Pergamon 
Geochimica et Cosmochimica Acta, Vol. 61. No. 3. pp. 623-637, 1997 
Copyright 0 1997 Elsevier Science Ltd 
Printed in the USA. All rights reserved 
0016-7037/97 $17.00 + .OO 
PI1 SOO16-7037( 96) 00359-6 
A petrologic and isotopic study of lodranites: Evidence for early formation 
as partial melt residues from heterogeneous precursors 
T. J. MCCOY, I,*,* K. KEIL, ?.+ R. N. CLAYTON,~ T. K. MAYEDA,? D. D. BOGARD,? D. H. GARRISON,* and R. WIELER~ 
?Hawai?i Institute of Geophysics and Planetology, School of Ocean and Earth Science and Technology, 
University of Hawai?i at Manoa, Honolulu, Hawai?i 96822, USA 
?Code SN4, NASA/Johnson Space Center, Houston, Texas 77058, USA 
3Enrico Fermi Institute, University of Chicago, Chicago, Illinois 60637, USA 
?ETH Ziirich, Isotopengeologie, NO C61, CH-8092 Ziirich, Switzerland 
(Received October 10, 1995; accepted in revised fom October 14. 1996) 
Abstract-We have conducted petrologic, chemical, and isotopic studies of lodranites in an attempt to 
constrain their genesis. Lodran, Gibson, Y-791491, Y-791493, Y-74357, Y-8002, Y-75274, MAC 88177, 
LEW 88280, EET 84302, FRO 90011, and QUE 93148 are classified as lodranites. Lodranites and 
acapulcoites are indistinguishable on the basis of oxygen isotopic compositions but are distinct in average 
grain sizes of their mafic silicates, with lodranites being significantly coarser-grained. Lodranites exhibit 
a diverse range of petrologic and mineralogic features: they range widely in mafic silicate compositions 
(Fa3_,3)r plagioclase (O-l 1.4 vol%), Fe,Ni metal (0.5-20 vol%), and troilite (0.2-5.3 ~01%) contents; 
and shock levels (S 1 -S4). They appear to have experienced high peak temperatures and rapid cooling 
in the temperature range recorded by metallographic cooling rates (i.e., 700-350?C). The only dated 
lodranite, Gibson, cooled to Ar closure temperatures at 4.49 2 0.01 Ga. 
Lodranites formed from chemically and isotopically heterogeneous precursors in which the mineral 
and oxygen isotopic compositions were correlated. Heating of their parent body to temperatures between 
- 1050- 1200?C resulted in formation of Fe,Ni-FeS and basaltic partial melts. Depletions of troilite and/ 
or plagioclase in most lodranites testify to the removal of some of these partial melts, although melt 
migration was complex. Lodranites appear to have experienced a complex cooling history of slow cooling 
at high temperatures, followed by rapid cooling at intermediate temperatures, possibly related to breakup 
of the parent body. Lodranites were liberated from their parent body during l-3 impact events, with 
most having cosmic ray exposure ages of 5.5-7 Ma. The acapulcoites are samples from the same parent 
bodv but were heated to lower temoeratures and, thus, experienced lower degrees of partial melting. 
Cop&ight 0 1997 Elsevier Science itd 
1. INTRODUCTION 
In recent years, considerable progress has been made through 
the study of meteorites in deciphering the igneous histories 
of differentiated asteroids (e.g., Taylor et al., 1993). We 
have studied acapulcoites (McCoy et al., 1996a) and here 
present the results of our studies of the closely related lodran- 
ites. Ever since the work of Bild and Wasson ( 1976)) it has 
been recognized that Lodran is probably the residue of a 
partially molten region of an asteroid. With the recent recov- 
ery of a diverse suite of lodranites and acapulcoites (mostly 
from Antarctica), our understanding of the partial melting 
history of the acapulcoite-lodranite parent body has in- 
creased dramatically. We find that in contrast to acapulcoites, 
lodranites appear to have experienced significantly higher 
degrees of partial melting, including silicate partial melting, 
as well as melt migration. Thus, these rocks provide insights 
into the physical and chemical processes that took place 
during partial melting and differentiation in relatively small 
asteroidal parent bodies. The present paper addresses the 
properties and genesis of lodranites. A third paper (McCoy 
*Present address: Department of Mineral Sciences. MRC 119, 
Smithsonian Institution, Washington, DC 20560, USA. 
+Also associated with the Hawai?i Center for Volcanology. 
623 
et al., 1997) will address details of partial melting and melt 
migration on the acapulcoite-lodranite parent body. 
Several criteria can be used to distinguish acapulcoites 
from lodranites, and authors have differed on the application 
of these criteria to establish these groupings. These criteria 
were discussed in detail by McCoy et al. (1996a) and are 
only briefly reiterated here. Both acapulcoites and lodranites 
have mafic silicates with compositions between those of en- 
statite and ordinary chondrites and oxygen isotopic composi- 
tions which cluster around 5?O -3.5%0 and 6?O -0.75%0. 
Note that both the mineral and oxygen isotopic compositions 
of these groups are heterogeneous and, thus, are not useful 
for discrimination between acapulcoites and lodranites. A 
fundamental distinction between these two groups is that 
lodranites are coarse-grained whereas acapulcoites are fine- 
grained (McCoy et al., 1992a, 1993). As we show later, we 
suggest that this difference in grain size stems from a basic 
difference in their petrogenesis, and we use grain size for 
distinction between the groups. Some, but not all, lodranites 
have high contents of metallic Fe,Ni and, thus, have been 
called stony-irons by some previous workers (e.g., Prior, 
1916). Based on our criteria, Lodran, Gibson, Yamato 
791491, Yamato 791493, Yamato 74357, Yamato 8002, Ya- 
mato 75274, MAC 88177, LEW 88280, EET 84302, FRO 
90011, and QUE 93 148 are lodranites. We recognize that 
other workers (e.g., Zipfel and Palme, 1993) have utilized 
624 T. J. McCoy et al. 
different criteria, particularly noting that lodranites are de- 
pleted in elements enriched in early partial melts (e.g., Al, 
Na, LREE) . By this definition, some plagioclase-rich lodran- 
ites (e.g., Gibson, Y-8002, EET 84302) would be classified 
as acapulcoites. While we include these in our lodranite 
group, we recognize that these meteorites are, in many ways, 
transitional between the two groups, a topic we return to 
later in this paper. 
2. SAMPLES AND ANALYTICAL TECHNIQUES 
We studied ten of the twelve known lodranites (except Yamato 
75274 and QUE 93148; the latter was only recently identified as a 
lodranite by Mason, 1995), including at least one sample from each 
of the paired samples. Pairing has been suggested for Yamato 8002 
with Yamato 75274 (Yanai et al., 1984) and is evident for Yamato 
791491 with Yamato 791493 from the mineralogical data of Naga- 
hara and Ozawa ( 1986) and Hiroi and Takeda (1991), consistent 
with our data. The following thin sections were studied: Gibson, 
UH 192: Lodran, USNM 481-4; Yamato 791491, 61-2; Yamato 
791493, 91-2; Yamato 74357, 62-3; Yamato 8002, 51-3; MAC 
88177,26; LEW 88280, 16; EET 84302, 12; and FRO 90011,l (UH 
= University of Hawaii; USNM = United States National Museum; 
Yamato = National Institute of Polar Research; MAC, LEW, and 
EET = Meteorite Working Group; FRO = European Meteorite 
Group). 
Several of the meteorites we studied have only recently been 
recovered and/or recognized as lodranites. Gibson was found as a 
complete stone of 67.1 grams in October 1991. It was recovered by 
a prospector surveying a sandy area 150 m from a highway -15 
km north of Esperance, Western Australia ( -33?41 ?S, 121?48?E). 
The main mass is in the possession of David New. FRO 90011 was 
recovered in the 1990-1991 Antarctic field season and originally 
classified by Folco ( 1992) as an acapulcoite, but is actually a lodran- 
ite (McCoy et al., 1993). EET 84302 was classified as a unique 
achondrite (Score and Lindstrom, 1990) but is a lodranite (McCoy 
et al., 1993). 
Polished thin sections were studied in transmitted and reflected 
light. Mineral compositions were measured on Cameca Camebax 
and Cameca SX-50 electron microprobes at the University of Hawaii 
and corrections were made using a manufacturer supplied PAP ZAF 
program. Natural and synthetic minerals of well-known composi- 
tions were used as standards. The maximum dimensions of twenty- 
five mafic silicate grains in a single section were measured to deter- 
mine the average grain sizes. Of these, twenty-three were randomly 
selected, and we also sought the largest and smallest grains in each 
meteorite. Inclusion of these grains did not significantly affect the 
average grain size. Olivines poikilitically enclosed in pyroxenes are 
not included in these measurements, since they could not have grown 
by the same process as other grains. 
Modal analyses were conducted using optical point counting tech- 
niques in reflected light at a magnification of 200-400X. At this 
magnification, plagioclase, Fe,Ni metal, troilite, and weathering 
products can be confidently discriminated from mafic silicates. Point 
counting of 500- 1000 points per section covered the entire thin 
section. Uncertainty (22~) in the modal abundance for a phase 
which comprises 5-10% of the rock is typically &0.2-0.3% and 
for phases which comprise 0.5-l%, typically *0.7-l%. As we 
show later, errors in determining the abundances within individual 
lodranites are small compared to differences in mineral abundances 
between lodranites. Unfortunately, meteorite slices on many thin 
sections of Antarctic lodranites are very small (e.g., Y-8002,51-3 
has an area of -8 mm*). While it is possible that these small samples 
may not be representative of a larger volume of rock, we show later 
that this is unlikely given the extensive studies of the 1 kg Lodran 
meteorite, the uniform enrichments of metal and depletions of pla- 
gioclase found by all workers, and correspondence of bulk chemistry 
calculated from modal data and mineral chemistry with that directly 
measured. 
Polished thin sections were etched with a mixture of 1% nitric acid 
and alcohol for 20-30 set to reveal the kamacite-taenite structure for 
the purpose of determining metallographic cooling rates. We did not 
etch Gibson, which is highly weathered, and MAC 88177, which 
is highly shocked, since these meteorites would not yield reliable 
metallographic cooling rates. Taenite was found in Lodran, LEW 
88280 and Yamato 791491, but not in Yamato 8002, Yamato 74357, 
EET 84302, and Yamato 791493. We used the method of Wood 
( 1967) and the revised cooling rate curves of Willis and Goldstein 
( 1981). Taenite grains typically have rims of strongly zoned taenite 
l-20 pm in thickness, surrounding regions of plessite. Due to their 
small sizes and low central Ni contents, Ni was measured on etched 
taenite grains. This may introduce small, albeit insignificant errors 
considering the uncertainties in the absolute metallographic cooling 
rates. Measurements were made in the approximate centers of semi- 
equant taenite grains, the majority of which are rimmed by kamacite. 
Point selection is based on identification of the minimum Ni content, 
using 1 set counting times on an LIF crystal centered on the Ni Kcr 
peak. These taenite grains have an average P content of <0.04 wt% 
and, thus, P content should not add uncertainties to the rates. 
Measurements of oxygen isotopic compositions were made on 
Gibson, Yamato 8002, Yamato 75274, Yamato 74357, EET 84302, 
Lodran LEW 88280. and MAC 88177 (see Table 2). usine tech- 
niques described by Clayton and Mayeda~ ( 1963, 1983 )I Weayhering 
contaminants are abundant in some lodranites and, therefore, samples 
of Yamato 74357, Yamato 8002, LEW 88280, and EET 84302 were 
acid-washed for 2-3 min with 6m HCl at 70?C prior to isotopic 
analysis. Data reported by Clayton et al. ( 1984, 1992) on lodranites 
were preliminary and have been adjusted slightly. Oxygen isotopic 
compositions of FRO 90011 were collected using a laser fluorination 
technique (Franchi et al., 1992). 
A sample of Gibson (44 mg) was irradiated in the Los Alamos 
Omega reactor (irradiation constant J = 0.03368 2 0.00026). Step- 
wise temperature extractions were made in a high vacuum RF fur- 
nace equipped with a thermocouple, and the argon isotopes released 
were measured on a mass spectrometer. Corrections were made for 
extraction blanks, radioactive decay, and reactor-produced interfer- 
ences. Errors reported for individual ages include uncertainties in 
these corrections plus analytical uncertainty in measuring the wAr/ 
39Ar ratios in the samples and the hornblende neutron flux monitors. 
Errors in individual ages do not include an additional uncertainty of 
about 1r1% due to the accuracy of determining the flux and the 
absolute age of the hornblende. 
Noble gases were analyzed by mass spectrometry in three different 
chips of Gibson using techniques described by Graf et al. ( 1990a). 
Argon was separated from Kr and Xe by freezing the latter two 
gases in a charcoal trap at -90 to -100?C. In two samples, the 
gases were extracted in a single temperature step at 1800?C and in 
the other (73.9 mg), we tried to remove possibly loosely bound 
atmospheric noble gases in a first step at about 300?C. This was not 
successful, however, since the 300?C step contains less than 1% of 
either noble gas, including Kr and Xe, which clearly are predomi- 
nantly atmospheric; the data of this step are, therefore, not given. 
3. RESULTS 
Lodranites have been the subject of numerous publica- 
tions, including broader reviews (e.g., Nagahara, 1992; 
Talceda et al., 1994; Mittlefehldt et al., 1996). Here, we 
focus on those properties of lodranites which help unravel 
their origins, and reference will be made to previous work as 
appropriate. We will also contrast the properties of lodranites 
with those of acapulcoites (McCoy et al., 1996a). Relevant 
data on lodranites are summarized in Tables 1 and 2. 
3.1. Petrography and Mineral Compositions 
3. I. 1. Texture 
Lodranites are coarse-grained, equigranular rocks and 
consist mostly of low-Ca pyroxene and olivine, minor 
amounts of chromite, troilite, and chromian diopside, vari- 
Petrogenesis of lodranites 625 
TABLE 1. Modes and mineral compositions of lodraaites. 
?v? Low-Ca PyfQ&ene e Plaeipdase 
Rev. Rev. Zonlne. ? Fe,Ni Fe.9 Weathering 
Fa Z&Xc? FS Fe0 CaO Fs Wo vol. 40 An (vol.%) (vol%) (vol.%) 
Gibson 3.1 (I) N (1) 5.8 (I) Y 
r Yam&o 8002 3.5 (2) N 
( 
(2) 3.7 (2) N 
Yamato 75274 3.9 (3) N (3) 4.0 (3) N 
Yamato 74357 7.9 (4) N (4) 13.8 (4) Y 
EET 84302 8.4 (I) W (I) 8.3 (I) N 
FR090011 9.4 Y (I) 12.6 (12) Y 
I 
(12) 
Yamato 791491 -12 (4) Y (4) II.8 (4) N 
\;m&o79?493 II.6 (5) Y (5) 12.2 (5) W 
12.6 (6) Y (6) 13.8 (6) N 
MAC 88177 13.3 (7) N (7,s) 12.2 (10) N 
LEW 88280 12.9 (I) N (I) II.5 (1) N 
(1) 
(2) 
(3) 
(4-9) 
(1) 
(1) 
(4) 
7;) 
(X) 
w (1) 4.3 45.1 (1) 
Y (2) I.1 45.3 (II) 
Y (3) 1.4 44.3 (II) 
Y (9) 6.3 42.8 (11) 
Y (1) 3 42 (13) 
Y (I) 5.2 44.9 (1) 
Y (4) 5.2 43.9 (4) 
Y (IO) 5.0 44.5 (10) 
W (I) 6.4 43 (6) 
Y (I) 6 42 (738) 
W (I) 4.8 45.0 (I) 
6.0 (I) 18.1 (I) 0.2 0.5 19.4 (I) 
10.3 (2) 30.9 (2) 6.3 0.3 2.8 (2) 
3.4 (II) 28 (II) 19.6 0.2 - (11) 
Tr.* (1,ll) I5 (4) 9.6 2.4 - (11) 
II.4 (I) 23 (13) 13.1 0.1 2.8 (I) 
Tr. (I) 12.3 (12) 7.7 3.0 4.7 (I) 
0.25 (4) 18.2 (4) 4.2 5.3 4.5 (1) 
1.6 (11) 18.3 (10) 9.0 3.0 - (11) 
0 (6) - 20 2.7 - (6) 
0 (7) - 0.5 I.9 - (8) 
0 (1) - 10.3 2.5 5.5 (1) 
? Reverse Zonmg: Y - Yes, N No (Homogeneous). W Weak. 
* Previous analyses report no plagioclase. We identified a 2.2 mm long plagioclase (PTS .62-3X 
( Possibly pair&d me&rites. 
References: (I) This Work (2) Nagahara. 1992 (3) Mori et al., 1984 (4) Hiroi and T&da, 1991 (5) Yanai and Kojima, 1982 (6) Bild and Wasson, 1976 (7) Takeda et 
al., 1994 (8) Prinz et al.. 1991 (9) Miyamoto and Takeda, 1991 (IO) Nagahara and Ozawa, 1986 (11) Yanai et al.. 1984 (12) F&o, 1992 (13) Mason. 1986. 
able amounts of Fe,Ni metal, plagioclase, and hydrated iron 
oxides of terrestrial origin, and traces of phosphates and 
schreibersite. Numerous 120? triple junctions between grains 
are indicative of extensive recrystallization. Average mafic 
silicate grain sizes for most lodranites are -500-600 pm 
(Table 2). Yamato 8002 has a larger average grain size (700 
pm) but only thirteen grains were present in the section 
examined and, thus, this average has a large uncertainty. 
EET 84302 is finer-grained (340 pm) than most lodranites 
but significantly coarser-grained than most acapulcoites 
( 150-200 pm; McCoy et al., 1996a). Figure 1 illustrates 
the textures of Acapulco and Lodran at the same scale, dem- 
onstrating the coarser-grained nature of the latter. 
3.1.2. Shock effects 
None of the lodranites are brecciated, and many are un- 
shocked (shock stage Sl of Stoffler et al., 1991), such as 
Lodran, EET 84302, FRO 90011, Yamato 791491, Yamato 
791493, and LEW 88280. Gibson is also unshocked (Sl) 
but exhibits numerous semi-parallel veins of hydrated iron 
oxides of terrestrial origin cutting silicates. There is no evi- 
dence to suggest that these veins are related to shock. One 
olivine grain in Yamato 8002 exhibits undulatory extinction. 
However, the small number of grains present in our thin 
section (,5 l-3) prevents definitive classification, and we, 
therefore, consider Yamato 8002 to be unshocked to very 
weakly shocked (S 1 -S2). Yamato 74357 is weakly shocked 
(S3), exhibiting undulatory extinction of olivines. It also 
has pm-sized troilite and, rarely, metal grains decorating 
planar fractures in mafic silicates. MAC 88 177 has olivines 
with multiple, intersecting sets of planar fractures and strong 
mosaicism. Shock veins of whole-rock melt, including 
finely-dispersed opaques, crosscut the specimen. Large 
( > 100 pm) troilite grains are polycrystalline and Fe,Ni 
metal-troilite composite grains are finely intergrown, indica- 
tive of rapid post-shock cooling. These features suggest that 
the rock is moderately shocked (S4) and that the shock event 
took place after recrystallization and partial melting. Some 
of these features in Yamato 74357 and MAC 88177 were 
TABLE 2. Petrologic, chemical and isotopic data for lodraaites. 
GrainSize Two-Px Shock 8180 6170 At70 
(Pm) (?c)t Stage 0% %) (So) 
Cooling Rate 
Method 
Cosmic-Ray 
Gibson 540 1020 Sl 3.57 0.53 -1.33 f (I) 6.0-6.4 (1) Yamato 8002 700 1030 Sl-2 3.85 0.51 -1.49 
I (1) Yamato 75274 - 1080 2.52 0.23 -1.08 
(I) 
Yamato 74357 580 1100 s3 3.44 0.48 -1.31 (I) 1.5~106 Diffusion (7) 17.5-20.5 (8) 
EET 84302 340 1150 Sl 3.31 0.53 -1.19 (I) 
FRO90011 540 1020 Sl 3.73 0.98 -0.96 l (4) 5.9-6.4 (6) Yamato 791491 560 1060 Sl - - - 
I 
-103 MCR** 
(1) 
6.1-6.6 
(6) Yamato 79 1493 570 1030 Sl - 
0.92 -0;s 
-I@ Chromitc Size 
(5) 
Lodran 580 1070 s1 3.41 (3) -I@ MCR** (1) 4.0-4.3 (6) 
MAC 88 I77 620 1140 s4 3.52 0.60 -1.23 (1) 7.2-7.5 (6) 
LEW 88280 610 1020 Sl 3.40 0.78 -0.99 (2) -103 MCR** (1) 4.0-4.3 (6) 
t Two-pyroxene closure temperatures calculated from pyroxene compositions given in Table 1, 
* Reference indicates source of data used in calculation 
** Muallographic cooling rate method. 
( Possibly paired meteorites. 
Reference% (1) This Work (2) Clayton et al., 1992 (3) Clayton et al., 1984 (4) Franchi et al., 1992 (5) Nagabara and Ozawa. 1986 (6) 
Eugster and Weigel, 1993 (7) Miyamoto and Takeda, 1991 (8) Takaoka et al., 1993 
626 T. J. McCoy et al. 
Fig. 1. Plane polarized, transmitted light photomicrographs of 
Lodran (a) and Acapulco (b) to the same scale (bar = 1 mm). 
Although acapulcoites and lodranites have similar mineralogies and 
mineral and oxygen isotopic compositions, they exhibit markedly 
different grain sizes. Lodranites experienced silicate partial melting 
and became coarse-grained, whereas silicates in acapulcoites did not 
melt and, hence, acapulcoites are finer-grained. 
noted by Takeda et al. (1994), although not specifically 
attributed to shock. 
3.1.3. Compositions of ma$c silicates 
Lodranites exhibit a broad range of mafic silicate composi- 
tions (Tables 1, 3). Olivine ranges from Fa3., (Gibson) to 
Fa13.3 (MAC 88177). Individual olivine grains in a number 
of lodranites exhibit reverse zoning (Fe0 decreasing from 
core to rim), consistent with reduction. Reverse zoning is 
prominent in lodranites of intermediate olivine composition 
(Fa8.4_1Z.6). FRO 90011, Yamato 791491, Yamato 791493, 
and Lodran have strong reverse zoning, with cores of grains 
enriched in Fe0 by up to 1 wt% relative to the rims. Weak 
reverse zoning is observed in EET 84302, with cores higher 
by -0.5 wt% relative to rims. No zoning is observed in 
Gibson, Yamato 8002, Yamato 75274, Yamato 74357, MAC 
88177, and LEW 88280. 
Pyroxene Fs (molar Fe/( molar Fe + Mg + Ca)) approxi- 
mately correlates with Fa of co-existing olivines. Pyroxenes 
also show a broad range of compositions (Fs~,,-~~.~) and 
reverse Fs zoning. In general, lodranites with higher Fs con- 
tents in low-Ca pyroxene than Fa contents in olivine (e.g., 
Gibson, Fax., , Fs~.~) also have inverse Fs zoning in the pyrox- 
ene. This is the case for Gibson, Yamato 74357, FRO 90011, 
and Yamato 791493. However, Bild and Wasson (1976) and 
Papike et al. (1995) found no reverse zoning in low-Ca 
pyroxenes of Lodran (Fa12.6, Fs,~.~), whereas Takeda et al. 
( 1994) note both exsolution and zonation in pyroxenes, al- 
though the zoning is slight. In many other lodranites (e.g., 
Yamato 8002, Yamato 75274, EET 84302, Yamato 
79 149 1) , olivine and low-Ca pyroxene have nearly equal Fa 
and Fs contents. The two lodranites with the highest Fa 
(MAC 88177, LEW 88280) have Fa 1.1-1.4 mol% higher 
than Fs. Ratios of Fa/Fs vary from 0.53 (Gibson) to 1.12 
(LEW 88280). All lodranites exhibit reverse CaO zoning in 
orthopyroxene. In addition, many lodranites show very fine 
scale ( -1-5 pm) lamellae of clinopyroxene in orthopyrox- 
ene, and vice versa (e.g., Yamato 79 1491, Nagahara and 
Ozawa, 1986; MAC 88177, Takeda et al., 1994). 
High-Ca pyroxene is a volumetrically minor phase (<3 
~01%: Takeda et al., 1994) and, thus, compositional data 
are frequently from a single grain (Table 1). All contain 
significant amounts of Cr203 (0.74- 1.74 wt%) (Table 3) 
and, hence, are properly referred to as chromian diopsides. 
Both our studies and published analyses (e.g., Nagahara and 
Ozawa, 1986) reveal that high-Ca pyroxenes are also re- 
versely zoned in Fs. 
3.1.4. Equilibration temperatures 
The coexistence of low-Ca pyroxene and chromian diop- 
side in lodranites allows estimation of two-pyroxene equili- 
bration temperatures. The use of a two-pyroxene geothermo- 
meter in zoned pyroxenes which may have experienced con- 
tinuous cooling (e.g., Nagahara, 1992) could introduce 
errors. It is important to note that these are closure tempera- 
tures and, thus, reflect minimum peak temperatures for these 
rocks. We, therefore, use these data only to constrain these 
minimum peak temperatures. We have used the transfer 
equations of Kretz ( 1982) to calculate two-pyroxene equili- 
bration temperatures utilizing the average compositions of 
low-Ca and high-Ca pyroxene as determined with the elec- 
tron microprobe. Uncertainties resulting from pyroxene het- 
erogeneities are estimated to be -~50?C, using ? 1 CJ composi- 
tional variabilities in the average compositions. Errors re- 
sulting from derivation of the Kretz ( 1982) equations are 
260?C and the effect of CrzOX on the equilibration tempera- 
tures is unknown. We also recognize that a variety of other 
two-pyroxene thermometer formulations are available (e.g., 
Lindsley and Frost, 1992; Frost and Lindsley, 1992), but 
we have not applied these because these authors also did 
not explicitly consider the possible influence of Crz03 on 
equilibration temperatures. 
We find that two-pyroxene equilibration temperatures 
range from 1020? to 1150?C with the highest for EET 84302 
(1150?C) andMAC 88177 (1140?C) (Table2). Wedidnot 
observe chromian diopside in EET 84302 (MWG PTS ,12) 
but use the analysis by Mason (1986) of chromian diopside 
given in Table 1. The equilibration temperatures of the paired 
meteorites Yamato 8002 and Yamato 75274 differ by 
Petrogenesis of lodranites 621 
SiO, 420 40.4 39.8 
037 025 o.l9 
IiO.2 b.d a.d. b.d 
Al2@ bd. n.d. b.d. 
b.d 
12.2 
E 
083 
463 
024 
b.d 
Na20 ad. 
K20 ad. 
n.d. 
8.17 
0.11 
E 
so.2 
031 
bd. 
rid. 
n.d 
ad. 
n.d 
Tud 100.36 99.39 
10 
a.4 
_- 
98.78 
10 
12.9 
__ 
__ 
37.2 
028 
n.d 
nd 
nd 
55.6 
0.72 
a12 
0.03 
tz 
E: 
7.88 
019 
E 
329 
011 
132 
036 
bd. 
nd nd 
99.06 
10 
13 
1.8 
99.32 
10 
33.6 
z 
0.02 
l.M) 
OLI9 
1.74 
OJ2 
E 
E! 
17.4 
0.41 
21.3 
0.60 
0.65 
OD2 
a.d. 
99.22 
3 
i.3 
45.1 
-_ 
__ 
33.9 
039 
lid. 
z 
nd. 
3.31 
OJ1 
037 
17.7 
22.1 
ltd. 
nd. 
97381 
I 
i.2 
44.9 
-_ 
52.1 
031 
1.07 
I.60 
298 
034 
17.7 
22.0 
0.71 
n.d 
98.81 
1 
Ca 
45.0 
__ 
-_ 
64.3 
ad. 
226 
ad. 
0.28 
rd 
bd 
::I?: 
9.27 
0.12 
0.66 
0.06 
loo.98 
10 
6.1 
3.7 
-SOY, consistent with our estimated uncertainties owing to 
zoning. Although lodranites exhibit a range in temperatures, 
no obvious correlation exists between these and other param- 
eters (e.g., mafic silicate compositions, plagioclase con- 
tents ). 
3.1.5. Plagioclase composition, abundance, and 
morphology 
Plagioclase plays a particularly important role in under- 
standing the genesis of lodranites, in large part because of 
the variability it exhibits in composition, abundance, and 
morphology between different meteorites. Compositions 
range from An12.A-30.9 (Table 1; Fig. 2), but we find no 
LODRANITES 
12 [ ?? ? I 
0 5 10 15 
Fa (mole%) 
Fig. 2. Modal plagioclase contents of lodranites (~01%) vs. Fa 
(mol%) of olivines. Except for EET 84302, lodranites with low Fa 
tend to have higher plagioclase abundances, whereas those with high 
Fa have low plagioclase abundances. 
correlation with mafic silicate compositions or plagioclase 
abundances, although Nagahara ( 1992) suggested that such 
a correlation existed, based on limited data for four lodranites 
(Yamato 8002, Yamato 75274, Yamato 791493, Yamato 
74357). Lodranites also vary widely in plagioclase contents 
(0- 11.4 vol%, Table 1 ), with a tendency for those with 
high contents of Fa in olivine ( Fa,,.6_13,3) to have very low 
plagioclase contents and those with lowest olivine Fa con- 
tents (Gibson and Yamato 8002) to have high plagioclase 
contents (6.0- 10.3 ~01%; Fig. 2 j. The latter are only slightly 
less than or equal to those typically found in ordinary chon- 
drites (9.6-10.3 wt%; McSween et al., 1991). EET 84302 
does not follow this trend and has 11.4 ~01% plagioclase, 
despite having intermediate mafic silicate compositions 
(Fa8.4, Fss.3). 
Bulk chemical analyses are available for some, but not 
all, lodranites. Our modal estimates of plagioclase content 
agree well with plagioclase contents calculated from bulk 
chemical analyses of lodranites. Fukuoka et al. (1978) re- 
ported that Lodran is highly depleted in Al, Ca, Na, and K, 
consistent with the absence of plagioclase in the mode. Zipfel 
and Palme ( 1993 ) confirmed the highly plagioclase-depleted 
nature of MAC 88177 (0% modal plagioclase) and FRO 
90011 (trace). Furthermore, Mittlefehldt et al. ( 1996) also 
confirmed the highly plagioclase-depleted nature of LEW 
88280 (0%) and MAC 88177 (O%), while demonstrating 
that EET 84302 contains near-chondritic abundances of Na 
and Ca, consistent with our modal analysis of 11.4 ~01% 
plagioclase. The good agreement in plagioclase content de- 
termined by modal analyses and calculated from bulk chemi- 
cal analyses suggests that differences in plagioclase modal 
628 T. .I. McCoy et al. 
Fig. 3. Photomicrograph in transmitted light with crossed polars of 
Yamato 8002. Plagioclase (pl; striated phase) is interstitial to, and 
sometimes surrounds, mafic silicates (si ) and Fe,Ni metal (m), sug- 
gesting that it crystallized from an interstitial, basaltic melt. Scale 
bar = 250 pm. The dark area to the lower right is epoxy. 
abundances between lodranites are real and do not simply 
reflect nonrepresentative sampling of heterogeneous rocks. 
In particular, numerous studies of different samples of the 
1 kg Lodran mass have consistently shown the lack of plagio- 
clase. Clearly, this rock does not contain plagioclase-rich 
areas comparable to EET 84302. 
Plagioclase often occurs as mm-sized interstitial grains 
which surround multiple mafic silicate and opaque mineral 
grains (Fig. 3). The best example of this type is in Yamato 
8002, although a single 2.2 mm long plagioclase crystal 
observed by us in Yamato 74357 (PTS ,62-3) has the same 
texture. Plagioclase grains in Gibson and EET 84302 often 
surround single mafic silicate grains and pinch between mul- 
tiple mafic silicates. In Yamato 791493, which contains only 
1.6 ~01% plagioclase, mm-long plagioclase grains pinch be- 
tween mafic silicates, with multiple grains in contact and 
surrounding mafic silicates. 
Lodran, MAC 88 177, and LEW 88280 are free of plagio- 
clase but contain an Al-rich phase first noted in Lodran 
by Bild and Wasson (1976) that is close to stoichiometric 
(K,Na)AlSiS012. Prinz et al. ( 1991) report the occurrence 
of a SiO>-rich feldspathic glass of unknown origin along 
grain boundaries in MAC 88177. Prinz et al. (1978) argue 
that the phase in Lodran was trapped melt, whereas Bild and 
Tallant (1984) concluded that it is a crystalline mineral, 
based on Raman spectroscopy. The origin of this phase thus 
remains enigmatic. 
3.1.6. Opaque mineral occurrences and abundances 
Metallic Fe,Ni and troilite are present in all lodranites and 
occur predominantly as relatively large (typically 0.5 - 1 mm 
maximum dimension; 0.1-2.6 mm range) interstitial grains. 
Metallic Fe,Ni and, less frequently, Fe,Ni metal-troilite in- 
tergrowths or pure troilite also occur as volumetrically insig- 
nificant, lo-20 pm-sized blebs in the centers (but not the 
rims) of mafic silicate grains, primarily orthopyroxene. Simi- 
lar occurrences have been observed in acapulcoites (e.g., 
McCoy et al., 1996a). A third occurrence is in association 
with the enigmatic Al-rich material in Lodran, which Prinz 
et al. (1978) noted to contain some Fe-rich phase. We ob- 
served that the Fe-rich material consists of irregular, <5 
pm-sized grains of Fe-rich metal and troilite, intergrown 
with a darker (in reflected light), silicate material of uncer- 
tain composition and structure (Fig. 4). Due to its fine grain 
size and intimate intergrowths, we were unable to obtain 
reliable quantitative analyses of this material. TEM studies 
could elucidate these compositions and contribute to our 
understanding of their formation, but such analyses were 
outside the scope of this study. This material occurs mostly 
along grain boundaries and less frequently as diffuse areas 
at the edges of, and extending into, mafic silicate grains. In 
addition to the <5 brn-sized grains of Fe,Ni metal and troi- 
lite, distinct veins of Fe,Ni metal and troilite 52 pm wide 
and 5 160 pm in length also occur, with each phase forming 
segments of these veins. A similar occurrence is noted in 
the paired meteorites Yamato 791491 and 791493 and in 
FRO 90011, where pm-sized irregular blebs of troilite are 
observed at grain boundaries. This material is also in- 
tergrown with an unknown silicate phase which, while not 
darker in reflected light, does have low birefringence in 
transmitted light and crossed polars. The origin of these 
troilite-silicate intergrowths is uncertain, but it may be re- 
lated to reduction of mafic silicates as the result of interaction 
with a S-rich fluid. Papike et al. ( 1995) previously suggested 
that mafic silicate zoning in Lodran was due to interaction 
with a S-rich fluid. rather than a carbon-rich material, as in 
the ureilites. We note that the four lodranites (Lodran, Ya- 
mato 791491, Yamato 791493, FRO 90011), in which we 
observed these intergrowths, also show the most dramatic 
inverse Fe0 zoning in olivine. 
Lodranites contain variable abundances of Fe,Ni metal 
(0.5-20 ~01%; Table 1). This appears not to be an artifact 
due to terrestrial weathering, since many of the meteorites 
in Table 1 (except Gibson) are relatively fresh. No correla- 
Fig. 4. Reflected light photomicrograph of Lodran. A pm-sized 
intergrowth of troilite and a SiOz-rich material occurs at the edges 
of large metal (m) and troilite (tr) grains and along silicate (si) 
grain boundaries. Scale bar = 50 pm. 
Petrogenesis of lodranites 629 
tion is observed between metal abundances and mafic silicate 
compositions. 
Lodranites also contain variable abundances of troilite 
(0.1-5.3 ~01%). The apparent depletion of troilite (0.5%) in 
Gibson is undoubtedly an artifact due to its highly weathered 
nature ( 19.4% terrestrial weathering products), whereas the 
enrichment of troilite in Yamato 791493 (5.3%) may be the 
result of a bias due to extensive plucking of the silicates in 
the thin section studied. Other lodranites exhibit only modest 
effects of terrestrial weathering. Some lodranites (e.g., the 
paired Yamato 8002 and Yamato 75274) appear to be quite 
depleted in troilite relative to likely chondritic precursors, 
while others may have troilite contents only slightly less 
or comparable to ordinary chondrites (3.6-7.2 wt%; Keil, 
1962). This range in troilite abundances is supported by 
measurements of S and Se in lodranites by Dreibus et al. 
(1995). These authors found 2.00-2.56 wt% S and 8.85- 
9.76 ppm Se in ordinary chondrites. In comparison, the lo- 
dranites FRO 90011 (2.42% S, 5.85 ppm Se) and MAC 
88177 (1.49% S, 6.14 ppm Se) were inferred to be only 
slightly depleted in troilite, consistent with modal analyses 
of FRO 90011 (3.0 ~01% troilite) and MAC 88177 (1.9 
~01% troilite). Gibson is severely depleted in S and Se, 
which Dreibus et al. ( 1995) also attributed to its extensive 
terrestrial weathering. Mittlefehldt et al. ( 1996) used the Se/ 
Co ratio for LEW 88280, MAC 88177, EET 84302, and 
FRO 90011 to examine the removal of troilite. These authors 
found that LEW 88280, MAC 88177, and FRO 90011 have 
approximately the same Se/Co ratio as acapulcoites, consis- 
tent with moderate to high modal troilite abundances. How- 
ever, EET 84302 has a much lower Se/Co ratio, consistent 
with its 0.1 ~01% modal troilite content. As in the case of 
plagioclase. bulk chemical analyses for S and Se are broadly 
consistent with modal data and suggest that the wide range 
of troilite abundances in lodranites is not due to poor sam- 
pling but is real. 
3. I. 7. Cooling rates 
Previous work on lodranites has generated a wide range 
of cooling rates. Cooling rates on several lodranites by a 
variety of methods have yielded rapid cooling rates at moder- 
ate temperatures. Bild and Wasson ( 1976) used the metallo- 
graphic method of Wood (1967) to derive a rate of - 104?C/ 
Myr for Lodran in the range of 600-400?C. Miyamoto and 
Takeda ( 1991) determined a rate of 1.5 X 1 Oh?C/Myr for 
Yamato 74357, based on modeling of Fe-Mg and Ca zoning 
in orthopyroxene in the range of lOOO-600?C. The presence 
of CaO zoning of comparable magnitude in other lodranites 
also suggests rapid cooling. The presence of high-Ca pyrox- 
ene lamellae in low-Ca pyroxene and vice versa might pro- 
vide further constraints on the thermal history of lodranites, 
although we are unaware of any such studies (Takeda et al., 
1994). Nagahara and Ozawa ( 1986) derived a rate for Ya- 
mato 791493 of -103?C/Myr between 800-6Oo?C, based 
on spine1 grain size and spinel-olivine equilibration tempera- 
tures. 
Slower cooling rates have been inferred at lower tempera- 
tures. Bild and Wasson ( 1976) calculated a low temperature 
cooling rate of -lO-30?UMyr for Lodran, using the 
method of Short and Goldstein ( 1967 ) which involves the 
measurement of maximum Ni content at the edge of taenite 
grains. Prinz et al. (1978) determined a rate based on size 
and composition of schreibersite lamellae of lO?CA4yr using 
the technique of Hewins and Goldstein ( 1977). J. I. 
Goldstein (pers. commun.) suggested that the metallographic 
A. Lodran B. LEW 88280 C. Yamato 791491 
5 ? I I I I 
4 10 20 40 60 100 
I I I I I 
4 10 20 40 60 100 
Distance to nearest edge (pm) 
4 10 20 40 60 100 
Fig. 5. Plots of central Ni concentration vs. distance to nearest edge for taenite grains in lodranites. Curves for 
cooling at 0.1 - lOO?C/Myr from Willis and Goldstein ( 198 1). Curves for lOOO- lOOOO?C/Myr (dashed lines) are 
approximate. Our data suggest a rate of 
suggest 
-104-105T/Myr for Lodran, whereas those of Bild and Wasson ( 1976) 
--104?C/Myr. Our data suggest that LEW 88280 cooled at a rate of 
- lO~?C/Myr. 
-lO??C/Myr and Yamato 791491 at 
630 T. .I. McCoy et al. 
techniques used by Bild and Wasson ( 1976) and Prinz et al. 
( 1978) may have been inappropriate for use in the lodranites. 
which contain high-Ni metal. Thus, evidence for slow cool- 
ing at low temperature is, at best, tenuous. 
We measured metallographic cooling rates for those lo- 
dranites that contain taenite (Lodran, LEW 88280, Yamato 
791491; Fig. 5). Our data for Lodran scatter but suggest a 
rate of - 104-1050C/Myr. This is somewhat faster than that 
found by Bild and Wasson ( 1976) (- 104?CA4yr) using the 
same technique, although both groups find fast rates for the 
temperature interval of -7OO-4OO?C, to which metallo- 
graphic methods apply. We also find fast rates ( -103?C/ 
Myr) for LEW 88280 and Yamato 791491, although we 
detected only three taenite grains (35-40 pm diameter) in 
the latter, resulting in a greater uncertainty in the cooling 
rate estimate. 
3.2. Oxygen Isotopic Compositions 
Oxygen isotopic compositions of lodranites are similar to 
those of acapulcoites but differ from those of IAB and IIICD 
irons (the most abundant groups of silicate-bearing iron me- 
teorites; Choi et al., 1995)) winonaites, and ureilites (Table 
2; Fig. 6). Clayton et al. ( 1992) discussed possible relation- 
ships of acapulcoites and lodranites with other meteorite 
groups (e.g., Kakangari, CR chondrites, a group of carbona- 
ceous chondrites; Kallemeyn et al., 1994) but considered 
similarities in oxygen isotopic compositions fortuitous. 
Thus, acapulcoites and lodranites are from a different parent 
body than the other meteorite groups. Acapulcoites and lo- 
dranites cannot be distinguished from one another on the 
basis of oxygen isotopes but can be distinguished based on 
texture and composition (McCoy et al., 1996a). Clayton et 
al. (1992) also noted that acapulcoites and lodranites, as a 
group, exhibit variations beyond what could be attributed 
to mass-dependent fractionation and analytical uncertainty. 
Fig. 6. Three-isotope oxygen plot for primitive achondrites. Aca- 
pulcoites, lodranites, and the related, unique meteorite LEW 86220 
are readily distinguished from IAB irons, IIICD irons, and wino- 
naites, as well as ureilites, but acapulcoites and lodranites cannot 
be distinguished from one another on the basis of oxygen isotopic 
composition. Data from Clayton and Mayeda (1988, 1996), Clayton 
et al. (1983, 1984, 1992). and Franchi et al. (1992). 
-0.75 
Ledran H 
t 
FR090011 ??
-1.00 LEW 88280? 
#.o Y75274 0 
-4 EEI 84302 ??
-1.25 MAC 88177 0 
I Gibson ?? Y 74357 ??
Y8#2 
-1.50 1 
0 5 10 15 
Fs (mole%) 
Fig. 7. Fa in olivine and A ?0 for lodranites are weakly positively 
correlated. Non-Antarctic and acid-washed Antarctic lodranites are 
represented by solid squares; nonacid washed Antarctic lodranites 
(Yamato 75274, MAC 88177) are represented by open squares and 
have higher uncertainties. This correlation is of nebular origin and 
suggests that lodranites formed from a chemically and isotopically 
heterogeneous precursor. 
Nine analyzed lodranites have an average A?0 = -1.16 
-t 0.21%0. In contrast, equilibrated H chondrite falls have an 
average Al70 = 0.73 ? 0.09%0 (N = 22) (Clayton et al., 
1991) and, thus, lodranites exhibit an isotopic variation more 
than twice that of H chondrites. Note that 2 values refer to 
standard deviations of the analyses, not standard error of the 
means. 
Oxygen isotopic compositions of lodranites are weakly 
positively ( r2 = 0.378) correlated with Fa contents of olivine 
(Fig. 7). However, the Antarctic lodranites Yamato 75274 
and MAC 88177 were analyzed without acid-washing, which 
at the time was not generally applied as a means of removing 
terrestrial weathering, and new samples for acid washing 
and subsequent analyses were, unfortunately, unavailable. 
Clayton and Mayeda (1988) showed that acid-washing, 
which removes terrestrial contamination, causes small 
(-0.1%0) but significant, positive or negative shifts in A170. 
While such shifts do not affect our conclusions regarding 
the classification of these meteorites as lodranites, they ob- 
scure the subtle correlations examined here: When these 
two meteorites are excluded, the correlation coefficient r? 
= 0.795 and, thus, A?0 and Fa are then positively corre- 
lated at the 95% confidence level. Positive correlation be- 
tween these parameters is also observed for the ureilites: For 
the non-Antarctic ureilites Dingo Pup Donga and Dyalpur 
and the acid-washed Antarctic ureilites ALH 84136, Yamato 
74659, Yamato 791538, and LEW 85440, which have Fa 
concentrations between Fa25_16 and A?0 of - 1.29 to -2.42 
(oxygen isotopic data from Clayton and Mayeda, 1988)) we 
calculate a comparable r? = 0.764. 
Although we argue that acapulcoites and lodranites are 
related and are from the same parent body (e.g., McCoy et 
al., 1996a), acapulcoites do not show a clear relationship 
between A?0 and Fa, as do the lodranites: Although the 
low-Fa acapulcoites ALH A81 187 and ALH 84190 also have 
the most negative A?0, the overall Fa distribution of aca- 
pulcoites is bimodal and no clear correlation exists with 
oxygen isotopic composition (Note that the lodranites were 
Petrogenesis of lodranites 631 
Table 4. Argon isotopic data for G~hson. From left to right, columns are extraction temperature (?c), 39Ar concentration 
(1O-9 a&I-F/t& w ia Ga, K&a ratia, and conwted ?%/-h, %@?41, %+@A& and SArpAr ratios. The 
K/Ca and %rpAr ratios have been multiplied by 100. Uncertaintis were derived from those associated with 
spectrometer measurement& and decay, blank, and reactor corrections. They do not include the uncertainty in irradiation 
constant, J= 0.03363 kO.OtW26. 
Temp 3QAr AGE K/Q 40/39 38/39 37139 38139 
Ga Xl00 Xl00 
400 0.80 5.468 iO.028 4.88 io.10 585.4ic10.4 81.72 fl.11 11.33 k0.23 lQ8.8-+8.0 
500 0.71 3.702 iO.029 3.96 *o.OQ 201.4k4.1 34.78 to.74 13.31 *0.30 24.924.1 
850 2.18 4.159 iO.014 8.99 kO.10 268.1i2.8 15.27 tO.15 7.54 io.ll 13.5sz1.5 
750 5.39 4.481 ~~0.009 17.15 *0.20 328.3il.Q 3.18 kO.02 3.08 TO.04 2.320.5 
825 10.20 4.499 to.003 20.29 io.21 32Q.7*0.7 1.42 ?O.OO 2.80 io.03 1.1 to.2 
875 9.78 4.502 iO.003 18.98 kO.19 330.3sco.7 1.50 to.00 2.78 iO.03 1.8iO.3 
925 8.82 4.482 io.004 15.83 ico.17 328.4iO.Q 2.04 to.01 3.33 io.03 2.2f0.4 
1000 4.13 4.453 io.Oo7 8.48 kO.07 320.7kl.8 4.08 to.02 8.17 io.09 5.020.7 
1100 3.82 4.473 io.oo8 1.75 *0.02 324.7k1.2 2.4Qio.01 30.20 i0.32 8.7k0.8 
1200 2.14 4.498 20.023 0.13 kO.00 32Q.5*5.0 7.58 i0.12 394.23 k7.20 44.7k2.8 
1300 0.21 5.443 iO.128 0.05 i0.w 577.0*47. 21.90 ii.32 1084.8 M8.9 38Ot39. 
1450 0.15 5.439 io.150 0.09 io.01 575.5254. 21.13 ~~2.15 810.9 k58.4 477*5a. 
1800 0.03 8.031 io.533 0.18 io.08 2518t814 40.59 zt13.7 2Q8.5 iQ8.9 1388*510 
subject to the same pre-analysis acid wash as the acapul- 
coites). We selected A I70 as the measure of oxygen isotopic 
composition, since it is unaffected by mass-dependent frac- 
tionation. Such fractionation could have occurred in the lo- 
dranites as a result of plagioclase loss. Plagioclase is en- 
riched in ?0 relative to mafic silicates (Clayton, 1993) and, 
thus, plagioclase removal could move the residues to lower 
6 ?0 values along mass fractionation lines. We do not, how- 
ever, believe that this process determined the isotopic com- 
positions of lodranites, since all acapulcoites and some lo- 
dranites (e.g., Gibson, Y-8002) contain chondritic abun- 
dances of plagioclase but still differ significantly in 6r*O. 
3.3. j9Ar-??Ar Chronology 
Argon isotopic data for stepwise temperature extractions 
of an irradiated sample of Gibson are given in Table 4 and 
calculated 39Ar-40Ar ages and K/Ca ratios as a function of 
fractional release of 39Ar are shown in Fig. 8. If we discount 
4.8: O 
: .,.: 
: -1 
19 
Gibson s? ; -2 
AQe=4.4QGa 
WI =2a3 Ppm 
3.8 
i ..) 
-3 
0 0.2 0.4 0.8 0.8 1 
CUMULATIVE 39Ar RELEASE 
Pig. 8. 3yAr-40Ar age (rectangles) and K/Ca ratio (dotted lines) 
as a function of cumulative release of 39Ar for stepwise extractions 
of Gibson. Individual age uncertainties are indicated by the width of 
the rectangle. Potassium concentration and the temperature releasing 
50% of the total ?9.4r are also indicated. Gibson has a plateau age 
of 4.49 Ga. 
the first three extractions that indicate recent Ar diffusive 
loss and the three last extractions, then seven extractions 
(750- 1200?C) releasing 90% of the j9Ar define a 39Ar-40Ar 
age of 4.484 t 0.018 Ga (where the error is la of deviations 
from the mean). The measured K content for our sample of 
Gibson (203 ppm) is less than that for chondrites, indicating 
that our sample contained less feldspar than that present in 
the thin sections studied (Table 1). 
The Ar release spectrum for Gibson shows some age varia- 
tions that deserve closer examination. Most of the 39Ar re- 
lease is associated with a phase having a K/Ca ratio of 0.1 S- 
0.2 (Table 4; Fig. S), which is probably feldspar. The first 
-8% of the 39Ar release show a somewhat lower ratio and 
younger ages. These may represent effects of terrestrial 
weathering. Both the decrease in the K/Ca ratio (Fig. 8) 
and changes in the rate of release of Ar as a function of 
extraction temperature suggest that a different phase de- 
gasses 39Ar at temperatures of - 1000?C and above. We as- 
cribe the slight decrease in age at -8O-85% 39Ar release to 
weathering effects of grain surfaces of this second phase, a 
phenomenon we have observed in other weathered meteor- 
ites. Because of the coarse grain size of Gibson (Table 2) 
and the relatively large fraction of the total 39Ar released 
from this higher temperature mineral phase, we doubt that 
the slight decrease in age at -1000?C is caused by 39Ar 
recoil redistribution in the irradiation. However, we cannot 
totally rule out this explanation. Because of the uncertainty 
in interpretation of the Ar data from the 1000?C extraction, 
a preferred ?9Ar-40Ar degassing age of Gibson would be 
4.493 2 0.011 Ga, which considers only the 825-925?C and 
1 lOO- 1200?C extractions releasing 71% of the total ??Ar. 
The 1300- 1450?C extractions suggest substantially older 
ages of -5.4 Ga associated with a phase with quite low K/ 
Ca. It is unlikely that these older ages can be explained by 
uncertainties in 40Ar blank corrections, which are only -5% 
for these extractions. (The large blank correction for the 
1600?C extraction, which released much less 40Ar, does make 
its age very uncertain). Reactor-generated corrections are 
even less important. We note that these high temperature 
632 T. J. McCoy et al. 
TABLE 5. Noble gases in Gibson. 
. 
?v@s(mg) )He 4He 20Ne 20N&*Ne ZZN&lNe 36Ar 36ArPAr 
Cosm~on 
aAr *4Kr l32Xe 2lNe 38Ar 22N&*Ne 
66.1 8.31 193 2.34 1.348 1.313 8.24 5.14 3190 0.307 0.040 1.32 0.067 1.211 
0.40 35 0.09 0.007 0.007 0.33 0.03 130 0.020 0.003 
144.1 8.66?1090 1.74 1.010 1.253 2.21 4.39 1830 0.097 0.024 1.37 0.099 1.211 
0.40 45 0.07 0.005 0.006 0.09 0.02 70 0.008 0.002 
13.9 8.62 945 2.17 1.225 1.290 4.95 5.08 3470 0.238 0.042 1.37 0.054 1.211 
0.40 40 0.08 0.005 0.006 0.20 0.03 140 0.018 0.003 
Noble gas concentrations in 108 cm3SlWg. la errors are given in italics. 
extractions also release the overwhelming amount of 36Ar 
(Table 4)) most of which is trapped rather than cosmogenic 
gas (see next section). (However, the concentration of 36Ar 
in our sample is approximately an order of magnitude less 
than that in the unirradiated samples reported below). Thus, 
it is possible that these older ages reflect the presence of 
excess radiogenic 40Ar, which was produced prior to the time 
of strong metamorphism, but was not totally lost from this 
retentive phase because of significant burial in the parent 
body. 
The 39Ar-40Ar age uncertainty quoted above does not in- 
clude the uncertainty in the irradiation constant. If we include 
this additional uncertainty, the absolute degassing age be- 
comes 4.49 ? 0.04 Ga. This is the first age determination 
of a lodranite, although Torigoye et al. ( 1993) reported that 
whole-rock Rb-Sr systematics of the lodranite Yamato 8002 
are consistent with a 4.50 Ga age, assuming the same initial 
isotopic composition as for a CA1 in Allende. They did not, 
however, determine the age of the meteorite. We can also 
compare the age of Gibson with 39Ar-40Ar ages of two aca- 
pulcoites (McCoy et al., 1996a). Because Gibson and the 
acapulcoite Monument Draw were included in the same irra- 
diation, their 39Ar-40Ar ages can be compared using uncer- 
tainties that do not include those in the irradiation constant. 
The age of Monument Draw is precisely defined at 4.517 
2 0.006 Ga (lg) and Gibson at 4.493 ? 0.011 Ga. Thus, 
these ages agree within 2g, but not la, and, thus, it is not 
impossible that the Ar degassing time for Gibson was actu- 
ally -0.02 Ga later than that for Monument Draw. Such 
an age difference, if real, may reflect the somewhat higher 
metamorphic heating of the lodranites compared to the aca- 
pulcoites and may suggest a longer cooling time to Ar clo- 
sure temperatures for the former. 
3.4. Noble Gases 
3.4.1. Cosmic ray exposure age 
The concentrations of cosmogenic ?Ne and the shielding 
parameters (22Ne/2?Ne),,, are very similar in all splits of the 
Gibson lodranite (Table 5). The rather high (**Ne/?Ne),,, 
ratios indicate a low shielding of a few cm at most and 
presumably a small preatmospheric size of Gibson. With the 
nuclide production model of Graf et al. ( 1990b), we estimate 
a shielding-corrected ?Ne exposure age of 6.0-6.4 Ma. We 
assume a *?Ne production rate typical for L chondrites, which 
is appropriate in view of the Mg concentration ( - 15.9 wt%) 
estimated from modes and mineral compositions. We also 
make the reasonable assumption that the preatmospheric ra- 
dius of Gibson (which was found as a whole stone of only 
67.1 g) was less than 30 cm. 
Eugster and Weigel ( 1993) report noble gas data for five 
lodranites. The *?Necos and also the ( 22Ne/2?Ne),,, values in 
Yamato 791491, MAC 88177, and FRO 90011 are similar 
to those in Gibson and, thus, their nominal exposure ages 
are also similar (Yamato 791491,6.1-6.6 Ma; MAC 88177, 
7.2-7.5 Ma; FRO 90011, 5.9-6.4 Ma; all calculated with 
the Graf et al. (1990b) model and the same assumptions 
as stated above for Gibson). For Lodran and LEW 88280 
(Eugster and Weigel, 1993 ) , we deduce nominal shielding- 
corrected *?Ne exposure ages of 4.0-4.3 Ma. The differences 
between these ages and those of the former four lodranites 
seem to be significant, although errors are on the order of 
20%. Takaoka et al. (1993) report a 21Ne exposure age of 
11.7 +- 1.5 Ma for the lodranite Yamato 74357. However, 
with our method, we calculate a considerably higher 2?Ne 
age of 17.5-21.5 Ma, whereby the upper limits reflect the 
assumption that the preatmospheric radius of Yamato 74357 
did not exceed 70 cm. This discrepancy is due to the fact 
that Yamato 74357 has a low (**Ne/*?Ne),,, ratio of only 
- 1.075, which indicates rather high shielding. As is shown 
by Graf et al. ( 1990b), the shielding correction of Eugster 
(1988) used by Takaoka et al. (1993) overestimates *?Ne 
production rates at (22Ne/21Ne),,, < 1.08. In any case, Ya- 
mato 74357 has a higher cosmic ray exposure age than any 
of the other lodranites discussed here. Thus, at least two, 
and possibly three, ejection events are needed to account for 
the exposure ages of all lodranites studied. It is interesting 
to note that four of seven lodranites have exposure ages in 
the same -5.5-7 Ma range as all acapulcoites (e.g., McCoy 
et al., 1996a). This is consistent with a single ejection event 
on a common source body for most acapulcoites and lodran- 
ites for which cosmic ray exposure ages have been measured. 
3.4.2. Trapped noble gases 
Lodranites and acapulcoites usually contain relatively 
large amounts of trapped noble gases, similar to those in 
type 3 or 4 chondrites (Palme et al., 1981; Eugster and 
Weigel, 1993; Takaoka et al., 1993; McCoy et al., 1996a). 
The isotopic compositions of Kr and Xe are often identical 
Petrogenesis of lodranites 633 
to those of component Q found in carbonaceous chondrites 
(Wieler et al., 1992). In all three analyses of Gibson, how- 
ever, Kr and Xe have atmospheric composition (data not 
given). The only exception is lZ9Xe. The ratio of ?29Xe/ 
13?Xe ranges between 1.12- 1.27, indicating the presence 
of radiogenic lz9Xe. Otherwise, Kr and Xe in Gibson are 
obviously mostly atmospheric contamination, presumably 
trapped in the abundant weathering products. A generous 
upper limit for primordial 13?XeQ is 0.005 X lo-? cm3STP/g, 
which is roughly 20x less than the trapped Xe in, for exam- 
ple, Monument Draw and is also less than what is usually 
found in type 5 or 6 chondrites. Unfortunately, the concentra- 
tion of trapped 36Ar cannot be reliably determined because 
of the possible contamination with atmospheric Ar and be- 
cause the concentration of radiogenic 40Ar is difficult to esti- 
mate reliably. Since, at least in chondrites, a large fraction 
of the trapped Xe survives dissolution of the meteorite by 
HF/HCl, it seems unlikely that the primordial trapped Xe 
in Gibson could have been lost nearly completely during 
terrestrial weathering. 
4. DISCUSSION 
4.1. Properties of the Precursor Chondritic Material 
We have previously presented evidence that the acapul- 
coites and, by inference, the lodranites, formed by thermal 
metamorphism and partial melting from chondritic precursor 
materials on the same parent body (McCoy et al., 1992a,b, 
1996a). In the case of lodranites, all evidence for the original 
texture of this chondritic precursor has been erased by heat- 
ing and partial melting. However, we suggest that the min- 
eral and oxygen isotopic compositions of lodranites were 
largely inherited from the chondritic precursor. We further 
suggest that some properties of lodranites resulted from neb- 
ular processes. This is because similar ranges in mineral and 
isotopic compositions as observed in lodranites have been 
noted in other groups of meteorites and attributed to a nebu- 
lar origin. For example, ordinary H, L, and LL chondrites 
(Clayton et al., 1991) and ureilites (Clayton and Mayeda, 
1988) show a correlation between A ?0 and mafic silicate 
compositions, not unlike what is observed in lodranites. 
4.2. Heating and Melting 
Lodranites have experienced extensive modification by 
heating (as is evident from the equigranular textures and 
abundant 120? triple junctions) and partial melting. This has 
previously been suggested by Nagahara ( 1992), Takeda et 
al. ( 1994). and Mittlefehldt et al. ( 1996). 
4.2.1. Evidence for Fe,Ni-FeS partial melting 
Lodranites have experienced Fe,Ni-FeS cotectic melting. 
The first partial melt in a chondritic system forms between 
9.50 and 98O?C, depending on composition, and consists of 
-85 wt% troilite (FeS) and - 15 wt% Fe,Ni metal (Kul- 
lerud, 1963; Kubaschewski, 1982). Thus, removal of a small 
amount of cotectic melt has a dramatic effect on the troilite 
abundance of the residue. It appears that Fe,Ni-FeS cotectic 
melting has occurred in the lodranites at 950- 1050?C. Some 
unweathered lodranites are highly depleted in modal troilite 
(e.g., Y-8002, Y-75274, EET 84302) relative to likely chon- 
dritic precursors (e.g., 3.6-7.2 wt% FeS in ordinary chon- 
drites; Keil, 1962). reflecting nearly complete removal of the 
Fe,Ni-FeS cotectic melt. Many other lodranites have modal 
troilite abundances and bulk S contents at chondritic or only 
slightly below chondritic levels. Evidence for Fe,Ni-FeS co- 
tectic melting and melt migration in the form of irregular 
patches and occasional veins of metal and troilite is found 
in some lodranites (e.g., Lodran, Y-791491, Y-791493, FRO 
90011) . McCoy et al. ( 1996a) argued that acapulcoites had 
experienced melting at the Fe,Ni-FeS cotectic, but not sili- 
cate partial melting. In the acapulcoites, the Fe,Ni-FeS cotec- 
tic melt is trapped in veins which range from pm- to cm- 
sized. In contrast, such veining is not common in lodranites. 
4.2.2. Silicate partial melting 
It is also apparent that lodranites have experienced silicate 
partial melting. The first melt in a chondritic system is basal- 
tic (Morse, 1980). Silicate partial melting of an olivine-rich 
rock in the system Fo-An-Si02 occurs at the peritectic point. 
This melt contains -55% plagioclase (Morse. 1980). We 
recognize that plagioclase in lodranites is albitic and the 
phase boundaries would shift in a Fo-Ab-SiO, phase dia- 
gram, which is currently unstudied. Examination of the Ab- 
Si02 binary phase diagram (Tuttle and Bowen, 1958) sug- 
gests that the boundaries would shift so that the initial melt 
would contain an even greater abundance of plagioclase. 
Thus, removal of this basaltic partial melt results in a sig- 
nificant depletion of plagioclase in the residual source rocks. 
In fact, most lodranites contain significantly less plagioclase 
than likely chondritic parental rocks (9.6- 10.3 wt% in ordi- 
nary chondrites; McSween et al., 1991). Yamato 74357, 
FRO 90011, Yamato 791491, Yamato 791493, Lodran, 
MAC 88177, and LEW 88280 contain 0- 1.6 ~01% plagio- 
clase. We interpret these lower than chondritic plagioclase 
contents as an indication of the removal of basaltic partial 
melts from lodranites. 
Some lodranites, however, contain abundant plagioclase. 
Plagioclase abundances in Gibson (6.0 vol%), Yamato 8002 
(10.3), and EET 84302 (11.4) are all equal to or only 
slightly less than plagioclase abundances in ordinary chon- 
drites. Evidence for silicate partial melting is present in these 
rocks as well. In Yamato 8002, single plagioclase crystals 
surround multiple mafic silicate grains. Interstitial textures 
for plagioclase grains are also observed in Gibson and EET 
84302. In addition, clinopyroxene is always associated with 
plagioclase in EET 84302 (Field et al.. 1993; Mittlefehldt 
et al., 1996). This association is typical of that expected in 
a basaltic partial melt and suggests to us that basaltic partial 
melting did occur, but that those melts did not migrate from 
the rock, but were trapped in situ, or could have migrated 
into the area sampled by these meteorites (see McCoy et al., 
1997). 
4.2.3. Formation of minor phases 
Redistribution of minor phases may also result from sili- 
cate partial melting. Takeda et al. ( 1994) noted the existence 
634 T. .I. McCoy et al. 
of a chromite-rich region in EET 84302. We did not observe 
this chromite-rich lithology in our section (,12). Records 
from other workers (e.g., M. M. Lindstrom, pers. commun.) 
indicate that only a small portion, perhaps an area 2 mm on a 
side, of the 59.6 g stone contains this chromite-rich lithology. 
Most of EET 84302 resembles normal, metal-bearing lodran- 
ites. Takeda et al. (1994) argued that this lithology formed 
by deposition of chromite in a small area as a result of 
movement of large amounts of Cr-rich silicate partial melts. 
They cite the absence of olivine and enrichment of pyroxene 
in the vicinity of the chromite-rich lithology as evidence of 
reaction between Cr203 in the melt and olivine to form chro- 
mite and pyroxene. If correct, this is another consequence 
of silicate partial melting in the lodranites. 
4.2.4. Unexplained features 
Some features of lodranites are likely due, at least in part, 
to the igneous processing they experienced. Yet the exact 
mechanisms which controlled these features are uncertain. 
First among these is the composition of plagioclase. Lodran- 
ites exhibit a broad range of plagioclase compositions 
(An,,,_,,,). This range is much wider than that exhibited 
by the acapulcoites (An_,,_,,), which are the likely direct 
precursors to the lodranites. It is possible to argue that the 
variability in the lodranites was inherited from the precursor 
chondrite, and we have yet to sample this full range of vari- 
ability in the acapulcoites. This seems unlikely to us. More 
likely, the partial melting experienced by the lodranites al- 
tered the plagioclase compositions from those of the precur- 
sors, probably by fractionation prior to melt removal or crys- 
tallization of other Ca-bearing phases (e.g., pyroxenes). 
However, plagioclase compositions show no correlation with 
other parameters (plagioclase abundances, mafic silicate 
composition), so these processes did not operate in any sys- 
tematic manner. 
A second unexplained feature relates to the relative abun- 
dances of Fe,Ni metal, FeS, and plagioclase. While it is 
clear that the abundances of these minerals were altered in 
response to removal of the Fe,Ni-FeS and silicate melts, 
how this removal affected individual lodranites is less clear. 
Lodranites exhibit a wide range of Fe,Ni metal abundances 
(0.5 - 19.6 vol%, excluding the highly-weathered Gibson). 
Unrepresentative sampling is not likely to be responsible for 
this wide range, given that the 1 kg Lodran has never been 
shown to contain metal-poor regions despite extensive study 
of this meteorite. This range is considerably greater than that 
for acapulcoites (14-23 wt%), the likely direct precursors 
of the lodranites. It also appears unlikely that the disparity 
in these two ranges reflects poor-sampling of the acapul- 
coites, since no metal-poor regions within acapulcoites have 
been noted in this study or in previous work. Rather, some 
lodranites (e.g., Lodran) must have gained metal during ig- 
neous processing, whereas other lodranites (e.g., MAC 
88177, 0.5 ~01% Fe,Ni metal) lost much of their original 
metal. We would note that it appears that oxidation of Fe0 
from silicates was not responsible for the increase in metal 
abundances, since the Fe,Ni metal-troilite irregular patches 
and veins are volumetrically insignificant. 
The relative abundances of troilite and plagioclase in indi- 
vidual lodranites are similarly puzzling. Lodranites tend to 
be depleted in these phases relative to likely precursor chon- 
dritic materials. However, some lodranites (EET 84302, Ya- 
mato 8002, Gibson, Yamato 75274) retain moderate to high 
abundances of plagioclase but are highly depleted in FeS. 
In contrast, other lodranites (MAC 88177. Yamato 74357, 
LEW 88280, Lodran, FRO 90011, Yamato 791491, Yamato 
79 1493) are highly depleted in plagioclase but contain mod- 
erate to abundant FeS. Thus, some lodranites appear to have 
lost only the Fe,Ni-FeS partial melt, whereas others appear 
to have lost primarily the silicate partial melt. We suggest 
that once melt migration begins, the coexistence of a dense, 
immiscible Fe,Ni-FeS melt in a silicate melt will lead to 
highly heterogeneous and complex assemblages, since it is 
not to be expected that these melts would be removed and 
deposited with equal ease. (A more complete discussion of 
melt migration in the acapulcoite-lodranite parent body is 
given by McCoy et al., 1997). Some authors (Nagahara, 
1992: Mittlefehldt et al., 1996) believe that one or more 
of these melts may have been introduced into the depleted 
lodranites. Nagahara (1992) used this mechanism to explain 
the textures of plagioclase in Yamato 8002, whereas Mittle- 
fehldt et al. (1996) argued for injection of an Fe,Ni-FeS 
melt to explain enriched Se/Co ratios in MAC 88177 and 
LEW 88280 relative to most acapulcoites. 
4.2.5. Peak temperatures 
The peak temperatures experienced by lodranites can be 
estimated an estimation of the degree of partial melting. 
Taylor et al. ( 1993) used the melting of peridotite (McKen- 
zie and Bickel, 1988) to model the degree of partial melting 
as a function of temperature for chondritic systems. Basaltic 
partial melts begin forming at - 1150?C in a chondritic sys- 
tem. Evidence for silicate partial melting in all lodranites 
suggests this as the minimum temperature to which lodran- 
ites were heated. The maximum temperature is estimated 
from the degree of partial melting necessary for complete 
melting and removal of plagioclase. Melting and removal of 
-10 ~01% plagioclase which is a typical content for chon- 
drites, would require - 15-20% silicate partial melting (de- 
pending on the composition of the plagioclase) and peak 
temperatures of the order of 1200?C. The estimate of 1200?C 
is relatively insensitive to the amount of melting in the range 
of 15 -20% (McKenzie and Bickel, 1988). Lodranites which 
experienced lower degrees of partial melting (e.g., the pla- 
gioclase-rich Gibson) might have experienced a temperature 
closer to 1150?C. Thus, a relatively small range of peak 
temperatures can result in a relatively wide range in the 
degree of silicate partial melting. 
It is interesting to note that these temperatures are consid- 
erably higher than the two-pyroxene geothermometer tem- 
peratures. However, the latter are closure temperatures and 
a period of relatively slow cooling might have occurred be- 
tween the peak temperature and pyroxene closure. Further, 
the two-pyroxene geothermometer temperatures show no 
correlation with degree of partial melting. We suggest that 
an additional factor may have influenced the degree of partial 
melting. It is known that Fe0 contents of mafic silicates 
strongly influence melting temperatures. Lodranites with low 
Petrogenesis of lodranites 635 
Fe0 contents exhibit relatively low degrees of silicate partial have the same mineralogy (opx + 01 + Fe,Ni metal -t FeS 
melting (as measured by plagioclase abundances, Fig. 2), ? plag) but differ in their mineral abundances. Lodranites 
whereas high-Fe0 lodranites exhibit relatively extensive sili- have heterogeneous and correlated mafic silicate and oxygen 
cate partial melting. We suggest that the wide range of Fe0 isotopic compositions which overlap those of acapulcoites, 
contents of these meteorites may also have strongly influ- although acapulcoites do not show a strong correlation be- 
enced the degree of silicate partial melting. There are, how- tween oxygen isotopic and mafic silicate compositions. Both 
ever, exceptions to this rule (e.g., EET 84302), and we groups are enriched in trapped noble gases. Complex thermal 
suggest that these exceptions were caused by real differences histories suggest breakup and reassembly of the acapulcoite- 
in peak temperatures. We also note that we cannot unequivo- lodranite parent body. In addition, four acapulcoites and five 
cally eliminate the possibility that increased plagioclase lodranites share a common cosmic ray exposure age (5.5- 
abundances in low-Fe0 lodranites reflect in situ crystalliza- 7 Ma). Further, EET 84302 is intermediate in many of its 
tion of partial melts, although we could offer no explanation properties (e.g., mafic silicate grain size, plagioclase abun- 
of why this crystallization would occur only in low-Fe0 dances) between the two groups (McCoy et al., 1993; Field 
lodranites. et al., 1993; Mittlefehldt et al., 1996). 
4.3. Chronology and Thermal History 
We are interested in two aspects of the chronology of the 
lodranites: their age of formation and their cooling history. 
The 4.49 Ga 39Ar-?oAr age of Gibson is the only measured 
age of any lodranite. Considering the prolonged thermal his- 
tory of the lodranites and its effect on the 39Ar-40Ar chronom- 
eter, this age clearly does not represent the time of formation 
of the lodranite parent body. That time is probably best given 
by the 4.556 Ga Pb-Pb age for Acapulco (GGpel et al., 
1992). The 39Ar-?oAr age for Gibson, along with similar 
ages for acapulcoites (McCoy et al., 1996a), thus place 
a lower limit on the time of significant cooling after the 
metamorphic heating experienced by the parent body. We 
note that the 4.49 Ga 39Ar-4?Ar age for Gibson lies within 
the measured range of 39Ar-40Ar ages of unshocked ordinary 
chondrites of 4.44-4.52 Ga ages that also presumably date 
the time of Ar closure at the end of parent body metamor- 
phism (Turner et al., 1978). 
The cooling history of lodranites appears to be complex. 
Several lines of evidence suggest that lodranites experienced 
a period of slow cooling at high temperatures. Two-pyroxene 
equilibration temperatures of lodranites are 100-200?C be- 
low peak temperatures inferred from the degree of partial 
melting. As two-pyroxene equilibration temperatures reflect 
the cessation of diffusion, relatively slow ( 1- 1OO?UMyr) 
cooling must have occurred between the peak temperatures 
and the establishment of the two-pyroxene equilibration tem- 
perature. A second argument for slow cooling comes from 
the -65 Ma difference between the likely formation age of 
the lodranite parent body and the -4.49 Ga ?9Ar-40Ar age. 
After this slow cooling, a period of very rapid cooling oc- 
curred between perhaps 1000?C and 4OO?C, as indicated by 
cooling rates of - IO?-10h?CNyr by a variety of methods 
(Table 2). This suggests that the lodranite parent body was 
at least partially disrupted by a large impact while at a tem- 
perature of ~900?C. The cooling rates in the middle temper- 
ature range suggest fragmentation iI,to pieces which range 
in radius from approximately 200 m to 6 km (using Eqn. 
12 of Haack et al., 1990). McCoy et al. (1996a) argued for 
breakup and reassembly of this parent body based on a more 
complete dataset for acapulcoites. 
There are, however, important differences between the 
two rock types, and these can be attributed to different de- 
grees of partial melting. We suggest that acapulcoites were 
heated to -9.50- 1050?C and experienced Fe.Ni-FeS cotectic 
melting (McCoy et al., 1996a). We note, however, that oth- 
ers argue for higher peak temperatures for at least some 
acapulcoites, without melt migration (Zipfel et al., 1995). 
A full discussion of these opposing viewpoints is given in 
McCoy et al. ( 1996a). Melt migration distances in acapul- 
coites were small (pm to cm), and chemical fractionation 
did not occur (McCoy et al., 1996a). In contrast, lodranites 
appear to have been heated to - 1 lOO- 125O?C, resulting in 
the formation of both Fe,Ni-FeS cotectic and basaltic partial 
melts. Melts migrated from the source rocks, leaving resi- 
dues (e.g., lodranites) which tend to be depleted in FeS and 
plagioclase. Another possible consequence of this difference 
in peak temperature may be the slightly younger 39Ar-40Ar 
age for Gibson compared to several acapulcoites. Thus, we 
conclude that acapulcoites and lodranites appear to represent 
the products of differing degrees of heating and partial melt- 
ing of common, isotopically and chemically heterogeneous 
precursors. 
The differences in grain sizes between the acapulcoites 
and lodranites could also be related to the presence or ab- 
sence of a silicate partial melt. Grain growth in these rocks 
probably occurs by Ostwald ripening, with the diffusion rate 
through the intergrain medium as the controlling factor in 
determining the grain sizes. Diffusion rates in silicate melts 
(summarized by Taylor, 1992) are up to six orders of magni- 
tude faster than those in crystalline rocks. Thus, the lodran- 
ites, which once may have contained silicate partial melts, 
became much coarser-grained than the acapulcoites, which 
lacked silicate partial melts. Because of the speed of grain 
growth in the presence of silicate melt, Taylor ( 1993) argued 
that the silicate melts must have migrated from their source 
regions in times of the order of 10? years. Otherwise, lodran- 
ites would be even coarser grained. 
4.4. Acapulcoites and Lodranites are from the Same 
Parent Body 
All available data suggest that acapulcoites and lodranites 
almost certainly originated on a common parent body. They 
Acknowledgmenrs-We thank David New, Roy S. Clarke, Jr., the 
MWG, NIPR, and EUROMET for providing samples for this study. 
We thank E. R. D. Scott, G. J. Taylor, H. Takeda, M. M. Lindstrom, 
J. I. Goldstein, and H. Haack for valuable discussions. Helpful com- 
ments by reviewers H. Palme, C. R. Neal, and M. I. Petaev signifi- 
cantly improved this manuscript. This work was supported in part 
by NASA grant NAGW-3281 (K. Keii, PI), NSF grant EAR- 
9218857 (R. Clayton, PI), and the Swiss National Science Founda- 
tion (R. Wieler 1. This is Hawai?i Institute of Geophysics and Plane- 
636 T. J. McCoy et al. 
tology Publication No. 923 and School of Ocean and Earth Science 
and Technology Publication No. 4168. 
Editorial handling: C. R. Neal 
REFERENCES 
Bild R. W. and Wasson J. T. (1976) The Lodran meteorite and its 
relationship to the ureilites. Mineral. Mug. 40, 721-735. 
Bild R. W. and Tallant D. R. (1984) Raman spectroscopy of the 
Lodran meteorite. Meteoritics 19, 190 (abstr.) 
Choi B.-G., Ouyang X., and Wasson J. T. (1995) Classification and 
origin of IAB and IIICD iron meteorites. Geochim. Cosmochim. 
Acta 59, 593-612. 
Clayton R. N. (1993) Oxygen isotopes in meteorites. Ann. Rev. 
Earth Planet. Sci. 21, 115 - 149. 
Clayton R. N. and Mayeda T. K. ( 1963) The use of bromine pen- 
tafluoride in the extraction of oxygen from oxides and silicates 
for isotopic analyses. Geochim. Cosmochim. Acta 27, 43-52. 
Clayton R. N. and Mayeda T. K. ( 1983) Oxygen isotopes in eucrites. 
shergottites, nakhlites, and chassignites. Earth Planet. Sci. Lett. 
62, l-6. 
Clayton R. N. and Mayeda T. K. ( 1988) Formation of ureilites by 
nebular processes. Geochim. Cosmochim. Actu 52, 1313- 1318. 
Clayton R. N. and Mayeda T. K. ( 1996) Oxygen isotope studies of 
achondrites. Geochim. Cosmochim. Actu 60, 1999-2017. 
Clayton R. N., Mayeda T. K., Olsen E. J., and Prinz M. (1983) 
Oxygen isotope relationships in iron meteorites. Earth Plnnet. Sci. 
Lett. 65, 229-232. 
Clayton R. N., Mayeda T. K., and Yanai K. ( 1984) Oxygen isotopic 
compositions of some Yamato meteorites. Mem. Natl. Inst. Polar 
Res. 35, 267-271. 
Clayton R. N., Mayeda T. K., Goswami J. N., and Olsen E. J. ( 1991) 
Oxygen isotope studies of ordinary chondrites. Geochim. Cos- 
mochim. Acta 55, 2317-2337. 
Clayton R. N., Mayeda T. K., and Nagahara H. (1992) Oxygen iso- 
tope relationships among primitive achondrites. Lunar Planet. Sci. 
XXIII, 231-232 (abstr.). 
Dreibus G., Palme H., Spettel B., Zipfel J., and Wlnke H. (1995) 
Sulfur and selenium in chondritic meteorites. Meteoritics 30,439- 
445. 
Eugster 0. (1988) Cosmic-ray production rates for 3He, 21Ne, 38Ar, 
83Kr, and lz6Xe in chondrites based on ?Kr-Kr exposure ages. 
Geochim. Cosmochim. Acta 52, 1649-1662. 
Eugster 0. and Weigel A. ( 1993) Xe-Q in lodranites and a hint for 
Xe-L. FRO 90011 another lodranite? Lunar Planet. Sci. XXIV, 
453 -454 (abstr.). 
Field S. W., Lindstrom M. M., and Mittlefehldt D. W. ( 1993) Petrol- 
ogy and geochemistry of Acapulco- and Lodran-like achondrites. 
Meteoritics 28, 347 (abstr.). 
Folco L. (1992) Meteorites from the 1990/91 expedition to the 
Frontier Mountains, Antarctica. Meteoritics 27,221-222 (abstr.) 
Franchi I. A., Akagi T., and Pillinger C. T. (1992) Laser fluorination 
of meteorites-Small sample analysis of 6 ?0 and 15~~0. Meteorit- 
its 27, 222 (abstr.). 
Frost B. R. and Lindsley D. H. ( 1992) Equilibria among Fe-Ti ox- 
The residue of limited partial melting of matter like ?a hybrid 
between H and E chondrites. Lunar Planet. Sci. IX, 356-358 
ides, pyroxenes, olivine, and quartz: Part II. Application. Amer. 
(abstr.). 
Min.&L 77, 1004-1020. 
Fukuoka T.. Ma M.-S.. Wakita H.. and Schmitt R. A. ( 1978) Lodran: 
Gopel Ch., Manhes G., and Allegre C. J. (1992) U-Pb study of the 
Acapulco meteorite. Meteoritics 27, 226 (abstr.). 
Graf Th. et al. (1990a) Cosmogenic nuclides and nuclear tracks in 
the chondrite Knyahinya. Geochim. Cosmochim. Acta 54, 25 ll- 
2520. 
Graf Th., Baur H., and Signer P. ( 1990b) A model for the production 
of cosmogenic nuclides in chondrites. Geochim. Cosmochim. Acta 
54,2521-2534. 
Haack H., Rasmussen K. L., and Warren P. H. ( 1990) Effects of 
regolith/megaregolith insulation on the cooling histories of differ- 
entiated asteroids. J. Geophys. Res. 95, 5 111-5 124. 
Hewins R. H. and Goldstein J. I. (1977) Cooling rates for lunar 
samples determined with a diffusion model for phosphide exsolu- 
tion. Proc. Lunar Sci. ConfI 8, 1625-1638. 
Hiroi T. and Takeda H. ( 199 1) Reflectance spectroscopy and miner- 
alogy of primitive achondrites-lodranites. Proc. NIPR Symp. Ant- 
arctic Meteorites 4, 163-177. 
Kallemeyn G. W., Rubin A. E., and Wasson J. T. ( 1994) The com- 
positional classification of chondrites: VI. The CR carbonaceous 
chondrite group. Geochim. Cosmochim. Acta 58, 2873-2888. 
Keil K. ( 1962) On the phase composition of meteorites. J. Geophys. 
Res. 67,4055-4061. 
Kretz R. ( 1982) Transfer and exchange equilibria in a portion of 
the pyroxene quadrilateral as deduced from natural and experi- 
mental data. Geochim. Cosmochim. Acta 46,4 1 l-42 1. 
Kubaschewski 0. ( 1982) Iron-Binq Phase Diagram. Springer Ver- 
lag. 
Kullerud G. (1963) The Fe-Ni-S system. Ann. Rep. Geophys. Res. 
67,4055-4061. 
Lindsley D. H. and Frost B. R. ( 1992) Equilibria among Fe-Ti ox- 
ides, pyroxenes, olivine, and quartz: Part 1. Theory. Amer. Min- 
eral. 77, 987-1003. 
Mason B. ( 1986) Thin section (,3) description of EET 84302. Ant- 
arctic Meteorite Newslett. 9, 18. 
Mason B. ( 1995) Thin section description of QUE 93148. Antarctic 
Meteorite Newslett. 18, 15. 
McCoy T. J., Keil K., Mayeda T. K., and Clayton R. N. (1992a) 
Monument Draw and the formation of the acapulcoites. Lunar 
Pkmet. Sci. XXIII, 87 l-872 (abstr.). 
McCoy T. J., Keil K., Mayeda T. K., and Clayton R. N. (1992b) 
Petrogenesis of the lodranite-acapulcoite parent body. Meteor&s 
27, 258-259 (abstr.). 
McCoy T. J., Keil K., Mayeda T. K., and Clayton R. N. ( 1993) 
Classificational parameters for acapulcoites and lodranites: The 
cases of FRO 90011, EET 84302 and ALH A81187/84190. Lunar 
Planet. Sci. XXIV. 945-946 (abstr.). 
McCoy T. J. et al. (1996a) A petrologic, chemical and isotopic 
study of Monument Draw and comparison with other acapulcoites: 
Evidence for formation by incipient partial melting. Geochim. 
Cosmochim. Acta 60, 268 l-2708. 
McCoy T. J., Keil K., Muenow D. W., and Wilson L. ( 1997) Partial 
melting and melt migration in the acapulcoite-lodranite parent 
body. Geochim. Cosmochim. Acta 61, 639-650 (this issue). 
McKenzie D. P. and Bickel M. J. (1988) The volume and composi- 
tion of melt generated by extension of the lithosphere. J. Petrol. 
29, 625-679. 
McSween H. Y., Jr., Bennett M. E., III, and Jarosewich E. ( 1991) 
The mineralogy of ordinary chondrites and implications for aster- 
oid spectrophotometry. Icarus 90, 107- 116. 
Mittlefehldt D. W., Lindstrom M. M., Bogard D. D., Garrison D. H.. 
and Field S. W. ( 1996) Acapulco- and Lodran-like achondrites: 
Petrology, geochemistry, chronology and origin. Geochim. Cos- 
mochim. Acta 60, 867-882. 
Morse S. A. ( 1980) Bus&s and Phase Diagrams. Springer-Verlag. 
Nagahara H. ( 1992) Yamato-8002: Partial melting residue on the 
?unique? chondrite parent body. Proc. NIPR Symp. Antarctic 
Meteorites 5, 191-223. 
Miyamoto M. and Takeda H. ( 1991) Cooling history of Antarctic 
primitive achondrite Yamato-74357. NIPR Symp. Antarctic Mete- 
orites 16, 165-167 (abstr.). 
Nagahara H. and Ozawa K. (1986) Petrology of Yamato-791493, 
?lodranite?: Melting, crystallization, cooling history, and rela- 
tionship to other meteorites. Proc. Symp. Antarctic Meteorites 10; 
Mem. NIPR Spec. Issue 41, 181-205. 
Palme H. et al. ( 1981) The Acapulco meteorite: Chemistry, mineral- 
ogy and irradiation effects. Geochim. Cosmochim. Acta 45, 727- 
752. 
Papike J. J., Spilde M. N., Fowler G. W., Layne G. D., and Shearer 
C. K. ( 1995) The Lodran primitive achondrite: Petrogenetic in- 
sights from electron and ion microprobe analysis of olivine and 
orthopyroxene. Geochim. Cosmochim. Acta 59, 3061-3070. 
Prinz M., Klimentidis R., Harlow G. E., and Hewins R. H. (1978) 
Petrologic studies bearing on me origin of the Lodran meteorite. 
Lunar Planet. Sci. IX, 919-921 (abstr.). 
Petrogenesis of lodranites 637 
Prinz M., Chatterjee N., Weisberg M. K., Clayton R. N., and Mayeda 
T. K. (1991) MAC 88177: A new type of achondrite? Lunar 
Planet. Sci. XXII, 1099- 1100 (abstr.). 
Prior G. T. ( 1916) On the genetic relationship and classification of 
meteorites. Mineral Mug. 18, 26-44. 
Score R. and Lindstrom M. M. (1990) Guide to the U.S. Collection 
of Antarctic Meteorites 1976-1988. Antarctic Meteorite Newslett. 
13, 57. 
Short J. M. and Goldstein J. I. ( 1967) Rapid methods of determining 
cooling rates of iron and stony iron meteorites. Science 156, 59- 
61. 
Stoffler D., Keil K., and Scott E. R. D. ( 1991) Shock metamorphism 
of ordinary chondrites. Geochim. Cosmochim. Acta 55, 3845- 
3867. 
Takaoka N., Nagao K.. and Miura Y. (1993) Noble gases in the 
unique meteorites Yamato-74063 and -74357. Proc. NIPR Symp. 
Antarctic Meteorites 6, 120- 134. 
Takeda H., Mori H., Hiroi T., and Saito J. (1994) Mineralogy of 
new Antarctic achondrites with affinity to Lodran and a model of 
their evolution in an asteroid. Meteoritics 29, 830-842. 
Taylor G. J. ( 1992) Core formation in asteroids. J. Geophys. Res. 
97, 14717-14726. 
Taylor G. J. ( 1993) Rapid magma migration in asteroids. Meteoritics 
28, 447-448 (abstr.) 
Taylor G. J., Keil K.. McCoy T., Haack H.. and Scott E. R. D. 
( 1993) Asteroid differentiation: Pyroclastic volcanism to magma 
oceans. Meteoritics 28, 34-52. 
Torigoye N., Yamamoto K., Misawa K., and Nakamura N. (1993) 
Compositions of REE, K, Rb, Sr, Ba, Mg, Ca, Fe and Sr isotopes 
in Antarctic unique meteorites. Proc. NlPR Symp. Antarctic Mete- 
orites 6, lOO- 119. 
Turner G., Enright M. C., and Cadogan P. H. (1978) The early 
history of chondrite parent bodies inferred from 40Ar-?9Ar ages. 
Proc. Lunar Planet. Sci. Con5 9, 989- 1025. 
Tuttle 0. F. and Bowen N. L. (1958) Origin qf granite in the light 
of experimental studies in the system NaAlSi,O.,-Si02-HzO. Mem- 
oir 74, Geolog. Sot. America. 
Wieler R., Anders E., Baur H., Lewis R. S., and Signer P. ( 1992) 
Characterization of Q-gases and other noble gas components in 
the Murchison meteorite. Geochim. Cosmochim. Acta 56, 2907- 
2921. 
Willis J. and Goldstein J. I. ( 1981 1 A revision of the metallographic 
cooling rate curves for chondrites. Proc. Lunar Planet. Sci. Conj 
12, 1135-1143. 
Wood J. A. (1967) Chondrites: Their metallic minerals, thermal 
histories and parent planets. Icarus 6, l-49. 
Yanai K., Kojima H., Prinz M., Nehru C. E., Weisberg M. K., and 
Delaney J. S. ( 1984) Petrologic studies of three primitive achon- 
drites from the Yamato meteorites collection, Antarctica. NIPR 
Symp. Antarctic Meteorites 9, 24-28 (abstr.). 
Zipfel J. and Palme H. (1993) Chemical composition of new acapul- 
coites and lodranites. Lunar Planet. Sci. XXIV, 1579- 1580 
(abstr.). 
Zipfel J., Palme H., Kennedy A. K., and Hutcheon I. D. (1995) 
Chemical composition and origin of the Acapulco meteorite. Geo- 
chim. Cosmochim. Acta 59, 3607-3627.