Document 271194

EARTH AND PLANETARY SCIENCE LETTERS 13 (1971) 61 - 70. NORTH-HOLLAND PUBLISHING COMPANY
PETROLOGY OF APOLLO 11 SAMPLE 10071.
A DIFFERENTIATED MINI-IGNEOUS COMPLEX
M.J. D R A K E and D.F. W E I L L
Center for Volcanology, University of Oregon, Eugene, Oregon 97403, USA
Received 6 September 1971
Revised version received 28 October 1971
Sample 10071, 33 is a thin section of Apollo 11 ferrobasalt showing an unusual dual texture. The major portion
of the sample is very similar to other fine grained Apollo 11 basalts, but the thin section also includes material with
a distinct variolitic texture. The two areas are separated by a sharp boundary and the mineralogy and composition of
the two textural types are quite distinct. The mineralogy and chemistry of the variolitic portion show it to be the
product of rapid cooling of a liquid, intermediate between the typical Apollo 11 ferrobasalt and the associated Si
and K-rich mesostasis. This liquid is the result of fractional crystallization of a magma of composition closely corresponding to the major portion of the 10071 system, followed by crystal-liquid separation. The sample provides
strong and direct evidence for igneous differentiation on the lunar surface.
1. I n t r o d u c t i o n
The discovery o f rocks w i t h unmistakably igneous
textures at f o u r locations on the lunar surface (Apollos
11, 12, 14 and L u n a 16) has p r o m p t e d a t t e m p t s to
explain their chemical variation in terms o f crystalliquid differentiation processes [2, 3 and m a n y others].
On line of investigation a t t e m p t s to relate the sampled
igneous rocks to parent material(s), as yet unsampled
and presuma.bly within the lunar interior, via differentiation processes such as partial melting. But the
inferred presence o f m a g m a at the surface makes it
likely that c r y s t a l - l i q u i d separation processes (gravitative settling, flow c o n c e n t r a t i o n mechanisms, filter
pressing, etc.) at or near the surface have also played
a role in the final stage o f d e v e l o p m e n t o f the igneous
rock suites. This is considered all the more likely w h e n
the low viscosity o f the liquids and the substantial
density contrast b e t w e e n lunar silicate liquids and
crystalline phases are taken into a c c o u n t [ 1 ]. Nevertheless, there is very little direct textural evidence, e.g.,
layering or banding, flow patterns, size sorting, etc., in
the returned igneous rocks to confirm such expectations. In addition, the relatively small range o f chemical variation o f the major elements in the igneous rocks
Table 1
Chemical composition of lunar igneous rocks (major elements)
from three mare locations. The standard deviation of individual rock analyses from the average concentration is given for
the Apollo 11 and Apollo 12 locations.
Apollo 11
(16 rocks)
Apollo 12
(17 rocks)
Luna 16
Avg.
S.D.
Avg.
S.D.
SiO 2
TiO 2
A1203
FeO
MgO
CaO
Na20
K20
MnO
Cr203
P2Os
40.44
11.15
9.57
19.00
7.36
11.07
0.46
0.18
0.26
0.33
0.13
0.96
1.03
1.20
0.78
0.71
0.66
0.06
0.12
0.03
0.08
0.05
45.02
3.35
9.19
20.55
10.94
9.80
0.29
0.06
0.27
0.49
0.10
1.34
0.87
1.38
1.04
3.37
1.31
0.11
0.01
0.01
0.20
0.03
Total
99.95
Refs.
[26-32]
100.06
[33-39]
43.8
4.90
13.65
19.35
7.05
10.40
0.33
0.15
0.20
0.28
100.11
[40]
from any one locality (table 1) seems to speak against
a major role for c r y s t a l - l i q u i d differentiation mechanisms on or near the lunar surface.
62
M.J.Drake, D.F. Weill, Lunar sample 10071
The conflict between theoretical expectations and
observations may be resolved if it is assumed that the
returned igneous rocks represent a sampling of only
the upper portions of lava ponds or flows. Thus, rapid
cooling and crystallization from supersaturated melts
would preclude any appreciable crystal liquid separation and the development of textures indicative of
flow or crystal settling. Alternatively, the igneous
rocks may be the result of small impact events producing much more limited volumes of melt which are
rapidly cooled. The difference between these two
alternative views is gradational and involves for the
most part a difference in the volume of liquid, and
hence in the cooling rate during crystallization. The
former hypothesis is usually favored. The most direct
answer to the problem will come with positive identification of a continuous igneous rock outcrop (such
as a lava flow unit or sill) on the lunar surface. Less
directly, the formation of the igneous rocks from
relatively large bodies of liquid is indicated when their
chemical variation is compatible with simple fractional crystallization of a common liquid. For example, most of the Apollo 12 igneous rocks can be related to each other in terms of olivine and/or pyroxene
separation [2, 3]. Unfortunately, this approach is
somewhat weakened by the relatively small range of
chemical variation of the major elements observed in
the igneous rocks. None of the late stage differentiates
often observed in terrestrial igneous complexes have
not yet been found among the suites of lunar igneous
rocks. In contrast, there is much evidence of intrasample fractional crystallization leading to a silicicpotassic residual liquid which forms a mesostasis in
nearly all of the rocks. The comparatively small volume of this residuum and its extremely high viscosity
[4] make it unlikely that it would be separated from
its parental material except under unusual conditions
favoring filter pressing. The granitic component of
sample 12013 [5,6] may have originated in such
fashion, but the evidence is only circumstantial.
Despite the theoretical likelihood of differentiation
by crystal liquid fractionation, no compelling evidence of such a process has yet been observed. Igneous
rock 10071 has been described in general terms as a
lunar basalt containing vesicles anf vugs [7]. It was also noted [7,8] that the rock contained two contrasting (coarse vs. fine grained) textural domains. The fine
grained portion of the rock was tentatively attributed
to recrystallization after flash heating [8], an essentially isochemical change, but no further petrographic
study of 10071 has been reported. A review of our
petrographic analysis of sample 10071 indicates that
the two textural domains represent subsystems that
are probably related by fractional crystallization. If
so, sample 10071 provides a direct indication of the
efficiency of fractional crystallization as a differentiation mechanism of the lunar surface.
2. General
description
Thin section 1007 i, 33 is separated by a sharp
boundary into areas of relatively coarse and fine
texture (fig. 1). Electron probe analyses reveal that
the two areas are also quite distinct chemically and
mineralogically; the coarser section being very similar
to many of the Tranquillity site basalts already described, whereas the finer grained portion is more
acidic and is logically interpreted as having crystallized from a liquid fraction which was residual after
partial crystallization of a magma of composition
closely corresponding to that of the coarser grained
material.
2.1. Coarse grained area
The coarse grained area has an average grain size of
1 0 0 - 2 0 0 #m. Ilmenite is the dominant opaque phase,
and it sometimes encloses armalcolite [9]. It occurs in
a variety of habits ranging from tabular to,very thin
and platy (often acicular in section). Troilite, usually
enclosing blebs of metallic iron, occurs in accessory
amounts. The most common silicate is titanaugite,
commonly zoned to ferroaugite. Olivine cores are
occasionally present within the pyroxenes. Plagioclase
is abundant and frequently tabular, probably indicating relatively rapid growth. Interstitial residual material, with its characteristic brown, mottled appearance and high K and Si content is common throughout this area. Silica is also present as a minor phase.
A modal analysis of the area is given in table 2.
2.2. Fine grained area
The texture of this area is dominated by sheaf-like
intergrowths of clinopyroxene and plagioclase and
may be classified as variolitic (fig. 1). Individual crystals in these intergrowths are acicular with typical
M.J.Drake, D.F. Weill, Lunar sample 10071
I m m
63
.
Fig. 1. Photomicrograph of 10071,33 showing the two distinct textural domains.
Table 2
Modal analyses (volume percent) of coarse and fine grained
areas of 10071, 33 obtained on electron microprobe.
Plagioclase
Augite
Ferroaugite and pyroxferroite
Pigeonite
Olivine
llmenite and other opaque phases
Silica
K and Si-rich residuum
Total
Number of points
Point interval (~m)
Coarse
Fine
23.5
38.5
13.0
2.0
1.0
15.5
0.5
6.5
100
400
100
35.5
14.5
29.0
1.0
0.0
4.0
7.5
8.5
100
400
20
dimensions of 1 - 5 / a m by 1 0 0 - 1 5 0 / a m . Some tubular
plagioclase and the more equant clinopyroxene of the
type found in the coarse grained area are also found
here. Much of the augite in this area is strongly zoned
to iron-rich compositions all the way into the pyroxferroite composition field, whereas such extreme zoning is u n c o m m o n in the coarse grained area. The small
grain size makes it difficult to detect optically any
sharp boundary between the clinopyroxene and
pyroxferroite, llmenite is the major opaque phase, but
it is much less abundant here than in the coarse grained
area and occurs only as thin plates. No olivine was
found, but the silica and potassium-rich residua are
appreciably more abundant in this area. The mode is
given in table 2.
M.J.Drake, D.F. Weill, Lunar sample 10071
64
3. Analytical results
4 (o)
Electron probe analyses for 8 or 9 elements have
been made of the major phases and also of the bulk
chemistry of the fine and coarse grained areas using a
defocussed beam technique. The details of our standard operating conditions and correction procedures
are described in [1]. The REE contents of the phosphates (table 3) were obtained from direct comparison with a rare earth lithium metaborate glass standard after only a background correction. Most of the
Table 3
Abundances of REE in apatites of 10071,33 (wt%).
Counting precision * [25, p. 279] is the major source of uncertainty for these low concentration analyses.
0
8
,
>Z
w
o
w
8
~
La
Ce
Pr
Nd
Sm
Eu
Gd
Dy
Er
2
3
4
5
6
0.12
0.31
0.05
0.26
0.09
0.01
0.16
0.12
0.06
0.42
1.02
0.22
0.79
0.25
0.10
0.39
0.34
0.30
0.40
0.82
0.16
0.89
0.26
0.02
0.44
0.35
0.25
0.29
0.76
0.14
0.76
0.22
0.04
0.33
0.19
0.09
0.25
0.58
0.14
0.60
0.17
0.01
0.22
0.12
0.08
0.12
0.19
0.03
0.23
0.08
0.02
0.12
0.16
0.13
1 - 3 , coarse grained area. 4-6, fine grained area.
* o = 5% relative at 1.0'~ concentration, 30% relative at 0.1%
concentration, 200% relative at 0.01% concentration.
compositional data graphically reported are partial
analyses, showing solid solution variation of the two
or three major cations. For each phase type the
elemental ratios in the partial analysis are obtained by
applying minor empirical correction factors to the
background corrected counts. The correction factors
are obtained directly from representative complete
analyses of the phase in question.
3.1. Opaque phases
Analytical evidence [20] indicates that the amount
of Mg entering ilmenite is controlled largely by the
Mg/Fe ratio in the coexisting liquid. This has been
confirmed experimentally for ilmenites grown at
successive stages of fractional crystallization of a liquid corresponding to an Apollo 1 l basalt composition
(G. McKay, personal communication). The large range
of Mg concentration and its strong correlation with
,
.
,
.............
(e)
o
4 . ( d ) ~
0
'" .............
'
'
2O
16
12
8
4
°o
1
nSn
IITTlIT~liNIIi
,
(b)
4
o
~ ,
02'd4'&'o'B'
ID
WEIGHT % MAGNESIUM
Fig. 2. Frequency distribution of Mg in ilmenites in 10071, 33:
(a) coarse grained area, ilmenites completely surrounded by
ferromagnesian silicates, 6 grains. (b) coarse grained area,
ilmenites largely surrounded by ferromagnesian silicates, 27
grains. (c) coarse grained area, ilmenites largely surrounded
by plagioclase feldspar, 20 grains. (d) coarse grained area,
ilmenite laths less than fifteen microns across, 10 grains.
(e) fine grained area, 20 grains. (4) indicates the average Mg
concentration.
the crystal habit of ilmenite in sample 10022, 28 was
used to demonstrate the extended crystallization
range of ilmenite from Apollo 11 basalts [ 1]. Fig. 2
shows the frequency distribution of Mg analyses and
the strong correlation with the various modes of
occurrence of ilmenite in 10071. Throughout the
coarse grained area the ilmenite shows a continuous
decrease in Mg from those grains which are included
in pyroxene (and presumably crystallized prior to
their hosts) to the late ilmenite occurring as thin
plates. The ilmenites of the fine grained area, on the
other hand, are uniformly low in Mg, indicating that
they crystallized from a differentiated liquid with
lower Mg/Fe ratio.
Reid et al. [i 1] reported a similar behavior in the
Ni and Co content of iron metal from Apollo 12
basalts. The Ni content of iron in 10071 was investigated with this observation in mind, but no systematic
difference between the two textural areas was detected. This was not unexpected since most of the iron in
both areas of 10071 is associated with troilite and
M.J.Drake, D.F. Weill, Lunar sample 10071
probably crystallized from an iron sulfide liquid which
separated from the silicate liquid late in the overall
crystallization sequence of both systems [ 12].
3.2. Plagioclase feldspars
Plagioclase exhibits little zoning and has essentially
the same composition range in the two areas (An73_ 8 l)Some analyses of plagioclase from the pyroxeneplagioclase sheafs in the fine grained area range down
to An6s , but the very thin nature of these crystals
makes them difficult to analyze precisely. The potassium content is relatively low in both areas (Or~.o_l. s)In addition to the obvious distinction in crystal habit
(fig. 1), the main difference between plagioclase in the
two areas is its markedly greater abundance in the fine
grained area (table 2). This suggests a differentiation
sequence in which the liquid that solidified to give the
fine grained texture was separated from the main portion of the system prior to the onset of plagioclase
crystallization.
Feldspars from a number of Apollo 11 and 12
igneous rocks depart systematically from ideal stoichiometry [ 1 , 4 and 13]. In fig. 3 we plot the analytical results for plagioclase in sample 10071. Analyses
of Stillwater feldspars are also shown for comparison
because they have grown from slowly cooling liquids
and should conform to the usual stoichiometry. They
0.16
0ll4
012
OllO
~
008
006
I
~, 0 0 4
002
-0.0
IDEAL
STOICHIOMETRY
0
~...~
~g~,~.k'~,oo~2,2s
I
0.02 004 0.06 0 0 8 0.10 012 014
I- (AI- C o - F e - M g )
ASE
016 0.18 0.20 0.22
Fig. 3, Departure of plagioclase analyses from stoichiometry.
Chemical symbols represent number of atoms relative to eight
oxygens. Based on counting precision, an error of -+0.02 (1 o)
is associated with each point for both axes. Symbols:
×, 10071, 33; u, Stillwater Intrusion.
65
were analyzed at the same time as the plagioclase of
10071 with exactly the same set of reference standards and correction procedures. The plot of the Stillwater feldspars indicates the scatter about ideal stoichiometry that can be expected from the analytical
technique. The departure from stoichiometry for the
10071 plagioclase is seen to be comparable to that
found in other Apollo 11 samples. For example, the
average of 30 plagioclase analyses from sample 10022
corresponds to the chemical formula (Cao.744Feo.o32
Mg0.o22, Nao.E0sKo.o 14)1.02o(Si2.32A11.62)3.9408 .
On fig. 3 this analysis would plot at coordinates
(Si-Na-K)-2 = 0.098 and 1-(A1-Ca-Fe-Mg) = 0.178. No
systematic differences in plagioclase stoichiometry
were found between the two textures.
The observed composition trends reveal a substantial departure from the normal plagioclase chemistry.
The norm in this context is taken as a tetrahedral
framework, derived from the silica structures, where
from one fourth to one half of the tetrahedral cation
sites are occupied by A13+. Incorporation of larger
univalent or divalent cations (predominantly Na ÷ and
Ca 2+) in sites interstitial to the tetrahedral framework
is associated with the A13+~ Si4+ substitution to preserve electroneutrality. Within this context, stoichiometry requires that: Na + Ca = 1, A1 + Si = 4, S i - Na = 2,
and A I - C a = 1 (the numbers of atoms are normalized
to 8 oxygens, and each element symbol is generalized
to include all cations of equal valency which enter
equivalent structural sites, e.g., 'Na' = Na + K, 'Ca' =
Ca + Mg + Fe 2÷, etc.). It is common practice in evaluating chemical analyses of feldspars to calculate only
the first two sums above, but this is not a complete
test of stoichiometry since three out of the four
equations are independent. Several workers [ 1 4 - 1 9 ]
have discussed certain aspects of non-stoichiometry
observed in terrestrial and synthetic feldspars. An
excess of Si over A1 call be accommodated in feldspars
according to Si 4+ + [] = A13++ Na ÷ or 2Si 4+ + [] =
2A13+ + Ca 2+, i.e., by creation of vacancies in the large
cation sites. Such a mechanism results in a feldspar
chemical formula which can be expressed either in
terms of excess dissolved SiO2 or the Schwantke
molecule, Cal/2AISi308. Chemical analyses [ 1] indicate that this is not the cause of non-stoichiometry in
the lunar plagioclases. Vacancies in large cation sites
is only a special case among the possible defects leading to feldspars which depart from ideal stoichiometry.
M.J.Drake, D.F. Weill, Lunar sample 10071
66
Vacancies in tetrahedral cation or in oxygen sites and
the possibility of limited stuffing of cations in nonstructural positions may also have to be considered in
attempting to explain the systematic departure from
stoichiometry shown by lunar plagioclases. It is impossible to attempt a solution to this intriguing problem with no more than analytical chemical data. The
latter can serve to indicate the problem and define
the chemical constraints, but only a structure-sensitive
technique (such as X-ray diffraction structural refinement) is capable of discovering the cause of the nonstoichiometry. The genetic implications are that this
phenomenon is perhaps related to rapid growth of
plagioclase under special lunar surface conditions.
3.3. Olivine, pyroxene and pyroxferroite
Analyses of these phases reveal a distinct contrast
between the two areas. Olivine is found in the coarse
grained portion, occurring as cores surrounded by
augite which zones progressively outward into ferroaugite or subcalcic ferroaugite. Occasionally, thin
lamellae of pigeonite are observed in the augite, and
infrequently pigeonite occurs as a core in place of
olivine. Figs. 4a and 4b show the partial analyses
(Mg, Fe and Ca) of the ferromagnesian silicates
recorded during the modal analysis of the coarse and
fine grained areas respectively. The average Fe/Mg
DI
.
.
.
.
ratio of these phases is greater in the fine grained portion than in the coarse grained portion. Extreme zoning from ferroaugite to compositions approaching
pyroxferroite is common in single grains of the finegrained area. A series of analyses along such a grain is
plotted in fig. 4c.
3.4. Residual phases
Both areas contain phases which can be classified
as late-forming or residual phases on the basis of
texture and chemistry. The composition of these
phases is similar in the two areas, but they are more
abundant in the fine-grained area. Much of the residual material is glassy or microcrystalline with over
70 wt % SiO2, and is very similar to the Si and K-rich
intergranular material found in many other Apollo 11
igneous rocks. Patches of essentially pure SiO2 are
also common. The intergranular material of both textures also contains common apatite grains and less
common zircon. Three apatite grains from each of the
two areas were analyzed for La through Er and the results are presented in table 3. The range of REE concentration indicates an enrichment o f several orders of
magnitude relative to chondritic abundances and is
similar to the range found in apatite from rock 12013
[20]. There is no systematic difference in the REE
concentrations between apatites from the two areas.
+5+4, ' ;
¢+b**
/,o,
o.........
\
~"~ ++:l:++
+~*,t +¢+, ~
++ ~+ ++
/
PC)
DI
HD
+++
++
++
.+
.
.
.
\
\
+%
.
.
.
\ FE
HD
i
MG
IO
+:5 \
~
.
~
FE
Fig. 4. Ferromagnesian silicates in 10071, 33: (a) Chemical variation in coarse grained area, 105 points. (b) Chemical variation in
fine grained area, 84 points. (c) Analyses at 10 # intervals along a composite ferroaugite-pyroxferroite crystal in fine grained area.
67
M.J.Drake, D.F. Weill, Lunar sample 10071
Table 4
Chemical analyses of coarse grained and fine grained areas of 10071.
Whole rock
1
SiO2
TiO 2
A1203
FeO
MgO
CaO
Na20
K/O
42.2
12.34
7.82
17.5, 16.3
0.41
Average
whole rock
2
40.93
11.68
7.93, 8.50
19.17
7.30
10.07
0.49
3
13.8
10.9
0.53
0.33
Total
4
Coarse grained
mode
probe
5
6
Fine
grained
7
41.6
12.6
8.1
17.7
7.3
10.5
0.48
0.33
43.1
9.2
9.6
16.8
7.5
13.0
0.55
0.26
42.2
10.2
9.7
16.5
7.7
11.3
0.64
0.44
46.9
4.8
14.5
15.4
3.3
10.8
0.66
1.02
98.61
100.01
98.68
97.38
1. Goles et al. [22]. 2. Goles et al. [21]. 3. Gast and Hubbard [23]. 4. Average of 1, 2, and 3.5. Calculated from mode of coarse
grained area. 6, 7. Defocussed beam electron microprobe analyses.
3.5. Major element chemistry
There are no complete major element chemical
analyses of sample 10071 in the literature, but partial
amalyses [ 2 1 - 2 3 ] have been assembled in table 4.
These data are labeled 'whole rock', but the sample
distribution scheme for 10071 analysis and the relative volumes of the fine and coarse grained areas indicate that the analyses are primarily representative of
the composition of the coarse grained area. The composition of this area was also calculated from its mode
by using average compositions of the major phases
(column 5, table 4). Further control on the composition of this portion of 10071 was obtained by defocussed beam electron probe analysis. Considering the
necessary approximations involved in arriving at the
composition of the coarse grained area by any of these
methods, the agreement (columns 4, 5 and 6 of
table 4) is satisfactory. The bulk composition of the
fine grained area was also obtained by the defocussed
beam technique (column 7, table 4). A comparison of
the major element chemistry of the two areas reveals
significant contrasts which can be summed up in terms
of various compositional parameters: SiO2, 46.9 vs.
42.2; TiO2, 4.8 vs. 10.2; FeO/MgO, 4.7 vs. 2.1; A1203/
CaO, 1.34 vs. 0.86; K 2 0 / N a 2 0 , 1.54 vs. 0.69 (fine
grained vs. coarse grained based on wt %).
4. Discussion
The crystallization sequence for Apollo 11 ferrobasalts has been determined experimentally and from
textural studies [1,8, 24 among others]. In general, it
has been found that ilmenite begins to crystallize at
relatively high temperatures and continues to crystallize to near solidus temperatures, its Mg content progressively decreasing during the sequence. Olivine also
forms early, but its crystallization is interrupted by a
peritectic reaction involving liquid and clinopyroxene.
The pyroxene is initially Ca-poor pigeonite while
augite forms at lower temperature. Although the details of the pyroxene relations are complicated and
not yet fully understood, it is well established that
there is a progressive decrease of the Mg/Fe ratio in
the crystallization sequence. Of the major phases,
plagioclase forms last in the sequence, followed b y
residual phases such as potassic feldspar, phosphates,
zircons, silica and glass. The extreme zoning of the
pyroxene indicates poor equilibration between crystal
phases and residual melt.
Within the context of the generalized crystallization
history described above, the mineralogical and chemical contrasts between the two textural types in rock
10071 are precisely what would be expected if the
68
M.J.Drake, D.F. Weill, Lunar sample 10071
fine grained, variolitic portion formed from a liquid
which had separated from a partly crystallized portion of 10071 magma. The coarse grained portion of
the sample now represents the solidified bulk 10071
magma. Separation of the liquid differentiate took
place after the olivine-pyroxene peritectic and prior
to the onset of plagioclase crystallization, when the
precipitating pyroxene had an intermediate Fe/Mg
ratio. Such a mechanism accounts for the absence of
olivine in the fine grained area, whereas all other
phases are present in both areas, llmenite and pyroxene had been crystallizing prior to separation and
although they would be expected in both areas, they
should occur in greater concentration and have a
higher average Mg content within the coarse grained
area. Plagioclase and residual phases would also crystallize in both systems, but they should be more concentrated in the fine grained area. Because of limited
liquid-solid reaction after crystals have formed, the
composition of plagioclase and residual phases would
be almost the same in both systems. The higher concentration of SiO2 in the differentiated, now fine
grained, system is almost entirely accounted for by
the large contrast in ilmenite (early forming) and silica
(probably the last phase to crystallize) content of the
two areas.
Rock 10071 is a composite sample presenting very
strong evidence for igneous differentiation by fractional crystallization followed by solid-liquid separation. The fine grained differentiate is intermediate in
composition between typical Apollo l 1 ferrobasalt
(typified by the coarse grained area) and the more
strongly differentiated mesostasis found in virtually
all of the Tranquillity basalts. A liquid corresponding
in composition to the mesostasis material has a viscosity in excess of 107 poises at 1100°C [4], and it
would require special conditions to effectively separate it from its almost completely solidified surroundings. It is unlikely that such a process was important
during the crystallization of surficial lava flows for
there is little direct evidence for it (e.g., veinlets or
other segregations of mesostasis material) in any of
the lunar igneous rocks collected so far. The only
possible exception is the granitic portion of Apollo 12
sample 12013 which may be interpreted as a segregation of strongly differentiated residuum by means of
filter pressing [5]. Separation of intermediate liquids
such as that which gave rise to the fine grained portion
of 10071 is perhaps easier to effect due to its greater
abundance and much lower viscosity of less than 104
poises at 1100°C [4]. The scale and extent of such
solid liquid separation is still a matter for conjecture.
There is strong evidence that the two portions of
10071 are related by crystal-liquid fractionation, but
the details are subject to several interpretations. The
separation of the intermediate liquid may have taken
place in some common interior (but most likely shallow) magma source or it might be the result of movements within a flow or lava pool at the surface. Certainly the two portions of 10071 need not have been
part of the same surface flow unit, although the simplest explanation would involve autoinjection of the
intermediate liquid into already solidified portions of
a lava flow or crusted lava pond. Most importantly,
however, sample 10071 provides us with strong and
direct evidence that differentiation via crystal-liquid
separation did occur during lunar igneous activity.
The sample also suggests that at least some of the
variations in texture, mineralogy, and chemistry observed in lunar rocks, breccias, and smaller fragments
need not always be related to very large scale differentiation or to different parental magmas. Finally, we
would like to suggest that an investigation of the
minor and trace element abundances in the two areas
of sample 10071 would be a rewarding companion
exercise.
Acknowledgments
The authors gratefully acknowledge the support of
National Aeronautics and Space Administration grant
NGL 38-003-022.
References
[ 1] D.F. Weill, I.S. McCallum, Y. Bottinga, M.J. Drake and
G.A. McKay, Mineralogy and petrology of some Apollo 11
igneous rocks, Geochim. Cosmochim. Acta, Suppl. 1,
Vol. 1 (1970) 937.
[2] I. Kushiro and H. Haxamura, Major element variation and
possible source materials of Apollo 12 crystalline rocks,
Science 171 (1971) 1235.
[3] G.M. Brown, C.H. Emeleus, J.G. Holland, A. Peckett and
R. Phillips, Picrite basalts, ferrobasalts, feldspathic
norites, and rhyolites in a strongly fractionated lunar
crust, Geochim. Cosmochirn. Acta, Suppl. 2, vol. 1 (1971)
583.
M.J.Drake, D.F. Weill, Lunar sample 10071
[4] D.F. Weill, R.A. Grieve, I.S. McCallum and Y. Bottinga,
Mineralogy-petrology of lunar samples. Microprobe
studies of samples 12021 and 12022; viscosity of melts
of selected lunar compositions, Geochim. Cosmochim.
Acta, Suppl. 2, vol. 1 (1971) 413.
[5] M.J. Drake, I.S. McCallum, G.A. McKay and D.F. Weill,
Mineralogy and petrology of Apollo 12 sample no.
12013: a progress report, Earth Planet. Sci. Letters 9
(1970) 103.
[6] O.B. James, Petrology of lunar microbreccia 12013, 6,
U.S. Geol. Surv. Int. Rep. Astrogeology 23 (1970) 1.
[7] H.H. Schmitt, G. Lofgren, G.A. Swann and G. Simmons,
The Apollo 11 samples: Introduction, Geochim. Cosmochim. Acta, Suppl. 1, vol. 1 (1970) 1.
[8] E. Roedder and P. Weiblen, Lunar petrology of silicate
melt inclusions, Apollo 11 rocks, Geochim. Cosmochim.
Acta, Suppl. 1, vol. 1 (1970) 801.
[9] S.E. Haggerty, F.R. Boyd, P.M. Bell, L.W. Finger and
W.B. Bryan, Opaque minerals and olivine in lavas and
breccias from Mare Tranquillitatis, Geochim. Cosmochim. Acta, Suppl. 1, vol. 1 (1970) 513.
[10] J.F. Lovering and J.R. Widdowson, The petrologic
environment of magnesium ilmenites, Earth Planet. Sci.
Letters 4 (1968) 310.
[11] A.M. Reid, C. Meyer Jr., R.S. Harmon and R. Brett,
Metal grains in Apollo 12 rocks, Earth Planet. Sci.
Letters 9 (1970) 1.
[12] B.J. Skinner, High crystallization temperatures indicated for igneous rocks from Tranquillity Base, Geochim.
Cosmochim. Acta, Suppl. 1, vol. 1 (1970) 891.
[13] A.M. Reid, J.Z. Frazer, H. Fujita and J.E. Everson,
Apollo 11 samples: major mineral chemistry, Geochim.
Cosmochim. Acta, Suppl. 1, vol. 1 (1970) 749.
[ 14] K.H. Carman and O.F. Tuttle, Experimental study bearing on the origin of myrmekite (abstract), Ann. Meeting
Geol. Soc. Am. 29A (1963).
[ 15 ] J.H. Carman and O.F. Tuttle, Experimental verification
of solid solution of excess silica in sanidine from rhyolites (abstract), Ann. Meeting Geol. Soc. Am. 33 (1967).
[16] W.J. Duffin, Plagioclase reactions, Min. Mag. 33 (1964)
812.
[ 17] J. Wyatt and G. Sabatier, Reaction des feldspaths alcalins avec des solutions hydrothermales de CaC12, Comptes
Rendus 260 (1965) 1681.
[ 18] K. Perry Jr., Representation of mineral chemical analyses
in 11-dimensional space: I. Feldspars, Lithos 1 (1968)
201.
[19] T.F.W. Barth, Feldspars (Wiley-lnterscience, 1969) 261 pp.
[20] Lunatic Asylum, Mineralogic and isotopic investigations
on lunar rock 12013, Earth Planet. Sci. Letters 9 (1970)
137.
[ 21 ] G.G. Goles, K. Randle, M. Osawa, R.A. Schmidt, H. Wakita, W.D. Ehmann and J.W. Morgan, Elemental abundances by instrumental activation analysis in chips from
27 lunar rocks, Geochim. Cosmochim. Acta, Suppl. 1,
vol. 2 (1970) 1165.
69
[22] G.G. Goles, K. Randle, M. Osawa, D.J. Lindstrom, D.Y.
Jerome, T.L. Steinborn, R.L. Beyer, M.R. Martin and
S.M. McKay, Interpretations and speculations on elemental abundances in lunar samples, Geochim. Cosmochim. Acta, Suppl. 1, vol. 2 (1970) 1177.
[23] P.W. Gast and N.J. Hubbard, Abundance of alkali metals, alkaline and rare earths, and strontium-87/strontium-86 ratios in lunar samples, Science 167 (1970) 485.
[24] J.V. Smith, A.T. Anderson, R.C. Newton, E.J. Olson,
P.J. Wyllie, A.V. Crewe, M.S. Isaacson and D. Johnson,
Petrologic history of the moon inferred from petrography, mineralogy, and petrogenesis of Apollo 11 rocks,
Geochim. Cosmochim. Acta, Suppl. 1, vol. 1 (1970) 897.
[25] H.A. Liebhafsky, H.G. Pfeiffer, E.H. Winslow and P.D.
Zemany, X-ray Absorption and Emission in Analytical
Chemistry (Wiley, New York, 1960) 357 pp.
[26] H.J. Rose Jr., F. Cuttitta, E.J. Dwornik, M.K. Carron,
R.P. Christian, J.R. Lindsay, D.T. Ligon and R.R. Larson, Semimicro X-ray fluorescence analysis of lunar
samples, Geochim. Cosmochim. Acta, Suppl. 1, vol. 2
(1970) 1493.
[27] W. Compston, B.W. Chappell, P.A. Arriens and M.J.
Vernon, The chemistry and age of Apollo 11 lunar material, Geochim. Cosmochim. Acta, Suppl. 1, vol. 2
(1970) 1007.
[28] J.A. Maxwell, L.C. Peck and H.B. Wiik, Chemical composition of Apollo 11 lunar samples 10017, 10020,
10072, and 10084, Geochim. Cosmochim. Acta, Suppl.
1,vol. 2 (1970) 1369.
[29] A.E.J. Engel and C.G. Engel, Lunar rock compositions
and some interpretations, Geochim. Cosmochim. Acta,
Suppl. 1, vol. 2 (1970) 1081.
[30] S.O. Agrell, J.H. Scoon, I.D. Muir, J.V.P. Long, J.D.C.
McConnell and A.Peckett, Observations on the chemistry, mineralogy and petrology of some Apollo 11
lunar samples, Geochim. Cosrnochim. Acta, Suppl. 1,
Vol. 1 (1970) 93.
[31] C. Frondel, C. Klein Jr., J. Ito and J.C. Drake, Mineralogical and chemical studies of Apollo 11 lunar fines and
selected rocks, Geochim. Cosmochim. Acta, Suppl. 1,
vol. 1 (1970) 445.
[32] H. Haramura, Y. Nakamura and I. Kushiro, Composition
of lunar fines, Geochim. Cosmochim. Acta, Suppl. 1,
vol. 1 (1970) 539.
[33] W. Compston, H. Berry, M.J. Venon, B.W. ChappeU
and M.J. Kaye, Rubidium-strontium chronology and
chemistry of lunar material from the Ocean of Storms,
Geochim. Cosmochim. Acta, Suppl. 2, vol. 2 (1971)
1471.
[34] J.A. Maxwell and H.B. Wiik, Chemical composition of
Apollo 12 lunar samples 12004, 12033, 12051, 12052,
and 12065, Earth Planet. Sci. Letters 10 (1971) 285.
[35] F. Cuttitta, H.J. Rose Jr., C.S. Annell, M.K. Carron,
R.P. Christian, E.J. Dwornlk, L.P. Greenland, A.W.
Helz and D.T. Ligon Jr., Elemental composition of some
Apollo 12 lunar rocks and soils, Geochim. Cosmochim.
Acta, Suppl. 2, vol. 2 (1971) 1217.
70
M.ZDrake, D.F. Weill, Lunar sample 10071
[36] J.H. Scoon, Chemical analyses of lunar samples 12040
and 12064, Geochim. Cosmochim. Acta, Suppl. 2,
vol. 2 (1971) 1259.
[37] A.E.J. Engel, C.G. Engel, A.L. Sutton and A.T. Myers,
Composition of five Apollo 11 and Apollo 12 rocks
and one Apollo 11 soil and some petrogenetic considerations, Geochim. Cosmochim. Acta, Suppl. 2, vol. 1
(1971) 439.
[38] A.A. Smales, D. Mapper, M.S.W. Webb, R.K. Webster,
J.D. Wilson and J.S. Hislop, Elemental composition of
lunar surface material (Part 2), Geochim. Cosmochim.
Acta, Suppl. 2, vol. 2 (1971) 1253.
[39] J.P. Willis, L.H. Ahrens, R.V. Danchin, A.J. Erlank, T.J.
Gamey, P.K. Hofmeyr, T.S. McCarthy and M.J. Orren,
Some interelement relationships between lunar rocks
and fines, and stony meteorites, Geochim. Cosmochim.
Acta, Suppl. 2, vol. 2 (1971) 1123.
[40] A.P. Vinogradov, Preliminary data on lunar ground
brought to earth by automatic probe "Luna-16", Geochim. Cosmochim. Acta, Suppl. 2, vol. 1 (1971) 1.