Published in Meteoritics, 29, pp. 214-221.
Abstract: ALH 84001, originally classified as a diogenite, is a coarse-grained, cataclastic, orthopyroxenite meteorite related to the martian (SNC) meteorites. The orthopyroxene is relatively uniform in composition, with a mean composition of Wo3.3En69.4Fs27.3. Minor phases are euhedral to subhedral chromite and interstitial maskelynite, An31.1Ab63.2Or5.7, with accessory augite, Wo42.2En45.1Fs12.7, apatite, pyrite, and carbonates, Cc11.5Mg58.0Sd29.4Rd1.1. The pyroxenes and chromites in ALH 84001 are similar in composition to these phases in EETA 79001 lithology A megacrysts but are more homogeneous. Maskelynite is similar in composition to feldspars in the nakhlites and Chassigny. Two generations of carbonates are present, early (preshock) strongly zoned carbonates and late (postshock) carbonates. The high Ca content of both types of carbonates indicates that they were formed at moderately high temperatures, possibly ~700°C. ALH 84001 has a slightly LREE-depleted pattern with La 0.67× and Lu 1.85× CI abundances and with a negative Eu anomaly (Eu/Sm 056× CI). The uniform pyroxene composition is unusual for martian meteorites, and suggests that ALH 84001 cooled more slowly than did the shergottites, nakhlites, or Chassigny. The nearly monomineralic composition, coarse-grain size, homogenous orthopyroxene and chromite compositions, the interstitial maskelynite and apatite, and the REE pattern suggest that ALH 84001 is a cumulate orthopyroxenite containing minor trapped, intercumulus material.
Introduction: The SNC meteorite clan (shergottites, nakhlites, Chassigny) is a small clan of igneous rocks widely thought to be samples of the martian crust (e.g., McSween, 1985). Because of this, discovery of a new SNC meteorite always generates interest in the meteoritic community. Interest is intensified if the sample is of unusual petrologic type, as an increase in diversity holds promise of uncovering new insights into martian petrologic evolution. In this paper, I report the unmasking of an unusual meteorite related to the SNC clan that had been masquerading as a diogenite in the Antarctic meteorite collection since 1985: ALH 84001. A small grain mount and bulk sample of this meteorite were examined as part of my study of diogenite meteorites, but it was only recently after I studied a thin section of ALH 84001 that the petrography of the sample revealed its true nature. Subsequent oxygen isotopic measurements by Clayton (1993) have confirmed its relationship to the SNCs; ALH 84001 has an oxygen isotopic composition indistinguishable from that of the nakhlites.
The purpose of this report is to provide a petrographic description of thin section ,64 and analysis by INAA of split ,20. A brief description of ALH 84001 was presented by Berkley and Boynton (1992). The sample described by Sack et al. (1991) was not bona fide ALH 84001; rather it was a mislabeled sample of the EETA 79002 diogenite (Score and Lindstrom, 1993).
Petrography and Mineralogy: ALH 84001 is a coarse-grained, cataclastic orthopyroxenite. Much of the original magmatic/metamorphic texture is preserved (Berkley and Boynton, 1992). Thin section ,64 consists of coarse, generally anhedral orthopyroxene up to 3.5 mm across exhibiting patchy extinction (Fig. la, b). Berkley and Boynton reported orthopyroxene grain sizes of at least 6 mm in their section. The orthopyroxene grains commonly join in 120° triple junctures. Euhedral to subhedral chromites up to 0.5 mm are poikilitically enclosed in or interstitial to orthopyroxene. The coarse orthopyroxenes contain transecting crushed zones of fine-grained anhedral orthopyroxene and chromite with grain sizes on the order of a few tens of µm. The coarse orthopyroxene and chromites frequently have grain boundaries offset along fine fractures (Fig. 1c). Maskelynite, typically a few hundred micrometers in size, occurs interstitial to coarse orthopyroxene and in the crushed zones. Some maskelynite is partially enclosed in coarse orthopyroxene.
Accessory phases are augite, apatite, pyrite, and Mg-Ca-Mn-Fe carbonate. Augite occurs as interstitial grains ~10 µm in size associated with apatite. Apatite occurs as interstitial grains up to ~300 µm. Pyrite grains are generally ~10 µm and are usually associated with interstitial chromite, maskelynite, and/or carbonate, or in the crushed zones associated with anhedral chromite. Carbonates occur as interstitial grains ~100 µm in size associated with maskelynite and pyrite (Fig. 1d). Fine-scale compositional zoning is present in these carbonates and sometimes offset along fractures, indicating that they were formed before the last shock event. Small carbonates, ~10 µm in size, are frequently found in the crushed zones or in fractures in orthopyroxene, indicating that they were formed after a shock event, or were remobilized by that event. Some of these carbonates appear to be replacing orthopyroxene.
Average compositions of orthopyroxene, augite, chromite maskelynite, apatite, and carbonate are given in Table 1. The orthopyroxene and chromite grains are relatively homogeneous in composition, both for major and minor elements. Figure 2 shows individual orthopyroxene and augite compositions for major and minor elements. The average orthopyroxene composition is Wo3.3En69.4Fs27.3, and there is little variation (Fig. 2b). Similarly, the contents of Al2O3, TiO2, and Cr2O3 generally show only limited variation, although several analyses are high in either Al2O3, TiO2, or Cr2O3 (Fig. 2c).
Chromite analyses exhibit minor variation, with correlate enrichments in Al2O3 and TiO2 and depletion in Cr2O3 (Fig. 3). There is no apparent systematic difference between cores and rims when evaluated in toto (Fig. 3b). Some individual grains do show slight zoning, but this zoning is not consistent from grain to grain. Some grains exhibit enrichments in Al2O3 and TiO2 and depletions in Cr2O3 in rims compared to cores, while other grains show the opposite. One apparently consistent difference among the chromites is that the small, anhedral grains in the crushed zones are Cr2O3-poor and Al2O3- and TiO2-rich compared to euhedral chromites (Fig. 3b). This is likely a metamorphic effect of the shock event, as textural evidence shows that the anhedral chromites are fragments of crushed euhedral chromites.
Maskelynite compositions cluster at about An34Ab62Or4, with a few analyses extending to more potassic and sodic compositions (Fig. 4a, b). The average of all maskelynite analyses is An31.1Ab63.2Or5.7 (Table 1). This composition is more sodic than that reported by Berkley and Boynton (1992), An39Ab56Or5, but overlaps the range of analyses performed by MacPherson (1985), An35- 59Ab57-61Or4-43. I found that these maskelynites were very susceptible to Na loss under electron bombardment. The analyses I performed were done at 15 kV potential, 3 nÅ beam current, and with the beam rastered over approximately a 16 × 16 µm area. These are conditions that minimize Na loss. Many of the maskelynite analyses are non-stoichiometric, exhibiting excesses in SiO2 compared to that required for their CaO, Na2O, and K2O contents (Fig. 4c). This suggests that a silica phase may have been present along with plagioclase as interstitial material. The more Na- and K-rich maskelynites are not more deviant than are the typical maskelynites (Fig. 4c).
Carbonate analyses are shown on Fig. 5. The average of all carbonate analyses is Cc11.5Mg58.0Sd29.4Rd1.1 (Table 1), which is similar to the average reported by MacPherson (1985), Cc11Mg60Sd29. Early carbonates, those that are coarser-grained, strongly zoned, and have textural evidence for predating shock, exhibit zoning from Fe- and Ca-rich compositions to almost pure magnesite and are followed by a narrow band of Fe-, Ca-rich compositions zoning again to magnesite (Figs. 1d, 5). This indicates that early carbonates were formed from multiple influxes of fluid. The association of euhedral pyrite with the early carbonates (Fig. 1d) suggests that zoning to Fe-poor compositions may have been facilitated by sulfide crystallization. Zoning was not detected in the small carbonates that occur in the crushed zones and in fractures. These late (postshock) carbonates show a narrower range in composition than found for zoning in the early carbonates (Fig. 5).
The siderite-magnesite limb of the solvus with dolomite-ankerite at greenschist (~250°C) and granulite (~700°C) facies conditions, as estimated by Anovitz and Essene (1987), are shown for comparison in Fig. 5. Stable carbonates will plot on or to the Ca-poor side of these lines at the indicated temperatures. Thus, unless the ALH 84001 carbonates were formed metastably below the solvus (i.e., with compositions in the two-phase field), their compositions indicate that they were formed at moderately high temperatures, perhaps around 700°C. The presence of euhedral pyrite associated with the carbonates (Fig. 1d) suggests the maximum temperature was <=742°C (e.g., Vaughan and Craig, 1978) if the pyrite and carbonate formed at the same time. The early carbonates, therefore, are products of hydrothermal alteration, rather than low-temperature weathering products. The inferred high-temperature origin for both the early and late carbonates is incompatible with genesis by Antarctic weathering and supports textural evidence indicating that they are preterrestrial.
Geochemistry: A bulk interior sample of ALH 84001 was analyzed by INAA for a suite of major, minor, and trace elements (Table 2) using procedures described previously for diogenites (Mittlefehldt, 1993a). The REE pattern for ALH 84001 (Fig. 6) exhibits a depletion in LREE relative to HREE and a negative Eu anomaly as would be expected for a cumulate orthopyroxenite. However, the depletion in LREE is not as great as would be expected for a cumulate from a melt with chondritic REE ratios The chondritic normalized La/Lu ratio of ALH 84001 is 0.36, compared to ~0.007 expected for a cumulate based on opx/melt partition coefficients (calculated after Colson et al. (1988), using the 1200°C Eg experimental melt composition of Wasylenki et al., 1993). For comparison, a mean of five orthopyroxene clasts from the Johnstown diogenite shows the REE pattern of a cumulate orthopyroxenite from a melt with chondritic REE ratios (Mittlefehldt, 1993a).
Discussion: Classification of ALH 84001. Originally, ALH 84001 was classified as a diogenite based on preliminary examination (MacPherson, 1985) that showed it was an orthopyroxenite containing minor chromite, maskelynite of sodic composition, and Mg-Ga-Fe carbonates. Subsequent petrographic study of ALH 84001 as part of a larger study of diogenites and orthopyroxenites from howardites showed that the chromites were unusually rich in Fe3+ (Berkley and Boynton, 1992). The contents of minor elements in ALH 84001 orthopyroxenes are within the ranges for diogenite orthopyroxenes (Mittlefehldt, 1993a), although ALH 84001 orthopyroxene has a distinctly higher TiO2/Al2O3 ratio (Mittlefehldt, 1993b), and the FeO/MnO ratio of ALH 84001 (36) is higher than the average for diogenites (29 ± 1; based on data in Mittlefehldt, 1993a). Numerous petrologic features indicate that ALH 84001 is not a diogenite, however. Pyrite has not been reported from any diogenite (nor any HED meteorite); rather, troilite is the HED sulfide phase. Plagioclase in diogenites is much less sodic than those of ALH 84001; the most sodic plagioclase from typical diogenites is about Ab26, but most are <Ab18 (Gooley, 1972), compared to Ab63 for ALH 84001. The most sodic interstitial plagioclase thought to have crystallized from trapped melt in the ferroan Yamato-75032-type diogenites is Ab35 (Takeda and Mori, 1985). Carbonate has not been found in diogenites or any other HED meteorites, not even as Antarctic weathering products (Gooding, 1986; Mittlefehldt and Lindstrom, 1991). (Regardless, textures and compositions demonstrate that the carbonates are preterrestrial in ALH 84001.) Diogenite spinels typically contain <1 wt% TiO2 (Berkley and Boynton, 1992; Mittlefehldt, 1993a), except for Yamato-75032- type diogenite spinels, which contain ~1.8 wt% (Mittlefehldt and Lindstrom, 1993), and Fe2O3 from stoichiometry is <1 wt%, whereas ALH 84001 spinels contain 2.2 wt% TiO2 and 7.7 wt% Fe2O3 (Table 1). Berkley and Boynton (1992) calculate 5.2 wt% Fe2O3 for their average chromite analysis, which has slightly higher Cr2O3 and Al2O3 than those reported here. Most diogenites are monomict or polymict breccias composed dominantly of orthopyroxene clasts in a matrix of comminuted orthopyroxene, and the original grain size was of the order of several centimeters, not millimeters (Mason, 1963). The cataclastic texture of ALH 84001 is rare in other diogenites. Tatahouine does exhibit patchy extinction but does not contain crushed zones. Berkley and Boynton (1992) reported, and my own observations support, that the texture of the diogenite ALHA 77256 is similar to that of ALH 84001. Maskelynite is rare in HED meteorites, although much of the plagioclase in the eucrites ALHA 81313 and Padvarninkai is maskelynite.
The mineralogic and petrographic features of ALH 84001 are more in accord with those exhibited by the SNC meteorites. Pyrite is present in the nakhlites (Berkley et al. 1980; Boctor et al., 1976; Bunch and Reid, 1975) and Chassigny (Floran et al., 1978). The composition of maskelynite in ALH 84001 is similar to that of feldspars in other SNCs, particularly Nakhla (Bunch and Reid, 1975) and Chassigny (Floran et al., 1978). Plagioclase in shergottites is typically maskelynite (e.g., McSween, 1985), but in the nakhlites and Chassigny it is not (Bunch and Reid, 1975; Floran et al., 1978). Calcium carbonates have been identified as preterrestrial weathering products in the shergottite EETA 79001 (Gooding et al., 1988), and are inferred to exist in Nakhla based on carbon abundance and isotopic analyses by step-wise heating (Carr et al., 1985). Chromites in SNC meteorites commonly contain substantial calculated Fe3+ (e.g., Floran et al., 1978; McSween and Jarosewich, 1983) and magnetite is commonly present (e.g., Bunch and Reid, 1975; Stolper and McSween, 1979). The SNC meteorites show a wide variety of shock textures, including patchy extinction of coarse-grained orthopyroxene in EETA 79001 (McSween and Jarosewich, 1983), similar to that in ALH 84001, although crushed zones similar to those in ALH 84001 have not been described.
Although petrologically ALH 84001 is undoubtedly a member of the SNC clan, it is a unique member of this clan. ALH 84001 is the only orthopyroxenite member. The mineral compositions for ALH 84001 are much more homogeneous than is typical for SNC meteorites. Complex zoning of major and minor elements in pyroxenes is common in shergottites (e.g., McSween and Jarosewich, 1983; Stolper and McSween, 1979; Treiman et al., 1994). Pyroxenes in Chassigny are more homogeneous than those in shergottites, but they do show a wider range in CaO contents than do those of ALH 84001 (cf. Floran et al., 1978, Fig. 3). Only the nakhlites have relatively homogeneous pyroxenes. In them, however, homogeneous augite cores are zoned to narrow FeO-rich, CaO-poor rims (Harvey and McSween, 1992). Although I intentionally performed analyses within a few micrometers of the interstitial phases (maskelynite and apatite) in ALH 84001, no zoning in major or minor elements was detected. Similarly, chromite compositions in ALH 84001 are much more uniform than those of shergottites or Chassigny (Floran et al., 1978; McSween and Jarosewich, 1983). Hence, ALH 84001 records conditions of formation not encountered by other SNCs, and study of this meteorite will increase our understanding of martian petrologic evolution.
Petrogenesis of ALH 84001. The petrology and trace element composition of ALH 84001 can be used to evaluate its likely petrogenesis. Because these samples were taken from a limited region of this 1.9-kg stone, this should only be considered a preliminary interpretation; more detailed sampling and study could reveal petrologic or geochemical diversity that would necessitate significant revision.
The original texture of the protolith was coarse-grained, equigranular with commonly occurring 120° triple junctures, indicating slow cooling either during magmatic crystallization, or metamorphic recrystallization, or both. The uniform major and minor element compositions of the orthopyroxenes similarly indicate slow cooling. Pyroxene phase equilibria (Lindsley and Anderson, 1983) indicate that the composition of the orthopyroxene was quenched in at ~1050°C. Because ALH 84001 is nearly monomineralic, and texture and mineral compositions indicate slow cooling at high temperatures, it is likely that ALH 84001 is a cumulate.
The presence of interstitial sodic maskelynite suggests that a trapped liquid component is present. Many of the maskelynites are not stoichiometric plagioclases; they contain excesses of SiO2 over that needed to balance the CaO, Na2O and K2O (Fig. 4c). This suggests that the interstitial regions were composed of feldspar plus a SiO2 phase that were mixed by shock.
The REE pattern of ALH 84001 is also consistent with the presence of trapped, intercumulus melt. For example, the higher than expected La/Lu ratio of ALH 84001 can be explained by inclusion of ~7-9% melt component assuming the melt has a flat REE pattern at ~9-7× CI chondrites but with a negative Eu anomaly. A negative Eu anomaly is unusual for SNC parent melts in general (cf. Longhi, 1991), especially for ALH 84001 as the petrography suggests that plagioclase crystallization did not precede that of orthopyroxene. The necessity for a negative Eu anomaly in this model parent melt is robust; even assuming all Eu in the bulk rock is from the trapped melt, that melt cannot have had a Eu/Sm ratio >0.65 and still match the La/Lu ratio of the whole rock. Because the composition of the parent melt of ALH 84001 is not known, the amount of trapped melt in this meteorite cannot be rigorously quantified. In particular, the parent melts of nakhlites and Chassigny were LREE enriched (Boynton et al., 1976; Longhi, 1991), and it may be that the ALH 84001 parent melt was similarly LREE enriched.
The bulk Na content and the average maskelynite and orthopyroxene Na contents can be used to estimate that the bulk sample contains only about 1% maskelynite. The trapped melt was likely basaltic in composition, and the amount of normative plagioclase and pyroxene were probably subequal. Therefore, it seems likely that the amount of trapped melt in sample ,20 is more nearly 2% than 7-9%. In this case, the parent melt would have been LREE enriched, with a La/Lu ratio ~4× CI roughly one-half that inferred for the nakhlite parent magma (Longhi, 1991). This calculated parent melt also has a negative Eu anomaly, with Eu/Sm ~0.5-0.7. However, in this case, a Eu anomaly is not mandated by the calculation if the Eu partition coefficient is less than that used. Hence, the parent melt may have a REE pattern very similar to that estimated for the nakhlites (Longhi, 1991). This is a more likely parent melt for ALH 84001, both because it fits the petrographic constraints and because negative Eu anomalies are not present in SNC basalts (e.g., Fig. 6), nor are they inferred for other SNC cumulates (Longhi, 1991). Ion probe measurements of orthopyroxene and interstitial apatite should resolve this issue.
A third possibility is that the REE pattern of ALH 84001 was affected by late infiltration metasomatism, as has been suggested for the nakhlites (Berkley et al., 1980). The early, zoned carbonates provide petrographic evidence for multiple fluid fluxes in ALH 84001 and may be taken to indicate infiltration metasomatism. However, the early carbonates observed in ,64 are discrete grains in interstitial regions, and the textural relations do not indicate that they formed as metasomatic replacements of earlier phases. Nevertheless, an infiltrating fluid phase rich in REE could cause preferential enrichment in the LREE in the cumulate. In this case, the parent melt REE pattern could only be inferred from REE analyses of orthopyroxene cores, if they were unaffected by metasomatism.
The interstitial carbonates and pyrite show that hydrothermal alteration affected the protolith. The fine-scale zoning of some of the carbonates shows that they were formed under varying conditions, including multiple influxes of the fluid phase. Some of the carbonates were clearly formed prior to arrival on Earth as the fine compositional zoning in them is offset along fine-scale fractures as a result of shock (Fig. ld). Some of the carbonates were clearly formed after shock, as shown by their presence in the crushed zones and along shock-induced fractures in pyroxenes. These carbonates may have been early carbonates that were mobilized by the shock event, rather than a new generation. If the carbonates did not form metastably, that is, that they were not formed at low temperature with compositions in the two-phase field below the solvus, then the compositions of the early and late carbonates indicate that both were formed at moderately high temperatures (Fig. 5). In this case, the compositions of the late carbonates would then suggest that the protolith may have reached ~700°C after the shock event.
Relation to Other SNC Meteorites. ALH 84001 has several mineralogic and petrographic features in common with nakhlites and Chassigny. The nakhlites and Chassigny are cumulates with minor interstitial melt (Bunch and Reid, 1975; Floran et al., 1978), as is ALH 84001. The more generally homogeneous pyroxenes in nakhlites and Chassigny (Floran et al., 1978; Harvey and McSween, 1992), compared to those of the shergottites (e.g., McSween and Jarosewich, 1983; Stolper and McSween, 1979; Treiman et al., 1994), also indicate slower cooling for the former, similar to that inferred for ALH 84001. The maskelynite compositions of ALH 84001 are within the range of feldspar compositions in nakhlites and Chassigny (Bunch and Reid, 1975; Floran et al., 1978) but are more sodic than the majority of those in the shergottites (McSween and Jarosewich, 1983; Stolper and McSween, 1979; Treiman et al., 1994). However, the typical maskelynites in ALH 84001 (Fig. 4b) are similar to some of the silica plus plagioclase glass mesostasis compositions of Shergotty and Zagami, but with lower SiO2 contents (cf. Stolper and McSween, (1979), Table 3 and Fig. 6). The parent melt inferred above for ALH 84001, assuming no infiltration metasomatism, is LREE enriched as were those for the nakhlites and Chassigny. Last, pyrite is present in ALH 84001, nakhlites, and Chassigny but not in the shergottites.
However, petrologic models for the petrogenesis of nakhlites and Chassigny are not compatible with the petrology of ALH 84001. The estimated parent melts for nakhlites and Chassigny are saturated in olivine and augite, not orthopyroxene (see Treiman, 1993, for discussion). The orthopyroxenes in Chassigny, although compositionally identical to those of ALH 84001 (Fig. 2a, b), are exsolution lamellae in magmatic augites (Floran et al., 1975). Therefore, it does not seem likely that ALH 84001 is directly related to Chassigny or the nakhlites.
In contrast, although the shergottites contain strongly zoned minerals, and, in three of them, textures indicative of rapid crystallization, ALH 84001 shows several petrologic similarities with the shergottites. Orthopyroxene compositions in ALH 84001 are similar to those of the megacrysts from EETA 79001 lithology A (Fig. 2a, b). Similarly, the chromite compositions are similar to chromites in these megacrysts and to chromite cores of lithology A (Fig. 3). Hence, a parent melt like that which formed the megacrysts in lithology A might be a suitable parent (in terms of major element composition only) for ALH 84001. The orthopyroxene compositions suggest that the ALH 84001 parent is slightly more ferroan (Fig. 2b). However, the EETA 79001 megacrysts contain olivine (McSween and Jarosewich, l983), which is absent in ALH 84001. Wasylenki et al. (1993) have shown that olivine is in reaction relation with a melt composition thought to be parental to the EETA 79001 megacrysts. Hence, fractional crystallization processes could form an orthopyroxenite layer in a layered pluton. Plagioclase compositions are still a severe problem; the first plagioclase to form in the Wasylenki et al. (l993) experiments is An70, compared to An39 for the most calcic maskelynite in ALH 84001. Maskelynite compositions in ALH 84001 are also much more sodic than the majority of those in the shergottites. Last, the calculated parent melts for shergottites are LREE depleted (Longhi, 1991), which appears incompatible with the trace-element composition of ALH 84001, unless infiltration metasomatism affected the protolith.
Thus, ALH 84001 exhibits petrologic affinities for both the nakhlites/Chassigny on one hand and the shergottites on the other, as well as petrologic differences. The parent melt of ALH 84001 may have been intermediate between those of the nakhlite/Chassigny and shergottite groups of the SNC clan.
Nomenclature of SNC Meteorites. ALH 84001 does not fit well into any of the three named groups of SNC meteorites. Moreover, as these meteorites are widely believed to come from Mars, an undoubtedly petrologically diverse planet, we can expect (hope) that very different rock types from the SNC parent body might find their way into our meteorite collections. Hence, there is good reason to attempt to improve the nomenclature used for the SNC meteorites. Continuing to force differing lithologies into the closest petrologic group, as was done for ALHA 77005 and LEW 88516, obscures petrologic diversity and may prove to be wrong if continued investigation shows that the meteorites are not as closely related as originally thought. There is also little benefit to creating new group names for every unique meteorite that is found. In addition, because ALH 84001 (and really, ALHA 77005 and LEW 88516) is not an S, N, or C, I suggest that it is appropriate to consider dropping the SNC clan name entirely. Retaining the SNC name may have the disadvantage of devaluing meteorites like ALH 84001 (and ALHA 77005 and LEW 88516) that do not fit into those classes.
It is widely believed in the meteoritic community that the SNC meteorites are samples from Mars, and many meteoriticists and planetary scientists use compositional and petrologic data on SNC meteorites to infer geochemical and petrologic processes and properties for Mars (e.g., Bertka and Holloway, 1988, Dreibus and Wänke, 1985; Warren, 1987), or use martian geologic information to constrain petrogenetic models for SNC meteorites (e.g., McCoy et al., 1992). Therefore, I suggest that it is time to come out of the closet and openly refer to those meteorites as martian whose petrology and oxygen isotopic composition indicate that they are from the same parent body as Shergotty. Standard petrologic names may be added to specify the lighology; Shergotty would be a martian basalt, ALH 84001 would be a martian orthopyroxenite, etc. This has the advantage that these meteorite names would be easily understood by all petrologists, rather than just those interested in meteorites. In addition, it would make it clear to the scientific community, and to the public, that samples from Mars are in our collections and available for study.
Regardless of whether the meteoritic community adopts martian as the sanctioned clan name, I believe revising the name of the SNC clan along the lines given above has considerable merit.
Conclusions: ALH 84001 is an orthopyroxene cumulate member of the martian meteorite clan. Orthopyroxene and chromite are relatively uniform in comparison and indicate that the protolith cooled slowly. Interstitial sodic maskelynite, probably originally a mixture of feldspar and silica, makes up about 1% of the rock. The trace-element contents of ALH 84001 support a cumulate origin but suggest that it contains a small amount, perhaps ~2%, of a trapped liquid component. Carbonates in ALH are preterrestrial as indicated by shock damage showing that they predate the last shock event, and by their compositions, which indicate moderately high temperature of formation, inconsistent with Antarctic conditions. ALH 84001 is a unique lithology among the martian meteorites. Its mineralogic and petrographic characteristics show similarities to both the nakhlites-Chassigny and to the shergottites, but ALH 84001 cannot be easily related to the petrogenesis of either of these groups.
Acknowledgments: I would like to thank A. Treiman for discussions on the petrography of martian meteorites in general and for pointing out the definitive textural evidence indicating a preterrestrial origin for the carbonates. Thanks are due J. Jones for providing unpublished data from the Eg set of experiments. The efforts of R. Martinez and S. Wentworth to take the visible light and BSE photomicrographs are greatly appreciated. Comments on a draft version of this manuscript by J. Jones, M. Lindstrom, and A. Treiman, and formal reviews by H. McSween and P. Warren resulted in considerable improvement and are appreciated. Special thanks go to D. Sears for setting up an accelerated review process and to H. McSween and P. Warren for breathtakingly speedy, but thorough, reviews. This work was supported by NASA RTOP #152-13-40-21 to M. M. Lindstrom.