Rb/Cs ratios, perhaps indicating mobilization of Cs at metamorphic temperatures. References: [1] Palme H. et al. (1988) in Meteorites and the Early Solar System, 436-461, Univ. of Arizona. [2] Humayan M. and Clayton R. N. (1993) LPSC XXIV, 685-686. [3] Davis A. M. et al. (1990) Nature, 347, 655-658. [4] Wulf A. V. and Palme H. (1991) LPSC XXII, 1527-1528. [5] McDonough W. F. et al. (1992) GCA, 56, 1001-1012. Figure 1 appears here in the hard copy. Zhai M.* Shaw D. M. A Revision to the Solar System Abundance and Condensation Temperature of Boron from Uncontaminated Falls We requested from participating museum curators interior fragments of chosen falls, never touched by water or other possible sources of B contamination. Thirty six were obtained, crushed, and analyzed for B by PGNAA (prompt gamma- ray neutron activation analyses) at McMaster University and at The National Institute of Standards and Technology. Boron concentrations are close to the sensitivity limit in both laboratories. Results agree well, but with slight systematic differences attributable to blank and background correction factors. Our results (Table 1) are similar to previous measurements on falls [1], but lower than in Antarctic meteorites [2,3], some of which are altered. To calculate the solar system abundance of B, the four carbonaceous chondrite analyses (Table 1) were used as follows. Since CM and CV meteorites contain 48% and 42% matrix [4], if the B ratio of inclusions/matrix is 0.17 [3], then the matrix of the four carbonaceous chondrites averages 0.97 ppm B. Taking the average Si abundance in CI to be 10.64% [5], the calculated solar system abundances from the four carbonaceous chondrites are 23.51, 22.54, 16.95, and 34.20, with a geometric mean of 23.5 B atom/10^6 Si atoms. For comparison, of 18 analyses of interior samples of falls and Antarctic carbonaceous chondrites [1,2,3], 12 have normalized matrix B between 17.6 and 31.4. A composite chondrite atomic composition was calculated for Mg, Na, Li, B, Ga, S, and Zn using their average abundances in H, L, LL, E, and CC meteorites, weighted by their fall frequencies [7] and normalized to Si and the CI abundance [5]. The values show a systematic decrease (Table 2). If this trend is related to volatility [8], then the condensation temperature of B should be between the condensation temperature of Li and Ga, at about 1125 degrees K. If the relative abundances of these elements are similar in the Earth's mantle except for Na [8], then Na appears to show a lower condensation temperature, similar to B. This difference may be due to different evolution paths of meteorites and the Earth. References: [1] Curtis D. B. and Gladney E. S. (1985) EPSL, 75, 311-320. [2] Shaw D. M. et al. (1992) Meteoritics, 27, 289. [3] Shaw D. M. et al. (1988) GCA, 52, 2311-2319. [4] Dodd R. T. (1981) Meteorites, 53, Cambridge Univ. [5] Anders E. and Grevesse N. (1989) GCA, 53, 197-214. [6] Mason B., ed. (1971) Handbook of Elemental Abundances in Meteorites, Gordon and Breach, New York. [7] Sears D. W. G. and Dodd R. T. (1988) in Meteorites and the Early Solar System (J. F. Kerridge and M. S. Matthews, eds.), 4, Univ. of Arizona, Tucson. [8] Jagoutz E. et al. (1979) Proc. LPSC 10th, 2031-2050. Table 1. Boron in meteorite falls. Number Boron Atom Ratio of content meteorites (ppm) s.d. B/10^6 Si Chondrites CM 2 0.52 0.02 10.3 CV 2 0.53 0.25 8.8 H chondrites 5 0.69 0.27 10.5 L chondrites 12 0.71 0.28 10.0 LL chondrites 3 1.07 0.97 14.7 E chondrites 3 0.87 0.31 12.5 Differentiated Aubrite 2 0.81 0.73 3.0 Diogenite 1 0.06 0.6 Eucrite 6 0.61 0.30 6.8 Table 2. Silicon and solar-system-normalized abundances of some elements in chondrites. Elements Abundance Condensation Temperature [8] (degrees K) Mg 0.89 [6] 1360 Na 0.73 [6] Li 0.66 [6] 1225 B 0.46 (this work) 1125 (estimated) Ga 0.33 [6] 1075 S 0.21 [6] Zn 0.12 [6] 700 Hua X.* Buseck P. R. Fayalite Formation in Some Primitive Chondrites Fayalite (up to Fa(sub)99.9) occurs in the Kaba and Mokoia CV3 carbonaceous chondrites. The grains can reach up to 100 micrometers in diameter and are commonly associated with magnetite and sulfides. Since the fayalite is essentially pure, it must have formed in an environment where Mg is absent. We believe it formed by a two-step gas-solid reaction: (i) the decomposition of enstatite to release SiO (g); (ii) reaction of SiO (g) with magnetite or sulfides to form fayalite. Production of SiO (g) occurs via a reaction such as: 2 MgSiO3 (s) + H2 (g) = Mg2SiO4 (s) + SiO (g) + H2O (g) (1) for which the equilibrium constant is K = the equation which appears here in the hard copy. Using the JANAF tables, we can calculate Delta G for different temperatures, followed by a calculation of PSiO for various H2O/H2 ratios: equation appears here. For fayalite formation, we considered the following reactions: 2 Fe3O4 (s) + 3 SiO (g) + H2O (g) = 3 Fe2SiO4 (s) + H2 (g), and (2) 2 FeS (s) + SiO (g) +3 H2O = Fe2SiO4 (s) + H2(g) + 2 H2S (g) (3) Since sulfides will first react with H2O to form magnetite, reaction (3) will proceed via reaction (2), so our calculation is only for reaction (2). Following the same procedure as for reaction (1), we obtain: the equation, which appears here in the hard copy. After plotting the P(sub)SiO vs. temperature for different H2O/H2 ratios we found: (i) If the H2O/H2 ratio is 5.1 X 10^-4 (solar), fayalite starts to form at a temperature of 300 K, at which the vapor pressure of SiO is only 10^-86 atm. We presume under these conditions the reaction is kinetically prohibited. (ii) If H2O/H2 = 100, fayalite forms above 3000 K, which is also unlikely, because at such a high temperature all silicates will decompose. (iii) If H2O/H2 = 1, the formation temperature of fayalite is ~ 640 K, and the vapor pressure of SiO is 10^-36 atm. (iv) If H2O/H2 = 10, fayalite begins to form at 1100 K, and the vapor pressure of SiO is 10^-19 atm. After taking the reaction rates into consideration, we prefer the combination of H2O/H2 = 10 and starting temperature of 1100 K, because for these conditions P(sub)SiO is 17 orders of magnitude greater than that at ~640 K. If we put additional constraints on the temperature based on the coexistence of magnetite, then a maximum temperature of 1200 K for the reaction can also be determined [1] (Fig 4), because above this temperature magnetite is not stable when the H2O/H2 ratio equals 10. In an environment with H2O/H2 = 10 atm and T = 1100 or 1200 K, the vapor pressure of SiO released from the evaporation of enstatite is 10^-17 and 10^-14.8, respectively. The decomposition of forsterite under identical conditions will be negligible [2 log P(sub)Mgo + log P(sub)SiO = 10^-57.4 (at 1100 K) and 10^-50.3 (at 1200 K)]. References: [1] Larimer J. W. (1967) GCA, 1215-1238; JANAF Thermodunamical Tables (1971). Yoneda S.* Simon S. B. Sylvester P. J. Hsu A. Grossman L. Large Siderophile-Element Fractionations in Murchison Sulfides Five sulfide-coated, rounded lumps (240-440 micrometers in largest dimension) were recovered from the dense fraction after freeze-thaw disaggregation of the Murchison CM2 chondrite. Each was split into a fraction studied by SEM and EPMA and a fraction analyzed by INAA. Petrographic study shows that S1 and S5 are composed predominantly of FeS enclosing ~10% olivine. In S1, olivine is Fa38 and occurs isolated or in clusters of 20-50-micrometer, rounded grains, some with glass inclusions. S6 consists mostly of 50-100-micrometer forsterite crystals, lesser enstatite, interstitial anorthite, and minor, 10-micrometer blebs of low-Ni metal. The entire object is surrounded by a 5-20-micrometer- thick rind of FeS and permeated by a series of subparallel veins of FeS. Both S3 and S4 contain silicates enclosed in massive troilite, with lesser pentlandite in the case of S4. In S4, the silicates include euhedral, zoned olivine (Fa18-40) crystals up to 30 micrometers in size; polycrystalline, masses of finer-grained (2-5 micrometers) olivine (Fa45) and Ca-pyroxene. S3 has subhedral forsterite crystals up to 20 micrometers across and irregular, fine-grained clasts up to 100 micrometers in size that resemble the Murchison matrix. Sulfide proportions in the samples studied by INAA were estimated by comparing their Mg contents with EPMA data for silicates in their splits: 86%, 63%, 51%, 83%, and 28% in S1, S3, S4, S5, and S6 respectively. Refractory siderophiles (W, Re, Os, Ir, Mo, and Ru) are unfractionated in S6, each having a concentration within 20% of C1 chondrites. In S1, S4, and S5, Ru and Ir abundances are 1.8-3.1 x C1 and 0.024-0.25 x C1, respectively, leading to C1- normalized Ru/Ir ratios of 115 +- 7, 7.0 +- 2.6, and 117 +- 13, respectively, as also seen in Allende pentlandite [1]. S3, which is depleted in both Ru and Ir, also has a superchondritic Ru/Ir ratio, 1.8 +- 0.4. Ru and Mo are strongly positively correlated in all samples. Au, As, Ga, Zn, Ni, and Co are close to or below C1 levels in S1, S3, S5, and S6, except that the C1-normalized enrichment factor for Ga in S3 is 3.8 +- 0.2 and that for Co in S1 is 2.1. S4 is quite distinct, however, with As at 5.4 +- 0.2 x C1, Ni at 5.5 +- 0.2 x C1, and Co at 4.7 +- 0.1 x C1. Except for S6, Se is above C1 levels in all samples, with enrichment factors of 4.3-5.3. As suggested by groups IIAB and IIIAB irons, fractional crystallization of metal from a low-S liquid containing chondritic proportions of siderophiles leads to residual liquids with Ru/Ir and Mo/Ir ratios that are greater than chondritic [2]. Formation of sulfides with these characteristics in S1, S3, and S5 from metal made by this process seems unlikely, however, as it would also lead to C1 chondrite-normalized Au/Ir ratios that are greater than those of Ru/Ir, while the Au/Ir ratios in S1, S3, and S5, 0.5-1.2 x C1, are far less than Ru/Ir ratios. Mo and Ru are the most chalcophile of the refractory siderophiles studied here, as seen from experimentally determined partition coefficients between solid metal and solid FeS for Mo (1.4 [3]), Ru (0.1-13 [4]) and Ir (12-240 [4]), and from the fact that Ru/Ir and Mo/Ir ratios in troilite nodules in Odessa and Canyon Diablo are factors of 2.4-4.0 and 42-55, respectively, higher than in coexisting metal [5]. Thus, if an alloy enriched in refractory siderophiles were equilibrated with FeS and then segregated from the FeS, the resulting sulfide would have many of the chemical features of S1, S3, and S5. Fremdlinge in CAIs are interpreted as high-temperature condensate alloys (e.g., [6,7]). In many of these objects, Ru and Mo are depleted relative to other siderophiles compared to C1 chondrites, though not in concert, due to volatility and phase separation effects accompanying condensation. Furthermore, some entire CAIs in Leoville and Efremovka are strongly depleted in both Mo and Ru relative to Ir compared to C1 chondrites [8], implying that the processes that made Fremdlinge resulted in significant amounts of Ru and Mo being left in the gas phase after formation of CAIs in some nebular regions. Thus, another possibility is that the sulfides seen here formed from metal that condensed from such a fractionated gas before Au condensed. References: [1] Davis A. M. et al. (1978) Meteoritics, 13, 438-439. [2] Scott E. R. D. (1972) GCA, 36, 1205-1236. [3] Jones J. H. et al. (1993) GCA, 57, 453-460. [4] Fleet M. E. et al. (1991) JGR, 96, 21949-21958. [5] Hermann F. et al. (1971) Mikrochim. Acta, 1971, 225-240. [6] Bischoff A. and Palme H. (1987) GCA, 51, 2733-2748. [7] Sylvester P. J. et al. (1990) GCA, 54, 3491-3508. [8] Sylvester P. J. et al. (1993) GCA, in press. Metzler K.* In Situ Investigation of Preirradiated Olivines in CM Chondrites Most CM chondrites are breccias that contain fragments of primary rock representing densely packed agglomerates of chondrules, CAIs, etc., all of which are mantled by thick layers of fine-grained mineral dust [1]. These dust mantles seem to be the result of dust sampling by the various components during their isolated existence in the solar nebula prior to the formation of the CM parent body [1]. Metzler et al. [1] concluded that these rock fragments are well-preserved remnants of the freshly accreted CM parent body(ies). There is an opposing hypothesis that favors an origin of the dust mantles in an active regolith on the CM parent body [e.g., 2]. A list of arguments against the latter view is given by Metzler et al. [1], including a hint at the absence of solar-wind-implanted gases in dust mantles and in fragments of primary rock. In analogy to brecciated ordinary chondrites and lunar breccias, the most probable residence of the solar gases in CM chondrites is their clastic matrix. The same holds for track-rich olivines that were observed in CM chondrites. The occurrence of these grains in the clastic matrix and their absence in the primary rock would give an additional argument for the idea of a dust mantle origin in the solar nebula rather than in a planetary regolith. To answer this important question, mosaics of backscattered electron images of several large polished thin sections of Murchison and Cold Bokkeveld were prepared. The thin sections (1.5-5 cm^2 each) were etched in a WN solution [3] for about 4 hr to reveal the heavy ion tracks in olivines. Results: The background GCR track density produced during meteoroid transit is on the order of 10^4 tracks/cm^2, as was previously observed by [4]. Following the definition given by Goswami and Lal [4], olivines with track densities >10^5 tracks/cm^2 were classified as preirradiated grains and were found in both meteorites in a very small quantity. In both meteorites, 39 preirradiated isolated olivine grains were found in the clastic matrix, whereas the investigated fragments of primary rock do not contain preirradiated olivines. In Murchison about 1.8% (15 out of 850 investigated grains) of the isolated olivines in the clastic matrix show high track densities in the range between 1.9 x 10^6 and >5 x 10^7, comparable to the results of Goswami and co-workers [4,5]. Both Fe-poor and Fe-rich olivines with grain sizes between 40 and 710 micrometers were found to be preirradiated. Track gradients were found in 33% of these olivines, which is very similar to the values obtained by Goswami and Lal [4] and identical to those obtained by MacDougall and Phinney [6]. About 0.4% (2 out of 530) of the investigated olivine-bearing chondrules and chondrule fragments are preirradiated. In the case of Cold Bokkeveld, 3.7% (24 out of 650) of the isolated olivines show high track densities. Thirteen of these 24 grains were found to be concentrated in a distinct inclusion (1 x 4 mm) that is characterized by its elongated appearence and clastic fabric. The track densities of its preirradiated olivines show a very narrow range, indicating a common irradiation history of these grains. The petrography of this inclusion is currently under investigation. Conclusions: Track-rich (preirradiated) olivines in CM chondrites occur exclusively in the clastic matrix of these meteorites, comparable to observations in brecciated ordinary chondrites. Fragments of primary rock in CM chondrites do not contain solar-wind-implanted gases [1] or preirradiated grains. This confirms the view that the dust mantles around various components of these rocks are the products of dust accretion in the solar nebula rather than of regolith processes on the parent body surface. References: [1] Metzler K. et al. (1992) GCA, 56, 2873. [2] Kerridge J. (1992) personal communication. [3] Krishnaswami S. et al. (1971) Science, 174, 287. [4] Goswami J. N. and Lal D.(1979) Icarus, 40, 510. [5] Goswami J. N. and MacDougall J. D. (1983) Proc. LPSC 13th, in JGR, 88, A755. [6] MacDougall J. D. and Phinney D. (1977) Proc. LSC 8th, 293. Zanda B.* Hewins R. H. Bourot-Denise M. Metal Precursors and Reduction in Renazzo Chondrules The positive Co-Ni correlation and Cr, P contents of metal in CR chondrites have generally been taken to indicate their primitive nature, probably inherited from condensation [1,2]. Si in the metal of primitive chondrites has also been reported and interpreted as a condensation heritage [3,4]. However, Cr, P, and Si (dissolved or in the form of inclusions) in metal of any CR chondrule generally fall within a +-10% range, though large interchondrule variations exist [5]. We have shown that Cr and Si in metal are in equilibrium with Fo and En in silicates, due to the reducing conditions that prevailed during chondrule formation [6]. In the present paper, we show that the Co-Ni trend was also established during chondrule formation out of heterogeneous precursor material with a variable Co/Ni ratio. Chondrules in Renazzo are classified as highly molten (HM), in which metal has been expelled to form a mantle outside the chondrule, medium molten (MM), with metal inside and at the periphery, and with evidence for grain coalescence, and little melted (LM), in which metal is only present in the form of small blebs dispersed among the silicates. In HM chondrules, Ni and Co concentrations are extremely homogeneous, comparatively low and in the cosmic ratio. In LM chondrules, quite the opposite: Ni and Co spread over a large range and the amount of scatter increases with decreasing degree of melting of the chondrule. In addition, they do not correlate along the cosmic ratio, but show a negative correlation if any. This heterogeneity is present not only from grain to grain in these chondrules, but also in individual metal grains. Such a heterogeneity is also exhibited in Cr and P abundances that span a much larger range than the +-10% found in the other chondrules. These results indicate that chondrule formation is responsible for the homogenization of Co and Ni contents of metal grains through coalescence and mixing. The less melted objects give an idea of the nature of metal in chondrule precursors, extremely heterogeneous and fine grained (each small heterogeneous metal bleb might be the result of partial melting of one or of coalescence and imperfect mixing of a few such grains). Co and Ni in these individual grains were not in the cosmic ratio, but wide sampling of dust in each chondrule precursor insured that this ratio was attained after mixing and homogenization, as seen in HM chondrule metal grains and from mean values of Co and Ni in LM chondrules. In MM chondrules, scatter of Ni and Co data are, as expected, intermediate between those of HM and LM chondrules, but Co and Ni are close to the cosmic ratio. The scatter is mostly due to addition of variable quantities of iron in the reduction during chondrule formation, which is responsible for Cr and Si integration into metal. Further evidence of such a process can be found in the less molten of these objects, in which metal grain coalescence is limited and peripheral grains are still different from inside grains. In these cases, Co and Ni distributions are clearly bimodal, high in inside grains, low in peripheral grains. Co/Ni in these two populations are somewhat scattered around the cosmic ratio, but their means (Ni: 7.75 = +- 0.24, Co: 0.36 +- 0.04, and Ni: 4.39 +- 0.34, Co: 0.23 +- 0.02, e.g., in the case of chondrule AL1) are very close to the cosmic ratio. This is in good agreement with the low values found in the homogeneous mantle grains of HM chondrules and, as noted by Lee et al. [7], indicates that the reducing agent was external to the chondrule. Cr abundances of these peripheral metal grains, however, match Cr abundances of the interior ones in these chondrules. This indicates that the redox state of all these grains was attained simultaneously and controlled by equilibrium with chondrule silicates. Slightly more extensive reduction of the latter close to the chondrule surface that added more Fe to peripheral metal grains resulted in only a minor variation of the Cr partition coefficient: it consequently also induced Cr addition, the Cr/Fe ratio varying only marginally. Therefore, we believe unlike [7] the process to have been nebular, and the reducing agent the nebular gas, although equilibrium with this gas was clearly not attained. References: [1] Weisberg M. K. et al. (1993) GCA, 57, 1567-1586. [2] Grossman L. and Olsen E. (1974) GCA, 38, 173-187. [3] Grossman L., et al. (1979) Science, 206, 449-451. [4] Rambaldi E. R. et al. (1980) Nature, 287; Nature, 293, 558-561. [5] Zanda B. et al. (1991) LPSC XXII, 1543-1544. [6] Hewins R. H. and Zanda B. (1992) Meteoritics, 27, 233. [7] Lee M. S. et al. (1992) GCA, 56, 2521-2533. Geiger T.* Spettel B. Clayton R. N. Mayeda T. K. Bischoff A. Watson 002--The First CK/Type 3 Chondrite The CK chondrites studied so far are all of petrologic types 4-6. In 1991 Watson 002, a petrologic type 3 carbonaceous chondrite, was found in the Nullarbor Region, South Australia and was classified as "CK3 anomalous" [1]. The supplement "anomalous" was added because it shows some features that do not fit into the mineralogical [2] or into the bulk chemical characteristics [3] of the CK chondrite group. Here we report on the petrography and mineralogy as well as on the bulk chemistry and on the oxygen isotopic composition of Watson 002. Petrograpy and mineralogy: The piece we studied has a 3-5 mm wide margin that shows a less compacted texture, which seems to be caused by terrestrial weathering. Watson 002 has a chondritic texture characterized by abundant chondrules, inclusions, and fragments embedded in a matrix. Matrix is the most comprehensive textural unit making up 70 vol% of the meteorite. Chondrules, inclusions, and fragments (grain size >100 micrometers) are present in modal portions of about 10 vol% each determined by point counting methods. The mean apparent diameter of 43 detected chondrules is 870 +- 380 micrometers ranging from 160 to 2100 micrometers. The mineral chemistry of olivines, pyroxenes, and plagioclases is comparable to other CK chondrites. Olivines and pyroxenes in some chondrules and fragments are not equilibrated. Almost pure forsterites (FaO.3) and enstatites (Fs1.6) occur in the core. Olivines at the edge contain ~20 mol% Fa. Like in other CK chondrites chondrules are frequently rimmed by magnetite. Matrix olivines have a mean Fa content of 34.4 mol% (range 31-38 mol%), and contain an average of 0.35 wt% NiO. Pyroxenes are less equilibrated than olivines; the composition of the clinopyroxenes vary from 6.7 to 19.9 mol% Fs. Two low-Ca pyroxenes with Fs 23.1 and 24.4 were found. Plagioclases occur with An contents from 26.8 to 50.1 mol%. Unlike the other known CK chondrites, Watson 002 contains abundant CAIs that consist of An-rich plagioclase, olivine, and green pleonast-spinel as the main phases. Similar spinel phases were reported from Maralinga 001 [4], and Karoonda [5]. The pleonasts are containing small amounts of Cr and Ti (0.25 wt% Cr203 and TiO2, respectively). Like in other CK chondrites the most abundant opaque phase is magnetite containing about 3.1 wt% Cr203 and 1 wt% TiO2, but we did not find the exsolution of ilmenite and spinel [6]. Sulfides are rare--only a few micron-sized pentlandite grains were found. Bulk chemistry and oxygen/isotopes: Watson 002 shows an unusual enrichment of Na, K, Ba, and light REE. Different INA-analyses of the core and the weathered margin show enrichment of these elements in the margin up to 50 x CI, whereas the core shows a "normal" CK pattern. Therefore we conclude that the high abundance of these elements in the weathered margin is due to terrestrial weathering. Besides that, there is also a strong enrichment of Ca (4.7 x CI) and Ti (~9 x CI) in the core, which can be explained in the case of Ca by the high abundance of CAls compared to other CK chondrites. Ti is predominantly located in magnetite and spinels. The oxygen isotopic composition differs from that of other CK chondrites. Previously analyzed CK chondrites plot within the range of the CO chondrites [7] but Watson plots at the uppermost end of the CV chondrite range and the lowermost end of the CM range. The oxygen isotopic ratios are delta^18O = +5.12 and delta^17O = -0.55. The composition of a single chondrite was found to be delta^18O = +0.56,delta^17O = -3.31, which lies within the range of CV chondrules, and is unequilibrated with the bulk meteorite. References: [1] Wlotzka F. (1993) Met. Bull., 74, Meteoritics 28, 1. [2] Geiger T. and Bischoff A. (1991) Meteoritics, 26, 4, 337. [3] Kallemeyn et al. (1991) GCA, 55, 881-892. [4] Keller L. P. et al. (1991) Meteoritics, 27, 1, 87-91. [5] MacPherson G. J. and Delaney J. S. (1985) LPSC XVI, 515-516. [6] Geiger T. and Bischoff A. (1990) LPSC XXI, 409-410. [7] Clayton R. N. and Mayeda T. K. (1989) LPSC XX, 169-170. Weisberg M. K.* Prinz M. Clayton R. N. Mayeda T. K. Grady M. M. Franchi I. A. Petrology and Stable Isotopes of LEW 87232, A New Kakangari-type Chondrite The discovery of new chondrite groups is an important step in widening our understanding of the primitive asteroidal materials on which models of early solar system processes are based. LEW87232 was tentatively classified as a CR chondrite [1] and our interest in the CR group and its diversity [2] led us to study this meteorite. This petrologic and stable isotope study shows that LEW87232 is, in fact, a new member of the rare Kakangari-type chondrite grouplet. Kakangari was recognized as the first member of a new chondrite group with petrologic, bulk chemical, and oxygen isotopic characteristics that sharply distinguish it from other chondrites [3-7]. Lea Co. 002 was found to be a second member [8]. Texturally, LEW87232 consists of chondrules, fragments, and metal spheres (chondrules) set in a fine-grained matrix. The chondrule mean diameter is 0.4 nm (some up to 1.6 mm). Most chondrules are porphyritic pyroxene, and olivine is poikilitically enclosed in the pyroxene. Rarely, chondrules are olivine rich. Metal chondrules consist of kamacite with exsolved taenite and are rimmed by, and enclose, lath-shaped pyroxene that is similar in size and morphology to the matrix pyroxene; accessory apatite and schreibersite are associated with the metal. The matrix consists mainly of low-Ca-pyroxene laths 1-3 micrometers wide, up to 15 micrometers long, and it is intermixed with an Fe oxide, possibly ferrihydrite. Ferrihydrite was identified in Kakangari [9]. Mineral compositions in chondrules, fragments, and matrix are fairly homogeneous, and similar, with pyroxene Wo0.2-0.5Fs2.6-3.6, olivine Fa0.5-2.9, Ca-pyroxene Wo45Fs0.8, and plagioclase An~60. Kamacite (Ni ~ 5.6%) and taenite (Ni ~ 27%) are homogeneous. Kakangari has similar mineral compositions [7]. Bulk compositions of the chondrules and matrix are strikingly similar, reflecting similarities in their modes and mineral compositions. Stable Isotopes: LEW87232 nitrogen, total delta-15N = +10.6 permil, [N] = 80.6 ppm, is closest to that of ordinary chondrites and differs from that of Kakangari, which has lighter N (total delta-15N = -20 permil). Total [C] = 1989 ppm and is also closest to ordinary chondrites. Kakangari total [C] = 864 [10]. Combustion temperatures indicate the presence of some organic component with delta-15N ~ +4 to +8 permil released at low T. N released above 1000 degrees C may be a combination of spallogenic N, with N possibly from SiC. The oxygen isotope compositions of Kakangari-type chondrites are shown in the figure. Whole rock LEW87232 plots close to the other Kakangari-type chondrites. Chondrule compositions are similar to those in Kakangari, but are displaced toward lower delta-18O values perhaps, in part, due to weathering. Chondrules from Kakangari-type chondrites generally have oxygen compositions similar to enstatite chondrite chondrules (shown by the loop) and some extend toward more 16O-rich compositions. Conclusions: LEW87232 is shown to be a Kakangari-type meteorite and it further defines this distinct chondrite grouplet. Characteristics that distinguish the Kakangari-type grouplet from other chondrite groups include (1) the oxygen isotope composition of the chondrules and matrix, (2) the high metal and pyroxene abundances and low FeO content of the silicates that indicate an oxidation state between H and E chondrites, (3) the Mg- and pyroxene-rich nature and similarity of the chondrules and matrix, (4) the unique intergrowths of matrix pyroxene within and rimming metal chondrules, suggesting that abundant Mg-rich pyroxene crystals formed in the nebula and were present during chondrule formation. References: [1] Mason B. (1992) Ant. Met. News., 15, 24. [2] Weisberg M. K. et al. (1993) GCA, 57, 1567-1586. [3] Graham A. L. and Hutchison R. (1974) Nature, 251, 128-129. [4] Clayton R. N. et al. (1976) LPSC, VII, 160-162. [5] Clayton R. N. et al. (1976) EPSL, 30, 10-18. [6] Davis A. M. et al. (1977) Nature, 265, 230-232. [7] Prinz M. et al. (1989) LPSC, XX, 870-871. [8] Prinz M. et al. (1991) LPSC, XXII, 1097-1098. [9] Brearley A. J. (1989) GCA, 53, 2395-2411. [10] Grady M. M. and Pillinger C. T. (1986) GCA, 50, 255-263. [11] Clayton R. N. and Mayeda T. K. (1985) LPSC, XVI, 142-143. Rubin A. E.* Kallemeyn G. W. Carlisle Lakes Chondrites: Relationship to Other Chondrite Groups Although chondrites are all solarlike in their abundances of nonvolatile elements, there are appreciable differences among chondrite groups in texture and mineralogical, chemical, and O-isotopic composition. There are now 12 chondrite groups, each containing at least five members: CI, CR, CM, CO, CV, and CK carbonaceous chondrites; H, L, and LL ordinary chondrites; EH and EL enstatite chondrites; and Carlisle Lakes chondrites. Eight Carlisle Lakes chondrites have been identified: Carlisle Lakes, ALH 85151, Y 75302, Y 793575, Y 82002, Acfer 217, PCA 91002, and PCA 91241; the latter two may be paired. The primary petrographic characteristics of the group include abundant matrix (42 +/- 11 vol%) and chondrules averaging 400 micrometers in apparent diameter. Secondary petrographic characteristics include moderate metamorphic recrystallization and, in most members, extensive brecciation. As a family, carbonaceous chondrite groups (excepting CI in some cases) are characterized by (1) group/CI mean-refractory-lithophile/Si abundance ratios of 1.00-1.35, (2) a moderate to high degree of Fe oxidation, (3) high fine- grained-matrix/chondrule modal abundance ratios (0.5-7), (4) an appreciable abundance of refractory inclusions (~0.5-5 vol%), (5) whole-rock O-isotopic compositions significantly below the terrestrial fractionation (TF) line, (6) siderophile and chalcophile abundance patterns that decrease monotonically with increasing volatility (with low to moderate CI-normalized Se/Sb concentration ratios, 0.6-0.9), (7) relatively abundant opaque-mineral-rich porphyritic chondrules, and (8) where present, plagioclase with high molar An. Ordinary chondrites (OC) are characterized by (1) group/CI mean-refractory- lithophile/Si abundance ratios of 0.77-0.82, (2) a low to moderate degree of Fe oxidation, (3) low fine-grained-matrix/chondrule modal abundance ratios (~0.3, excluding metamorphosed OC), (4) a negligible abundance of refractory inclusions, (5) whole-rock O-isotopic compositions above the TF line, (6) moderate CI-normalized Se/Sb concentration ratios (0.9-1.1), (7) relatively few opaque-mineral-rich porphyritic chondrules, and (8) where present, plagioclase with low molar An. Enstatite chondrites resemble OC in many respects. They are characterized by (1) group/CI mean-refractory-lithophile/Si abundance ratios of ~0.60, (2) a very low degree of Fe oxidation, (3) essentially no fine-grained matrix, (4) a negligible abundance of refractory inclusions, (5) whole-rock O-isotopic compositions on the TF line, (6) moderate CI-normalized Se/Sb concentration ratios (1.0-1.2), (7) a low to moderate abundance of opaque-mineral-rich porphyritic chondrules, and (8) plagioclase with very low molar An. Carlisle Lakes chondrites are characterized by (1) a CI-normalized mean- refractory-lithophile abundance ratio of 0.85, (2) a high degree of Fe oxidation (as indicated by a molar ratio of [FeO/(FeO + MgO)] x 100 of 37 and an atomic ratio of metallic-Fe/Si of 10^-4), (3) a high matrix/chondrule modal abundance ratio (1.6 +/- 0.9), (4) essentially no refractory inclusions, (5) whole-rock O-isotopic compositions appreciably above the TF line (e.g., delta ^17O = 5.12 per mil), (6) a high mean CI-normalized Se/Sb concentration ratio (1.5 +/- 0.2), (7) few opaque-mineral-rich porphyritic chondrules, and (8) plagioclase with low molar An. Our previous INAA data for Carlisle Lakes and ALH 85151 [1] showed high Ca and low Au, which we attributed to weathering and nebular processing respectively. However, new data for PCA 91002 reveal only a small Ca enhancement and no Au depletion. We now suspect that the low Au abundances in Carlisle Lakes and ALH 85151 are also weathering artifacts resulting from oxidation of the Au carrier in the matrix and transport to new sites. The carbonaceous chondrite family comprises three clans (CI-CR?, CM-CO, and CV-CK). These chondrites are characterized by a generally high oxidation state and, except in CI chondrites, the presence of abundant, isotopically heterogeneous refractory inclusions; it seems likely that this family formed relatively far from the Sun because higher temperatures at small heliocentric distances are typically associated with reduction and isotopic homogeneity. The other major family of chondrites, here dubbed the inner solar system family, also comprises three clans (H-L-LL, EH-EL, and Carlisle Lakes). In both chondrite families, individual groups within each clan differ in mean- refractory-lithophile/Si abundance ratios by <=8%, indicating that they must have agglomerated from similar precursor components at similar times and similar heliocentric distances. Within each family, interclan agglomeration locations were more widely separated in space and time but were still within the same general region of the solar nebula. Reference: [1] Rubin A. E. and Kallemeyn G. W. (1989) GCA, 53, 3035-3044. Schulze H.* Otto J. Rumuruti: A New Carlisle Lakes-type Chondrite We report here preliminary results of the investigation of a meteorite that fell on January 28, 1934, at 10:45 p.m. at Rumuruti, Kenya. The stone, weighing originally about 75 g, was part of a shower of a few pounds. It was picked up immediately after the fall and has been in the collection of the Museum fur Naturkunde in Berlin since 1938, but it has never been investigated. The stone has a black crust. A cut exhibits a nice light-dark structure typical of regolithic breccias. The numerous clasts are light-grey and reach up to 7 mm. The groundmass is dark grey. The portion of chondrules in the meteorite is rather small. They are often broken or irregularly shaped. The mineralogical investigation revealed a quite equilibrated olivine with a high fayalite content ranging in composition from Fa37 to Fa42 and averaging Fa39 (PMD 2.4; n = 66). The grains are up to 400 micrometers and often appear to be fragments of larger lithologies. Low-Ca pyroxene is much less abundant. It is unequilibrated with a mean of approximately Fs26 (n = 2). It is smaller than olivine (up to ~100 micrometers) and often shows polysynthetical twinning. A Ca-rich pyroxene was measured having En43Fs17Wo40. The common plagioclase reaches several micrometers and is mostly of oligoclasic composition with a mean of Ab85An11Or5 (n = 15), similar to OCs [1]. Also whitlockite of a composition similar to OCs [2] has been observed. Common sulfides comprise pentlandite (~35 wt% Ni) and low-Ni iron sulfide, which is pyrrhotite according to powder diffraction patterns. They occur individually or intergrown as grains of up to 1 mm. Pentlandite partly forms flamelike exsolutions in pyrrhotite. Also chalcopyrite, which is otherwise a rare mineral in meteorites [3], can be observed in grains of up to 50 micrometers. It is generally intergrown with the other sulfides. The common chromian spinel is Ti-rich (TiO2 ~6 wt%), Cr2O3 ranges from 32 to 48 wt%, and FeO from 37 to 53 wt%. For charge balance a high Fe3+-content is required (12-51 mol% of the iron). The mean composition of this spinel phase can be expressed as a mixture of the end members chromite (55 mol%), ulvospinel (17 mol%), magnetite (15 mol%), and spinel (9 mol%). Only the magnetite (4-25 mol%) and the chromite component (46-68 mol%) are strongly variable, obviously substituting each other. Chromian spinel occurs intergrown with the sulfides, as xenomorphic or chondrulelike individual grains (up to 200 micrometers) or as inclusions in the olivine. Nickel-iron is a rare phase. Only four grains of up to 30 micrometers have been observed. It seems to be associated with pentlandite and is very rich in Ni (67 wt% Ni). The homogeneity of the olivine and the grain size of plagioclase indicates a classification as a type-4 chondrite, whereas some glass in a chondrule points also to type 3. A refined investigation of clasts and groundmass will provide more clarity. Rumuruti is only mildly shocked (S2 according to [4]), but a vein restricted to one of the light clasts indicates that components of the meteorite experienced higher shock pressures. The unusual assemblage of fayalite-rich olivine (Fa39), Ti- and Fe3+-rich chromite, pentlandite, pyrrhotite, and chalcopyrite is comparable to the highly oxidized Carlisle Lakes-type meteorites [3,5]. Rumuruti now brings this group, together with Carlisle Lakes, ALH85151, Y75302 [3], and Acfer 217 [6], to five meteorites where Rumuruti is the first observed fall. References: [1] Van Schmus W. R. and Ribbe P. H. (1968) GCA, 32, 1327-1342. [2] Van Schmus W. R. and Ribbe P. H. (1969) GCA, 33, 637-640. [3] Rubin A. E. and Kallemeyn G. W. (1989) GCA, 53, 3035-3044. [4] Stoffler D. et al. (1991) GCA, 55, 3845-3867. [5] Weisberg M. K. et al. (1991) GCA, 55, 2657-2669. [6] Bischoff A. et al. (1993) Meteoritics, submitted. Prinz M.* Weisberg M. K. Clayton R. N. Mayeda T. K. Ordinary and Carlisle Lakes-like Chondrite Clasts in the Weatherford Chondrite Breccia Weatherford is similar to Bencubbin [1], which was shown to be a highly unusual chondrite breccia [2]. These meteorites have similar oxygen and nitrogen isotope compositions [3,4] and are members of the CR clan [5], which includes ALH 85085, Acfer 182, LEW 85332, and CR chondrites. Both meteorites contain various chondritic clasts, but ordinary chondrite xenoliths have previously been found only in Bencubbin [6]. We have now found an ordinary chondrite xenolith, as well as a Carlisle Lakes-like [7,8] clast in Weatherford. Herein we discuss the findings of our petrologic and oxygen isotope study of these clasts and their significance. Ordinary Chondrite Clast: A 1.7 x 1.0-cm clast, in AMNH sample 4713, has sharp contacts with the host chondrite and one edge extends off the edge of the sample. It has about 85% chondrules, 1% matrix, and 14% metal and troilite. The average size of chondrules and chondrule fragments is 0.4 mm; H- chondrite chondrules are estimated at 0.3 mm, and L chondrites at 0.6-0.8 mm [9]. Modally, the chondrite has (in vol%) 39 olivine 36.9 opx, 4.1 cpx, 6.2 plag, 0.3 chromite, 7.6 FeNi, and 5.9 FeS, very similar to that of H chondrites [10]. Mineralogically, olivine is homogeneous at Fa(sub)17, orthopyroxene is zoned from Wo(sub)0.5-0.8 Fs(sub)7-16, clinopyroxene (often rimming opx) is Wo(sub)31 Fs(sub)20, plagioclase is glassy to devitrified and albitic, and FeNi is homogeneous, with 10.5% Ni, 0.45 Co, <0.02 P. The oxygen isotope composition of the clast is delta ^18O = 3.40, delta ^17O = 1.86, Delta ^17O = 0.09, giving it lighter oxygen than any other ordinary chondrite (Fig. 1). The clast is classified as an H>3.5 ordinary chondrite with unusual oxygen isotopic composition. Carlisle Lakes-like Clast: Small clasts of Fe-rich, olivine-rich, material are found in the Mg-rich Weatherford host chondrite. Clasts range from 50 micrometers to 2.5 mm are completely recrystallized and equilibrated, and no chondrules were observed. Modally, they have (in vol%) 74.4 olivine, 5.2 clinopyroxene, 13.5 plagioclase, 0.7 chromite, and 6.2 pyrrhotite/pentlandite; no orthopyroxene was found. Olivine is Fa(sub)41, clinopyroxene is Wo(sub)43 Fs(sub)14, plagioclase is albitic, and chromite has high TiO2 (5.5%) and low Al2O3 (6%). Petrologically, clasts are very similar to Carlisle Lake-type chondrites, except they are highly equilibrated (type 6) and contain no chondrules. They could not be derived from the ordinary chondrite clast (with olivine at Fa(sub)17) or the host chondrite (with Fa(sub)3). Clasts large enough for oxygen isotopic analysis have not yet been found. Conclusions: Weatherford is a remarkable chondrite breccia. The ordinary chondrite clast in Weatherford differs from the one in Bencubbin in petrologic grade (Bencubbin clast is H<3.5) as well as oxygen isotope composition (Fig. 1). Both clasts have unusual oxygen that differs from any other ordinary chondrite. Carlisle Lakes-like clasts have not yet been found in Bencubbin. The presence of three kinds of unusual chondritic materials in Bencubbin/Weatherford improves our knowledge of primitive chondritic materials by allowing us to study meteorites not represented in our current sampling. References: [1] Mason B. and Nelen J. (1968) GCA, 32, 661-664. [2] Weisberg M. K. et al. (1990) Meteoritics, 25, 269-279. [3] Clayton R. N. and Mayeda T. K. (1978) GCA, 41, 1777-1790. [4] Prombo C. A. and Clayton R. N. (1985) Science, 230, 935-937. [5] Prinz M. et al. (1993) LPS XXIV, 1185-1186. [6] Lovering J. F. (1962) In Researches on Meteorites (C. B. Moore, ed.), 179-197. [7] Rubin A. E. and Kallemeyn G. W. (1989) GCA, 53, 3035-3044. [8] Weisberg M. K. et al. (1991) GCA, 55, 2657-2669. [9] Grossman J. N. et al. (1988) In Meteorites and the Early Solar System (J. F. Kerridge and M. S. Matthews, eds.), 619-659. [10] Van Schmus W. R. (1969) Earth Sci. Rev., 5, 145-184. Fig. 1 appears here in the hard copy. Bridges J. C.* Franchi I. A. Hutchison R. Pillinger C. T. A New Oxygen Reservoir? Cristobalite-bearing Clasts in Parnallee Three alpha-cristobalite-bearing clasts up to 1.0cm in diameter (CB1, CB2, CB3) have been identified in Parnallee (LL3). CB1 and CB2 consist of <30 micrometer cristobalite grains rimmed by minor augite in a 'roof tile' mosaic texture [1]. In CB3 cristobalite and augite together form 0.2-1.5mm long blebs enclosed by clinoenstatite, as in a cristobalite-bearing clast from Jilin [1]. CB3 contains glassy areas beside some of the cristobalite- bearing blebs. SEM examination and EDS analysis of the glassy areas show sub-micron high-Ca pyroxene grains in feldspathic mesostasis. The glassy areas have a bulk composition including 52-53 wt% SiO2, 12-13 wt% CaO and 22wt% Al2O3. The surrounding clinoenstatite is En95.7-98.5 with CaO <0.12 wt% and Al2O3 <0.15 wt%. It contains occasional lamellae of high-Ca pyroxene (Wo33- 45, En54-61) with 3.6-7.1wt% Al2O3. Augite, which rims the cristobalite is Wo31-40, En45-52. The Parnallee clasts do not contain the secondary high-Ca pyroxene, which surrounds a cristobalite grain in ALHA 76003 (L6) [2]. Replicate analyses gave delta^17O and delta^18O values of 8.60, 11.50 for CB1, and 8.56, 11.48 for CB2 (Fig. 1). The corresponding Delta^17O values [3] are 2.62 and 2.59. These unusually high delta^17O excesses preclude an origin through planetary [4] or impact-induced differentiation of ordinary chondrite material. CB1 and CB2 plot on an extension of a mixing line defined between the cristobalite grain from ALHA 76003 and its ordinary chondrite host [2]. The Parnallee clasts have escaped the metamorphism associated with secondary pyroxene crystallization in ALHA 76003 and so have undergone less isotopic exchange. Our values are probably close to the endmember of the mixing line. The Delta^17O values do fall within the range of the Carlisle Lakes-type chondrites, whole samples of which have Delta^17O values of 2.47 to 2.91 [5]. A line of slope 0.5 through the Carlisle Lakes field (Fig. 1) passes close to the Parnallee clasts. It has been suggested that early crystallized olivine in chondrules from Carlisle Lakes-type meteorites reacted with nebular vapors leading to the nucleation of a second generation of olivine [5]. Such extended olivine condensation would enhance the likelihood of SiO2 saturation in the residual nebular gas [6]. Gas-solid separation models predict that when nebular gas starts condensing SiO2 it should be extremely depleted in Al and Ca [1]. However their petrography suggests that the Parnallee clasts crystallized from an Al2O3 and CaO-bearing melt, which was presumably derived from SiO2-rich solid containing these oxides. The Parnallee clasts' origin is unclear but might be related through unknown fractionation processes to a larger O-isotope reservoir with which the Carlisle Lakes-type meteorites are associated. References: [1] Brigham C. A. et al. (1986) GCA, 50, 1655-1666. [2] Olsen E. J. et al. (1981) EPSL, 56, 82-88. [3] Clayton R. N. et al. (1991) GCA, 55, 2317-2337. [4] Ruzicka A. and Boynton W. V. (1992) Meteoritics, 27, 283. [5] Weisberg M. K. et al. (1991) GCA, 55, 2657-2669. [6] Grossman J. N. and Wasson J. T. (1983) GCA, 47, 759-771. Langenauer M.* Krahenbuhl U. Halogen Enrichments in Antarctic Meteorites and Their Relation to the Recovering Site We present the distribution of the elements F, Cl, Br, and I in 14 different H5 and H6 chondrites from Allan Hills, Elephant Moraine, Lewis Cliff Ice Tongue, and Thiel Mountains. The depth distribution was measured in stepwise removed layers from the surface into the interior. The dimensions of the investigated pieces from the individual meteorites were about 2 x 2 x 2 cm. All meteorites show higher halogen concentrations on their surfaces than in the interior, especially for I (enrichment factor up to 850 in the first few millimeters). This contamination happens mainly when the meteorites are lying on the surface of the Antarctic ice [1]. A contamination during the time the meteorite resides enclosed in the ice is not very likely because the ice has low halogen concentrations [2] and the temperatures are always below the freezing point of water. The different halogen species deposited onto the surface of a meteorite may diffuse into the interior; this process can be enhanced by water. The degree of contamination with F, Cl, Br, and I increases linearly with time for a meteorite and can be correlated neither with its degree of weathering nor with its terrestrial age. The major contamination source for F, Cl, and probably Br is airborne sea spray. It seems to be the same for all meteorites (same ratio of enrichment). The major contamination source for I is the biogenically CH(sub)3I produced in the sea [3,4]. This molecule and its products of oxidation in the Antarctic atmosphere (e.g., I(sub)2) can be transported over longer distances to the inner regions of Antarctica than airborne sea spray (inorganic aerosols) [3,5]. H chondrites found toward the center of Antarctica (e.g., Lewis Cliff or Thiel Mountains) are therefore less contaminated relative to I by F, Cl, and Br than those found near the coast (e.g., Allan Hills or Elephant Moraine). References: [1] Langenauer M. and Krahenbuhl U. (1993) Meteoritics, 28, 98- 104. [2] Legrand M. R. and Delmas R. J. (1985) Ann. Glaciol., 7, 20-25. [3] Heumann K. G. et al. (1987) GCA, 51, 2541-2547. [4] Heumann K. G. et al. (1990) GCA, 54, 2503-2506. [5] Legrand M. R. and Delmas R. J. (1988) JGR, 93D, 7153-7168. Fig. 1, which appears here in the hard copy, shows a map of Antarctica showing the four places from which H5 and H6 chondrites were studied. Numbers in parentheses give the quantity of analyzed meteorites. Krahenbuhl U.* Langenauer M. Allan Hills 82102: A Chondrite Collected Embedded in the Antarctic Ice From the over 10,000 meteorites collected in Antarctica only 4 have been found still partially embedded in the ice. Allan Hills 82102 was found in the Far Western ice field of the Allan Hills, and was collected together with the surrounding ice. Two splits (120 and 170 mg) of this meteorite were obtained for the investigation of the enrichments of the halogens F, Cl, Br, and I. One split originated from the part embedded in ice, and the other from the area exposed to the atmosphere. Both splits reveal almost identical enrichment profiles, assuming a pristine concentration as measured for other H chondrites [1]. This finding indicates that this meteorite was not embedded in ice throughout its terrestrial history. In comparison to other H chondrites it can be demonstrated that the observed enrichment must be the result of prolonged exposure to the Antarctic atmosphere. In addition, it has been estimated that small Antarctic meteorites (<100 g) can be transported on the ice by strong winds [2]. Therefore, it is not possible that the terrestrial age of Allan Hills 82102 of 11,000 years (determined by C-14 measurement [3]) is a measure for the age of the surrounding ice. References: [1] Langenauer M. and Krahenbuhl U. (1993) Meteoritics, 28, 98- 104. [2] Cassidy W. et al. (1992) Meteoritics, 27, 490-525. [3] Nishiizumi K. et al. (1989) Nature, 340, 550-552. Douglas C.* Wright I. P. Grady M. M. Romanek C. S. Pillinger C. T. Carbon Isotopic Measurements of Third-Generation Salts from LEW 85320 Preterrestrial salts associated with SNC meteorites indicate that water-based chemical activity probably took place on the parent body of these samples. One way of trying to understand the nature of these weathering processes is to look at situations on Earth where analogous effects might occur. For an initial investigation, an ordinary chondrite collected from Antarctica, LEW85320 (H5), has been selected. When this sample was found it was extensively coated with carbonate deposits (subsample ,39). After their removal, a second generation of minerals grew while the sample was in storage at Houston under an atmosphere of dry nitrogen (sub sample ,15). Subsequently, a third crop has formed (subsample ,103). Subsamples ,39 and ,15 have been identified as the magnesium carbonate nesquehonite with (<10%) hydromagnesite and barringtonite [1,2]. The carbonates in LEW85320 have been analyzed for their carbon content and stable isotopic compositions using stepped heating analysis and static mass spectrometry. The first generation carbonate (,39) was found to contain 9.7 wt% carbon, which is close to the theoretical value for pure nesquehonite (Mg (HCO3) (OH).2H2O). The carbon release profile over the temperature range 200-600 degrees C shows two well resolved peaks. The first (300- 425 degrees C) reached a maximum yield at 375 degrees C with a delta^13C = +4.4 per mil, while the second (450-550 degrees C) had a delta^13C of +5.4 per mil. A similar experiment with the second generation carbonate (,15), affords what appears to be a single release of carbon, although a small partially resolved second component may be present (400-550 degrees C). The overall release seen in subsample ,15 reaches a delta^13C = +3.6 per mil and accounts for 9.2 wt% of the sample. The third generation of carbonate (,103), which formed only after several years of storage, has an overall carbon content of 6.8 wt%. Upon stepped heating a double release of carbon is observed, similar to ,39, but with the highest delta^13C value observed being -0.8 per mil. The double carbon release seen in ,39 and ,103 may be associated with dehydration of the nesquehonite followed by decrepitation of the anhydrous salt, or it may be related to the salt's formation conditions since both ,39 and,103 formed over a period of years in fairly dry conditions, while ,15 was formed over a period of months in what could have been a more water-rich environment. Assuming the carbon in the carbonate derives from atmospheric CO2 (-7 to -8.0 per mil) and using the fractionation factor for calcite (no value is known for nesquehonite) carbonate formation temperatures can be calculated [2]. Results obtained in the current study give temperatures of 8.7 +- 4 degrees C, 24.5 +- 4 degrees C, and 57 +- 5 degrees C respectively for the three generations of carbonate. The first two temperatures, corresponding to formation in Antarctica and Houston respectively, although high, are believable while the third, also for Houston, is not. During prolonged storage in a dry nitrogen atmosphere, the CO2 source could have been something other than atmospheric and its availability could have been variable. The weathering products observed in this study are of a complex nature and may be more complicated by collection and storage conditions. Analyses on other weathered samples from different collection sites, both on Antarctica and elsewhere, together with a study of material stored in an environment controlled for the specific purpose of investigating the above problems are necessary. The understanding of different weathering regimes could ultimately lead to better interpretation of the martian climate, past and present, through the study of SNC meteorites. References: [1] Gooding (1992) Icarus, 99, 28-41 [2] Grady et al. (1989) Meteoritics, 24, 1-7 [3] Velbel et al. (1990) GCA, 55, 67- 76. Tuesday, July 20, 1993 Nucleosynthesis and Extinct Radioactivities Special Session with DPS 2:00 p.m. Theater Chair(s): G. W. Lugmair Leising M. D.* Endangered and Extinct Radioactivity Gamma ray spectroscopy holds great promise for probing nucleosynthesis in individual nucleosynthesis events, via observations of short-lived radioactivity, and for measuring global galactic nucleosynthesis today with detections of longer-lived radioactivity. Many of the astrophysical issues addressed by these observations are precisely those that must be understood in order to interpret observations of extinct radioactivity in meteorites. It was somewhat surprising that the former case was realized first for a Type II supernova, when both 56Co [1] and 57Co [2] were detected in SN 1987A. These provide unprecedented constraints on models of Type II explosions. Live 26Al in the galaxy might come from Type II supernovae and their progenitors, and if this is eventually shown to be the case, can constrain massive star evolution, supernova nucleosynthesis, the galactic Type II supernova rate, and even models of the chemical evolution of the galaxy [3]. Titanium-44 is produced primarily in the alpha-rich freezeout from nuclear statistical equilibrium, possibly in Type Ia [4] and almost certainly in Type II supernovae [5]. The galactic recurrence time of these events is comparable to the 44Ti lifetime, so we expect to be able to see at most a few otherwise unseen 44Ti remnants at any given time. No such remnants have been detected yet [6]. Very simple arguments lead to the expectation that about 4 x 10^-4 M(sub)solar mass of 44Ca are produced per century. The product of the supernova frequency times the 44Ti yield per event must equal this number. Even assuming that only the latest event would be seen, rates in excess of 2 century^-1 are ruled out at >=99% confidence by the gamma ray limits. Only rates less than 0.3 century^-1 are acceptable at >5% confidence, and this means that the yield per event must be >10^-3 M(sub)solar mass to produce the requisite 44Ca. Rates this low are incompatible with current estimates for Type II supernovae and yields this high are also very difficult to understand for any standard supernova models. This situation is puzzling. Searches for 60Fe gamma rays have also produced only upper limits, corresponding to a limit of 1.7 M(sub)solar mass in the present interstellar medium. Given the usual assumption of steady state between production and decay, the current rate of synthesis of 60Fe is less than 1.7 M(sub)solar mass/2.2 m.y. It has been suggested that a neutron-rich NSE occurs in small regions in both Type Ia supernovae supernovae and in core-collapse supernovae [7]. Either type might eject significant quantities of 60Fe. If we know the frequency of a particular type of 60Fe-producing event in the past few million years, then we can limit the mean 60Fe mass ejected per event. We have M(sub)ej (60Fe) <= 8 x 10^-5/R(SN) M(sub)solar mass where R(sub)SN is the frequency of the supernovae that eject 60Fe, in number per century. Type Ia supernovae might eject roughly 10^-4 M(sub)solar mass of 60Fe [8], which is very close to this limit. References: [1] Leising M. D. and Share G. H. (1990) Astrophys. J., 357, 638. [2] Kurfess J. D. et al. (1992) Astrophys. J. Lett., 399, L137. [3] Clayton D. D. et al. (1993) Astrophys. J. Lett., submitted. [4] Nomoto K. et al. (1984) Astrophys. J., 286, 644. [5] Woosley S. E. (1988) Proc. Astron. Soc. Aust., 7, 355. [6] Leising M. D. and Share G. H. (1993) Astrophys. J., submitted. [7] Hartmann D. H. et al. (1985) Astrophys. J., 297, 837. [8] Woosley S. E. (1991) In Gamma-Ray Line Astrophysics (P. Durouchoux and N. Prantzos, eds.), 270-290, AIP Conf. Proc. No. 232, New York. Meyer B. S.* R-Process Extinct Radioactivities Type II supernovae are thought to be the result of the core collapse of massive stars. These catastrophic events disrupt much of the presupernova star, but often leave a roughly 1.4-solar-mass neutron star as a remnant. Several seconds after the collapse of the core, a high-entropy wind begins to blow from the surface of the nascent neutron star [1]. This wind is an ideal site for the r process of nucleosynthesis [2]. The nuclear abundance distribution produced in this wind agrees well with the solar-system r-process abundance distribution. In addition, the r-process yield in this wind is some 10^-4 solar masses, in good agreement with the amount expected from galactic chemical evolution arguments. One important implication is that the r process is not rare--most type II supernovae probably produce r-process nuclei. This contrasts with low-entropy supernova sites for the r process that typically produce 0.1 solar masses of r-process material per event and therefore must be rare [e.g., 3]. Several short-lived radioactive isotopes are produced in the r process. These are ^129I (half-life 15.7 m.y.), ^244Pu (half-life 80.8 m.y.), ^247Cm (half- life 15.6 m.y.), and ^107Pd (half-life 6.5 m.y.). Clear evidence exists for the presence of live ^129I [4], ^244Pu [5], and ^107Pd [6], while only an upper limit exists for ^247Cm [7]. Because of the short lifetimes of these nuclei, knowledge of their abundances in the early solar system and of their production in nucleosynthetic events yields important information about the chronology of the first few million years of the solar system's history and the last nucleosynthetic events contributing to the solar abundances. I have computed the production ratios of these short-lived radioactive isotopes in one particular model of the high-entropy r process. The results are ^107Pd/^110Pd = 0.66, ^129I/^127I = 2.03, ^244Pu/^238U = 0.36, and ^247Cm/^238U = 0.23. These ratios do not differ greatly from those already present in the literature, and the discrepancy between the free-decay timescales inferred from ^129I and ^244Pu (roughly 100 m.y.) and that from ^26Al (a few million years) remains. The production ratios presented above are for one particular r-process model. The ratios are sensitive to the details of the astrophysical model, for example, the velocity of the wind. Also, in the r process several components of differing degrees of neutron richness add together to give the final r- process abundances. It is not clear that the weighting of these different components is unique. The production ratios will be sensitive to the weighting of the components. The ratios are also sensitive to the properties of very- neutron-rich nuclei that are only known from theoretical nuclear-structure models [e.g., 8]. References: [1]} Duncan R. C. et al. (1986) Astrophys. J., 309, 141-160. [2] Meyer B. S. et al. (1992) Astrophys. J., 399, 656-664. [3] Hillebrandt W. et al. (1976) Astron. Astrophys., 52, 63-68. [4] Jeffery P. M. and Reynolds J. H. (1961) JGR, 66, 3582-3583. [5] Alexander E. C. et al. (1971) Science, 172, 837-840. [6] Kaiser T. and Wasserburg G. J. (1983) GCA, 47, 43-58. [7] Chen J. H. and Wasserburg G. J. (1981) EPSL, 52, 1-15. [8] Meyer B. S. et al. (1989) Phys. Rev. C., 39, 1876-1882. Howard W. M.* Radioactive Nuclides and the Astrophysical P Process The astrophysical p-process is the conversion via photodisintegration reactions and proton-capture reactions of a solar-system-like distribution of s- and r-process nuclei into the proton-rich p-nuclei [1,3]. This conversion can only take place on a hydrodynamical timescale when the radiation temperature is extremely high (T > 10^9 K). Type II supernovae are probably major contributors to the bulk of the solar-system p-nuclei because they contain zones with enrichments of s-process elements that are heated to such high temperatures by the expanding supernova shock wave. Type Ia supernovae may also contribute [1,2] if the surface composition of the exploding white dwarf is enriched in s-process elements. The p-processs produces in significant quantity several interesting radioactive nuclides with relatively long half-lives, including ^92Nb (tau(sub)1/2: 3.6 10^7 yr), ^97Tc (tau(sub)1/2: 2.6 10^6 yr), ^98Tc (tau(sub)1/2 4.2 10^6 yr) and ^146Sm (tau(sub)1/2: 1.08 10^8 yr). In principle, if the production rates of these radioactive nuclides are known, the measurement of their extinct radioactivity in meteorities can have them serve as chronometers for the astrophysical p-process and for supernovae nucleosynthesis. We will discuss the details of the production of these radionuclides in the astrophysical p-process and the implications for obeservation of their extinction in meteorites. Of all the possible p-process chronometers, ^146Sm is the most interesting, since evidence for its decay has been observed in meteorites. We will discuss in detail the production of ^146Sm and its dependence on the astrophysical environment and on nuclear physics quantities. For example, the production of ^146Sm critically depends on the competition between (gamma,alpha) and (gamma,n) reactions on ^148Gd and ^150Gd. We will discuss the implications of the measurements of the extinct ^146Sm in meteorites for the astrophysical pprocess. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48. References: [1] Howard W. M. et al. (1991) Ap. J. Lett., 373, L5. [2] Howard W. M. and Meyer B. S. (1992) Nuclei in the Cosmos, Bristol, 607-612. [3] Lambert D. L. (1992) Astron. Astrophys. Rev., 3, 201,. [4] Woosley S. E. and Howard W. M.(1978) Ap. J. Suppl., 36, 285, 1978. Clayton D. D.* Extinct Radioactivity and Evolution of the Galactic Disk To understand the meaning of extinct radioactivity for the origin of the solar system, it is first necessary to compare the observed initial meteoritic concentrations of those nuclei to the concentrations that are expected in the mean interstellar medium. Any differences are attributed to special circumstances of solar birth. Traditionally one estimates the concentration ratio of extinct activity Z* to stable nuclide Z by Z*/Z = (y*/y)(tau/T(sub)G), where y* and y are the stellar yields of the two nuclei, tau is the mean lifetime of the extinct nucleus, and T(sub)G is the age of the Galaxy. However, by considering the history of growth of the mass of the Galactic disk by metal-poor infall, I have demonstrated by analytic solutions [1,2] that the mean radioactivity is enhanced relative to stable nuclei by the past infall. For the analytic family f(t)/M(sub)G (t) = k/(t+ delta) relating infall rate f(t) to gas mass MG(t) this increase is by a factor (k+1). The mean interstellar ratio becomes Z*/Z =(k+1 )(y*/y) (tau/T(sub)G). This enhancement impacts every known case. It becomes possible to account for the observed interstellar 26Al gamma emission by supernova nucleosynthesis if the infall parameter k=3-5. But by the same token the required free decay interval for ^129I is increased by about 40Myr. I will try to clarify this theoretical development. References: [1] Clayton D. D. (1985) in Challenges and new developments in nucleosynthesis, (W. D. Arnett and J. W. Truran, eds.), Univ. Chicago. [2] Clayton D. D. (1988) Monthly Notices R. Astron. Soc., 234, 1-36. Cameron A.* Extinct Radioactivities and Star Formation No abstract available. Arnould M.* Paulus G. Meynet G. Wolf-Rayet Stars and the Isotopic Anomaly Connection Isotopic anomalies are now known to be carried by high-temperature inclusions of primitive meteorites that formed from solar reservoirs out of equilibrium with the rest of the solar nebula, as well as by various types of grains (diamond, graphite, SiC) that are considered to be of circumstellar origin, and have survived the process of incorporation into the solar system (see e.g. [1] for a recent review). Such anomalies provide new clues to many important astrophysical problems, and raise the question of their nucleosynthetic origin. In fact, they offer the exciting perspective of confronting abundance observations with nucleosynthesis models for a very limited number of events, even possibly a single one. This situation is in marked contrast with the one encountered when trying to understand the bulk solar system composition. Up to now, Red Giant stars, massive mass loosing objects (of the Wolf-Rayet type), novae or supernovae have been proposed as possible contributors to the observed anomalies. In this paper, we revisit the role that could possibly be played in that respect by Wolf-Rayet (WR) stars. Wolf-Rayet stars are appealing isotopic anomaly contributors for many reasons. In particular (1) they are observed to loose mass at very large rates that can exceed 10^-5M solar masses yr^-l, the ejected material being contaminated with the products of hydrogen and helium burning, and (2) certain WR stars are known to make dust episodically in their winds [e.g., 2]. In addition, the role of WR stars would be well in line with the "bing-bang" model for the isotopic anomalies promoted by Reeves [3]. The aim of this contribution is to extent and update previous calculations [4,5] of the isotopic anomalies that could be carried by the wind of WR stars of various masses and initial compositions during different phases of their evolution, those anomalies possibly loading circumstellar WR grains. The calculation of the WR wind composition is performed on grounds of detailed stellar evolutionary models that incorporate extended nuclear reaction networks, as well as recent improvements in our knowledge of various basic physical ingredients, like mass loss rates, opacities, or nuclear reaction rates. Results will be presented for various radionuclides with lifetimes in excess of ~10^5 yr, which are considered to be responsible for certain observed anomalies, or which could lead to anomalies that remain unobserved at present. Isotopic patterns for the elements ranging from carbon to lead will also be presented. Those predictions will be confronted with existing data, or will help unravel cases of potential interest for further laboratory quest. References: [1] Harper C. L. Jr. (1992) In Nuclei in the Cosmos II (F. Kappeler and K. Wisshak, eds.), 113-126, IOP Publ. Co. [2] Williams P. M. et al. (1992) Mon. Not. R. Astron. Soc., 258, 461-475. [3] Reeves H. (1978) In Protostars and Planets (T. Gehrels, ed.), 339-426, Univ. of Arizona. [4] Arnould M. and Prantzos N. (1986) In Nucleosynthesis and Its Implications on Nuclear and Particle Physics, (J. Audouze and N. Mathieu, eds.), 363-372, Reidel. [5] Meynet G. and Arnould M. (1993) In Origin and Evolution of the Elements (N. Prantzos et al., eds.), Cambridge, in press. Tuesday, July 20, 1993 Achondrite-Lunar Melange 2:00 p.m. Cascade Ballroom Chair(s): M. M. Grady K. Marti Torigoye N.* Misawa K. Tatsumoto M. U-Th-Pb and Sm-Nd Isotopic Systematics of the Goalpara Ureilite One of the interesting features of ureilites is the light REE-enriched component that is dissolved by HNO3 leaching [1,2]. In this work, we performed acid-leaching of several mineral fractions from Goalpara ureilite for U-Th-Pb and Sm-Nd analyses. Olivine and pyroxene grains were hand-picked from 150-300- micrometer-sized fraction. Because they still contained carbon and metal sulfide they were further crushed to <63 micrometers and metal was removed with a hand magnet. These separates and whole-rock powders were washed by ethanol, and leached in 0.01N HBr, 1N HNO3, and in some cases, 7N HNO3. Concentrations of U, Th, and Pb in residues are 0.05-0.3 ppb, 0.1-0.7 ppb, and 5-100 ppb, respectively, corresponding to <=0.01X CI chondrites. Lead isotopic compositions of the residues are less radiogenic and close to Canon Diablo troilite (CDT) Pb [3] (Fig. 1). The U-Pb and Th-Pb ages of all the fractions are older than 4.5 Ga, indicating terrestrial Pb contamination (MT). Because of low concentration of U, Th, and Pb, a small amount of Pb can have a significant effect on the U-Pb and Th-Pb model ages. 238U/204Pb (mu) value of the least contaminated residue is 3, which is higher than mu (0.14-0.5) value of carbonaceous chondrites [3,4]. The higher mu value may be due to either volatile depletion by nebula fractionation or to depletion of Pb during segregation of sulfide that occurred prior to the formation of ureilite as an ultramafic cumulate. The Sm and Nd abundances in the residues are also extremely low; 0.4-2 ppb and 1-2.5 ppb, respectively, corresponding to 0.002-0.01X CI chondritic abundances. All the residues show high 147Sm/144Nd ratios (0.23 ~ 0.44), and the fraction with the highest Sm/Nd plots on the 4.55 Ga chondritic isochron (Fig. 2). The 1N HNO3 leachates do not contain light-REE-enriched components, except for the samples containing black metal-carbon phases, which also contain a large amount of terrestrial Pb in the residual fractions. Therefore, interstitial carbon-metal phases may have adsorbed terrestrial contamination of the incompatible elements, which are significantly depleted in the ureilites. References: [1] Boynton W. V. et al. (1976) GCA ,40, 1439-1447. [2] Goodrich C. A. et al. (1991) GCA, 55, 829-848. [3] M. Tatsumoto et al. (1973) Science, 180, 1278-1283. [4] Tatsumoto M. et al. (1976) GCA, 40, 617-634. Nyquist L. E.* Wiesmann H. Bansal B. Shih C.-Y. Isotopic Studies of Angrite LEW 86010 and the Early History of Its Parent Body We summarize Mn-Cr, Rb-Sr, and Sm-Nd investigations of LEW 86010 (LEW) and Angra dos Reis (ADOR) and present a synthesis, including data for other isotopic systems. Samarium-neodymium investigation of LEW 86010 showed some surficial terrestrial contamination and/or Antarctic weathering products that could be removed by leaching in 2N HCl. A conventional Sm-Nd isochron anchored by data for pyroxene and phosphate, containing ~97% of the REE, yields an age of 4.538 +/- 0.018 Ga (2 sigma) using the Williamson [1] regression; the error limit increases to +/-0.033 Ga for the York [2] regression. Including data for leached plagioclase and olivine doubles the range in Sm/Nd, lowers the age to 4.532 Ga, but increases the uncertainty to +/-0.040 Ga. The isochron age agrees with Sm-Nd ages of ~4.55-4.56 Ga previously reported for ADOR [3,4] and LEW [5]. Manganese-chromium isotopic studies show excess ^53Cr from extinct ^53Mn* (t(sub)1/2 = 3.7 Ma) in the olivine of both LEW [6,7] and ADOR [7] corresponding to initial ^53Mn*/^56Mn = 1.3-1.4 x 10^-6. Angrite olivine closed to Cr isotopic homogenization ~18 Ma after Allende inclusions formed with ^53Mn*/^55Mn = 4.4 +/- 1.1 x 10^-5 [8]. The Mn-Cr formation interval is ~2x longer than the ~8-Ma difference between the single stage Pb-Pb ages of Allende and the angrites [5,9,10]. Angrite PbPb ages of 4.551 +/- 0.004 [9] to 4.5578 +/- 0.0005 [5] have been interpreted as limiting the time between major volatile element loss from the angrite parent body (APB) and crystallization of the angrites to <=2 Ma [5]. The Mn-Cr and Sm-Nd ages both date element partitioning among crystallizing mineral phases, the end point for the Mn-Cr formation age, and the starting point for the Sm-Nd age. Their sum should equal the age of Allende, and is 4.556 +/- 0.018 Ga using our preferred Sm-Nd age (4.538 Ga). This agrees adequately with Allende Pb-Pb ages of 4.559 +/- 0.015 Ga [9] to 4.566 +/- 0.002 Ga [10]. A ^146Sm-^142Nd isochron gives initial ^146Sm/^144Sm = 0.0076 +/- 0.0009 for LEW, corresponding to solar system initial (^146Sm/^144Sm)(sub)o = ~0.0080 to ~0.0086, for LEW crystallization 8 and 18 Ma after Allende respectively. The latter value is especially consistent with "high" values of ^146Sm/^144Sm for the eucrite Ibitira (0.0090 +/- 0.0010 [11]) and a silicate inclusion from the Caddo IAB iron (0.0099 +/- 0.0021 [12]). Strontium-87/strontium-86 measurements for angrite minerals with low Rb/Sr give I^87(sub)Sr = 0.698972 +/- 8 and 0.698970 +/- 18 for LEW and ADOR respectively, relative to ^87Sr/^86Sr = 0.71025 for NBS987, in agreement with Lugmair and Galer [5]. I^87(sub)Sr for the angrites is thus ~0.00011 higher than ALL, measured for Allende inclusions [13,14], corresponding to ~9 Ma of growth in a solar nebula with a CI chondrite value of ^87Rb/^86Sr = 0.91, or 5 Ma in a nebula with ^87Rb/^86Sr = 1.51, as in the solar photosphere [15]. The interval of ~18 Ma between formation of Allende and closure of the Mn-Cr system in angrites implies an average Rb/Sr ratio ~2-3x lower than the nebular value, probably reflecting an episode of volatile loss from the APB during this time. A "best average" for four eucrite clasts analyzed in our laboratory gives I^87(sub)Sr (4.558 Ga) = 0.699002 +/- 16, corresponding to growth of radiogenic ^87Sr* for 2.4 +/- 1.4 Ma in a solar nebula with a CI Rb/Sr ratio, or 1.4 +/- 0.9 Ma in a nebula with the solar photospheric Rb/Sr ratio. References: [1] Williamson J. H. (1968) Canadian J. Phys., 46, 1845-1847. [2] York D. (1966) Canadian J. Phys., 44, 1079-1086. [3] Lugmair G. W. and Marti K. (1977) EPSL, 35, 273-284. [4] Jacobsen S. B. and Wasserburg G. J. (1984) EPSL, 67, 137-150. [5] Lugmair G. W. and Galer S. J. G. (1992) GCA, 56, 1673-1694. [6] Nyquist L. E. et al. (1991) LPS XXII, 989-990, and unpublished data. [7] Lugmair G. W. et al. (1992) LPS XXIII, 823-824. [8] Birck J.-L. and Allegre C. J. (1988) Nature, 331, 579-584. [9] Chen J. H. and Wasserburg G. J. (1981) EPSL, 52, 1-15. [10] Gopel C. et al. (1991) Meteoritics, 26, 338. [11] Prinzhofer et al. (1989) Astrophys. J., 344, L81-L84. [12] Stewart B. W. (1993) LPS XXIV, 1359-1360. [13] Gray et al. (1973) Icarus, 20, 213-239. [14] Podosek F. A. et al. (1991) GCA, 55, 1083-1110. [15] Anders E.and Grevasse N. (1989) GCA, 53, 197-214. Yanai K.* Angrites LEW 87051 and Asuka 881371: Similarities and Differences Both angrite meteorites Lewis Cliff (LEW) 87051 (U.S. collection) and Asuka (A) 881371 (Japanese collection) were collected from Antarctica. The collecting sites of the angrites are Lewis Cliff Ice Tongue, 84 degrees 17 minutes S and 161 degrees OO minutes E in the Transantarctic Mountains, and Asuka Station, 72 degrees 50 minutes S and 24 degrees 30 minutes E, in Queen Maud Land, respectively. Therefore the two localities are separated by almost 2500 km. LEW87051: LEW87051 is 0.6 g in original weight and 1 x 0. 7 x 0.5 cm in diameter; it is a tiny individual achondritic meteorite, completely covered with a black fusion crust [1]. Petrographically, this specimen shows the typical porphyritic texture of olivine with subequal amount of groundmass plagioclase laths and interstitial pyroxene with little opaque. Plagioclase laths, 0.02 x 0.3 mm, are in a subparallel arrangement (Fig. 1). Pyroxene is titanian fassaite showing weak pleochroic of purplish tint, the average composition of which is Wo50.2, containing 6-9% A1203 and 2-6% TiO2, and range En1-29, Fs21-55, and Wo44-53. Olivine contiains variable composition, average Fo21, and ranged Fo8-91 correspond with Fe-rich rim to Mg-rich core. Plagioclase is almost pure anorthite (An99-100). Asuka-881371: A-881371 is 11 g in original weight and 2.0 x 1.6 x 1.6 in diameter. It is a rounded stone almost completely covered with a dull black (not shiny) fusion crust. Pale green, relatively coarse olivine crystals can be seen on the exposed interior surface. Petrographyically, the Asuka angrite shows an unbrecciated and typical ophitic texture with less porphyritic olivine crystal (xenocryst?), and consists mostly of euhedral plagioclase, integranular fassaite, and olivine with traced opaque and spinel (Fig. 2). Pyroxene is titanium fassaite with pleochroic brown in the rim, which is average composition Wo52 containing high CaO (over 22%), A12O3 (3.5-9.9%) and TiO2 (1-5%), and ranged EnO-29, Fs18-50, and Wo48-55. Olivine contains variable composition ranged Fo2.8-90 with Fe-rich rim and extremely Mg-rich core; however, most of the olivine is in the range Fo57-72. Plagioclase is almost pure anorthite (An97-100). Conclusion: Macroscopically, LEW87051 and Asuka angrites are clearly recognized as individual specimens for they are completely covered with a black fusion crust. The great distance between the two localities strongly supports the fact. Therefore, it seems that both angrites have individually fallen on the Antarctic continent and they are not a pair. Nevertheless, both angrites are petrographically very similar, especially their mineralogy and chemical composition, except for some differences in their texture (Figs. 1 and 2). Two angrites LEW87051 and A-881371 seem to have a close genetic relationship on the parent body. LEW87051 and A-881371 angrites are quite difference from Angra dos Reis (ADOR) and LEW86010 angrites from their petrography, mineralogy, and compositions [2,3]. References: [1] Mason B. (1989) Antarctic Meteorite Newsletter 12, No. 1, 15. [2] Printz N. et al. (1977) EPSL, 35, 317-330. [3] Mason B. (1988) Antarctic Meteorite Newsletter 10, No. 2, 32 . Kim Y.* Marti K. Acapulco Nitrogen Isotopic Systematics and Genetic Relationships Among Meteorites Nitrogen shows a large isotopic variation in bulk meteorites. Extreme isotopic variations in some carbonaceous chondrites probably reflect incomplete mixing in the solar nebula, since carbon-bearing presolar grains were found in meteorites with unusual nitrogen isotopic signatures [1]. As nitrogen isotopic characteristics of solar system materials are studied, the strengths and limitations of a single isotope ratio have to be explored in discussions of origin. Several recent investigations were directed toward understanding the distribution and behavior of nitrogen in different classes of meteorites. We report nitrogen isotopic structures in separated phases of chondrites, a primitive achondrite, and an iron meteorite and implications regarding genetic relationships. The texture of the Acapulco meteorite reveals extensive solid-state recrystallization at higher temperature. Metallic Fe-Ni has a fine structure akin to the Widmanstatten structure. Nitrogen measurements on individual mineral separates of Acapulco reveal that the metal phase is the major carrier of N (13 +/- 3ppm) with a very light N component (delta ^15N <= -150 per mil), while all silicates carry the heavy component (delta ^15N + 10 per mil) [2], demonstrating that N was not equilibrated between metal and silicates. Furthermore, the opaque mineral chromite also shows isotopic disequilibrium with isotopically light N (delta ^15N <= -70 per mil). Is this a distinct feature of the Acapulco parent body, or can we recognize possible relationships with other parent bodies? Possible genetic links with H chondrites were explored in a study of metal separates of three ordinary H chondrites: Dhajala (H3.8), Foreast Vale (H4), and Ste. Marguerite (H4). However, their N concentrations and isotopic signatures are very different from those of Acapulco metal. In the case of Dhajala, which is assumed to be a primitive object but has no "presolar" signatures [3], the N concentration in the metal is very low (1.7ppm) and no isotopic variations are observed between bulk (mean delta ^15N + 1.1 per mil) and metal (0.0 per mil) except for a spallation component in the melt step. For a study of possible relationships of Acapulco metal with iron meteorites, nitrogen was measured on the etched metallic phases of Acapulco and Cape York (IIIAB), which is considered to be a equilibrated meteorite. The results show that nitrogen is enriched in the Ni-rich phase (taenite and/or plessite) with 690 ppm N in Cape York and 87 ppm N in Acapulco. The large nitrogen concentration in these samples may indicate the presence of N phases in etched metals. However, the mean delta ^15N values of N-enriched phases in Cape York (-80 per mil) are different from that of Acapulco taenite (-130 per mil), and Cape York metal separates do not reveal any lighter nitrogen components. Oxygen isotopic signatures have been used to recognize possible genetic relationships between different classes of meteorites. Unfortunately, oxygen systematics are not available in metal phases, and if silicate inclusions are used for this purpose, the question of phase equilibration needs to be considered. We have shown that, based on nitrogen in Acapulco, an association of chromite and metal is indicated. Therefore, chromite may provide a useful link in this approach. Based on oxygen isotopes an association between IIIA and B irons and pallasites was observed [4] and their nitrogen isotopic signatures are very light [5]. Oxygen isotopes suggest a genetic link between IIE iron meteorites and H chondrites, and the measured nitrogen isotopic signatures in metals of three H chondrites are indeed close to those in IIE iron [5]. References: [1] Zinner E. et al. (1989) GCA, 53, 3273-3290. [2] Kim Y. and Marti K. (1993) LPS XXIV, 801-802. [3] Alexander C. M. O' D. et al. (1990) EPSL, 99, 220-229. [4] Clayton R. N. et al. (1986) LPS XVII, 141. [5] Prombo C. A. (1984) Ph.D. thesis, Univ. of Chicago. Grady M. M.* Franchi I. A. Pillinger C. T. Carbon and Nitrogen Chemistry of Lodranites: Relationship to Acapulco? Recent studies on the mineralogy, petrology, and oxygen isotopic composition of lodranites and acapulcoites indicate that these meteorites are probably derived from a common parent body, but experienced different degrees of partial melting [1,2]. Ar-Ar chronometry implies that lodranites were heated ca. 100 degrees C higher than acapulcoites, and cooled more slowly [3], however measurement of nitrogen and xenon in Acapulco [4,5] shows that volatiles are not equilibrated between different phases within the meteorite, hence its thermal history has been complex. The aim of this study is to determine the carbon and nitrogen chemistry of lodranites, for comparison with Acapulco, to indicate the effect that differing thermal histories might have had on the volatile inventories of these meteorites. The carbon chemistry of Acapulco has been described previously [6]. The meteorite contains ca. 400 ppm indigenous carbon, distributed between two major phases: graphite and carbides. Graphite has been identified petrographically in Acapulco [7], where it is intimately associated with metal. In contrast, both Lodran and MAC 88177 contain much lower quantities of indigenous carbon: approximately 100 ppm and 38 ppm respectively, released in decreasing amounts up to 1200 degrees C. In Lodran, delta^13C rises almost monotonically, from -25 per mil at 600 degrees C to -12 per mil at 1200 degrees C; total delta^13C is ca. -23 per mil. Neither meteorite shows evidence for the occurrence of graphite. Nitrogen released by pyrolysis of Acapulco totals ca. 2.8 ppm [4,5], and is resolvable into two components, with delta^15N ca. +10 per mil and -120 per mil [8]. The first component is, as yet, unidentified, but the second is believed to be associated with the metal fraction [8]. The procedure used herein, of several combustion steps below 500 degrees C to remove contaminants, followed by high resolution combustion up to 1200 degrees C, would also resolve discrete nitrogen-bearing components, if present. Analysis of whole-rock Lodran yielded 17.0 ppm nitrogen, with delta^15N ca. +4 per mil. A prominent release of nitrogen occurred between 650 degrees C and 900 degrees C, 50% of the total, with delta^15N varying between ca. +3 per mil and +9 per mil. Lodran is very different from Acapulco in both its carbon and nitrogen chemistry. There is little evidence for the presence of graphite in the former meteorite and it does not appear to contain the component of isotopically light nitrogen that is so abundant in Acapulco. However, Lodran does manifest a nitrogen- bearing component with intermediate isotopic composition, the location of which, in the metal or silicate portion of the meteorite is, as yet, unknown. Since Lodran has apparently experienced an elevated temperature regime compared with Acapulco, it might be possible that nitrogen has been remobilized and mixed in the former meteorite, leading to erasure of the characteristic isotopically light signature. Lodran has a higher whole-rock nitrogen abundance than Acapulco, thus it is unlikely that the presence of heavier nitrogen in Lodran is simply a result of fractionation of a reservoir during open system heating. References: [1] McCoy T. J. et al. (1993) LPS XXIV, 945-946. [2] Clayton R. N. et al. (1992) LPS XXIII, 231-232. [3] Bogard D. D. et al. (1993) LPS XXIV, 141-142. [4] Sturgeon G. and Marti K. (1990) LPS XXI, 1220-1221. [5] Becker R. H. (1991) LPS XXII, 69- 70. [6] Grady M. M. and Pillinger C. T. (1986) GCA, 50, 255-263, [7] Palme H. et al. (1981) GCA, 45, 727-752. [8] Kim Y. et al. (1992) LPS XXIII, 691-692. Field S. W.* Lindstrom M. M. Mittlefehldt D. W. Petrology and Geochemistry of Acapulco- and Lodran-like Achondrites Primitive achondrites are meteorites that have mineral and bulk compositions similar to chondrites, but have non-chondritic textures. These achondrites were metamorphosed at high temperatures, perhaps up to that of the Fe-FeS eutectic or chondrite silicate solidus [1]. We have initiated geochemical and petrologic study of several Acapulco- and Lodran-like achondrites in order to test petrogenetic models based largely on petrologic arguments. We have studied the following meteorites: ALHA81187 and ALHA81261 (Acapulco-like), LEW88280 and MAC88177 (Lodran-like) and EET84302 (transitional) [1]. Of our Acapulco-like achondrites, we have finished petrologic characterization only on ALHA81187. Our thin section is distinct from that studied by [1] in that orthopyroxene is the dominant silicate, and we found no plagioclase. The low plagioclase content is like that of Lodran-like achondrites, but our INM data (below) suggest that the thin section is unrepresentative. LEW88280 and MAC88177 are medium-grained, granular rocks with metal and troilite occurring as inclusions in silicates, as thin veins cutting silicates and as discrete grains of intergrown kamacite and troilite. Our thin section of EET84302 is metal, sulfide, and chromite. Clinopyroxene and plagioclase occur in minor amounts. Orthopyroxene grains contain abundant metal inclusions in linear trains. The texture is similar to cumulate sulfide textures found in some terrestrial igneous rocks. Acapulco-like achondrites have been suggested to be high grade metamorphic rocks in which partial melting of Fe-FeS and phosphates occurred [2], although the melts may not have left the parent rock. Our and literature [3-6] INAA data generally agree with this interpretation: Sm/Sc ratios of Acapulco-like achondrites are between 0.8-2 times H chondrites, and Na/Sc ratios are between ~0.7-1.2 times H chondrites indicating that neither silicate nor "phosphate" partial melts were lost from the rocks. Siderophile and chalcophile elements are fractionated. Y-74063 has high Se/Co and low Ir/Ni ratios [6], while ALHA81187 has low Se/Co and high Ir/Ni ratios. This variation is consistent with either fractionation by partial melting in the Fe-Ni-S system, or with heterogeneous distribution of metal and troilite in these achondrites. This can be tested through additional analyses of the meteorites. If the samples of Y-74063 and ALHA81187 are representative of these achondrites, then the results suggest that mobilization of Fe-FeS eutectic melts occurred. The Lodran-like achondrites are believed to be partial melting residues [7]. The trace lithophile element data on Lodran [8], LEW88280, and MAC 88177 are compatible with this interpretation. Highly incompatible elements are depleted relative to more compatible elements such as Sc: Sm/Sc ratios are ~0.1-0.5 times, Na/Sc ratios are 0.05-0.1 time, and Eu/Sc ratios are 0.05-0.4 times H chondrites. EET84302 is classified as a Lodran-like achondrite, but is recognized as being transitional to the Acapulco-like achondrites [1]. Our INAA data show that EET84302 has not lost a silicate partial melt: ratios of Sm/Sc and Na/Sc are ~1 time H chondrites, and Eu/Sc is about 1.8 times H chondrites. In lithophile trace element contents, EET84302 is identical to the Acapulco-like achondrites. Our sample of EET84302 was metal-rich (~40% metal) and chromite-rich (~3% based on INAA and EMPA data), and in this regard is distinct from Acapulco-like achondrites. These modal differences will have no effect on lithophile element ratios, however. References: [1] McCoy et al. (1993) LPS XXIV, 945. [2] McCoy et al. (1992) Meteoritics, 27, 361. [3] Palme et al. (1981) GCA, 45, 727. [4] Schultz et al. (1982) EPSL, 61, 23. [5] Kallemeyn and Wasson (1985) GCA, 49, 261. [6] Kimura et al. (1992) Proc. NIPR Sym. Ant. Met., 5, 165. [7] Bild and Wasson (1976) Min. Mag., 40, 721. [8] Fukuoka et al. (1978) LPS IX, 356. Herzog G. F.* Xue S. Klein J. Juenemann D. Middleton R. 26Al and 10Be Activities of Lodranites and Winona Noble gas measurements by [1] indicate that four lodranites LEW 88280, Lodran (a fall), MAC 88177, and Yamato 791491 have the same cosmic ray exposure age of a few million years. The elevated ^22Ne/^21Ne ratios of these lodranites, from 1.22 to 1.28 [1], suggest that shielding was light and production rates appreciably lower than in average chondrites. Cosmic-ray irradiation in space for, say, 4 My would bring ^26Al and ^10Be to within 2% and 16% of their respective saturation values. Thus measurement of ^26Al may provide information about production rates and shielding and ^10Be about exposure age. We separated magnetically metal- and silicate-rich material from the four lodranites mentioned above and from Winona. The ^26Al and/or ^10Be activities (Table 1) were measured by accelerator mass spectrometry [2] with the statistical 1-sigma precision shown; the activities are thought to have an overall accuracy of 6-8%. Although the metal phases were etched with HF, they retained some silicate. To get a quantitative indication of the amounts of silicate present, the Mg concentrations in aliquots of the dissolved metal samples (Table 1) were measured by ICP/MS. The Mg, Al, Ca, Ti, Mn, and Fe contents of the silicate phases were determined by DCP emission spectrometry [3]. The measured activities in silicates from LEW 88280, Lodran, and Y 791491 resemble one another closely: The average ^26Al and ^10Be activities are 50.9 and 16.7 dpm/kg compared to estimated production rates of about 55 and 23 dpm/kg. These results lead to an exposure age of ~3.3 My, but do not indicate substantial lowering of production rates. The ^26Al and ^10Be contents of MAC 88177 are about half the values expected at saturation under normal shielding and are lower than those in the other three lodranites. These results are consistent with the very light shielding inferred from the exceptionally high ^22Ne/^21Ne ratio of 1.28, and perhaps with some lowering due to terrestrial age. Kirsten et al. [4] found a ^21Ne content of 25.2 x 10^-8 cm^3 STP/g and a low ^22Ne/^21Ne ratio of 1.071 for Winona, a find of uncertain age with heavily weathered metal. The measured ^10Be activities are also low, about half the estimated production rates. A ^21Ne production rate of about 0.314 x 10^-8 cm^3 STP/g- My would be expected under normal shielding in a body with the bulk composition of Winona [5,6]. If we assume a short terrestrial age and a constant ratio of ^10Be to ^21Ne production [7], then an exposure age on the order of 150 My is implied. Use of the measured ^26Al activity in the same way gives a shorter but more uncertain exposure age of ~110 My. The high ^26Al activity in Winona "metal" may indicate the presence of sulfide [5]. Table 1, which appears here in the hard copy, shows ^10Be and ^26Al (dpm/kg) in silicate- and metal-rich samples from lodranites and Winona. References: [1] Eugster O. and Weigel A. (1993) LPS XXIV, 453- 454. [2] Middleton R. and Klein J. (1986) Proc. Workshop Tech. Accel. Mass Spectrom., England, 76-81; Middleton R. and Klein J. (1987) Phil. Trans. R. Soc. London, A323, 121-143. [3] Feigenson and Carr (1985) Chem. Geol., 51, 19-27. [4] Schultz L. and Kruse H. (1989) Meteoritics, 24, 155-172. [5] Mason B. and Jarosewich E. (1967) GCA, 31, 1097-1099. [6] Eugster O. (1988) GCA, 52, 1649-1659. [7] Graf Th. et al. (1992) GCA, 54, 2521-2534. Zipfel J.* Palme H. Are Acapulcoites and Lodranites Genetically Related? Petrological and oxygen isotopic studies suggest that acapulcoites and lodranites are closely related. Meteorites of both groups have essentially achondritic equilibrated textures and are similar in mineralogy except that lodranites are coarser grained and have lower plagioclase abundances. The Acapulco meteorite and other acapulcoites have bulk chemical compositions close to those of ordinary chondrites. Compositions of lodranites are different from acapulcoites, primarily reflecting plagioclase fractionation. We performed bulk chemical analyses by instrumental neutron activation analyses of the acapulcoites Monument Draw (M), ALHA 81261 (81), and Acapulco (A) and the lodranites Gibson (G), MAC 88177 (88), and FRO 90011 (F). Data for ALHA 81261, Monument Draw, Mac88177 and FRO 90011 are given in [1]. Additional analyses were done on a new sample of Acapulco. Published data for Lodran (L; [2]), Y-791493 (Y; [3]), ALHA 77081 (77; [4,5]), and Acapulco [5] were considered. Acapulcoites have a very narrow compositional range. Larger variations are only found for the compatible element Cr and the volatile Zn. Both elements are largely hosted in chromite. Similar variations are observed within a single meteorite. Cr in Acapulco bulk samples ranges from 3500 ppm up to 7420 ppm, reflecting inhomogeneous local distribution of chromite. High and variable U abundances and enhanced LREE contents in Acapulco indicate high and variable modal abundances of phosphates since apatite is the main U and LREE carrier. Lodranites are depleted in all plagioclase elements (K, Na, Eu, Ca) and this readily distinguishes them from Acapulcoites (see Fig. 1). Abundances for the compatible Mn are similar in both groups. Se, representing modal sulfide content, is only slightly depleted in lodranites. Similar Cr and Zn variations as in acapulcoites are observed in lodranites. Refractory elements such as Sc and V are lower in lodranites than in acapulcoites (Sc/Mg ratio is about 30% lower). The composition of metal blebs enclosed in opx and olivine of Acapulco suggests that acapulcoites and lodranites passed through a stage of partial or complete melting early in their history [1]. During this process early crystallized chromite grains may have locally accumulated in the partial melt producing an inhomogeneous Cr,Zn distribution. Lodranites suffered additional fractionation(s). Slow cooling (coarse grain size) may provide conditions favorable for removal of residual melt rich in plagioclase elements. However, equilibrium fractional crystallization would not produce the observed negative Eu anomaly in the residual solid. The negative Eu-anomaly in lodranites requires loss of solid plagioclase grains or loss of a nonequilibrium melt of plagioclase composition. Sulfides were much less effectively extracted from lodranites than plagioclase. Summary: Acapulcoites mark the transition from chondrites to differentiated achondrites. Although they experienced igneous processes they largely retained their primitive composition. Lodranites have lost a residual melt fraction. Details of this process are unclear. A common early history of acapulcoites and lodranites is possible. However, the more evolved lodranites require formation conditions different from those of acapulcoites. References: [1] Zipfel J. and Palme H. (1993) LPSC XIV, 1579-1580. [2] Fukuoka T. et al. (1978) LPSC IX, 356-358. [3] Haramura H. et al. (1983) Mem. Natl. Inst. Polar Res., Spec. Issue, 30, 109-121. [4] Schultz L. et al. (1982) EPSL, 61, 23-31. [5] Kallemeyn G. W. and Wasson J. T. (1985) GCA, 49, 261-270. Garrison D. H.* Rao M. N. Bogard D. D. Reedy R. C. Determinations of Solar Proton Spectrum in Lunar Rock 68815 Oriented lunar rock 68815, with 2 x 10^6 years of surface exposure, has become the most widely studied detector of solar cosmic ray (SCR) products. Radio-nuclides ^10Be, ^26Al, ^53Mn, ^14C, and ^81Kr measured in 68815 by several investigators have been used to estimate the long term average solar proton flux, J(sub)>10(4 pi, E > 10 MeV) and rigidity, R(sub)o (MV) [1-4]. Reported ^26Al and ^53Mn depth profiles were consistent with J = 70 p/cm^2/s and R(sub)o = 100 MV [1], whereas depth profiles of SCR ^81Kr suggested J(sub)>10 = 160 p/cm^2/s and R(sub)o = 85 MV [2]. The nearly flat depth profile for ^10Be in 68815 suggested a softer proton spectral shape or a higher erosion rate [3]. Profiles of SCR ^14C, measured in the same 68815 column used for our noble gas analyses, were consistent with J(sub)>10 ~144 p/cm^2/s at R(sub)o = 85 MV, or with J(sub)>10 ~91 p/cm^2/s at R(sub)o = 100 MV [4]. As noted by [3], such combinations of spectral parameters are not unique and may vary widely. SCR profiles, therefore, typically yield comparable spectral parameters to those determined from track studies and contemporary spacecraft measurements. In this study we determined depth profiles of SCR ^21Ne, ^22Ne, and ^38Ar produced by nuclear interactions of energetic (~10-100 MeV) solar protons in documented depth samples of 68815. In addition, we resolved SCR Ne from GCR Ne by their distinct isotopic ratios, a resolution that has been obtained only for neon. We derived the most probable spectral values for solar protons from the measured datasets, placing no prior restrictions on the values of spectral parameters or of the rock erosion rate. We evaluated all combinations of J(sub)>10, R(sub)o, and erosion (Q) using a least squares techniques to measure the "goodness of fit" between measured and calculated SCR abundances. By fixing two parameters at a time (R(sub)o and Q), the third (J(sub)>10) was iterated to determine an optimal value for each set. Comparisons between parametric sets were based on the standard deviation of the profile fit, behavior of the residuals, neon isotopic ratios, and implied GCR exposure ages. No values of R(sub)o lower than 70 MV gave reasonable fits to the data, and R(sub)o values greater than 100 MV implied more SCR neon production at depth than permitted by either the measured isotopic ratios or the resulting GCR exposure ages. Our quantitative data analysis yielded an optimal fit combination (i.e., set with highest probability) of J~110 p/cm^2/s and R(sub)o~85 MV with an erosion of ~2 mm/Myr for ^21Ne, ^22Ne, and ^38Ar. Recently published SCR neon and argon data for lunar rock 61016 [6] have been reevaluated under the same criteria described above. Unique, self-reliant determinations of J(sub)>10, R(sub)o, or Q are difficult. Virtually all combinations of R(sub)o (70 to 125 MV) and Q (1-3 mm/Myr) resulted in SCR profiles of equal goodness of fit given some optimal flux between 50 and 135 p/cm^2/s. We have also applied this statistical treatment to SCR radionuclide data sets for 68815; results of this analysis will be compared with SCR Ne and Ar. References: [1] Kohl et al. (1978) LPSC XIV, 2299-2310. [2] Reedy and Marti (1991) The Sun in Time, 260-287. [3] Nishiizumi et al. (1988) LPSC XVIII, 79-85. [4] Jull et. al. (1992) LPSC XXIII, 639-640. [5] Garrison et al. (1993) LPSC XXIV, 521-522. [6] Rao et al. (1993) JGR, (in press). Wieler R.* Baur H. Signer P. Parentless Fission and Radiogenic Xe in Lunar Breccia 14301 Studied by Closed-System Etching Several lunar samples contain fission Xe in excess of what may reasonably be attributed to in situ fission of U or Pu [1-3]. Many of these samples also contain noble gases implanted by the solar wind (SW). In stepped heating, the parentless fission Xe emerges at lower temperatures than the solar Xe, indicating that the fission component resides even closer to the grain surface than SW-Xe. Grain-size-suite data confirm a surface siting of the fission Xe and reveal also surface-correlated radiogenic 129Xe [2]. The presence of these two components strongly suggests that the solar noble gases were also trapped very early, which makes these samples very attractive for studying properties of the ancient solar corpuscular radiation. We explore here the suitability of the closed system stepped etching (CSSE) technique [4] to separate parentless and solar components. We report preliminary data on a bulk sample of breccia 14301 gently crushed and sieved to 25-150 micrometers. By the time of this writing, roughly 40% of the total Ne and 20% of the Xe have been released in 20 etch steps. Trapped Ne shows the familiar two-component structure SW-SEP (solar energetic particles; [4]), with 20Ne/22Ne ratios ranging between ~13.6 and ~11.6. This indicates that we also need to consider the presence of two solar Xe components with slightly different isotopic compositions [5]. In a diagram 134Xe/132Xe vs. 136Xe/132Xe all data points fall in between the two straight lines that connect the 244Pu fission Xe point on the one hand with the SW-Xe and SEP-Xe points, respectively, on the other. The highest 136Xe/132Xe ratio of ~0.42 is observed in one of the first steps. The last etch steps analyzed so far are essentially devoid of fission Xe, since the data plot in between the SW-Xe and SEP-Xe points. The data pattern thus clearly confirms that 244Pu is the source of the parentless fission Xe and that this component is released even more easily than the SW-Xe [2]. All steps so far release radiogenic 129Xe, including those that are devoid of fission Xe. This corroborates that at least part of the 129Xe(sub)rad is sited in places more resistant to etching than the fission Xe [2]. The noble gas data reveal the existence of several phases of different etchability, as expected for a bulk sample. It is conceivable that the 129Xe(sub)rad resides more deeply in the grains than the fission Xe or that the more acid resistant phases (e.g., mineral grains?) may contain 129Xe(sub)rad but no fission Xe. CSSE analyses of mineral separates may help to decide between the alternatives. Acknowledgments: This work was supported by the Swiss National Science Foundation. References: [1] Drozd R. et al. (1972) EPSL, 15, 338-346. [2] Bernatowicz T. J. et al. (1979) Proc. LPSC 10th, 1587-1616. [3] Hohenberg C. M. et al. (1980) Proc. Conf. Lunar Highlands Crust, 419-439. [4] Wieler R. et al. GCA, 50, 1997-2017. [5] Wieler R. et al. (1992) LPSC XXIII, 1525-1526. Nier A. O.* Schlutter D. J. Extraction of He and Ne from Individual Lunar Ilmenite Grains by Pulse Heating The pulse-heating technique employed for extracting helium and neon from individual interplanetary dust particles [1] has been extended to a similar study of individual lunar grains. A succession of 5-s constant power pulses is applied to the oven holding the particle. The power is increased in 0.25-W increments until all the gas is removed. The peak temperature reached during a pulse lasts about 2 s and increases by roughly 75 degrees C for each 0.25-W increment in power. In the present investigation six individual ilmenite grains of lunar soil 71501 and of breccia 79035 were studied. It was felt that this method of extracting the gas might help in distinguishing between surface embedded solar wind (SW) particles and more deeply embedded constituents such as solar energetic particles (SEP) [2], or gas of trapped or primordial origin. Although only six particles of each type have been studied to date, interesting results are beginning to emerge. For example, for both types of particles, for the initial low power pulses where the maximum pulse temperature does not exceed 500 degrees C, the ^3He/^4He ratio falls near 4 x 10^-4, as expected, if the helium is primarily unfractionated solar wind implanted near the surface. As the pulse temperature is increased to around 1000 degrees C and the solar wind gas presumably has been removed, the ^3He/^4He ratio falls to around 2.5 x 10^-4, in rough agreement with the layer etching results [2]. Likewise, the ^20Ne/^22Ne ratio falls from around 14 to a value near 12, as in the etching experiments [2]. In the case of ^4He/^20Ne ratios there appears to be a real difference between the particles from the two ilmenites. For the 79035 grains, the ratio falls from around 600 for the surface gas to around 150 for the later high-temperature extractions. On the other hand, for the 71501 grains, the ratio starts somewhat lower, near 400, and drops below 100 as the pulse temperature is raised. A qualitatively similar difference was observed in the total gas released by laser beam extractions performed on single grains from the same lunar ilmenite samples [3]. While there is considerable scatter in the data, the overall results are gratifying, and should become more definitive as more particles are investigated. The initial releases, almost certainly from the surfaces of the particles, come closer to the solar wind values [4] than generally reported for lunar grains. It will be interesting to see whether or not the differences observed are real and have a bearing on the general problem of the variation of the solar wind with time [5]. Acknowledgment: We are indebted to R. Wieler for the ilmenite grains used in the investigation. References: [1] Nier A. O. and Schlutter D. J. (1993) LPS XXIV, 1075-1076.[2] Wieler R. et al. (1986) GCA, 50, 1997-2017. [3] Olinger C. T. et al. (1990) Meteoritics, 25, 394. [4] Geiss J. et al. (1972) Apollo 16 Prelim. Sci. Rept., 14-1 to 14-10, NASA SP 315. [5] Becker R. H. and Pepin R. O. (1989) GCA, 53, 1135-1146. Kim J. S.* Marti K. Experimental Artifacts in Nitrogen Isotope Measurements of Meteorites Several research groups have studied contamination problems and molecular interferences in nitrogen isotope measurements, but some problems still require clarification. Protocols adopted for nitrogen isotope measurements generally consider questions such as CO interference, removal of hydrocarbons, and N2O and NO conversion [1]. In the analysis of nanogram amounts of N, contamination, exchange reactions, and interferences are more visible than in large N samples. During nitrogen measurements we observed several potential problems and developed an improved protocol to achieve high-quality isotopic data: 1. Nitrogen loss and isotopic exchange were observed on the extraction system wall. The wall has active surfaces produced by vapor deposition (previous samples) that absorb many molecules, including nitrogen. This absorbed nitrogen releases or exchanges nitrogen with sample N in the following extraction steps. Therefore the losses need to be calibrated and the extent of isotopic exchange determined at the nanogram level. A continuous adsorption during sample extraction of the gas phase onto zeolite at liquid nitrogen temperature reduces nitrogen loss and amount of exchange. 2. We also found nitrogen isotopic memory effect by CuO. During sample gas cleaning by CuO, nitrogen exchanges with residual nitrogen in the CuO, and losses to CuO by solubility and/or uptake of nitrogen during oxygen uptake. This effect is clearly visible after analysis of large amounts of nitrogen. In such cases the CuO blank showed traces of previously measured isotopic signatures. Therefore, the isotopic signature of the CuO blank must be assessed before proceeding. 3. NO interference was recognized. In measurements of N in bulk H chondrites, the steps above 900 degrees C show anomalous contribution to the mass 30 peak, which decreases rapidly with time in the mass spectrometer. Using the ratio mass 30 to mass 31 and the corresponding physical properties of the interfering compound, we identified the NO molecule. NO is produced during heating of the meteorites, and this molecule interacts with metal surfaces (e.g., valves and system metal). It is then released slowly from a metal surface and added to sample nitrogen during N transfer to the inlet volume of the mass spectrometer. Similar effects were reported last year [2], in addition to a rapid change of the measured 29/28 ratio. Hashizume and Sugiura concluded that curious phenomena indicate nonequilibria between two components, and thus the silicates in ordinary chondrites would not contain trapped nitrogen, which is in contradiction with their data. To eliminate the NO effect on mass 30, we made two modifications in the protocol. One is a final cleaning step of the gas phase using a glass finger at liquid nitrogen temperature; the other is the closing of the inlet valve after admitting the sample gas to the mass spectrometer. This protocol eliminates NO interference when the mass spectrometer is not contaminated by NO. 4. There are also nitrogen calibration issues. Last year nitrogen data for metal separates and bulk samples of some H chondrites were reported to reveal large isotopic variations (delta ^15N value from -44 to 119) [3]. Because Kung and Clayton [4] did not observe such variations, we measured nitrogen in Jilin (H5) and found a bulk average delta ^15N = 17 per mil. We also measured a metal separate from Forest Vale and observed a maximum value delta ^15N = 15 per mil. We were unable to confirm the value reported by [3]. We performed a series of calibrations against air nitrogen and NBS-steel standards to determine nitrogen loss and exchange, and against an internal meteorite standard (Cape York). Our analytical procedures are well reproduced. The NBS- steel and Cape York iron are therefore suitable as interlaboratory calibration standards for removal of experimental artifacts. References: [1] Boyd S. R. et al. (1988) J. Phys. E: Sci. Instrum., 21, 876- 885. [2] Hashizume K. and Sugiura N. (1992) GCA, 56, 1625-1631. [3] Hashizume K. and Sugiura N. (1992) Meteoritics, 27, 232. [4] Kung C. and Clayton R. N. (1978) EPSL, 38, 421-435. Fisher D. E. The 129Xe Anomaly in MORB: Gone with the Wind? I have performed replicate crushing experiments on two glass MORB (East Pacific Rise and Mid Atlantic Ridge) in which previous whole-rock melting experiments showed ^40Ar/^36Ar ratios ranging up to ~15,000, indicating efficient trapping of gases from the mantle [1,2]. I loaded nearly a gram of mm-sized pieces and crushed varying portions of them under vacuum, transferring the released gases directly into the mass spectrometer and obtaining more than 20 separate aliquots. Though the Xe/Ar ratio was higher than atmospheric in all aliquots, indicating the presence of mantle xenon, none of the data show any excess ^129Xe from the decay of ^129I early in earth history. It is clear that some terrestrial xenon contains the anomaly [3], but it is not at all clear that the MORB source region does, though some models of mantle and atmospheric evolution rely heavily on this result [4]. I have gone through the literature, and find a diversity of results. Four papers, all from the same laboratory, present clear evidence of the anomaly [4-7], five others do not [8-12], and one straddles the fence [1]. The situation is complicatedby the ubiquitous presence in MORB of a component withatmospheric-like rare gas isotopic ratios. I shall discuss the attempts of various workers to separate these components, and the probability that a true anomaly exists in all or in some MORB source regions, by comparing Xe isotopic data obtained through stepwise heating, total fusion, or crushing experiments, with other pertinent ratios. "I have forgot much, Cynara, gone with the wind; Have flung roses, roses riotously with the throng." References: [1] Fisher D. E. (1986) GCA, 50, 2531-2541. [2] Fisher D. E. (1985) JGR, 90, B2, 1801-1807. [3] Boulos M. S. and Manuel O. K. (1971) Science, 174, 837-840. [4] Allegre C. J. (1983) Nature, 303, 762-766. [5] Staudacher T. and Allegre C. J. (1982) EPSL, 60, 389-406. [6] Staudacher T. (1989) EPSL, 96, 119- 133. [7] Marty B. (1989) EPSL, 94, 45-56. [8] Fisher D. E., EPSL, (sub for publ.) [9] Hiyagon H. (1992) GCA, 56, 1301-1316. [10] Ozima M. and Zashu S. (1983) EPSL, 62, 24-40. [11] Takaoka N. and Nagao K. (1978) Nature, 276, 491-492. [12] Kirsten T. and Richter H. (1981) Meteoritics, 16, 341. Tuesday, July 20, 1993 Interstellar Grains I: Conditions of Formation 4:15 p.m. Theater Chair(s): G. R. Huss Lodders K.* Fegley B. Jr. Chemistry in Circumstellar Envelopes of Carbon Stars: The Influence of P, T, and Elemental Abundances Last year we reported major- and trace-element condensation chemistry in the circumstellar envelope (CSE) of the well known carbon-star IRC+10216 [1]. Here we present results of the most comprehensive study done to date for major- and trace-element chemistry in CSEs of C stars, considering wide ranges in pressure (P), temperature (T), and elemental abundances (s-process enhancements and variable C/O and C/N ratios). These calculations are helpful for interpreting astronomical observations of gas-phase abundances and dust formation in CSEs and the chemistry of graphite, TiC, and SiC grains found in meteorites. Parameters: The present results cover ranges of P = 10^-2 to 10^-15 bar and T < 3000 K. Carbon to oxygen (C/O) ratios of 1 to 10 are considered. Gow [2] reported C/O ratios of 1-10 in 61 C stars with a mean C/O ratio of 2. However, Lambert et al. [3] found C/O = 1.01-1.76 with a mean C/O ratio of 1.15 +/- 0.17 for 30 C stars that were also considered by Gow [2]. Major elements other than C have solar abundances [3], but s-process element abundances may be increased up to 100X solar [4]. Major-Element Condensates: Figure 1 illustrates graphite, TiC, and SiC condensation surfaces as a function of C/O ratio and P. The condensation sequence is very sensitive to C/O ratio and total pressure. At C/O > 2, over the whole pressure range considered, graphite condenses first. Then condensation temperatures of later condensates (e.g., TiC, SiC) are independent of the C/O ratio. However, at C/O = 2 and P < 3 X 10^-3 bar TiC condenses prior to graphite. At C/O = 1.05, the condensation sequence is more sensitive to pressure: At P < 3 X 10^-7 bar the sequence is C(sub)Gr, TiC, SiC, between 3 X 10^-7 < P < 3.4 X 10^-5 bar it changes to TiC, C(sub)Gr, SiC, and it becomes TiC, SiC, C(sub)Gr at P > 3.4 X 10^-5 bar. Trace-Element Condensation: At a given C/O ratio and P, the condensation temperatures of C(sub)Gr, TiC, and SiC provide boundaries for the classification of the condensation behavior of trace elements. For example, elements condensing prior to the first major condensate (at low C/O) or between the first and second major condensate behave coherently. In this respect Ta, Nb, W, Zr, and Hf are classified as extremely refractory. Highly refractory carbides (Mo, V) condense after the first condensate(s) but always prior to SiC. Refractory carbides (Y, Cr) condense after C(sub)Gr, TiC, and SiC. At constant C/O and P, condensation temperatures of s-process elements depend on the abundance enhancements (f(sub)abu) as: 1/T(sub)cond = A + B X log f(sub)abu. For example, an enhancement of 100X solar will increase the condensation temperatures of Zr, Mo, or Y by about 100-150 K (depending on total P). In this case, Mo behaves as an extremely refractory element and pure YC(sub)2 condenses closer to SiC. References: [1] Lodders K. and Fegley B. (1992) Meteoritics, 27, 250-251. [2] Gow C. E. (1977) Publ. Astron. Soc. Pac., 89, 510-518. [3] Lambert D. L. et al. (1986) Astrophys. J. Suppl., 62, 373-425. [4] Utsumi K. (1985) Proc. Japan Acad., 61B, 193-196. Fig. 1 appears here in the hard copy. Ozima M.* Mochizuki K. Origin of Nanodiamonds in Primitive Chondrites: (1) Theory Microdiamonds in primitive chondrites are characterized by Xe-HL, which supposedly formed in a type II supernova. Several models have been proposed for the origin of the microdiamonds. These include chemical vapor deposition (CVD) [e.g., 1], interstellar shock [2], and UV-annealing of small graphite particles [3]. However, it is difficult for any of these models to explain the unique association of Xe-HL with the microdiamonds. We have suggested that a diamond formation process, proposed by Kaminsky [4], for the origin of a particular terrestrial diamond, carbonado, may apply to the microdiamonds in primitive meteorites [5,6]: Kaminsky speculated that carbonado was formed from natural coal that was enriched in uranium and hence subjected to irradiation by high-energy particles produced from the uranium and thorium. The paper in this volume by Mochizuki et al. [7] reports nanometer-sized diamondlike clusters in a uranium-rich natural coal, in accordance with Kaminsky's hypothesis. Mochizuki et al. also report the possibility of the production of nanodiamonds in graphite that was irradiated with a 50-KeV argon beam. These experimental studies strongly suggest that microdiamonds can be produced by irradiation of carbonaceous matters with energetic particles. On the basis of these experimental results, we propose a scenario for the origin of the microdiamonds in primitive chondrites. The scenario gives a reasonable explanation for the unique association of Xe-HL with the microdiamonds as well as for their formation in a supernova envelope. We assume that carbonaceous materials (amorphous carbon, graphite, and hydrocarbon grains) in the outer envelope of a supernova was irradiated by energetic particles (including Xe-HL) emitted during supernova explosion. The energetic particles then interacted with the carbonaceous matter: Most of the energy was dissipated through electronic interaction, and at the end of the journey the particles produced cascade displacement of target atoms. Suppose that an xenon atom impinged into an amorphous carbon particle of 25 nm, which is a typical grain size for interstellar amorphous carbon [8]. In order for the xenon atom to stop within the target particle, the maximum energy would be less than 0.01 MeV. For an impinging energy E(sub)O, the number (n) of cascade-displacement atoms in the target is given by the simple relation [9] n = E(sub)O/E(sub)D, which gives n ~ 250 atoms for E(sub)O = 0.01 MeV. The disturbed region, if recrystallized to have formed a cubic-shaped diamond, would be about 1 nm in size. We can also show from a simple energetic consideration that diamond would be a more stable phase than graphite at room temperature for a size smaller than a few nanometers. The latter estimation is made by comparing the excess pressure induced inside a small particle with the surface tension above which diamond phase becomes stable. The maximum size thus estimated is about a few nanometers. A similar conclusion was drawn by Nuth [10], who used a different approach. Hence, if the displaced region were to recrystallize, the region would form diamond. It is easy to see that the process necessarily leads to the association of Xe-HL with microdiamonds formed in this manner. References: [1] Anders E. (19?7) Phil. Trans. R. Soc. Lond., A323, 287-304. [2] Tielens A. G. G. M. et al. (1987) Astrophys. J., 319, L109-L113. [3] Nuth J. A. III and Allen J. E. Jr., Astrophys. Space Sci., 196, 117-123. [4] Kaminsky F. (1987) Dokl. Akad. Nauk SSSR, 294, 439-440. [5] Ozima M. and Zashu S. (1991) Nature, 351, 472-474. [6] Ozima M. and Zashu S., Meteoritics, 26 382. [7] Mochizuki K. et al., this volume. [8] Sorrell W. H. (1990) Mon. Not. R. Astron. Soc., 243, 570-587. [9] Lehmann C. (1977) Interaction of Radiation with Solids and Elementary Defect Production, 172, North-Holland, Amsterdam. [10] Nuth J. A. III (1987) Astrophys. Space Sci., 139, 103-109. Mochizuki K.* Ozima M. Tuchiyama A. Kitamura M. Shimobayashi N. Origin of Nanodiamonds in Primitive Chondrites: (2) Experiment Ozima and Mochizuki [1] suggested that microdiamonds in primitive meteorites were formed by irradiation of carbonaceous matters such as graphite, amorphous carbon, or hydrocarbons with energetic particles emitted from supernova. To test this hypothesis, we carried out the following experiments. 1. We investigated a uranium-rich coal from Cluff Lake, Canada. Electron microprobe analysis of this sample showed that there are numerous uranium oxide grains of about 10-20 micrometers almost uniformly distributed in hydrocarbon matrix. A small amount of PbS was also identified by the EPMA analysis. If the U,Th-induced radiation were to produce diamonds, they must be found in radiation-damaged regions around the uranium oxide grains. Hence, we very carefully searched for microdiamonds in the radiation-damaged regions by TEM (transmission electron microscope). We observed many crystalline particles of about 20 nm, of which concentration in the radiation damaged region is about 500 ppm. Electron diffraction analysis with a TEM on the crystalline particles gave a powder ring pattern. Because of the limited resolution of the TEM, the electron diffraction was taken over an area (about 0.5 micrometers x 0.5 micrometers) that contained several grains. In Table 1, we show the spacing characteristics deduced from the diffraction analysis, where the observed dspacings (denoted as D) are normalized to the table values corresponding to diamond, graphite, and PbS (denoted as d) respectively. If diamond is chosen for the normalization, the D/d ratios become almost constant for major diamond spacings, including the three most intense ones (shown by bold letters). One intense spacing at D = 1.887 angstroms, however, cannot be attributed to diamond, but may be due to PbS. We conclude that the observed electron diffraction pattern is primarily due to diamond, but other components such as PbS may also be present. Hence, the experiment seems to confirm Kaminsky's hypothesis [2] that high-energy particles derived from U,Th-decays interacted with hydrocarbon (i.e., natural coal ) to have formed microdiamonds in uranium-rich coals. 2. We irradiated 50-mesh graphite powder by a 50-MeV argon beam with a linear accelerator, and examined the irradiated sample with TEM. We observed a crystalline particle that showed a diffraction pattern similar to diamond. The diffraction pattern, however, may also be attributed to graphite. Since we could take the diffraction only from one direction, it is difficult to rule out the latter possibility. To resolve the ambiguity, it is essential to make the electron diffraction analyses of the crystalline particle at least from two directions, which is being carried out. Currently we are trying to see the irradiation effect on different target materials (amorphous carbon, graphite, and hydrocarbon) with different noble gas beams (Kr or Xe). References: [1] Ozima M. and Mochizuki M., this volume. [2] Kaminsky F. (1987) Dokl. Akad. Nauk SSSR, 294, 439-440. Table 1, which appears here in the hard copy, shows electron diffraction spacing characteristics of nanograins in natural coal. Cassidy W. A.* Kern C. M. Primordial Mineral Growth in a Plasma Diamonds, SiC, TiC, corundum, and graphite have been found in primitive meteorites [1] as crystalline components existing with, but far removed from equilibrium with, lower-temperature minerals and, in some cases, hydrocarbons. From anomalous isotopic ratios there is a presumption that the grains, many of them submicroscopic in size, had an extra-solar system origin [1] and may have formed within stellar atmospheres. If that were true, their environment of formation was a plasma. A plasma is an electrically neutral gas containing a high proportion of its component molecules in the ionized state, with charge balance satisfied by the presence of free electrons; it has been described as a fourth state of matter. The exact way in which the presence of a plasma would affect the Gibbs Phase Rule and alter the laws of thermodynamics is not understood, and condensation and vaporization in the presence of a plasma is affected in ways that cannot now be predicted. A good example of the truth of this assertion is the commercial process of growing diamonds in a plasma, where diamonds grow very rapidly in an atmosphere whose components before dissociation were 99.5% H2 + 0.5% CH4, at pressures ranging from 40 to 65 millibars and substrate temperatures ranging between 800 and 1000 degrees C. These P-T conditions, particularly the pressures, are very different from those necessary for equilibrium growth of diamond in the absence of a plasma. If non-shock meteoritic diamonds grew "metastably" in a plasma, perhaps other meteoritic minerals would also grow in such an environment: silicates, for example, have been detected in interstellar clouds. Such minerals might be incorporated later, along with diamonds, as minor constituents in the matrix of forming meteoroidal objects, and be unrecognized because of their similarities to the much more abundant intrasolar system minerals making up the bulk of primitive meteorites. References: [1] Nittler L. R. et al. (1993) LPS XXIV, 1087-1088. Kern C. M. Witkowski R. E.* Cassidy W. A. Stability of Minerals in a Plasma We report here the results of an experiment designed to determine whether a group of silicate, oxide, sulfide, and metal substrates would remain stable or decompose under the highly energetic conditions of a plasma in which diamonds can form metastably. Using a microwave plasma reactor, we exposed the substrates to a plasma consisting of 99.5% hydrogen and 0.5% methane for 24 hr. The temperature of the substrate was approximately 950 degrees C, the pressure of the vacuum chamber was approximately 45 mbar, and the flow rate of the gas mix was approximately 200 cm^3/s. The silicate substrates were olivine, plagioclase, augite, nepheline, pyrope, hornblende, serpentine, and phlogopite; oxide substrates were magnetite and hematite; sulfide substrates were troilite, pentlandite, and pyrrhotite; and metal substrates were iron, nickel, kamacite, and taenite. Metastable diamonds formed on all the silicate substrates except nepheline. There was no diamond formation on any of the oxide, sulfide, or metal substrates. In every instance where diamonds did not grow the substrates were covered with thick secondary deposits. A substantial amount of pitting was observed on the surfaces of all the silicate substrates except nepheline. There is evidence for the loss of a substantial amount of silicon from at least three silicate substrates (e.g., plagloclase, anorthite, and pyrope in a Ca-rich groundmass). A thick layer of platelike Al-rich secondary deposits (possibly metallic Al and/or corundum) formed on the nepheline substrate. The magnetite, hematite, troilite, pentlandite, pyrrhotite, iron, kamacite, and taenite substrates decomposed significantly. Thick, dark, highly porous secondary iron-rich deposits formed on the surfaces of these substrates. These iron-rich deposits may consist of metallic Fe, FeC, and/or FeS and may be reaction products resulting from the interaction of the plasma and substrates. Other secondary deposits common to the silicates were iron (possibly in the form of metallic Fe, iron carbide, and/or iron sulfide) and calcium (possibly in the form of elemental Ca and/or oldhamite). Secondary iron typically formed porous, loosely bound aggregates, often containing small (0.1-micrometer) spheroids. Fibers approximately 0.05 micrometers in diameter and 10-30 micrometers in length were observed on substrates containing a substantial amount of Ca (e.g., plagioclase, augite, anorthite, and pyrope in a Ca-rich groundmass). The initial presence of Ca in the substrates, loss of Si from the substrates, and growth of fibers on substrates may be related. Samples were run in groups of seven and secondary mineral formation, at least in part, resulted from cross contamination. We can infer from this cross contamination that Fe, Si, Ca, and S were fractionated from some substrates, incorporated into the plasma, and then deposited in different morphologic forms. These forms include fibers, fibrous networks, iron-rich, porous, loosely bound aggregates, and spheroids. The pressure and temperature conditions used in this experiment can be found in the atmospheres of red giant stars. The results of this experiment suggest that most of the minerals used as substrates would not be stable under conditions that favor the metastable growth of diamonds. However, morphologic structures similar to those produced on the substrates in this experiment may also exist in the atmospheres of red giant stars. Wednesday, July 21, 1993 CAIs 8:30 a.m. Theater Chair(s): G. Meeker A. Ruzicka Walker R. M.* LEONARD MEDAL ADDRESS: Searching for Interesting Needles in the Meteoroid Haystack It has been my great privilege and pleasure to participate in the study of extraterrestrial materials during an exceptionally interesting scientific period. In this talk, I first intend to reminisce a bit. Partly, my goal is to communicate to younger members of the Meteoritical Society the sense of excitement that accompanied certain scientific developments that they now take for granted. Partly, I intend to use the occasion to point out some first- order problems that remain unsolved in several older areas of research. Following the treatment of matters past, I will finish by discussing the location and study of interstellar grains in situ in primitive meteorites--a subject of great current interest to me and several colleagues at the McDonnell Center for the Space Sciences at Washington University. By and large, my work has not dealt with the formation of various types of extraterrestrial materials per se; rather, I have used extraterrestrial samples to learn about other aspects of the history of the solar system. My start in our field came from the thought, later realized in practice, that energetic, heavy nuclear particles might produce tracks in lunar rocks making them the equivalent of nuclear emulsions in recording galactic and solar radiations impinging on the lunar surface. The development of fission track dating and the discovery of excess fission tracks from the decay of 244Pu were outgrowths of that work. Similarly, the work on Brownlee particles (IDPs) was primarily motivated by the idea that some of the particles were probably from comets and that comets were a good place to look for interstellar grains. The desire to study IDPs stimulated the development of a number of microanalytic techniques including, notably, ion microprobe measurements of isotopes in small samples. This led, in turn, in our laboratory, to the identification of individual interstellar grains of SiC and graphite in acid residues prepared at the University of Chicago. The study of interstellar grains is an entirely new field of astrophysics that represents an increasingly important part of contemporary meteorite research. The location and study of interstellar grains in situ is accomplished using an X-ray mapping technique. By optimizing the mapping parameters in our EDS system, it is now possible for us to locate one interstellar grain every day or so in favorable cases. Several dozen SiC grains have been found in situ and typical examples will be shown. One of the ultimate goals of this work is to remove and study the surface properties of interstellar grains that have not been subjected to the harsh chemical treatments used to produce acid residues. Progress towards this goal will be reviewed at the meeting. The work I describe is a shared enterprise with many others and this is a great part of the joy derived from it. The capacity of science to transcend national boundaries, cut across cultural differences, and even to span age differences, is a source of continuing amazement and pleasure in a world otherwise much taken with conflict. El Goresy A.* Zinner E. K. Matsunami S. Palme H. Spettel B. Lin Y. Nazarov M. Efremovka 101.1: A Primitive CAI with Superrefractory REE Patterns and Enormous Enrichments of Sc, Zr, and Y in Fassaite and Perovskite A fragment (30 mg) consisting of two inclusions (101.1 and 101.2) was separated from the Efremovka (CV3) meteorite. 101.1 is an unusual Type A CAI, whereas 101.2 consists of Cr-spinel and fassaite. INAA of the whole fragment revealed 16% MgO reflecting significant contributions from 101.2. Refractory lithophile elements in the bulk fragment have CI-enrichment factors of ~14 with two times enrichment factors for Ca, Eu, and Yb. CAI 101.1 (1.6 mm) contains more than 90% gehlenitic melilite (Ak(sub)1- Ak(sub)32) in its core. It is surrounded by a 5 layer rim sequence (~40 micrometers thick) consisting of spinel -->Al- diopside + fassaite (<= 0.7% Sc2O3) -->forsterite (Fo(sub)97- Fo(sub)100) --> diopside --> forsterite. Two small complete CAIs with a two layer sequence (diopside + anorthite) are contained within the core. Numerous layered sinuous inclusions presumably rim sequence fragments also consisting of diopside + anorthite, are locally crowded in the core. The melilite core is sprinkled with fassaite, perovskite, FeNi, and OsRu-rich metal blebs. Fassaite grains (<= 30 micrometers) contain enormous concentrations in Sc (up to 12.9% Sc2O3) and Zr (up to 5.4% ZrO2). Fassaite rims around FeNi blebs are rich in V (up to 5.4% V2O3) and are zoned with decreasing Sc-, Zr-, and V-concents from the metal cores to the outer fassaite rims. Sc2O3 and ZrO2 concentrations in fassaite display a positive correlation with a correlation coeffient of 0.88. This coherent behavior is a result of a complex cation substitution involving Mg, Ti, Sc, Zr, and V. A coupled substitution is demonstrated by the excellent linear correlation between Mg^2++Ti^4+(y) and Sc^3++Zr^4++Ti^3++V^3+(x) satisfying the equation y = 0.70-0.66x and having a linear regression coefficient of 0.84. Ti^3+/Ti^tot varies between 0.27 and 1. In contrast to fassaites, perovskites are generally depleted in Sc and Zr and enriched in Y (<=1.4% Y2O3). The assemblage andradite+wollastonite+ Fe^degree/or FeNi metal was encountered inside the sinuous fragments and in the diopside-anorthite rim sequence of one of the captured CAIs. The texture is strongly suggestive of the reduction of andradite to wollastonite+Fe^degree (via Fe3O4 and FeO). We consider this texture as evidence that andradite was formed in an oxidizing solar gas before capture of the sinuous inclusions by the host CAI and that andradite reduction took place after capture in the highly reduced host CAI refractory liquid. REE concentrations in various minerals reveal a distinct superrefractory pattern with depletions in Tm and Yb. Zr, Y, and Sc abundances of individual phases in the CAI core are indicative of crystal-liquid fractionation during crystallization: fassaite is relatively depleted in Y and enriched in Sc and Zr. In contrast, perovskite displays a complimentary abundance pattern for these elements. The individual mineral layers of the captured CAIs, sinuous fragments, and the rim sequence of the host CAI have similar superrefractory REE patterns but do not show fractionation between Sc, Y, and Zr. This indicates that the rim sequences did not crystallize from the same liquid of the CAI, but condensed from a common gas reservoir with a distinct superrefractory REE signature. Melilite contains excess ^26Mg* with an inferred initial ^26Al/^27Al ratio of (4.4 x 10^-5). However, the Mg-Al system in the anorthites of the captured sinuous fragments and the rim layers of the small CAIs is disturbed. This strongly suggests that oxidation and alteration processes took place in the earliest stage of the solar system. MacPherson G. J.* Davis A. M. A Hibonite-Perovskite-rich Type A Leoville Inclusion Hibonite-bearing refractory inclusions preserve some of the most primitive chemical, isotopic, and petrologic properties from the earliest solar system [e.g., 1]. Among inclusions from CV3 chondrites, those in Leoville have been the objects of particular scrutiny ever since an early (now questionable) report [2] of hibonite with exceptionally high inferred initial 26Al/27Al of ~1 x 10-4, much higher than the generally accepted "canonical" value of 5 x 10-5 [3]. Leoville inclusions are interesting for other reasons as well, however. They contain little of the secondary mineralization (feldspathoids, gamet, etc.) that obscures primary textures and mineralogy in Allende CAIs, and many show the same pronounced flattening characteristic of other components in Leoville, leading to the possibility that the isotopic signatures of some of these inclusions may reflect the timing of the event causing the flattening [4]. Leoville 3535-3b is a compact and irregularly shaped Type A inclusion (maximum dimension >3 mm) with a very asymmetric structure, and is exceptionally rich in hibonite of two distinct textural types. The interior is mostly Al-rich melilite (Ak(sub)0.3-23). Enclosed in the melilite, but concentrated mostly toward one side of the inclusion, is a dense swarm of spinel grains (up to 20- 30 micrometers) and large (up to 230 micrometersm) angular fragments consisting of spinel and hibonite (0.6-1.8% V(sub)2O(sub)3, 0.3-3.7% MgO). At the center of this "swarm" is an aggregate of spinel-perovskite spherules (similar to ones in fine-grained spinel-rich inclusions) that are enclosed in a fine-grained mixture of secondary anorthite and pyroxene. This side of the inclusion is overlain by a relatively thin rim sequence of melilite, diopside, and olivine. On the opposite side of tbe inclusion the central melilite contains few inclusions of any kind but is overlain by an exceedingly thick and complex rim sequence. Within ~100-200 micrometers of the rim the melilite becomes crowded with myriad 10-20-micrometer-sized euhedral blue hibonite blades (<0.1-0.4% V(sub)2O(sub)3, 1.4-4.5% MgO), locally intergrown with spinel and/or perovskite. In places this zone expands into remarkable 200-300- micrometer thick intergrowths of mostly perovskite, together with hibonite and melilite. A complete Wark-Lovering type rim sequence is present only on the side of the inclusion containing the hibonite-spinel intergrowths, and terminates where the inclusion has apparently been broken. It consists (from innermost to outermost layers) of hibonite + perovskite, hibonite + spinel, melilite, Al-Ti pyroxene, and olivine. Broken surfaces on the inclusion are rimmed by a thin layer of pyroxene only. Ion microprobe analyses of spinel, melilite, and hibonite (both populations) yield an array of points on an Al-Mg isochron diagram that cluster tightly along a best-fit line with slope (initial 26Al/27Al) of (5.18 + 0.26) x 10-5 at 27Al/24Mg up to ~75, indistinguishable from the canonical value [3]. None of the phases in 3535-3b show significant Mg isotopic mass fractionation. Melilite in the interior of the inclusion has an unfractionated (Group I) REE pattern, but the hibonite-perovskite intergrowths in the thick rim sequence have a heavy-REE-enriched pattern (up to 400X Cl chondrites for Lu), with light-REE-enriched 100X. The spinel-hibonite clumps in the interior appear to be broken xenoliths enclosed by the later melilite, whereas the second-generation hibonite clearly formed as part of the rim-forming process. Yet the series of events leading to this complex structure occurred within a sufficiently short time that the two Al-Mg isotopic signatures are indistinguishable and "canonical." The thick hibonite-perovskite rim sequence probably did not form by volatilization of the melilite interior, because the trace-element enrichments in the rim relative to the interior would require a very large degree of evaporation that is not tenable in view of the lack of Mg isotopic fractionation. The very high abundance of Ti in the rim either requires addition of Ti (at least) during rim formation or else an unusual Ti-rich rim precursor. References: [1] Hinton R. W. et al. (1988) GCA, 52, 2573. [2] Lorin J. C. et al. (1978) Fourth Intl. Conf. on Geochron. Cosmochron. Isotope Geol., 257, U.S. Geol. Surv. Open-File Rept. 78-701. [3] Hutcheon I.D. (1982) Am. Chem. Soc. Symposium Ser. 176, 95. [4] Caillet C. et al. (1991) Meteoritics, 26, 326. Ireland T. R.* Normal Zirconium Isotopic Composition in Murchison Hibonite 13-13 Murchison hibonite 13-13 has a special place in isotopic astronomy because it has the largest isotopic anomalies in Ca and Ti as yet measured with delta ^48Ca at +105 per mil and delta ^50Ti at +273 per mil [1]. It has been suggested that these anomalies are the result of nucleosynthesis in neutron-rich supernova ejecta [2] and Ca and Ti in this hibonite grain have been the least diluted by normal solar system material. It would be expected that the isotopic compositions of the other Fe peak elements would also show the effects of the same nucleosynthetic process(es), but the abundances of the other elements are quite low. This is because of the refractory composition of hibonite and the relatively high volatilities of the other Fe peak elements. Zirconium is another refractory element that has recently been examined for its isotopic composition. Harper and coworkers [3] found a small ^96Zr enhancement (~2 epsilon) in Allende CAIs, and it was suggested that the ^96Zr anomaly was correlated with the ^50Ti enhancements (~10 epsilon) in the same inclusions and that both isotopes were produced by the same process [4]. The purpose of this report is to test whether a ^96Zr anomaly can be resolved in 13-13. If present at the same level as in the Allende CAIs relative to the ^50Ti anomaly, a ^96Zr enhancement of the order of 60 per mil should be observed. Zirconium isotopic compositions were measured by ion microprobe mass spectrometry. A mass resolution of 8000 (1% valley) was found to be sufficient to resolve molecular interferences as well as hydrides. However, in some hibonite analyses, tailing from ^40Ca(sub)2^160^+ became unacceptably large. Terrestrial zircon and Zr metal standards were also measured, and good agreement was found with terrestrial Zr isotopic compositions [5] after normalization to the ^94Zr/^90Zr ratio. The Zr isotopic composition of 13-13 is normal within expenmental uncertainty and the maximum anomaly at the 2- sigma level is well below 10 per mil (Fig. 1). No ^96Zr anomaly is associated with the large ^50Ti anomaly in this Murchison hibonite, and so it is unlikely that the nucleosynthetic model of Harper and coworkers for coproduction of ^96Zr and ^50Ti is valid. References: [1] Ireland T. R. (1990) GCA, 54, 3219-3237. [2] Hartmann D. et al.(1985) Astrophys. J., 297, 837-845. [3] Harper C. L. et al. (1990) Meteoritics, 25, 369. [4] Harper C. L. et al. (1991) LPSC XXII, 517-518. [5] Minster J. and Ricard L. P. (1981) Int. J. Mass Spec. Ion Phys., 37, 259-272. Fig. 1 appears in the hard copy here. Simon S. B. Grossman L.* Davis A. M. Beckett J. R. Chamberlin L. Evidence for Extremely-High-Temperature Melting in the Solar Nebula from a CaAl4O7-bearing Spherule from Murchison We have recovered a unique refractory spherule (B6) from the Murchison C2 chondrite. Approximately 140 micrometers in diameter, it is concentrically zoned, with an outer rim sequence, from outermost to innermost, of aluminous diopside (10 micrometers thick), anorthite (3 micrometers) and melilite (3 micrometers). Inside the melilite layer is a 7-micrometer-thick, nearly pure (except for a single, diverging-inward spray of hibonite crystals) layer of spinel. Inward from this layer is a 22-micrometer-wide zone of hibonite (~5.5 wt% TiO2) + spinel, in which hibonite laths, 1-4 micrometers across and up to 10 micrometers wide, are predominantly radially oriented and enclosed in spinel. Inward from this zone, presumably at the core of the inclusion, are CaAl4O7, occurring as anhedral grains ~10 micrometers across, and minor perovskite. Some of the hibonite laths protrude into the CaAl4O7. The sequence of mineral assemblages from the spinel shell inward parallels that expected for fractional crystallization of a melt of the composition of B6. Based on this, the inclusion's spherical shape, and its texture (radially oriented hibonite laths, including a diverging-inward spray; laths enclosed in spinel and protruding into CaAl4O7), we conclude that the oxide phases in B6 crystallized from a liquid. The spinel layer indicates that at least some of the spinel was molten; from the bulk composition, calculated liquidus phase relations in the system Al2O3-MgO-CaO [1], and the amount of spinel contained in the layer, we infer a melting temperature >2000 degrees C. This is >500 degrees higher than the maximum temperature at which any condensed major phase is stable at 10-3 atm in a gas of solar composition, but we see no evidence of evaporation. First, the inclusion has a Group II REE pattern, rather than a Group III or an ultrarefractory pattern, which could reflect devolatilization. Second, although evaporation of molten (but not solid) Mg2SiO4 leads to Mg isotopic mass fractionation [2], we found the Mg isotopic composition of spinel and hibonite in B6 to be essentially normal (DELTA 25Mg = 0 +- 2.5 permil). This means that no more than ~15% of the Mg could have evaporated, which, by analogy with experiments with forsterite at 2050 degrees C [2], suggests that the melt was exposed to the solar nebula for a very short time, perhaps as little as two minutes. This could indicate rapid formation of the spinel shell in B6, sealing off the molten interior from the solar nebula. Evaporation of solid spinel could have occurred, but would probably not fractionate Mg isotopes significantly. Evidence of an unusually high temperature history is preserved in the spinel of B6. It averages 1.7 +- 0.4 mol% excess Al2O3 relative to MgAl2O4, unlike the stoichiometric (within analytical error) spinel found in most CAIs. Much larger Al2O3 solubilities than observed in B6 spinel have been produced in synthetic systems at temperatures as low as 1300 degrees C [3]. In our crystallization experiments, excess Al2O3 ranges from 2 mol% in spinel equilibrated with melilite + hibonite + liquid at 1400 degrees C to 30 mol% in spinel equilibrated with liquid at 1499 degrees C. In corundum-bearing runs, excess Al2O3 in spinel increases from 12 mol% at 1349 degrees C to 24 mol% at 1450 degrees C, consistent with [3]. Excess Al2O3 in spinel is directly correlated with aAl2O3/aMgO based on experiments with solids [4]; it should also be correlated with aAl2O3/aMgO of coexisting liquids, and with temperature at constant aAl2O3/aMgO [1]. Spinels in our experiments have large excess Al2O3 contents because coexisting liquids have aAl2O3/aMgO >6 [1]. The bulk composition of B6 and residual liquids produced by crystallization of spinel from this composition have aAl2O3/aMgO ~1 [1], resulting in lower excess Al2O3 in B6 spinel than in our synthetic spinel. In type B inclusions, liquids with which spinel equilibrated also had aAl2O3/aMgO ratios ~1, but because equilibration temperatures were <~1500 degrees C, this spinel has negligible excess Al2O3, consistent with the results of [4]. The larger amounts of excess Al2O3 in B6 spinel indicate that its equilibration temperature was substantially higher than in type Bs (i.e., >~ 1500 degrees C), consistent with the above observations. References: [1] Berman R. G. (1983) Ph.D. thesis, U. British Columbia. [2] Davis A. M. et al. (1990) Nature, 347, 655-658. [3] Viertel H. U. and Seifert F. (1980) N. Jb. Miner. Abh., 140, 89-101. [4] Chamberlin L. et al. (1992) GSA Abs. with Prog., 24, A257. Steele I. M.* Silica-free, Refractory Inclusion with Al-rich Alteration and Perovskite Exsolution A hibonite-spinel-perovskite CAI from an Allende thin section is in many respects similar to other hib-sp-pv inclusions, especially one described in [1], but also shows several unusual alteration features. The inclusion is about 300 micrometers in long dimension, but with extended curved ends giving an indication of an original spherical object. A single angular 200-micrometer hibonite grain (TiO2: 1.2-1.6; MgO: 0.57-0.77; SiO2: <0.05; FeO: <0.10; Cr2O3 <0.01%), which includes several 20-micrometer perovskites (Al2O3 up to 2%) occurs at the inclusion center and is surrounded by a mantle of spinel zoned with FeO near 1% at hibonite boundary to 12% at inclusion edge. Within the spinel are blades of texturally and compositionally distinct hibonite (TiO2: 4.6-8.2; MgO: 2.4-4.4; SiO2: 0.13-0.35; FeO: 0.26-0.70; Cr2O3: 0.03-0.07%), which appear to have grown from the rim toward the inclusion center. Also within the spinel mantle are numerous micrometer-sized perovskite grains. A rim sequence surrounds the entire inclusion and includes Fe-olivine, diopside, scapolite-nepheline, and hibonite layers from edge toward spinel mantle. No melilite, forsterite, or fassaite is present in inclusion, i.e., no silicates are present. Evidence of low-temperature alteration occurs as (1) replacement of perovskite within spinel mantle and adjacent to rim by Mg-ilmenite (MgO: 3.0-7.8; Al2O3: 0.3-2.0; MnO: 0.25%; V present); (2) exsolution as oriented submicrometer lamellae of Al2O3 from the larger Al-rich perovskites within central hibonite; and (3) veins, especially within hibonite and adjacent to perovskite, of Al2O3 composition (Al2O3: 97.2-79.2; SiO2: 1.5-14.2; Na2O: 0.6-4.9%; minor Fe, Ca, K) and common fiberous texture. All analyses sum to 100%, and Na correlates with Si. The ratio of Na to Si implies a mix of Al2O3 and nepheline. The inclusion is very similar to one previously described [1], but in place of central melilite, the present inclusion has hibonite. The angular hibonite grain contrasts with the euhedral hibonite blades and generally spherical inclusion. While these features imply an original melt, the angular hibonite suggests a relic grain that may have formed a nucleus for a melt. Special care must be given to interpreting bulk analyses if some CAI contain two or more distinct components. The occurrence of Al2O3 in two textural forms suggests that at least some Al2O3 is secondary either due to exsolution or possibly from breakdown(?) of hibonite. Previous observations of corundum do not necessarily imply that Al2O3 is a primary phase. References: [1] Bischoff A. et al. (1982) LPSC XVIII, 81-82. Misawa K.* Fujita T. Kitamura M. Nakamura N. Yurimoto H. A Relict Spinel Grain in an Allende Ferromagnesian Chondrule It is suggested that one of the refractory lithophile precursors in CV-CO chondrules was a hightemperature condensate from the nebular gas and was related to Ca,Al-rich inclusions (CAIs) [1-3]. However, little is known about refractory siderophile precursors in chondrules [4]. Allende barred olivine chondrule R-11 consists mainly of olivine (Fa(sub)7- 18), pyroxene (En(sub)93Fs(sub)1Wo(sub)6, En(sub)66Fs(sub)1Wo(sub)33), plagioclase (An(sub)80), Fe-poor spinel, and alkali-rich glass. The CI- chondrite normalized REE pattern of the chondrule, excluding a spinel grain, are fractionated, HREEdepleted (4.6-7.8 x CI) with a large positive Yb anomaly. The REE abundances are humpshaped functions of elemental volatility, moderately refractory REE-enriched, suggesting that the refractory lithophile precursor component of R-11 could be a condensate from the nebular gas and related to Group 11 CAIs [1,2]. An interior portion of spinel is almost Fe-free, but in an outer zone (2040 micrometers in width) FeO contents increase rapidly. TiO(sub)2, Cr(sub)2O(sub)3, and V(sub)2O(sub)3 contents in core spinel are less than 0.5%, which is different from the V-rich nature of spinel in fluffy Type A CAIs [5]. The Fe-Mg zoning of spinel may have been generated by diffusional emplacement of Mg and Fe during chondrule-forming events. The spinel contains silicate inclusions and tiny metallic grains. The largest silicate inclusion is composed of Al,Ti-rich pyroxe