American Mineralogist, Volume 67, pages 14J7, 1982
Mineralogy of mafic and Fe.Ti oxide-rich differentiatesof the
Marcy anorthositemassif, Adirondacks, New york
LBwrs D. Asuwnl
Lunar and Plnnetary Institute
3303 NASA Road l, Houston, Texas 77058
Abstract
Rocks enriched in mafic silicates, Fe-Ti oxides and apatite form a minor but ubiquitous
facies of nearly all Proterozoic massif-typeanorthosite complexes.In the Marcy massif of
the Adirondack highlands,New York, the mafic rocks have two modesof occurrence: l) as
conformable segregationsinterpreted as cumulate layers in the border zones, and 2) as
dikes throughout the massif, interpreted as having crystallized from residual liquids
extracted at varying stagesduring differentiation. Primary mineral compositions in these
rocks are commonly preserved,despitethe effectsof subsolidusrecrystallizationin the
high pressure granulite facies. In the mafic rock suite, primary compositions of mafic
silicates reveal an iron enrichment trend which is broadly similar to that of basic layered
intrusions, but suggestive of somewhat higher crystallization temperatures. Textural
relationships suggest the following crystallization sequence: plagioclase, pigeonite +
augite, hemo-ilmenite * magnetite,apatite. Fe-rich olivine replacedpigeonitein the lateststage residual liquids. The mineralogy and field relationships of these mafic rocks are
consistent with their origin as diferentiates from the same melts which produced the
anorthosites. This conclusion agrees with recent geochemical data that suggest an
independentorigi-nfor the spatially associatedmangerite-charnockitesuite.
Introduction
For the past several decades,the controversy
over the origin of massif-type anorthosites has
centered around the question of the consanguinity
of anorthositic rocks and spatially associatedsuites
of orthopyroxene-bearinggranitic rocks (charnockite-mangerite series). Recent trace element geochemical studies focusing on, but not limited to,
rare earth elements (REE) (Philpotts et al., 1966;
Greenet al., 1969,1972;
Seifertet a|.,1977;Seifert,
1978; Simmons and Hanson, 1978; Ashwal and
Seifert, 1980)demonstratethat the acidic rocks do
not represent differentiates from the melts which
producedthe anorthosites.This was recognizedas
early as 1939by A. F. Buddingtonon the basis of
field relationships (summarized by Buddington,
1969).
In addition to huge volumes of pure anorthosite,
the typical massif anorthosite suite contains minor
facieswith larger proportions of mafic silicates,FeTi oxides and apatite. Extreme concentrations of
these minerals give rise to minor ultramafic rocks,
as well as the ilmenite-magnetite ore deposits
0003-004)v82t0r02-0014$02.00
whose associationwith anorthositemassifsis well
known (Rose, 1969).Itis thesemafic rocks, and not
the mangerite-charnockites,which complementthe
anorthosites,as demonstratedby REE geochemical
relationships(Ashwal and Seifert, 1980).The purpose of this paper is to describe the field relationships and mineralogical characteristics of mafic
rocks associatedwith the Marcy anorthosite massif
in the Adirondack highlands of northern New York
State, and to emphasize the importance of these
rocks in understandingthe fractionation histories of
this and other massif-typeanorthositecomplexes.
Field relations of mafic faciesof anorthosite
In the Marcy massif,the most felsic anorthosites
are found in the central regions, and more mafic
varieties occur near the margins. Becauseof the
domical form of the massif, the mafic border facies
overlies the anorthositic core zone (Buddington,
1939,1960,1969).In someplacesthere is a systematic increase in mafic silicates and oxide minerals
from coarse, core zone anorthosite (Marcy facies)
structurally upwards to finer-grained, border zone
14
t5
ASHWAL : OXIDE.RICH DIFFERENTIATES, MARCY ANORTHO SITE
/u"30'N
/u"15'N
4"OO'N
71"30'W
@
:
7/t4
-
ltons"rir"-rhornockire series
Porogneiss,morble, quortzite ond
othet melosedimsntqry ro<ks
(in(ludesminor gobbrc & ohphibolite)
O
5
l0
15 Miles
O
5
lO
15
leucogobbro ond
Ptedominontly
1'51
J. :.'i. .t leuGonorite (<ontocts doshed
wngte tnlelleol
Pr.do.inontly
I
,/
onoithosits
Mofi< or ultromoli<
<umulote loyer
,/
moti, an
)(
fe-fi oxide deposir
2O Kilometers
"
Fig. I Geologic map of the Marcy anorthosite massif, Adirondacks, New York, showing samplelocations. Base map taken from
lsachsenand Fisher (1970).
gabbroicanorthositer(Whitefacefacies) (Buddington, 1939; Davis, 1971).Elsewhere,the border
zonescontain local mafic to ultramafic layers (color
index > 40%) which are conformable to foliation in
host gabbroicanorthosites.Mafic rocks also occur
as dikes throughoutthe massif,as discussedbelow.
The distributionand structuralrelationships,where
'Rocks are classifiedon the basis of color index (volume 7o
mafic minerals), and the mineralogicalmodifiers gabbroic, noritic, or troctolitic are useddependingon whether the predominant
mafic phase is high-Ca pyroxene, low-Ca pyroxene, or olivine,
respectively(Buddington, 1939;Streckeisen,1976).The modifier
oxide-rich is added when modal ilmenite f maenetite exceeds
lOVo.
known, of these mafic layers and dikes are shown
on a geologicmap of the Marcy massifin Figure 1.2
In the border zones,local ultramaficlayers commonly have sharp lower contacts and are gradational upwards into more feldspathicrocks. These are
most easily interpretedas resultingfrom processes
involving gravitational settling and accumulation of
mafic silicatesand Fe-Ti oxide minerals. The layers
range in thickness from 1 cm to 1 m and are often
2Details of sample locations may be obtained by ordering
Document AM-82-182 from the Business Office, Mineralogical
Societyof America, 2000Florida Avenue, N.W., Washington,
D.C. 20009.Pleaseremit $1.00in advancefor the microfiche.
l6
ASHWAL : OXIDE.RICH DIFFERENTIATES, MARCY ANORTH)SITE
rhythmically repeated several times on the scale of
individual outcrops. The attitudes of the layers are
in most placesconsistentwith the domical shapeof
the massifs(Fig. 1), althoughtheir lateral extent is
difficult to estimatedue to the paucity of exposures.
There are also mafic rocks which occur as sharpwalleddikesof variedcomposition,crosscuttingthe
anorthositic foliation at high angles. These dikes
may have crystallized from residual liquids extracted at various stages of differentiation of the anorthosite suite (Ashwal, 1978a,b). In the Marcy
massif, mafic dikes are not spatially restricted to the
border facies, but may be found even in the central
regions(Fie. 1). The dikes are generallyless than
7 m thick, although Buddington (1939, p. 50-51)
described a mafic gabbro dike some 60 m wide in
the Carthage anorthosite massif in the western
Adirondacks.
Mineral constitution and petrography
Primary minerals of the Marcy anorthosite complex includeplagioclase,high-Capyroxene,low-Ca
pyroxene, olivine, ferrian ilmenite, titaniferous
magnetite, and apatite. Metamorphic minerals include garnet, hornblende, biotite and quartz, in
addition to recrystallized equivalentsof the primary
phases,suggestingequilibration in the high pressure
granulitefacies (deWaard, 1965).
Conformable occurrencesof mafic rocks
Modes of conformable mafic facies rocks from
the Marcy massif are given in Table 1. There is
generally a progressivedecreasein mafic and oxide
minerals with increasing stratigraphic height above
the basalcontact of a cumulate layer; the scale over
which this variation in modal mineralogy takes
place ranges from centimeters to meters. Cumulus
phasesinclude augite,inverted pigeonite,ilmenite,
magnetite,and, less commonly, Fe-rich olivine and
apatite. The primary mineralogies and textures of
theserocks, however,havebeenaffectedto varying
degreesby subsolidusrecrystallization,which has
resulted in the production of metamorphic garnet
and hornblende,and in the reconstitutionofcoarsely exsolvedpyroxenesinto smaller discretegrains
with different compositions. Zones of garnet 1-5
mm thick commonly occur betweenthe basal contacts of mafic layers and underlying leucogabbros
or leuconorites. Although metamorphic garnet,
hornblende, and pyroxenes constitrrte nearly 20Vo
of some samples,other rocks appearto have completely retained their primary mineralogiesand textures (Table 1, sample SA-3D). In general, it is
possible to see through the overprinting effects of
subsolidusrecrystallization,as discussedbelow.
The color indices of the lower, most mafic parts
of the layers vary between approximately 60Voand
100%(Table l). The layers generally lie on and are
Table l. Point counted modes of conformable mafic rich rocks from the Marcv massif
sA-3D
Plagioclase
Inverted
pigeonite
Orthopyroxene
Clinopyroxene
Ollvine
L.2r
63.7
SA-5
25.6
13.9**
-
SL-IIY
66. S
q q ) 1
IIrenite
Magnetite
Apatite
Garnet
Hornblende
I
3.2
t
_n
2.4
,:t
I.1
0.6
0.9
0.I
q
11
2
?
v.z
index
-
0.5
tr
_
-tr
Sulfides
Biotite
ScatrpLite
Other secondaries**r
Color
SA-3A
0.I
includes
nj.nor
75.6*
tr
3.4
13.4
t8.9
4.2
9.0
4.5
0.3
0.2
2.2
I.O
O.2
42.6*
7.L
4.8
19.0
I
I
SA-I5B
73.5*
3.2
2.L
II.4
nt
0.5
Mtr-568
MM-56A
SL-2IA
Sr,-21
20.9
L.7
74,O*
19,3*
51.5*
t.u
2A.9
J-b
II.2
0.9
48.6
1.9
t5.5
]5.3
I.O
-
4 .2
O.7
u.b
4.1
0.5
4.3
L6.2
10.5
L4.7
o.2
0.4
l. r
1,, f"'
L5.2
27.L*
cr
9.5
8.4
3.3
1 8 .O
0.5
0.5
100.0
100.0
94.8
36.3
oxiderich
pyrox-
leuconorite
host
iltiperthite;
2.4
4.1
7.!
52.3
SA-15A
0.3
100.0
Ieuconorite
host
enr!e
layer
*
SIJ-IIA
rr
includes
trace
r00.0
Ioo.0
100.0
100.0
100.0
r00.0
100.0
91.2
24.4
57.4
25.5
79.1
26.0
80.7
oxiderich
pyroxenrte
Iayer
Ieuconorite
host
oxiderich
gabbro
Iayer
oxiderich
mfic
gabbro
Iayer
leucogabbro
host
oxiderich
@f ic
gabbro
layer
arcut
of
pleonaste,
***
leucoga-bbro
host
includes
sericite
aal
haatite
alteration
0.5
loo.0
48.0
oxiderich
aabbro
host
100.0
72.9
oxiderich
mf lc
troqtolite
Layer
l7
ASHWAL : OXIDE-RICH DIFFERENTIATES, MARCY ANORTHOSITE
gradational structurally upwards into host rocks of
leucogabbroic or leuconoritic composition (color
index26Voto 36Vo),althoughin somecasesthe host
rock has a color index as high as 46% (Table l,
sampleSL-21).
In all layers, high-Ca pyroxene (augite) dominates over low-Ca pyroxene (inverted pigeonite
and/or orthopyroxene). The large high-Ca/low-Ca
pyroxene ratio in the highly recrystallized rocks
(e.9., MM-56B) may be due to the increase in
discrete augite grains from the recrystallization of
coarsely exsolved primary pyroxenes. The paucity
of low-Ca pyroxene in other specimens(e.9., SL21, SL-21A,5683d)is relatedto the onsetof Fe-rich
olivine crystallization,which replacedlow-Ca pyroxene during the later stages of solidification.
Olivine-bearinglayers such as representedby specimen 5683d do, however, contain minor inverted
pigeonite. The low-Ca pyroxene-deficientor olivine-bearinglayers, or both, contain abundant apatite, which also must be a late-stagecumulusphase.
In addition to their mineralogy, the location of these
layers at the extreme margins of the massifs (Fig.
1), and hence at the highest stratigraphiclevels,
suggeststhat they may represent some of the very
last accumulations from late stage mafic liquids
which had been segregatedto the upper regions of
the massifs.
All of the conformable mafic layers contain major
amounts(>10%) of ilmeniteand magnetite.In some
cases,the oxide mineral content is sufficiently high
that the rock can be consideredore grade. Conformable oxide mineral-rich layers within gabbro or
leucogabbro are particularly abundant in the Sanford Lake area where they are mined for their
ilmenite concentrations(Gross, 1968),but similar
layers of much lesser extent have been reported
from elsewherein the Marcy massif (Buddington,
1953, Table 1, sample no. 20). Ilmenite usually
predominates over magnetite in the conformable
layers, although there are examples where the reverseis true (Table 1, sampleSL-2IA).
Mafic dikes
Modes of mafic dikes from the Marcy massif are
given in Table 2. The dikes show wide ranges in
color index (approximately 40-90%), and mineral
constitutions.It is important to determinewhich, if
any, ofthese dike rocks representliquids. They are
finer grainedthan the mafic cumulate rocks; however, primary textures in the dike rocks are generally
not well preserved. Many contain abundant garnet
(Table 2). The profound effect ofgarnet production
on modal mineralogy (particularly color index) may
be seenby comparingthe modes of samplesSR-9
and MM-6 (Table 2), which despite their large
Table 2. Point counted modes of mafic rich dikes from the Marcy massif
MM-I A
8.0
Plagioclase
pigeonite
Invertedl
orthopyioxene
cLinopyroxene
olivlnett
Ilnenite
Magnetite
Apatite
33.8r
22.5
1.9
63.4
LZ.Z
10.9
I 3.0
z.o
5.4
ql
Garnet
Hornblende
Sulfiales
Other secondariesr*i
t:t
color
nock
92.O
index
* includes
alteration
66.2
pyroxenLte
tl4)e
flinor
50.0*
24.7
L7.5
17.0
23.3
11.r
14-4
1 0 .r
3.1
4.7
I
t7-2
I
2.3
9.7
6.8
5.8
4.3
42.r
13.r
9.2
6 .O
8.4
23.9
4-L
lo. 2
o.4
6.
/
42.3
6-4
14.3
4 3- 2
antiperthiter
oxitlerich
nafic
norite
**
includes
r00. 0
54.4
oxiderich
gabbro
iddlingsite
1 0 0 .0
5 4 .r
gabbro
alteration;
_n
?l
0.5
J.v
(7
L9.7
4.4
"-n
ET
1.5
1.9
J.O
loo.0
o
26.4
45.6*
no
10.8
t7.7
1.6
100.o
MM-44A
sA-14A
MM-13A
100.0
100.0
100,0
56.8
7L.3
50.0
oxiderich
gabbro
oxiderich
mafic
gabbro
oxiderich
gabbro
***
includes
selicite
100.0
98.r
oxiclerich
wehrlite
andl hffitite
l8
ASHWAL : OXIDE-RIC H DIFFERENTIATES, MARCY ANORTHOSITE
mineralogicaldifferences,have similar bulk compositions. Thin layers of garnet are also found at the
contacts between some of the dikes and the surrounding anorthositic rocks in a manner similar to
that of the garnet zones at the basal contacts of the
ultramafic cumulates. Because of these garnet
zones,it is impossibleto determinewhether these
dikes once had chilled margins.
Oxide-rich, olivine gabbro or mafic gabbro dikes
which lack low-Ca pyroxene (samples SR-9 and
MM-6, Table2) may have crystallized from the final
liquid differentiates of the anorthosite suite, since
their REE patterns (Ashwal and Seifert, 1980)are
characteized by high total abundancesand negative Eu anomalies,and relativelight REE depletion.
These patterns are complementary to anorthosites
and leucogabbros,which exhibit low REE abundances,positive Eu anomalies,and relative light
REE enrichment(Seifertet a|.,1977;Simmonsand
Hanson, 1978).Mineral compositionsof thesedikes
are consistentwith their origin as latest-stagedifferentiates,as discussedbelow.
Although the mafic dikes probably in large part
represent liquid compositions, the possibility of
mineral accumulation or loss in these liquids during
or after their emplacementas dikes must be considered. Additionally, there are rare examplesof ultramafic dikes includinghypersthenite(sampleMM-8)
and oxide-rich wehrlite (sample MM-44A) which
may represent remobilized cumulates (deWaard,
1970). These complications must be taken into
account before accurate liquid lines of descentcan
be determined.
textureis that the anorthositesrepresentaccumulations of pligioclase concentratedeither gravitationally or by flowage(Emslie, 1975;Martignole,1974;
Morse, 1969;Olmstead,1969).
Primary textures suggestthat the pyroxenescrystallized after plagioclase,and in turn were followed
by the Fe-Ti oxide minerals. It is not possibleon
textural grounds to determine whether augite or
pigeonite crystallized first, but it is likely that they
precipitated simultaneouslyduring most of the crystallization history. In the anorthosites,leucogabbros, and leuconorites,coarsely exsolvedprimary
pyroxenesor their metamorphicequivalentspartially surround earlier and larger plagioclasecrystals in
a subophitic relationship. More mafic rocks also
have subophitictextures;however, the presenceof
large,discretepyroxenecrystalsof probablecumulate origin suggestco-crystallization of plagioclase
and pyroxene. In contrast to the tabular shape of
plagioclase,cumulus mafic silicatesare equant in
shape, and the ultramafic cumulates (oxide-rich
pyroxenites) have an equigranular texture.
In a similar manner, ilmenite and magnetite partially or completely surround the pyroxenes, suggesting that the oxides crystallized later, and this
texture is particularly well developed in the oxiderich pyroxenite cumulate layers. The oxide minerals may constitute up to 40Voof these rocks (Table
1), and since the layers grade structurally upwards
into less oxide-rich rocks, the textures may be
interpreted to represent cumulus Fe-Ti oxides
which have been extended interstitially by postcumulus growth. Oxide rich pyroxenite layers with
similar textures have been reported from other
anorthosite massifs (Subramaniam,1956, Fig. 3)
Petrographicdetermination of crystallization
and are also present in zoned ultramafic complexes
(Taylor and Noble, 1969,Fig. 10).
Despite the effects of subsolidus recrystallizaApatite occurs as a cumulus phase only in the
tion, primary igneous textures are preserved in very late-stagelow-Ca pyroxene-deficientor olivmany cases,permitting the distinction of cumulus ine-bearingcumulates, or both. In these rocks it
and post-cumulusphasesand allowing the determi- occurs as intergrowths with the Fe-Ti oxides, alnation of crystallization sequences.Plagioclasein though aggregatesof very coarse apatite crystals
the Marcy massif and in other anorthosite complex- (up to 1 mm across) are also found. Olivine occurs
es was the first phase to crystallize. Evidence for as largecumulusgrainsup to 5 mm acrosswhich, in
this conclusionincludes large, euhedralor subhe- some cases,poikilitically enclosepyroxenes,apadral plagioclasecrystals with smaller, anhedral py- tite and plagioclase; it also occurs as smaller,
roxenes in interstitial areas. A common feature of roundedgrains within cumulus augite crystals (samanorthositic rocks in the central parts of massifs is ple 5683d).Sparse grains of inverted pigeonite are
the alignment of tabular plagioclase crystals with
invariably present in this layer, which represents
their longest dimensions subparallel,forming a pla- the only known coexistenceof three primary mafic
nar fabric in the rock. The implication of this silicatesin the Marcy massif.
ASHW AL : OXI DE-RICH DIF F ERENTIATES. MARC Y AN ORTHOSITE
l9
Mineral compositions
reflect higher initial crystallization temperatures
than those of the Skaergaardintrusion (Fig. 3).
Pyroxenesand olivines
The coexisting mafic silicates from cumulate layAll primary minerals in the anorthosite-mafic ers define a trend of increasingfe* with increasing
gabbro suite with the exception of olivine and stratigraphicheight. For example, sample SA-3D,
apatitecontain complex subsolidusexsolutionfea- an oxide-rich pyroxenite layer, contains the most
tures. Primary pyroxene compositions, however, magnesianaugite-invertedpigeonitepair yet found
can be reasonably ascertainedby bulk analysis in the Marcy massif (fe* of low-Ca pyroxene :
usingbroad-beamelectron microprobetechniques. 0.36). Samples SA-3A and SA-5 are more feldThis has been done for pyroxenes of the Adiron- spathic leuconoriteswhich are conformablewith,
dack Marcy massif using a combinationof energy and occur within a few meters stratigraphically
dispersive(ED) and wavelength dispersive (WD)
microprobetechniques.3
In the Marcy massif, many low-Ca pyroxenes
contain broad exsolution lamellae of augite inclined
to the (001)directionsof the presentorthopyroxene
hosts,and later fine augitelamellaeparallelto (100)
of their hosts (Fig. 2a). These textures suggest
initial crystallizationof pigeonite,followed by exsolution and inversion to orthopyroxenein response
to slow cooling. Coexistingaugitescommonly contain several generationsof pigeonite lamellae, as
describedin detail by Jaffe et al. (1975).
Primarycompositionsof coexistingaugite-inverted pigeonite and augite-olivine pairs from mafic
cumulatelayers and dikes are plotted in Figure 3a.
RepresentativeED and WD microprobe analyses
for one sampleare given in Table 3.aThe compositions show a wide rangein fe* [molar Fe/(Fe+Mg)],
and define a fractionation trend which is broadly
similar to that of basic layered intrusions. In the
Marcy massif, however, the coexistingpyroxenes
are consistent with a narrower solvus, and may
3Primarycompositions of coarsely exsolved pyroxenes were
determined using an ED microprobe technique, whereby the
electronbeamwas broadened,in somecasesup to 100microns,
in order to incorporate entire grains. This method is successful
for major elements;ED analysesof host and lamella phasesare
generally within 5 relative Vo of WD analyses, and calculated
ratios of mineral moleculesshow excellent agreement(Table 3).
However, the poor system resolution and low peak/background
ratios make quantitative ED determination of minor element
oxides essentiallyimpossible,as illustrated by comparisonof
valuesfor MnO by WD and ED techniques(Table 3). Becauseof
this, primary pyroxene compositions were determined by reintegratingfine-beamWD analysesof host and lamella phasesin
the proportions given by the ED analysesin terms of Wo, En and
Fs (Table3).
aA complete table of pyroxene and olivine microprobe analysesmay be obtainedby ordering Document AM-82-183from the
BusinessOffice, Mineralogical Society of America, 2000Florida
Avenue,N.W., Washington,D.C. 20009.Pleaseremit $1.00in
advancefor the microfiche.
illustrating progressive
Fig.2.Photomicrographs
recrystallization of primary pyroxenes (crossed nicols). (A):
Primary inverted pigeonite surrounded by primary augite.
Coarse augite lamellae (white) are inclined to the (001)direction
of host orthopyroxene (at extinction). A later set of fine (100)
augite lamellae (nearly vertical) is also present. Length of
photograph: I mm. (B): Partially recrystallized inverted
pigeonite illustrating the migration and coalesccenceof coarse
augite lamellae at grain boundaries. Length of photograph: 1.5
mm. (C): Granular mosaic of augite and orthopyroxene
illustrating final result of recrystallizatron. Length of
photograph: 2mm.
ASHWAL : OXIDE-RIC H DIFFERENTIATES, MARC Y ANORTHASITE
Table 3. Representative.microprobeanalysesof pyroxene and olivine from sample 5683d
Meta.
+
opx. I
Meta.
opx.5
Meta.
cpx.'
Meta.
cpx.9
si02
5I.03
50.6
50.92
51 .8
Ti02
0 .t r
Nzog
1.38
ct2o3
Fe0*
Mn0
0.05
3 0 .5 9
0.45
Total
structural
51 .18
Olivinef
3 2 .5 3
na
o -27
5.UO
I .8r
2.AO
0. 00
0.00
0.04
0.00
na
26.27
L7.29
0.1
o.39
o.2a
o.44
na
na
r3.79
LS-Z
0.6
tz.fu
1 6 .0
11.0t
1 1. 3
L5.26
1I.98
14.96
0.55
0.6
19.84
1 9 .4
5.51
L6.L2
0.01
0.00
na
na
o.20
100 .69
ro0 .44
M9o
Na2O
29.6
'I.UU
Pri$ary
cpx.f
0.17
0.34
2-6
Primary
pig**
1 0 0 .8 9
1 0 0 .0
o-77
99.94
Formula based on 6 oxygens (pyroxene)
1 0 0. 0
100.65
na
or 4 oxygens (olivine)
si
1.960
L.952
1.935
r.946
1.953
r.940
0.988
Alrv
o.o4o
o.o4g
0.065
0.054
o.o4?
0.060
0.o00
2.000
2.000
2.000
2.000
2.000
2.000
0. 988
o.o22
0.003
0.002
0.983
0. oI5
0. 958
o.023
0.000
2.006
o.o72
o . o 72
0.010
0.000
0,438
0.007
o -624
0.808
o. 057
2.015
0.134
.
o,4r4
0.002
0.633
o.77A
0.035
0.005
0.001
0.841
0.013
0.871
0.226
0.065
0.008
o.000
0. 548
0.o09
o.677
0.65s
0.057
2.019
I .334
0.0I1
o. 678
0. 000
-,vl
Ti
Fe*
Mn
t{s
Ca
Na
Wo
En(Fo)
Fs (Fa)
L.2
48.8
50.0
0.950
0.021
0. 9I9
o.026
r.988
L.4
48.5
50.0
43.2
33.4
2 3- 4
I.961
42.6
34.7
22.7
* total
** cornposition
na
not aalyzed
forr
of
Fe as FeOt
by conrbining
wDA analyses
in proportions
of host and lmella
T wavelength
(wDA),
dispersive
analysis
5 energy dispersive
cooparison
with
tlD analysls
above, this ultrarnafic cumulate layer, and contain
progressivelymore Fe-rich pyroxene compositions
with increasing distance above the basal contact
(Fig. 3a). On a larger scale, sample5683d,a mafic
cumulate layer from near the outer margin of the
massif(Fig. 1), containsFe-rich olivine in addition
to augite and inverted pigeonite. The compositions
of thesephasesare consistentwith the idea previously discussedthat this samplemay representan
accumulationof mafic crystals from very late-stage
liquid. The mineral assemblagein this rock representsthe last appearanceof low-Ca pyroxene (inverted pigeonite), which had apparently reacted
with the liquid to form Fe-rich olivine.
Coexisting augites and olivines from mafic dikes
SR-9 and MM-6 are among the most Fe-rich mafic
silicatesof anorthositesuite rocks. Their compositions are consistent with the hypothesis that these
dikes crystallizedfrom the latest-stageresidualliquids of the anorthosite suite.
2.OO7
11.7
45.0
43.3
34.8
36.O
29.2
2.O23
(33.7)
(66.3)
pigeonite
calculated
prinary
inverted
given
EDA analysest
by broad beil
to 1O0t for
notmalized
analysis
With increasingdegreesof metamorphicrecrystallization,the pyroxenetexturesand compositions
change in a regular fashion. Augites become progressively more calcic as they are transformed, by
means of exsolution of low-Ca pyroxene and accompanyingrecrystallization,from coarse,primary
crystals into finer grained aggregateswith mosaic
texture (Fig. 2). For the low-Ca pyroxenes, the
coarse augite lamellae, originally exsolved parallel
to the (001)directions of host pigeonites(Fig. 2a),
have gradually migrated to the borders of the
grains, where they have coalescedinto polygonal
(Fie. 2b). The final result ofthis process
aggregates
is a fine-grained,granular mosaic of hypersthene
andaugite(Fig. 2c) whosecompositionscorrespond
to thoseof the host and lamellarphases,respectively, of the original primary grains. Tie-linesjoining
these metamorphic augites and orthopyroxenes
have a shallower slope than those joining the
primary augite-pigeonite pairs (Fig. 3b), con-
ASHWAL : OXIDE-RICH DIFFERENTIATES. MARCY AN ORTHOSITE
2l
A
oLrvlNEs
Fs
Fig. 3 (A) Primary pyroxene compositionsas determinedusing broad beamenergydispersivetechniques,and olivine compositions
as determinedusing wavelengthdispersiveanalyses.Solid tie lines: mafic cumulatelayers. Dashedtie lines: mafic dikes. Skaergaard
trend after Wager and Brown (1968).(B) Compositionsofprimary igneouSpyroxene and pyroxene-olivine pairs (short dashedlines)
and reconstitutedmetamorphic pairs (solid lines) plotted with respectto the augite-pigeonitesolvus (dashedcontours) and the augite
orthopyroxenesolvus (solid contours) of Ross and Huebner (1975).
sistent with the higher K, : (Fe/Mg)ro*-c. o",o*"."/
(Fe/Mg)6g6-capyroxene
values of metamorphicallyreequilibratedpyroxenes(Kretz, 1963).When applied
to the pyroxenesolvusof Rossand Huebner(1975),
the igneous pyroxene pairs yield temperatures in
excessof 1100"C, whereasthe metamorphicpairs
yield temperaturesbetween 600' C and 900" C (Fig.
3b). Both igneousand metamorphictemperatures
must, however, be consideredas minimum values
becauseof the uncertain effectsof increasedpressure on the pyroxene solvus. Similar relationships
for pyroxenes from the Marcy massif have been
reportedby Bohlen and Essene(1978).
Plagioclase
Plagioclasein anorthositic rocks from most massif-type complexesgenerally rangesin composition
from Ana6 to An65 (mole Vo) (Anderson, 1969).
Becauseof the extremely large grain size and the
rather irregular distribution of exsolved antiperthitic lamellae,the primary compositionsof anorthositic plagioclasesare difficult to determine, particularly by microprobe analysis.For the Marcy massif,
however, Isachsenand Moxham (1969)have analyzed bulk megacrystcompositionsfrom two extensive vertical sections using spectrographicmeth-
22
ASHWAL : OXIDE-RICH DIFFERENTIATES, MARCY ANORTHOSITE
ods. Their measuredcompositionalrangeamong45
specimensanalyzedwas Ana2.6to An55.2and Or1.7
to Or1s.6,with averagevalues of Anae.7and Or5.2.
(These values, given in mole 70 were converted
from the weight percentagesreported by Isachsen
and Moxham,1969.)The high K contentof anorthositic plagioclaseis in accordancewith relatively
high temperaturesof crystallization(Sen, 1959),but
precipitation of high-K plagioclasefrom melts richer in K than typical unfractionatedbasaltsmust also
be consideredas a possibility (Morse, 1981;Ranson, 1981).
In addition to antiperthiticlamellae,anorthositic
plagioclasescommonly contain minute rodlets and
particles of Fe-Ti oxide minerals, also thought to
have originatedby meansof exsolution(Anderson,
1966).Minor element analysesof plagioclaseconcentratescontainingthese oxide inclusions(Isachsenand Moxham, 1969)reveal substantialamounts
of Fe and Ti (wt.% FeO : 0.27-1.48,avg. : 0.51'
wt.VoTiO2 : 0.04-0.85,zy5. : 0.28), consistent
either with high crystallization temperatures
(Smith, 1975),or with high contentsof Fe and Ti in
the parental magma.
In the mafic rocks, the determination of primary
plagioclasecomposition is further complicatedby
the presenceof abundant garnet, which formed by
means of a subsolidus reaction involving a net
consumptionof anorthite component (Mclelland
and Whitney, 1977).In many casesthis resultedin
strongly zoned plagioclase,enriched in albite and
orthoclase components towards the garnet. For
1
Ab
An
0r
2
3
4
5
6
7
8
9
l0
7 6 , 8 7 2 . 9 7 0 , 4 6 3 , 7 7 0 , 2 6 4 , 0 6 l , 5 6 1. 5 6 t . 9 7 2 . 1
20.7 25,0 27,8 34,6 27.4 34,2 36.6 36,7 36,3 25,2
'l
'l
2,5 2,1 |.8
2,7
,7 2.4 I,8
L8
.9 I,8
Fig. 4 Photomicrograph illustrating variability of plagioclase
(white) composition (mole Vo) inrelation to lamellar or coronitic
garnet (dark, high reliefl. Sample MM-6. Length of photograph is
approximately 6 mm.
example, microprobe analyses of plagioclase in
oxide-richmafic gabbro dike MM-6 yield compositions adjacent to lamellar or corona garnet which
are 11to 15mole Volowerin An and approximately
I mole %higher in Or comparedto compositionsfar
removed from garnet (Fig. a). Representativemicroprobe analysesof plagioclasein mafic rocks are
givenin Table4. The analysesin the centralregions
of grains have the best chance of representing the
primary igneouscompositions,and are in general
somewhat more sodic than anorthositic plagioclases,as would be expectedfor later differentiates.
However, the effects of the garnet-producingreactions on plagioclasecompositions in these rocks
makes determination of plagioclase fractionation
trends by direct measurementrather difficult.
Fe-Ti oxides
Fe-Ti oxide minerals are a ubiquitous constituent
of anorthositesuite rocks, and are presentin major
proportions in the mafic rocks and associatedore
deposits. Like the pyroxenes, the oxides exhibit
complex subsolidusexsolutionfeatures.However,
in addition to exsolutionin the strict sense(hematite lamellaein ilmenite, pleonastelamellaein magnetite), textural and compositional relations among
the Fe-Ti oxides may be further complicated to
varying degreesby subsolidusoxidation-reduction
reactions.This has produced,for example,ilmenite
oxy-exsolutionlamellaein magnetite.Both ED and
WD microprobe techniques were used in analyzing
the compositionsof these Fe-Ti oxide minerals.
For the ED analyses,a defocussedbeamofup to 20
microns in diameter was used in an attempt to
incorporate hematite lamellae in ilmenite and pleonaste lamellae in magnetite. No attempt was made
to integrateoxy-exsolution lamellae, but in some
casesthesecould not be avoided.Textural relations
and ED compositionsof Fe-Ti oxides are summarized in Table 5. Average ED compositions are
plotted on a TiO2-FeO-Fe2O3diagram in Figure 5.
Primary igneous compositions of Fe-Ti oxides
are rarely preservedin plutonic rocks becauseof
the extensivedegreeto which coexistingtitanomagnetite and ferrian ilmenite react upon cooling according to the equation
Hem"" + IJsp"" : M&9., * Ilm"..
(1)
This results in oxidation of the ulvospinel component in titanomagnetiteand correspondingreduction of the hematite component in ferrian ilmenite,
formingilmeniteand magnetite,respectively,either
23
AS HWAL : OXIDE.RICH DIFFERENTIATES. MARCY ANORTHO SITE
Table 4. Representativewavelength dispersivemicroprobe analysesof plagioclases
ru
ran
59,.?0
rim
62.20
si02
56.05
6 0 .1 7
55-72
56.89
5A.77
60.77
5A.92
TlO2
o. 00
o. 00
0 .04
0. 0I
o.o0
0.00
0.00
0.o4
0.00
0.02
Nzog
27.79
27.77
26.8A
26.27
24.63
26.29
24-24
25.01
24.2r
Fe0*
0.39
o-24
0. l1
U.IO
0.06
0.14
o.r2
0.13
o.0I
0.15
!4gO
0. 06
0.00
0.00
0.00
o.00
0.00
o.00
0.00
0.00 .
0.oo
b- 5v
9 .6 0
8.33
7. 7 6
5.88
7.55
5.29
6.38
5.O9
25.59
ca0
6r.88
sR-9
core
Na20
6.00
7.60
5. 69
6 .t 5
t.Lt
d-zo
7.01
4.37
7.64
8.61
*zo
0.57
0.59
0.55
o.62
0.38
0.29
0.33
0.47
0.34
0.42
99.63
l-00.78
99.49
1o0.41
99.99
100.46
99.08
100.70
Tota1
Structural
si
A1
Formula
2.530
L.479
4. 009
based
2.665
1.336
4. OOl
-
Ti.
Fer
0.015
M9
0.004
99.O4
LOO.22
on 8 oxygens
2.52L
1.481
4 l,O2
z-tI)
r.434
4. O09
2.6LA
r.380
3.998
2.705
z.oz)
2.736
2.6A2
2.742
L.293
3.998
I.38I
4. O05
L.2t6
4.OO2
L.324
4.006
1.258
4.000
0.001
0.009
0.004
o.001
-
0.001
-
0.006
0.006
0.002
0.005
0. 005
0.005
Ca
o.424
U.JIJ
U.4bf,
o-404
0.370
0.280
0.360
0.250
0.307
0.24L
NA
0. s25
0.653
0.499
0. 540
o.619
0.715
0.606
0.718
0.665
0.736
K
o.o33
1.001
0.033
I.008
0.032
1.001
0.036
0.986
0.021
r.012
0.0r7
r.0r7
o.0r9
0.990
0.027
r.001
0.019
0.991
0.o24
l-.008
Ab
53.5
65.4
50.1
5 5 .r
6r.3
70.6
61. 5
72.t
6 7. L
73.6
An
43.2
31.3
46.7
4L.2
Jb.b
zt.t
36.6
25.2
31.0
24.O
3.3
3.2
3.7
z.l
r.v
4.1
Or
* Total
3'.3
Fe as Feor
r*
See Fig.
1.9
4
as discrete grains (external granule "exsolution")
or as thin lamellae in the host grains (Buddington
and Lindsley, 1964). The overall result is an increasein the ilmenite and magnetitecomponentsof
the rhombohedraland spinel seriesphases,respectively, causing a counterclockwise rotation of their
tielines on the TiO2-FeO-Fe2O3diagram.
While hematitelamellae in ilmenite and pleonaste
lamellae in -magnetite were incorporated in the
defocussedbeam ED microprobe analyses, oxyexsolution lamellae in magnetite and magnetite reduction-exsolutionlamellae in ilmenite were avoided, and as a result, many of the compositionsof
coexisting Fe-Ti oxides shown in Figure 5 representthosewhich resultedafter subsolidusreconstitution. Attempts to re-integrate the oxidation or
reduction lamellae to obtain primary compositions
would not be possible because of the unknown
extent of previous external granule "exsolution"
(Buddington and Lindsley, 1964).
Despite the effects of subsolidus recrystallization, there is a perceptible compositional trend
among the Fe-Ti oxide minerals. In the anorthositesand other early cumulates,ilmenites tend to be
richer in hematite component, and coexisting spinels richer in magnetitecomponent, comparedto the
oxides in later cumulates and mafic gabbro dikes.
Similar relationships in other anorthosite massifs
(Hargraves,1962;Anderson, 1966;Duchesne,1972)
have been interpreted as an indication of progressive reduction of the melts during differentiation.
However. becauseof the effects of counter-clockwise rotation of Fe-Ti oxide tie lines in responseto
reaction(1) duringcooling,as previouslydiscussed,
is dominatedby
rocks whosebulk oxide assemblage
ilmenite [ilm/(ilm + mag) = 0.9] will tend to retain
near-primary ilmenite compositions, but magnetite
compositions will be widely different (much more
Ti-poor) than primary values. Conversely, magnetite-rich rocks [ilm/(ilm + mag) < 0.75] will contain
ASHWAL : OXIDE.RICH DIFFERENTIATES, MARCY ANORTHO SITE
threefold increase in V2O3 among magnetites, to a
maximum of 3.0 wt.Voin sample SA-5 (Table 6).
Discussion
.
Mog
Fig. 5 Average ED microprobe compositions of Fe-Ti oxide
minerals from the Marcy massif plotted in terms of TiOrFeOFe2O3.Dots represent bulk compositions of oxide assemblages
givenfrom modes.Enlargedarearepresentsshadedareain inset.
magnetiteswith compositions closer to primary
valuesthan coexistingilmenites. This relationship
is supportedby the distinct correlationbetweenFeTi oxide compositionsand the modal proportionsof
ilmeniteand magnetite(Table5, Fig. 5). The geothermometer-oxygen barometer of Buddington and
Lindsley (1964) thus cannot be used to obtain
estimatesof primary temperature andfOz.
WD microprobe analyses of coexisting Fe-Ti
oxidesfrom a layered cumulatesequencewhich is
gradationalupwards from a basal oxide-rich pyroxenite to leuconorite are given in Table 6. There is a
slight but progressive increase in hematite component in ilmenite and a decreasein ulvospinelcomponent in magnetite stratigraphically upward in this
sequence,but these may not reflect primary variations, as discussedabove. However, minor elements may be used as reliable indicators of primary
fractionationtrends (Himmelbergand Ford, 1977).
Both magnetitesand ilmenites from this sequence
show a progressivedecreasein Al2O3, MnO, and
MgO, and a progressiveincreasein Cr2O3and V2O3
with stratigraphic height. Noteworthy is the nearly
The mafic cumulates and dikes of the Marcy
massif are interpreted as late-stage differentiates
from a magma that had undergone extensive fractionation of plagioclase, followed sequentially by
crystallization and accumulation of pyroxenes, FeTi oxides, apatite, and Fe-rich olivine. Crosby
(1968)and Buddington (1969)have proposed flow
differentiation models by which plagioclasecrystals
migrated to the central regions of ascendingmagmas,to form a core zone of relatively pure, coarsegrained anorthosite. The liquid in the outer regions
became progressively enriched in mafic constitutents, and subsequentlycrystallizedas a somewhat
finer-grainedsheath of more mafic material. Local
incorporation of earlier-formedplagioclaseaccumulations by this mafic-enrichedliquid could account
for the block structures common in most anorthosite massifs (Isachsen, 1969).Although flow differentiation appearsto be a satisfactory mechanismto
account for the formation of the pure anorthosites
and their restriction to the central regions of the
massifs.the nature of the conformable mafic-ultramafic layers describedhere suggeststhat they were
formed by gravitational accumulation. It is suggested here that differentiation by gravitational means
became the dominant mechanism subsequentto
emplacementof the partially solidified melts. In the
case of the Marcy massif, final emplacement is
inferred to have taken place at a stage in the
crystallization history at least as early as the time
when the pyroxenes first precipitated, allowing the
formation of the early, Mg-rich ultramafic cumulates in a relatively static environment. The occurrence of the mafic facies in the border zones of the
massif, and hence structurally above the earlierformed anorthosites of the core zone. could be
interpreted as evidence that plagioclase sank, at
least in the early stagesof crystallization.
The relatively minor volume of mafic rocks compared to the associatedanorthositesis consistent
with a rather feldspathic parental magma composition, possiblyin the gabbroicanorthosite(77.5-90%
normative plagioclase) range. The exact composition of the parental magma is difficult to determine
becausean accurate field estimation of the volume
of mafic rich rock is not possible. Although highly
aluminous magmas such as gabbroic anorthosite
ASHWAL : OXIDE.NCH DIFFERENTIATES, MARCY AN ORTHOSITE
25
Table 5. Microtextures and compositions (EDA) of coexisting Fe-Ti oxides
S a m pel
No.
RockType
M o d a lC o m p o s i t i o n s
I l m / ( I l m + M a)g
M ic r o t e x t u r e s
I I flrrMag
tional
Cornposi
Range
Average
Comp.
H*ro.q - r5.6
H"tr
No. of
Analyses
CUMULATES:
5654'A
Anorthosite
SL-l'lY
O x i d er i c h
pyroxenite
S L - 'llA
Leuconorite
SA-3D
SA-3A
SA-5
!f'1-568
SA-l5A
5683d
7
6
t.a
0 .9 0
I'lm:
Mag:
?3.1
0.82
coarse hem. lamellae,
depleted near mag.
Mag: honogeneous
H"tIr.l - r7.0
3.3
UtPo.
o
H"tr
4.5
u t P o . o-
UtPo.2
8
4.5
0 .9 0
coarse hemlamellae,
depleted near mag.
Mag: homogeneous
H e m g . -g
Hem
, ^.
tz.a
6
u ' P o . o- 0 . 3
uspO..l
b
O x i d er i c h
pyroxenite
0 .6 5
Ilm: homogeneous
Mag: 3 sets of pleonaste
lamellae, discrete
p l eonaste
H " t z .9 - 4 . 0
Hem- .
J
u t P r a . S- r 4 . g
u'P4
r .5
L e u c o n o trei
0.65
Ilm:
Mag:
Hem^
,
z.t - +.t
u t P o . 6- q . o
Hem^ ^
I t m : fine hem. lamellae,
depleted near mag.
M a g: homogeneous
H"t8.3- r2.4
"'Po.o - 3.0
H"tr
coarse hem. lamellae,
depleted near mag.
M a g : few tiny pleonaste
I amelI ae
H"t]g.2 - r8.2
H"r.l5. 5
utPt .l
.
Leuconorite
oxide rich
mafic aabbro
O x i d er i c h
gabbro
1.0
'r5.
3
'r3.6
0xide rich
ll.7
mafic troctol ite
0.90
0 .9 4
0.90
0.72
coarse hem. lanellae
homogeneous
Ilm:
Iln:
nomo9ene0us
few pleonaste lamel lae,
few i lm ox'id, lamellae
I lm:
u t P o . 3-
0.5
r6.2
r.5
0.9
Uspo.8
Hen,
^ .
tv.o
Ilm: homogeneous
Mag: 2
ee
t sl l o
' l asm
a fe ,t i n y p l e o n a s t e
a b u n d a n ti l m
oxid. IamelIae
Hem,,
u t P r o .-o2 r . 5 u t P 5r . 7
Ilm: homogeneous
Mag: fine pleonaste
I amelI ae
Htts.B- 9.r
u " P o . 2- ' l . 5
H"t7.
-
5-z
6
u s p l. 9
n e m l 7 . 9- 2 0 . 9
I l m : c o a r s eh e ml a m e l l a e ,
d e p l e t e dn e a r m a g .
M a g : f i n e p l e o n a s t el a m e l l a e u t P o . o- 0 . 8
Heml r
t.r
4
utp0. .l
9
9
6
6
I
3
l4
D IK E S :
Ml"l-l4A
0xide rich
mafic norite
13.5
0.8t
5
UtPo.
B
9
lz
I't'l-6
0xide rich
mafic Aabbro
0.72
I lm: homogeneous
Mag: few pleonaste
I a m e l I a e ,f e w i l m
oxid. Iamel'lae
H"tr.o- 2.6
" ' P 2 1 . 4- 2 2 . 3
H"rl.8
u t p 2 2o.
2
3
SR-g
0xiderich
gabbro
0.s4
Ilm: homogeneous
Mag: feli p]eonaste
l a m ella e
H"to.8- 2.4
utPz6.g - 27.7
H"tl.8
4
u r p 2 7I .
3
have been suggestedas much as forty years ago
(Buddington,1939),their existencehas been questioned and their possiblemodesof origin have been
debated.The recent discovery of sills and dikes of
gabbroic anorthosite composition provides the first
direct evidence for the existence of these magmas
(Crosby,1969;Husch et a|.,1975; Isachsenet al.,
1975;Simmons,et a|.,1980;Wiebe,1979),although
their origin remains somewhatof a problem. Emslie
(1978)has proposed that the aluminous magmas
which gave rise to massif-typeanorthositeswere
produced by high-pressurefractionation of olivine
and orthopyroxene from mantle-derived olivine
tholeiite magmaswhich accumulatednear the base
of the continental crust. The aluminous melts were
then emplacedinto higher levels in the crust, where
they precipitatedonly plagioclaseat the liquidus.
This model accounts for the relatively high fe* of
the anorthositicrocks (Emslie, 1973;Simmonsand
Hanson, 1978),and is consistentwith recent SmNd isotopestudiesof the Adirondack Marcy massif
which indicate an ultramafic source area for the
26
AS HWAL : OXIDE-RIC H DIFFERENTIATES. MARC Y ANORTHO SITE
Table 6. Representative WD microprobe analyses of Fe-Ti
oxides
I I m e ntie s
lilagneti tes
SA-3A
SA-5
5A-3D
SA-3A
sA-5
si0.
0.03
0.07
0.05
o.o2
0.01
0.01
T i0 ^
3.34
0.67
0.05
52.61
50.92
47.86
A l^ 0 ^
l.5t
0.67
0.54
0.06
0.04
0.03
cr^0^
0.37
t.t6
3.84
0.01
0.03
0.15
8.34
Fe^0.
58t5
61.83
6l.12
2.22
3.03
Fe0
33.65
3 0 .B B
3t.31
42.77
4?.23
i4n0
0.07
0.46
0.27
0.26
0 .1 8
0.02
0.07
0.01
ltlS0
0.01
2.30
I .85
0.94
v^0.*
t.r8
2.10
3.00
0 . 13
0.22
0.59
Total
98.48
97.47
99.93
I00.58
98.60
99.29
4l.ll
S t r u c t u r a l f o r m u l ab a s e do n 3 c a t i o n s ( m a q n e t i t e )o r 2 c a t i o n s ( i l m e n i t e )
Si
0.001
0.003
0.002
0.000
0.000
0.000
Ti
0.097
0.020
0.001
0.977
0.968
0.91?
AI
0 069
0.031
0.024
0.002
0.001
0.001
Cr
0 . 0 ]t
't.689
0.036
0 . 11 6
0.000
0.001
0.003
1.325
0.041
0.058
0.159
Fe-
r.085
1.014
1.761
,l.003
0.882
0.892
0.871
Mn
0 002
0.00]
0.000
0 . 0 10
0.006
0.006
!19
0 010
0.004
0.001
0.085
0.070
0.036
0.036
0.066
0.092
0.003
0.004
0 . 0 12
3.000
3 000
3.000
l 4 a s( I l n ) 8 9 . 7 0
usp (Hem)10.30
97.BB
99 84
2.12
0 t6
Fe"'
2.000 2.000
2.000
(e7.71) (e6.87) (et.63)
(2.2s) (3.r3) (8.37)
* V 2 0 3a b u n d a n c ewse r e c o r r e c t e d f o r T i K B o v e r l a p b y s u b t r a c t i n ga n
e m p i r i c a l l y d e t e r m i n e dl % o f T i K o c o u n t s f r o m v K o c o u n t s p r i o r t o f u r t h e r
natrix corrections.
melts parental to the anorthosites and related mafic
differentiates (Ashwal et al., 1980).
Acknowledgments
This studywasdonein partialfulfillmentof the requirements
for the degreeof Doctorof Philosophyat PrincetonUniversity.
LinFinancial
supportcamefromNesaGrantNGL-31-001-283,
coln S. Hollister,principalinvestigator,and from the DepartPrincetonUnivermentofGeologicalandGeophysical
Sciences,
sity. Some of the data were collected while the author held a
National ResearchCouncil Resident ResearchAssociateshipat
r.rasA,/Johnson
SpaceCenter. A portion of the researchreported
in this paper was done while the author was a ResidentResearch
Scientistat the Lunar and Planetary Institute, which is operated
by the Universities SpaceResearchAssociation under Contract
No. N,c.sw-3389with the National Aeronautics and Space Administration. The manuscript was critically reviewed by R. B.
Hargraves,E. Dowty, G. Ryder, J. L. Warner, R. F. Emslie,
and S. R. Bohlen.Assistancein the field was providedby F. B.
Nimick, R. Warren, and M. A. Arthur. Maps were provided by
Y. W. Isachsenand P. R. Whitney of the New York State
Museumand ScienceService.C. G. Kulick, R. W. Brown and
D. W. Phelpsl provided assistance with the electron microprobes.This is L.P.I. ContributionNo. 447.
References
Anderson, A. T., Jr. (1966)Mineralogy of the Labrieville anorthosite, Quebec. American Mineralogist 51, 167l-l7 ll.
Anderson, A. T., Jr. (1969)Massif type anorthosite:a widespread Precambrian igneous rock. In Origin of Anorthosite
and Related Rocks, Y. W. Isachsen,ed., New York State
Museumand ScienceServiceMemoir 18,47-55, N. Y.
Ashwal, L. D. (1978a) Crystallization history of massif-type
anorthosite (abstract). EOS: Transactions of the American
GeophysicalUnion 59, 393.
Ashwal, L. D. (1978b)Petrogenesisof massif-typeanorthosites:
crystallizationhistory and liquid line of descentof the Adirondack and Morin complexes.Ph.D. thesis, Princeton University, 136pp.
Ashwal, L. D. and Seifert, K. E. (1980) Rare earth element
geochemistryof anorthosite and related rocks from the Adirondacks,New York, and other massif-typecomplexes.Bulletin of the GeologicalSociety of America 91, 105-107,659-684.
Ashwal,L. D., Wooden,J. L. and Shih,C. -Y. (1980)Nd and Sr
isotope geochronologyof the Marcy anorthosite massif, Adirondacks, New York (abstract).GeologicalSociety of America Abstractswith Programs,12, 380.
Bohlen,S. R. and Essene,E. J. (1978)Igneouspyroxenesfrom
metamorphosedanorthosite massifs. Contributions to Mineralogy and Petrology 65,433-442.
Buddington, A. F. (1939) Adirondack igneous rocks and their
metamorphism. Geological Society of America Memoir 18,
343 pp.
Buddington, A. F. (1953)Geology of the SaranacQuadrangle,
New York. New York State Museum Bulletin 346, 100pp.
Buddington, A. F. (1960)The origin ofanorthosite reevaluated.
RecordsGeological Survey of India 86, Paft 3, 421-432.
Buddington, A. F. (1969) Adirondack anorthosite series. In
Y. W. Isachsen, ed., Origin of Anorthosite and Related
Rocks,New York StateMuseumand ScienceService,Memoir 18,215-231.
Buddington,A. F. and Lindsley, D. L. (1964)Iron titanium
oxide mineralsand synthetic equivalents.Journal ofPetrology
5,310-357.
Crosby, P. (1968) Igneous differentiation of the Adirondack
anorthositeseries.XXXIII International GeologicalCongress
2, tt-48.
Crosby, P. (1969) Petrogenetic and statistical implications of
modal studiesin Adirondack anorthosite. In Y. W. Isachsen,
ed., Origin of Anorthosite and RelatedRocks, New York State
Museumand ScienceServiceMemoir 18, 289-303.
Davis, B. T. C. (1971)Bedrock geologyof the St. RegisQuadrangle, New York. New York State Museum and Science
Service, Map and Chart SeriesNumber 16, 34 pp.
deWaard, D. (1965) A proposed subdivision of the granulite
facies. American Journal of Science 263,455-461.
deWaard,D. (1970)The anorthosite-charnockitesuite ofrocks of
Roaring Brook valley in the Eastern Adirondacks (Marcy
Massi0. American Mineralogist 55, 2063-2075.
Duchesne, I.-C. (1972) Iron-titanium oxide minerals in the
Bjerkrem-Sogndalmassif, southwestern Norway. Journal of
Petrology13,57-81.
Emslie, R. F. (1973)Some chemical characteristicsof anorthosite suites and their significance.Canadian Journal of Earth
Sciences10,54J1.
Emslie, R. F. (1975) Nature and origin of anorthositic suites.
GeoscienceCanada2, 99-104.
ASHWAL : OXIDE.RICH DIFFERENTIATES. MARC Y ANORTHO SITE
Emslie, R. F. (1978)Anorthosite massifs,rapakivi granites, and
late Proterozoic rifting of North America. precambrian Research7,6l-98.
Green,T. H., Brunfelt, A. O. and Heier, K. S. (1969)Rare earth
element distribution in anorthosite and associatedhigh grade
metamorphicrocks, Lofoten-Vesteraalen,Norway. Earth and
Planetary ScienceLetters 7, 93-98.
Green,T. H., Brunfelt,A. O. and Heier, K. S. (1972)Rareearth
elementdistribution and K/Rb ratios in granulites,mangerites,
and anorthosites,Lofoten-Vesteraalen,Norway. Geochimica
et CosmochimicaActa 36,241-257.
Gross, S. O. (1968) Titaniferous ores of the Lake Sanford
district, New York. In J. D. Ridge, ed., Ore Deposits of the
United States, American Institute of Mining, Metallurgical,
and PetroleumEngineers,N. Y., 140-154.
Hargraves,R. B. (1962)Petrology of the Allard Lake anorthosite
suite, Quebec. In, Petrologic Studies, Geological Society of
AmericaBuddingtonVolume, 163-189.
Himmelberg, G. R. and Ford, A. B. (1977)Iron-titanium oxides
of the Dufek intrusion, Antarctica. American Mineralogist 62,
623-633.
Husch,J. M., Kleirispehn,K. and Mclelland, J . (1975)Anorthositic rocks in the southern Adirondacks: basementor nonbasement? Geological Society of America Abstracts with
Programs7, 78.
Isachsen,Y. W., Mclelland, J. and Whitney, p. R. (1975)
Anorthosite contact relationshipsin the Adirondacks and their
implicationsfor geologichistory. GeologicalSociety of America Abstracts with Programs7, 78.
Isachsen,Y. W. and Fisher,D. W. (1970)Geologicmap of New
York, Adirondack Sheet. University of the State of New
York, The State Education Department, Albany, N. Y.
Isachsen,Y. (1969)Origin of anorthosite and related rocks-a
summarization.In, Y. W. Isachsen,ed., Origin of Anorthosite
and Related Rocks, New York State Museum and Science
ServiceMemoir 18, 435-445.
Isachsen,Y. W. and Moxham, R. L. (1969)Chemical variation
in plagioclasemegacrysts from two vertical sections in the
main Adirondack metanorthositemassif. In, Y. W. Isachsen,
ed., Origin of Anorthosite and RelatedRocks, New york State
Museum and ScienceService, Memoir [E,255-265.
Jaffe,H. W., Robinson,P. and Tracy, R. J. (1975)Orientationof
pigeoniteexsolution lamellae in metamorphic augite: correlation with composition and calculated optimal phase boundaries. American Mineralogist 60, 9-28.
Kretz, R. (1963) Distribution of magnesiumand iron between
otthopyroxene and calcic pyroxene in natural mineral assemblages.Journal of Geology 71, 773-:785.
Martignole, J. (1974)MeOhanismof differentiation in the Morin
anorthositecomplex. Contributions to Mineralogy and Petrology 44,99-107.
Mclelland, J. M. and Whitney, P. R. (1977)The origin of garnet
in the anorthosite-charnockitesuite ofthe Adirondacks. Contributions of Mineralogy and Petrology 60, 16l-181.
Morse, S. A. (1969)Layered intrusions and anorthositegenesis.
In, Y. W. Isachsen,ed., Origin of Anorthosite and Related
Rocks, New York State Museum and ScienceService. Memoir 18. 175-187.
27
Morse, S. A. (1981)Kiglapait geochemistry III: potassium and
rubidium. Geochimica et CosmochimicaActa 45. 163-180.
Ohnstead,J. F. (1%9) Petrology of the Mineral Lake intrusion,
northwestemWisconsin.In, Y. W. Isachsen,ed., Origin of
Anorthosite and RelatedRocks, New York State Museum and
ScienceServiceMemoir 18, 149-161.
Philpotts,J. A., Schnetzler,C. C. and Thomas, H. H. (1966)
Rare earth abundancesin an anorthosite and a mangerite.
Nature 212, 805-806.
Ranson, W. A. (1981)Anorthosites of diverse magma types in
the PuttuaalukLake area, Nain complex, Labrador. Canadian
Journal of Earth Sciences18, 26-41.
Rose, E. R. (1969)Geology of titanium and titaniferous deposits
of Canada. Geological Survey of Canada Economic Geology
ReportNo. 25,177 pp.
Ross, M. R. and Huebner, J. S. (1975) A pyroxene geothermometer based on composition-temperaturerelationships of
naturally occurring orthopyroxene, pigeonite and augite. International Conference on Geothermometry and Geobarometry, Extended Abstracts, Pennsylvania State University,
University Park, PA.
Seifert, K. E. (1978)Anorthosite-mangeriterelations on Baker
Mountain, New York, Bulletin of the Geological Society of
America 89,245-250.
Seifert, K. E., Voight, A. F., Smith, M. F., and Stensland,
W. A. (1977)Rare earths in the Marcy and Morin anorthosite
complexes. Canadian Journal of Earth Sciences 14, 10331045.
Sen, S. K. (1959)Potassium content of natural plagioclaseand
the origin of antiperthites. Journal of Geology 67, 479495.
Simmons,E. C. and Hanson, G. N. (1978)Geochemistryand
origin of massif-type anorthosites. Contributions of Mineralogy and Petrology66, l19-135.
Simmons,E. C., Hanson, G. N. and Lumbers, S. B. (1980)
Geochemistryof the Shawmereanorthosite complex, Kapuskasingstructural zone, Ontario. PrecambrianResearchI l, 4371.
Smith, J. V. (1975) Some chemical properties of feldspar. In,
P. H. Ribbe, ed., Feldspar Mineralogy, Mineralogical Society
of America Short Course Notes 2. SM-18-SM-29.
Streckeisen,A. (1976)To each plutonic rock its proper name.
Earth ScienceReviews 12, l-33.
Subramaniam,A. P. (1956)Petrology of the anorthosite-gabbro
massat Kavadur, Madras, India. GeologicalMagazine93, No.
4,287-300.
Taylor, H. P., Jr., and Noble, J. A. (1969)Origin of magnetitein
the zoned ultramafic complexes of southeasternAlaska. In,
H. D. B. Wilson, ed., MagmaticOre Deposits,Monograph4,
Economic Geology Publishing Co., 209-230.
Wager, R. L. and Brown, G. M. (1968)Layered Igneous Rocks,
Oliver and Boyd, Ltd., Edinburgh, 588 pp.
Wiebe, R. A. (1979)Anorthositic dikes, southernNain complex,
Labrador. American Journal of Science279,394-410.
Manuscript received, January 9, l98l;
acceptedfor publication,August 3I, 1981.