Content uploaded by Darrell James Henry
Author content
All content in this area was uploaded by Darrell James Henry on Jan 16, 2014
Content may be subject to copyright.
American Mineralogist, Volume 90, pages 316–328, 2005
0003-004X/05/0203–316$05.00/DOI: 10.2138/am.2005.1498 316
INTRODUCTION
The concentration of Ti in biotite has long been considered
to be a function primarily of changing temperature conditions in
metamorphic rocks and has been suggested as a potential geo-
thermometer (Engel and Engel 1960; Kwak 1968; Robert 1976;
Dymek 1983; Patiño Douce 1993). However, the factors that
inß uence the incorporation of Ti in biotite are more multifaceted
than simple temperature change, and involve a relatively complex
interplay among temperature, pressure, biotite crystal chemistry,
and coexisting mineral assemblages (Guidotti et al. 1977, 1988;
Dymek 1983; Labotka 1983; Guidotti 1984; Tracy and Robinson
1988; Guidotti and Sassi 2002; Henry and Guidotti 2002).
Experimental investigators have examined how temperature,
pressure, and composition control the amount of Ti in biotite.
The temperature effect appears to be the most inß uential. In
phlogopite, the incorporation of Ti is relatively dramatic at
high temperatures, particularly higher than 800 °C. For in-
stance, Robert (1976) found that Ti solubility is relatively low,
0.07 Ti atoms per formula unit (apfu), at 600 °C and 1 kbar, but
increases signiÞ cantly to 0.2 Ti apfu at 800 °C and 1 kbar, and
to 0.7 Ti apfu at 1000 °C and 1 kbar. Increasing pressure has the
opposite effect, with Ti concentrations decreasing substantially
with increasing pressure (Forbes and Flower 1974; Robert 1976;
* E-mail: dhenry@geol.lsu.edu
The Ti-saturation surface for low-to-medium pressure metapelitic biotites: Implications
for geothermometry and Ti-substitution mechanisms
DARRELL J. HENRY,1,* CHARLES V. G UIDOTTI,2 AND JENNIFER A. THOMSON3
1Department of Geology and Geophysics, Louisiana State University, Baton Rouge, Lousianna 70803, U.S.A.
2Department of Geological Sciences, University Of Maine, Orono, Maine 04469, U.S.A.
3Department of Geology, Sci 130, Eastern Washington University, Cheney, Washington 99004, U.S.A.
ABSTRACT
The Ti content of biotite can serve as a geothermometer for graphitic, peraluminous metapelites
that contain ilmenite or rutile and have equilibrated at roughly 4–6 kbar. The relationship between Ti-
content, temperature, and Mg/(Mg + Fe) value was calibrated empirically using an extensive natural
biotite data set (529 samples) from western Maine and south-central Massachusetts in combination
with the petrogenetic grid of Spear et al. (1999). The calculated Ti-saturation surface is curved such
that for a given Mg/(Mg + Fe) value, Ti concentration increases as a function of temperature in a
nonlinear fashion, and for a given temperature Ti concentrations decrease with an increase in Mg/(Mg
+ Fe). The Þ t to the Ti-saturation surface can be reformulated as the geothermometric expression: T =
{[ln(Ti) – a – c(XMg)3]/b}0.333, in which T is temperature in degrees Celsius, Ti is the number of atoms
per formula unit (apfu) normalized on the basis of 22 O atoms, XMg is Mg/(Mg + Fe), a = –2.3594, b
= 4.6482 × 10–9 and c = –1.7283. The calibration range for this expression is XMg = 0.275–1.000, Ti =
0.04–0.60 apfu, and T = 480–800 °C. Precision of the Ti-in-biotite geothermometer is estimated to be
±24 °C at the lower temperature range and improves to ±12 °C at higher temperatures. Application of
the Ti-in-biotite geothermometer to ilmenite- or rutile-bearing, graphitic, peraluminous metapelites
equilibrated at 3–6 kbar is generally consistent with independent temperature determinations, but with
some deviations that represent local reequilibration. Consequently, the Ti systematics in biotite can
also serve as the basis of a very sensitive indicator of chemical equilibrium, or lack thereof. Applica-
tion of the geothermometer to metapelites not containing the requisite mineral assemblages can lead
to minor-to-signiÞ cant errors in estimated temperatures.
Biotite Ti-substitution mechanisms are controlled by several factors. Based on the biotite calibra-
tion data set, magnesian biotites (XMg > 0.65) incorporate Ti in accordance with the exchange vector
TiAl2R–1Si–2, where R is the sum of the divalent cations Mg + Fe + Mn. This substitution mechanism
is primarily a response to misÞ t of the octahedral and tetrahedral layers in magnesian biotites. Inter-
mediate biotites (XMg <0.65), particularly at higher temperatures, exhibit enhanced Ti concentrations,
most consistent with the Ti-deprotonation TiO2R–1(OH)–2 exchange vector. Dominance of Ti-deprot-
onation substitution is largely a function of reduction of H2O activity at higher metamorphic grades.
Supplementary biotite data from metaluminous amphibolites and maÞ c granulites, metamorphosed
isothermally with variable H2O activities, reveal that low-Al biotite incorporates signiÞ cantly higher
concentrations of Ti relative to peraluminous biotite as a result of a combination of the exchange vec-
tors TiO2R–1(OH)–2 and RSiAl–2 substituting in roughly an 8:1 ratio.
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY 317
Arima and Edgar 1981; Tronnes et al. 1985). Robert (1976) noted
that at 1000 °C, a pressure increase from 1 to 7 kbar results in a
decrease in Ti concentration in phlogopite from 0.7 Ti apfu to 0.2
Ti apfu. Early experiments also demonstrated a compositional
inß uence such that Ti content in biotite generally increases with
an increase in Fe content (Arima and Edgar 1981; Abrecht and
Hewitt 1988). Combinations of these parameters can have a pro-
found effect on Ti content in biotite. For instance, Patiño Douce
and Johnston (1991), examining Ti concentrations in intermediate
Fe-Mg biotite from a series of partial-melting experiments of
natural peraluminous metapelites, found that at 825–975 °C and
7–13 kbar, Ti increases with temperature in a non-linear fashion.
Based on these experiments, Patiño Douce (1993) calculated the
pressure effect on Ti concentrations in intermediate Fe-Mg (XMg
= 0.5) biotite coexisting with orthopyroxene, K-feldspar, and
quartz. He estimated that a pressure increase from 5 to 15 kbar
causes a decrease of Ti by 0.24 apfu at 900 °C, by 0.10 apfu at
800 °C, and by 0.02 apfu at 700 °C.
Despite these previous studies, it is difÞ cult to apply ex-
perimental Þ ndings to the spectrum of P-T conditions, bulk
compositions, and mineral assemblages found in metamorphic
and igneous rocks. A better understanding of this complex re-
lationship is needed. An important result of the present study
is the illustration of how careful investigations of biotite in
well-constrained natural systems can further elucidate aspects
of the experimental studies and provide an independent means
to investigate additional factors that inß uence Ti concentrations
in biotite.
Henry and Guidotti (2002) studied Ti concentrations in bio-
tite from metapelitic rocks in a geologically and petrologically
well-characterized terrane associated with the Acadian Orogeny
of western Maine. They investigated biotite in metapelitic rocks
in the garnet through sillimanite-K-feldspar zones at roughly
isobaric conditions of about 3.3 kbar, based on the petrogenetic
grid of Spear and Cheney (1989). Their extensive natural biotite
data set (>450 analyses) from thoroughly equilibrated metapelitic
schists of western Maine exhibits systematic variations in Ti
contents over a continuum of metamorphic grade and biotite
mineral chemistry. To limit the number of competing substitu-
tions in biotite, Henry and Guidotti (2002) focused on samples
with selected mineral assemblages. They studied pelitic schists
that contained quartz, aluminous minerals (chlorite, staurolite
or sillimanite), Ti minerals (ilmenite or rutile), and graphite.
In such schists, Si, Al, and Ti in biotite are near or at satura-
tion levels for a given temperature, and the effect of variable
bulk composition is limited. The occurrence of graphite in the
metapelitic schists restricts the metapelites to low and fairly
constant fO2. Consequently, biotites contain low and roughly
constant amounts of Fe3+ (~12% of Fetotal) (Guidotti and Dyar
1991). At the same time, the biotites studied by Henry and
Guidotti (2002) have a relatively wide range of Mg/(Mg + Fe)
ratios controlled by sulÞ de-silicate interactions (Henry 1981;
Guidotti et al. 1988). Henry and Guidotti (2002) estimated the
equilibration temperature for each sample by noting its position
relative to mapped isograds. Isograd temperatures themselves
were derived from the Spear and Cheney (1989) petrogenetic
grid for metapelites assuming 3.3 kbar. Based on the inferred
temperature and biotite chemistry, they were able to calculate a
3.3 kbar Ti-saturation surface for natural biotite as a function of
temperature and Mg/(Mg + Fe) values.
One implication of the Henry and Guidotti (2002) investiga-
tion is that a Ti-in-biotite geothermometer can be formulated
from the surface-Þ t expression of the Ti-saturation surface. How-
ever, this geothermometer is strictly valid only for biotite from
metapelites that: (1) have mineral assemblages similar to those of
the calibration samples; (2) equilibrated at roughly 3.3 kbar; and
(3) fall within the range of biotite compositions and temperatures
of the calibration samples. Because of the ubiquitous nature of
biotite in many rock types, further development of this potential
single-mineral thermometer is warranted. So, this paper consid-
ers geothermometric aspects of the biotite Ti-saturation surface
in a more general sense, deÞ nes compositional and textural
manifestations of biotite disequilibrium or local equilibrium,
and examines likely Ti-substitution mechanisms in biotite in a
crystallochemical and petrologic context.
BIOTITE DATA SETS
The biotite data considered in this study include the original
Henry and Guidotti (2002) data set as well as supplementary
data from Maine and south-central Massachusetts. Biotite data
from Massachusetts are especially useful because they developed
at higher temperature and slightly higher pressure conditions.
For all biotites, we redetermined metamorphic pressures and
temperatures using the updated petrogenetic grid of Spear et
al. (1999). The updated petrogenetic grid was used because
it provides improved consistency among a variety of natural
metapelitic mineral assemblages (Pattison et al. 2002).
Maine data set
The biotite-bearing samples from western Maine are primar-
ily from metamorphosed Ordovician-Devonian strata that were
extensively metamorphosed in the Siluro-Devonian Acadian
Orogeny (Osberg et al. 1985; Guidotti 1989). The Acadian meta-
morphism involves regionally developed recrystallization with
grades ranging from subgreenschist-facies to upper amphibolite-
facies (Guidotti 1989). The polymetamorphic history includes
an initial greenschist-facies regional metamorphism (M1), and
regional/contact metamorphic overprints (M2 and M3) that
are largely postkinematic. Further details of the metamorphic
history are presented in Guidotti (1970a, 1970b, 1974, 1985,
1989), Henry (1981), Holdaway et al. (1982, 1988), deYoreo et al.
(1989), Pressley and Brown (1999); Henry and Guidotti (2002),
and Johnson et al. (2003). Of special importance is the observa-
tion that in the speciÞ c area considered by Henry and Guidotti
(2002) (see their Fig. 1), the M3 metamorphism is the dominant
metamorphic overprint and that minerals in the metapelites across
that region thoroughly equilibrated under nearly isobaric condi-
tions (Guidotti 1989; Henry and Guidotti 2002).
These western Maine samples come primarily from the
Rangeley, Oquossoc, Rumford, and Old Speck Mtn. 15' quad-
rangles. In this region, the isograd pattern in metapelitic schists
roughly follow the outlines of the intruding pluton borders, and
developed during the M3 metamorphic overprint. Based on the
isograds, Guidotti et al. (1988) and Guidotti and Dyar (1991)
deÞ ned 4 major zones (garnet, staurolite, sillimanite, and silli-
manite-K-feldspar) (Table 1). They divided several of the zones
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY318
into lower, middle, and upper portions based on progressive
changes in textures and compositions of muscovite, biotite, and
chlorite from speciÞ c limiting assemblages (Guidotti et al. 1988,
1991). This approach resulted in 10 mappable metamorphic zones
that were used to assign temperatures to individual samples by
calibrating the isogradic reactions against the Spear et al. (1999)
petrogenetic grid at 4 kbar and interpolating temperatures based
on the relative position with metamorphic zones (Table 1).
Besides the Henry and Guidotti (2002) data, literature
sources (see below) provided additional metapelitic biotite data
(29 analyses) from south-central Maine so we could extend the
temperature and compositional range of our study. A subset of
biotite data came from Ferry (1981), who examined graphitic
sulÞ de-rich schists that equilibrated under similar pressure condi-
tions to those of western Maine. The data of Ferry include several
Mg-rich biotites from pelites similar to those of the western
Maine samples, but for lower grade (biotite and garnet zone)
conditions (480–520 °C). Additional data came from Dutrow
(1985), who studied biotite to upper sillimanite zone metapelites
that have appropriate mineral assemblages, and include some
relatively Fe-rich biotites [Mg/(Mg + Fe) = 0.275].
South-central Massachusetts data set
Biotite data (50 analyses) from south-central Massachusetts
broaden our study to higher temperatures of 680–800 °C (Tracy
1978; Tracy and Robinson 1988; Peterson 1992; Thomson 1992,
2001). Several of the metapelite samples exhibit sulÞ de-silicate
reactions, thereby extending the biotite range to more magnesian
compositions [Mg/(Mg + Fe) = 0.43–1.00] (Tracy and Robin-
son 1988). The Massachusetts samples are from the Silurian
and Devonian metasedimentary rocks of the Kearsarge-Central
Maine Synclinorium of south-central Massachusetts that were
intensely deformed and metamorphosed during the Acadian
Orogeny. Pelitic gneisses, cordierite-garnet pegmatites, and
leucocratic garnetiferous melt segregations are found in this
upper amphibolite to granulite-facies terrane. The metamorphic
zones include the upper sillimanite, lower sillimanite-K-feldspar,
upper sillimanite-K-feldspar, muscovite melt, and biotite-garnet-
cordierite melt zones (Table 1). The latter zone is a consequence
of a discontinuous biotite dehydration melting reaction (DM2).
The Massachusetts biotite data were used for calibration purposes
only if the biotite data were derived from graphitic, peraluminous,
and Ti-saturated metapelitic rocks similar to the western Maine
metapelites. Additionally, because of the possibility of signiÞ cant
reequilibration associated with ß uid released from local crystal-
lization of melts, biotite data were restricted to samples that do
not exhibit signiÞ cant intergranular compositional heterogeneity
among biotite grains in the same sample. The most recent pres-
sure estimate for the peak metamorphism is 6–7 kbar following
a counterclockwise P-T path (Thomson 2001). We estimated
peak metamorphic temperatures using the petrogenetic grid of
Spear et al. (1999) assuming a pressure of 6 kbar.
RESULTS
Ti-saturation surface
To obtain the optimal Þ t of the biotite data, we normalized
the biotite formulae on a 22 O-atom basis assuming all Fe as
Fe2+ to calculate Ti atoms per formula unit (apfu). We then Þ t
the entire data set, containing 529 biotite analyses, to equations
that relate Ti in peraluminous biotites at 4–6 kbar to T °C and
Mg/(Mg + Fe) values. More than 500 candidate surface-Þ t equa-
tions were tried using the TableCurve3D software. All data were
given equal weighting. Candidate Þ t equations were sorted by the
F-statistic criterion (ratio of the mean squared regression to the
mean squared error) to optimize the surface Þ t while maintain-
ing simplicity of form of the equation. Using this criterion, the
optimal surface-Þ t equation is:
lnz = a + bx3 + cy3 (1)
where x = T °C, y = Mg/(Mg + Fe), and z = Ti (apfu).
FIGURE 1. Titanium saturation surface for natural peraluminous
biotites at 4–6 kbar. Perspective view of the surface highlights the
inß uence of Mg content on Ti. Vertical lines connect individual data
points above and below the Þ tted surface. This biotite data set does not
include values of Ti > 0.6 apfu and Mg/(Mg + Fe) < 0.275.
TABLE 1. Temperature constraints for graphitic, aluminous, biotite-
bearing metapelite samples from western Maine and
south-central Massachusetts
Zone no. Zone Isograd T °C T °C
name reaction (4 kbar)* (6 kbar)*
1 Lower Grt (495)
2 Middle Grt (515)
3 Upper Grt (535)
Grt + Chl = St + Bt 545
4 Lower St (555)
5 Upper St (570)
6 Transition St + Chl = Bt + Sil 580 615
7 Lower Sil (600)
St = Grt + Bt + Sil 620 655
8 Upper Sil (640) (665)
9 Lower Sil-Kfs Ms = Kfs + Sil 660 680
10 Upper Sil-Kfs (685) (705)
DMZ† no. Zone name Dehydration melt reaction
DM1 Ms melt Ms + Ab = Sil + Kfs + melt 650 650
DM2 Bt-Grt-Crd melt Bt + Sil = Grt + Crd + melt 715 775
DM3 Bt-Opx-Crd melt Bt + Gr t = Opx + Crd + melt 795 825
* Temperatures are based on locations of isograd reactions and dehydration
melting reactions at 4 and 6 kbar calibrated against the Spear et al. (1999)
petrogenetic grid. The temperatures in parentheses represent the estimated
temperature in the middle of each zone. Abbreviations: Ab = albite; Bt = biotite,
Chl = chlorite, Crd = cordierite, Grt = garnet, Kfs = K-feldspar, Ms = muscovite,
Sil = sillimanite and St = staurolite.
† Dehydration melting zone number. This represents the temperature of major
discontinuous dehydration melt reactions in metapelitic rocks as defi ned by
Spear et al. (1999).
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY 319
Table 2 gives values for a, b, and c and several statistical
parameters for the Þ t. Figure 1 is a three-dimensional projection
of the data and Þ tted surface. The Þ t to the data is excellent, hav-
ing a correlation coefÞ cient (r2) of 0.924. The additional data
and updated temperature estimates used in this study resulted
in a signiÞ cantly improved Þ t relative to the previous results
of Henry and Guidotti (2002) who obtained an r2 of 0.866 for
the Þ t. An encouraging aspect of this Þ t is that the Ti-saturation
surface, although being derived from natural data, is consistent
with the general Ti trends observed in the biotite experiments
discussed earlier.
Figure 2 shows isotherms projected on a Ti vs. Mg/(Mg + Fe)
diagram for biotite. This projection provides an instructive way
of viewing the inß uence of temperature on biotite chemistry in
Ti-saturated metapelites. Isotherms are relatively closely spaced
at temperatures below about 600 °C, and at more magnesian
compositions. Because our data only covers Ti values ranging
from 0.04 to 0.60 apfu, Mg/(Mg + Fe) values from 0.275 to 1.000,
temperatures of 480 to 800 °C, and pressures of 4 to 6 kbar,
extrapolation of isotherms beyond these ranges may introduce
signiÞ cant error. By calculating a temperature using Equation 2
(below) for each of the calibration data points and determining its
deviation relative to the Þ tted saturation surface, we are able to
estimate the accumulated uncertainties for different segments of
the surface. The calculated temperatures are within ±24 °C of the
saturation surface in the range of 480–600 °C, ±23 °C in the range
600–700 °C, and ±12 °C in the range 700–800 °C. The standard
deviations in the different ranges are considered measures of
precision of the Þ t of the surface, being accumulated uncertainties
associated with analytical precision and assignment of correct
temperature and pressure to individual samples. Although the
accuracy of the Ti-saturation surface is related to the quality of
the Þ tted surface, it is principally a function of the accuracy of
the positions of the isogradic reactions in the Spear et al. (1999)
petrogenetic grid. However, slight shifts in the locations of the
isogradic reactions will not have signiÞ cant effects on the general
shape of the biotite Ti-saturation surface.
Single mineral Ti-in-biotite geothermometer
Using Equation 1, we can derive a single-mineral Ti-in-biotite
geothermometer for peraluminous biotites in metapelites equili-
brated at 4–6 kbar. Temperatures can be determined either by
plotting biotite Ti and Mg/(Mg + Fe) values on Figure 2 or by
calculating temperatures directly from the expression:
T = {[ln(Ti) – a – c(XMg)3]/b}0.333 (2)
where T is temperature in °C, Ti is the apfu normalized to 22 O
atoms, XMg is Mg/(Mg+Fe), and the a, b, and c parameters are
given in Table 2. This expression is valid in the range XMg =
0.275–1.000, Ti = 0.04–0.60 apfu and T = 480–800 °C. Based
on the accumulated uncertainties of the determination of the Ti-
saturation saturation surface, the precision of the Ti-in-biotite
geothermometer is estimated to be ±24 °C at lower temperatures
(<600 °C), improving to ±12 °C at high temperature (>700 °C).
The accuracy of the Ti-in-biotite geothermometer, being primar-
ily a function of the accuracy of the Spear et al. (1999) petroge-
netic grid, is difÞ cult to assess quantitatively, but Ti-in-biotite
temperatures generally agree with independently determined
temperature estimates to within 25 °C (see below).
DISCUSSION
The systematics of Ti in peraluminous biotite not only permits
calibration of a Ti-in-biotite geothermometer, but also gives us
convenient criteria for detecting deviation from chemical equi-
librium of individual biotite grains. Furthermore, the calibration
biotite data set will give us insights into Ti-substitution mecha-
nisms across the range of peraluminous biotites at different tem-
peratures and will reveal the manner in which biotite chemistry
is inß uenced by metamorphic conditions. Finally, we will use
an additional biotite data set from metaluminous metamorphic
rocks to examine the way that Ti substitution responds to dif-
ferent bulk compositions.
Ti-in-biotite geothermometer: application to natural
peraluminous metapelites
The accuracy of the Ti-in-biotite geothermometer can be
assessed if well-constrained temperature estimates, obtained
by independent geothermometers or petrogenetic grids, are
available. SigniÞ cant disagreements between the temperature
estimates may reveal either: (1) inadequacy of the equation used
to Þ t our calibration data; (2) a problem with the independent
temperature estimate; (3) inappropriate extrapolation of the Ti-
in-biotite geothermometer outside of its calibration range; or (4)
alteration, disequilibrium, local equilibrium, or reequilibration of
the biotite following the peak temperature conditions. The latter
factor can be particularly subtle.
Figure 3 compares temperature estimates from the original lit-
TABLE 2. Summary of surface-fi t equation coeffi cients and statistical
parameters* for the full biotite data set using the equation
lnz = a + bx3 + cy3
Coeffi cient Value Standard error 95% confi dence limits
a –2.3594 0.0141 –2.3870 to –2.3317
b 4.6482e–9 5.1970e–11 4.5461e–9 to 4.7503e–9
c –1.7283 0.0584 –1.8432 to –1.6134
* This surface-fi t equation has a coeffi cient of determination (r2) of 0.924 for
these data, where r2 = 1 – (SSE/SSM). SSE is the sum of the squares due to error
and SSM is the sum of the squares about the mean.
FIGURE 2. Temperature isotherms (°C) calculated from the surface-Þ t
equation on a Ti vs. Mg/(Mg + Fe) diagram. The dashed curves represent
the intermediate 50 °C interval isotherms.
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY320
erature source with temperatures obtained using the Ti-in-biotite
geothermometer. There is a favorable agreement between these
independent temperature estimates, but there are some notewor-
thy deviations. For discussion purposes it is convenient to divide
the data into 3 general groupings of metamorphic environments:
low-pressure contact metamorphic aureoles, lower-to-middle
amphibolite-facies regional metamorphic settings, and upper
amphibolite-to granulite-facies regional metamorphic settings.
Low-pressure contact metamorphic aureoles. Several
relatively low pressure terranes (3–3.5 kbar) contain graphitic,
biotite-bearing peraluminous metapelites associated with contact
metamorphic aureoles. Biotite from a series of 26 metapelites
containing ilmenite or rutile is reported from the 3 kbar Balla-
chulish contact metamorphic aureole, Scotland (Pattison 1985,
1987, 1989, 1991, unpublished data; Pattison and Harte 1991;
Pattison et al. 2002). Application of the Ti-in-biotite geother-
mometer gives temperatures that average 16 ± 23 °C higher
than the estimated temperatures of 550–630 °C. Symmes and
Ferry (1995) characterized 4 representative biotites from the an-
dalusite-cordierite to leucocratic-veining zones in the high-level
Onawa Contact Aureole, Maine. Over the temperature range of
519–643 °C, the Ti-in-biotite temperatures are within ±24 °C
of their estimated temperatures. Porphyroblastic peraluminous
schists in the thermal aureole of the Victor Harbor Granite, South
Australia, preserve a record of mineral growth consistent with a
counterclockwise P-T path. Maximum temperatures of 580 °C
were reached at ~3 kbar before pressure increased to ~4 kbar
with a concurrent decrease in temperature to ~560 °C, as sug-
gested by late mineral growth (Alias et al. 2002). The Ti-in-biotite
geothermometer applied to 6 samples gives similar temperatures
with an average value of 570 ± 16 °C.
Lower-to-middle amphibolite facies regional metamor-
phic settings. Peraluminous schists from the Bronson Hill
anticlinorium of New Hampshire equilibrated at 3.8–6.6 kbar
and 490–575 °C according to Florence et al. (1993). Ten biotite
analyses give Ti-in-biotite temperatures slightly deviating (–8 ±
25 °C) from the Florence et al. values. Biotites from amphibo-
lite-facies peraluminous schists from Itivdlinguaq, Greenland
(Dymek 1983) give Ti-in-biotite temperatures of 566 ± 24 °C
consistent with original estimates of 575 °C. Peraluminous
metapelites from the staurolite zone and transitional sillimanite
zone of the contact aureole associated with the Harney Peak
Granite of the Black Hills, South Dakota equilibrated at about
550–570 °C and 3.5–4.5 kbar (Shearer et al. 1986), and have
biotites that yield Ti-in-biotite temperatures of 536 ± 16 °C (stau-
rolite zone) and 553 ± 16 °C (transition zone). Eight samples
from the garnet-cordierite zone (~720 °C, 4–5 kbar), the high-
est-grade zone in a regional-contact metamorphic aureole from
the Ryoke metamorphic rocks in the Yanai district of SW Japan
(Ikeda 1998a), have biotites with Ti-in-biotite temperatures of
712 ± 20°C. Biotite inclusions in K-feldspar in this high-grade
zone are consistently Ti-richer than matrix biotites and give Ti-
in-biotite temperatures 29–35 °C higher than matrix biotites,
implying continued reequilibration of the matrix on cooling from
higher temperature conditions.
Upper amphibolite- to granulite-facies regional
metamorphic settings. Application of the Ti-in-biotite geother-
mometer to high-grade metapelites generally agrees with the
independent temperature estimates, but there are some signiÞ cant
deviations reß ecting retrogression of biotite or local departures
from equilibrium. Garnet-cordierite-sillimanite gneisses from the
El Tormes Thermal Dome, Spain are estimated to have equili-
brated at 695 °C, 4.3 kbar (Ibarguchi and Martinez 1982) and
have matrix biotites that have Ti-in-biotite temperatures of 687
± 12 °C. The high-grade metapelites from south-central Mas-
sachusetts not only contain samples that displayed signiÞ cant
intergranular homogeneity that were used as part of the cali-
bration data set, but also have samples that exhibit signiÞ cant
intergranular heterogeneity and were excluded from the calibra-
tion (Thomson 1992, 2001; Peterson 1992). The results of their
entire biotite data set, including matrix brown biotite and brown
biotite proximal to other maÞ c phases, are plotted on Figure. 3.
Data from 112 individual analyses illustrate that deviations can
be signiÞ cant, signaling some compositional resetting. Similar
deviations appear relatively common in many high-grade, melt-
bearing metapelites (e.g., Graessner and Schenk 2001; Moraes
et al. 2002). The wide range in biotite chemistry commonly
found within individual high-grade metapelite samples will be
discussed further when biotite re-equilibration is considered.
Peraluminous metapelites without a Ti-saturating mineral
and/or graphite.As expected, application of the Ti-in-biotite
geothermometry from peraluminous metapelites without Ti-satu-
rating minerals (ilmenite or rutile) generally results in moderate
to large temperature underestimates. For example, biotite data
from 4 samples of magnetite-bearing, biotite-sillimanite schists
and garnet-cordierite gneisses from South Africa yield Ti-in-
biotite temperatures 61 ± 12 °C below the 750 °C estimate of
Waters and Whales (1984). Additionally, melting experiments
(6 kbar, 750–900 °C) of Patiño Douce and Harris (1998), using
natural peraluminous schists with no reported Ti-saturating min-
FIGURE 3. Calculation of Ti-in-biotite temperatures for graphitic,
peraluminous metapelites with ilmenite and/or rutile for areas that
are 3–6 kbar. T (reference) represents the temperature (°C) from the
literature source. T (calculated) represents the calculated Ti-in-biotite
temperature (°C). The solid line is reference line with a slope of 1.
The literature sources include Ibarguchi and Martinez (1982), Dymek
(1983), Pattison (1985, 1987, 1989, 1991, unpublished data), Shearer et
al. (1986), Pattison and Harte (1991), Peterson (1992), Thomson (1992,
2001), Florence et al. (1993), Symmes and Ferry (1995), Ikeda (1998a
), Alias et al. (2002), and Pattison et al. (2002).
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY 321
erals in the starting materials or experimental products, contain
biotites that yield Ti-in-biotite temperatures that underestimate
the experimental temperatures by 80 ± 27 °C.
Ti-in-biotite temperatures from peraluminous metapelites with
Ti-saturating minerals, but without graphite, can produce reason-
able temperatures but they can also be erratic. For instance, biotite
from lower-to-upper sillimanite zone (580–650 °C) peraluminous
metapelites from western Maine (Guidotti unpublished data) con-
taining magnetite and ilmenite, but not graphite, yield Ti-in-biotite
temperatures that underestimate the calibrated temperatures by
only 20 ± 12 °C. However, biotite from 31 non-graphitic, ilmenite-
bearing, peraluminous metapelites from the Ballachulish contact
metamorphic aureole of Scotland will, on average, underestimate
the reference temperatures (560–820 °C ) by 9 °C, but with a much
greater standard deviation (±50 °C) of the calculated temperatures
(Pattison 1985, 1987, 1989, 1991, unpublished data; Pattison and
Harte 1991; Pattison et al. 2002).
Monitor of chemical equilibrium and/or re-equilibration
To use reactions that involve biotite and/or application of the
principles of thermodynamics to make inferences on tempera-
ture, pressure, or ß uid evolution, we need to know that chemical
equilibrium involving biotite and coexisting minerals has been
maintained. However, establishing that a given biotite, or portion
of a biotite grain, has reached equilibrium with a particular set of
minerals can be equivocal if not impossible. Equilibrium textures
do not necessarily imply chemical equilibrium has been attained
or retained (Guidotti et al. 1991). Nonetheless, if equilibrium
textures are combined with other chemical equilibrium indica-
tors it is much more likely that equilibrium biotite compositions
can be assumed. Chemical criteria such as systematic element
partitioning involving exchangeable cations, such as Mg and Fe2+,
among coexisting phases have long been considered one of the
strongest arguments (Deb and Saxena 1976; Stephenson 1979).
However, observations of this sort require compositional data for
several coexisting minerals in each of a number of samples that
equilibrated at roughly the same temperature and pressure.
Whereas demonstrating equilibrium is generally difÞ cult, dem-
onstrating disequilibrium may be quite straightforward. In particular,
changes in Ti concentration in biotite may result from disequilibrium
and/or development of local equilibrium. Biotite compositions can
change quite readily, particularly if a late ß uid phase is present.
Biotite may be altered or replaced, it may undergo retrogressive
cation exchange or it may experience reequilibration due to thermal
overprint after cooling. Titanium contents of biotite grains and the
degree of homogeneity therefore provide a means to evaluate at-
tainment, retention, and extent of chemical equilibrium.
Minor biotite alteration. Sometimes, retrograde reactions
involving Ti can be inferred from textural features. For example,
it is common for biotite with minor degrees of chloritization to
develop oriented sagenitic rutile (Fig. 4). The Ti concentration
of the remaining biotite decreases while the Fe content generally
increases at the same time (Veblen and Ferry 1983; Shau et al.
1991; Pitra and Guiraud 1996). Application of the Ti-in-biotite
geothermometer to such biotite grains will always result in
anomalously low temperatures.
Development of biotite with different colors due to retrogression.
A common observation in upper amphibolite- and granulite-fa-
cies rocks is the presence of biotites with two distinct colors in
the same thin section, typically red-brown (Ti-rich) and green (Ti-
poor) (Edwards and Essene 1988; Thomson 1992, 2001; Ikeda
1998b; Brown 2002; Waters and Charnley 2002). Texturally, it is
not always obvious which color-type of biotite is the most likely
equilibrium composition. However, based on this study it is clear
that the brown (Ti-rich) biotite most likely represents higher-
temperature metamorphic conditions versus the (Ti-poor) green
biotite. For example, Þ ve samples of granulite-facies cordierite
pegmatite from south-central Massachusetts have coexisting
red-brown matrix biotite and green biotite next to cordierite
(Thomson 1992, 2001). The conditions of metamorphism based
on mapped isograds, are 745–775 °C at 6–7 kbar. For three of the
red-brown matrix biotites, the Ti-in-biotite geothermometer gives
temperatures of 758–771 °C (Fig. 5). Two samples have lower
temperatures of 711–715 °C, suggesting some compositional
resetting. In contrast, the green biotites next to cordierites yield
temperatures well below the 450 °C isotherm, are likely artifacts
of retrogression. Garnet-biotite Mg-Fe exchange geothermom-
etry (TWQ 2.02b program, Berman 1991) using red-brown and
green biotites result in temperature differences of roughly 50
°C for the red-brown biotites (735–746 °C) and green (692–697
°C). Waters and Charnley (2002) attribute similar compositional
relations between red and green biotite to development of local
equilibrium on cooling in the presence of a ß uid.
Intergranular and intragranular compositional heteroge-
neity of biotite. Cryptic variations in Ti contents of biotite are
common when biotite is near other ferromagnesian minerals.
These features are manifest in various ways. In low- to medium-
grade metapelites, biotite rims next to chlorite can be enriched
in Ti and Fe, possibly due to incipient chloritization of the biotite
without developing rutile (Henry 1981; Zen 1981). In contrast,
distinct local depletion of Ti in biotite near garnet being replaced
by biotite is also common in high-grade metapelites (e.g., Ikeda
1998b; García-Casco et al. 2001). Additionally, biotite replacing
cordierite in metapelites commonly has higher XMg and lower Ti
content than the biotite located away from the cordierite (Ikeda
1998b; Waters and Charnley 2002). Waters and Charnley (2002)
argued that phenomena of this type represent development of
FIGURE 4. Partially chloritized biotite with development of sagenitic
rutile needles. The sample is a granitic gneiss from the Beartooth
Mountains, Montana (Henry, unpublished data).
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY322
local equilibrium.
A more subtle feature is intragranular and intergranular varia-
tion in biotite chemistry without obvious textural or color distinc-
tions, particularly in high-grade metamorphic rocks. This type of
variation is particularly apparent in the biotite data of Thomson
(1992, 2001). Thomson (1992) analyzed 5–10 separate red-brown
biotite grains with a range of Mg/(Mg+Fe) contents in 15 samples
of peraluminous gneiss from south-central Massachusetts. Within
each sample, if the highest Ti analysis is considered, the trend
of the biotite data roughly follows the calculated Ti-in-biotite
isotherms at ~760–780 °C (Fig. 6). However, if all analyses
are considered, each sample shows an array of T – Mg/(Mg
+ Fe) points trending across the Ti-in-biotite isotherms. The
trends involve a decrease in Ti content with a slight increase
in Mg/(Mg + Fe). If the trends are extrapolated, they lead to
Ti-depleted and signiÞ cantly Mg-enriched compositions similar
to the green biotite data plotted in Figure 5. Besides the data of
Thomson (1992), the biotite data of others have similar cryptic
intergranular relations in biotites from high-grade metapelitic
terrains (Graessner and Schenk 2001; Moraes et al. 2002; Waters
and Charnley 2002). All of them likely represent compositional
resetting and/or development of localized equilibrium.
Ti substitution mechanisms
Based on several crystallographic, crystal-chemical, experi-
mental, and theoretical studies, many potential Ti substitution
schemes have been proposed (Forbes and Flower 1974; Robert
1976; Holdaway 1980; Dymek 1983; Tronnes et al. 1985; Abrecht
and Hewitt 1988; Ahlin 1988; Burt 1988; Brigatti et al. 1991;
Cruciani and Zanazzi 1994; Waters and Charnley 2002). All sub-
stitutions involve quadrivalent Ti substituting for a lower-charge
cation in the M2 site of the octahedral layer requiring coupled
substitutions involving multiple cations and possibly anions. The
four most widely advocated models for substituting Ti into biotite
(Table 3) involve three general types of coupled substitutions: (1)
within-octahedral site substitutions resulting in either accumulation
of octahedral vacancies via the Ti!R–2 substitution or loss of oc-
tahedral Al via the TiRAl–2 substitution; (2) octahedral-tetrahedral
site coupled substitution resulting in enhancement of tetrahedral Al
via the TiAl2R–1Si–2 substitution; or (3) deprotonation (dehydrogen-
ization) resulting in loss of H via the TiO2R–1(OH)–2 substitution.
Recently, the latter deprotonation substitution has been strongly
favored as the dominant Ti substitution mechanism in high-tem-
perature metapelitic biotites with intermediate Mg/(Mg+Fe) ratios
(Waters and Charnley 2002; Cesare et al. 2003). However, it is
not clear whether this mechanism is operative across the entire
compositional and temperature spectrum.
In theory, each Ti substitution mechanism is inß uenced by
bulk composition, mineral assemblage, biotite crystallochemical
constraints, and/or metamorphic ß uid compositions in different
ways. The Ti!R–2 exchange vector will be most likely enhanced
by factors that increase octahedral vacancies in biotite such as
the presence of muscovite that may act as a dioctahedral-vacancy
saturating phase via the exchange vector Al2R–3 (Labotka 1983;
Henry and Guidotti 2002). Titanium introduced with a TiRAl–2
exchange mechanism will be affected by rock compositions, such
that low-Al rocks, those without coexisting Al-saturating phases,
should have higher Ti contents in biotite. Conversely, if Ti is
introduced via the TiAl2R–1Si–2 substitution mechanism, low-Al
rocks should have lower Ti contents in biotite. The TiO2R–1(OH)–2
exchange vector will be affected by the local activity of H2O in
the metamorphic ß uid such that lower aH2O ß uids will lead to
more deprotonation and, consequently, higher Ti concentrations
in the biotite. Additional biotite substitutions that may indirectly
inß uence the amount of octahedral Ti include Al2R–1Si–1, Al2R–3,
AlOR–1(OH)–1 and MgFe–1. The MgFe–1 exchange is particularly
signiÞ cant for Ti substitution. In general, an increase in Fe2+
in biotite is usually correlated with higher Ti levels, all other
factors being constant (Abrecht and Hewitt 1988; Henry and
FIGURE 5. Ti-in-biotite geothermometry relations for coexisting
red-brown and green biotite from cordierite pegmatites from south-
central Massachusetts. Tie-lines connect coexisting red-brown and
green biotite from Þ ve separate samples (data of Thomson 1992, 2001).
Note the Mg-enrichment of the green biotites relative to the coexisting
red-brown biotite.
FIGURE 6. Cryptic variations of biotite chemistry in individual
samples based on multiple biotite analyses among 15 different samples
of garnet-cordierite gneiss from south-central Massachusetts (Thomson
1992, 2001).
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY 323
Guidotti 2002).
Normalization procedures. Different biotite normalization
procedures establish the number of ions calculated in a biotite
formula and, consequently, may inß uence the interpretation of
likely substitution mechanisms. The optimal situation is to nor-
malize biotite on the basis of 24 anions, provided that the biotite
is completely analyzed for all cations, anions, and oxidation
states. However, because of the inability of the microprobe to
measure H contents and the oxidation states of Fe, most biotite
analyses are normalized using different assumptions. Four biotite
normalization procedures have been proposed for microprobe
analyses. (1) The most common procedure is to normalize
analyses on the basis of 22 O atoms and assume all Fe is Fe2+.
The inherent supposition is that (OH)1– + F1– + Cl1– = 4.0. If the
biotite is partially deprotonated, the 22 O normalization will
underestimate the total number of cations proportional to the
amount of deprotonation, and result in a proportional number of
apparent octahedral vacancies. Conversely, if some Fe is Fe3+, the
22 O normalization will overestimate the total number of cations
proportional to the number of Fe3+ cations. (2) Dymek (1983)
considered that Ti and excess octahedral Al was linked to octa-
hedral vacancies and suggested an iterative procedure be used
in which the sum of the tetrahedral cations, octahedral cations,
Ti and excess octahedral Al is normalized to 14 cations. (3) In
contrast, Waters and Charnley (2002) suggested that the number
of octahedral vacancies is minor and advocated the deprotonation
substitution as the dominant Ti substitution. They use an iterative
normalization procedure on the basis of 22 + Ti O atoms. (4)
Finally, based on complete biotite analyses, Bohlen et al. (1980)
found that there are minor or no octahedral vacancies (<0.05
pfu) in biotites from a granulite-facies sample. They suggest
that normalization based on tetrahedral + octahedral cations =
14 may be the most reasonable. All cation-based normalization
procedures are independent of H2O contents and oxidation states
of transition elements.
To examine the inß uence of normalization on interpretation
of substitution schemes, biotite data were normalized on the
basis of all four of the normalization schemes. For biotites with
<0.3 Ti apfu, the different normalization procedures produce
minor changes in the absolute values of cations. When plotted
on binary diagrams that are used to test possible substitution
schemes there are insigniÞ cant changes in the relative positions
of the data. For those biotites with >0.3 Ti apfu, there is a pro-
portionately greater shift in the calculated values of the cations at
the higher Ti contents, although the general compositional trends
are maintained. Consequently, for the following discussion of Ti
substitution mechanisms, the 22 O normalization procedure is
used for the lower-Ti peraluminous biotite of the staurolite/lower
sillimanite zone metapelites, but in the higher-Ti biotite of the
amphibolite/granulite-facies metaluminous rocks, a 14 cation
normalization is used.
Ti substitutions in peraluminous biotite. The biotite Ti-
saturation surface has 3 general regions that display notewor-
thy changes in slope of the Ti-XMg-T relations: Region 1, the
steeper-sloped Mg-richer region that starts at roughly XMg>0.65;
Region 2, the relatively shallowly sloped region at the lower
temperature (<600 °C) and intermediate XMg portion of the
surface; and Region 3, the higher temperature region (>600 °C)
(Fig. 7). Rather than relying on the surface Þ t with any biases
that might be introduced, we used a subset of the actual biotite
data from western Maine to assess likely Ti substitutions in each
of these regions. For this more detailed examination, biotite data
are limited to staurolite zone and transition/lower sillimanite
zone samples. These zones were chosen because they exhibit
a signiÞ cant transition in the character of the Ti systematics in
biotite, they contain the largest numbers of analyses, and they
have equilibrated in a relatively restricted temperature range,
545–580 °C for the staurolite zone and 580–620 °C for the
transition/lower sillimanite zone.
Figure 8 shows four binary composition diagrams and the
trends that would result from different exchange vectors. In
Region 1, the Mg-rich portion of the diagram, there is a general
inverse relationship between Ti and Mg/(Mg + Fe) ratios of
biotite in each metamorphic zone (Fig. 8a). In addition, Ti is
negatively correlated with the sum of the total divalent cations
(Mg + Fe + Mn) and Si contents, and is positively correlated with
the Altotal (Figs. 8b, 8c, 8d). This trend in data is most consistent
with variation due to the exchange vector TiAl2R–1Si–2. These
relationships are likely controlled by crystal-chemical factors in
that substitution of relatively small Ti4+ cations in the octahedral
layer and Al3+ cations in the tetrahedral layer compensates for any
octahedral-tetrahedral layer misÞ t resulting from substitution of
larger Fe2+ cations for smaller Mg2+ cations in the octahedral layer
(Guidotti et al. 1977). This produces a decrease in the average
layer distortion and results in the M1 and M2 octahedral sites
becoming more distinct, with M1 being more distorted and larger
than M2 (Brigatti and Guggenheim 2002). In Region 2, staurolite
zone biotite data at intermediate XMg, inverse relationships among
TABLE 3. Most likely substitution schemes and exchange vectors for
incorporation of Ti into biotite
Site substitution Exchange vector
2VIR2+ = VITi + VI
■
■* Ti
■
■ R-2
2VIAl = VITi + VIR2+ TiRAl-2
VIR2+ + 2IVSi = VITi + 2IVAl TiAl2R-1Si-2
VIR2+ + 2(OH)1- = VITi + 2O2- TiO2R-1(OH)-2
* VIR2+ represents the sum of the divalent cations in the octahedral sites and
VI
■
■represents the octahedral site vacancies.
FIGURE 7. Titanium saturation surface for natural peraluminous
biotites at 4–6 kbar highlighting different regions on interest.
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY324
Ti and XMg, sum of the total divalent cations (Mg + Fe + Mn) and
Si contents persist, but are much less pronounced (Figs. 8b, 8c,
8d). This relationship is most compatible with the TiAl2R–1Si–2
exchange. However, the principal variation of the data is in the
total divalent cations (Mg + Fe + Mn), Si, and Altotal contents
and is most consistent with the exchange vectors Al2R–1Si–1,
AlOR–1(OH)–1 and Al2R–3. In Region 3, transition and lower sil-
limanite zone biotite data at intermediate XMg, Ti is dispersed to
higher values and is best explained by a signiÞ cant amount of
substitution due to deprotonation described by TiO2R–1(OH)–2
(Figs. 8b, 8c, 8d).
The shift in the Ti substitution mechanism above the stau-
rolite zone suggests that there may be a fundamental change in
the metamorphic environment at higher metamorphic grades.
The increasing importance of the TiO2R–1(OH)–2 exchange
implies that biotite above the staurolite zone is being exposed
to metamorphic ß uids with lower aH2O. In fact, there is direct
evidence for deprotonation in higher-grade biotites from western
Maine. Dyar et al. (1991) measured H contents in a subset of
biotites from the calibration data set and found that biotites from
higher-grade zones, particularly above the staurolite zone, were
signiÞ cantly depleted in H. The lower aH2O environment can
be understood by considering the likely metamorphic ß uids in
equilibrium with graphite in the COHS system. Graphite-bearing
metapelites that undergo dehydration reactions generally produce
ß uids that have H/O ratios of 2:1 i.e., they approach the H2O
maximum for a given temperature, with the balance of the ß uid
being composed of other C- and S-bearing species (Ohmoto and
Kerrick 1977; Connolly and Cesare 1993). Calculation of the
proportions of species in a COHS metamorphic ß uid demon-
strate that as temperature increases, the amounts of other species
reduce the maximum amount of H2O. As such, at temperatures
below 550 °C at 4 kbar there are only 3–5% other species at the
H2O maximum (Connolly and Cesare 1993). However, above
550 °C, the maximum XH2O decreases signiÞ cantly from 0.92 at
600 °C to 0.82 at 700 °C to 0.73 at 800 °C. In theory, this de-
crease in XH2O will enhance the deprotonation exchange reaction
TiO2R–1(OH)–2 and greater amounts of Ti should be incorporated
in the biotite. Further, if there are aqueous metamorphic ß uids
with high salinity, the aH2O will also be signiÞ cantly reduced
(Aranovich and Newton 1998) and Ti-deprotonation exchange
will likely occur in the coexisting biotites. In contrast, if the
FIGURE 8. Biotite data from western Maine for the staurolite and transition/lower sillimanite zones normalized on a 22 O basis. (a) Ti – Mg/(Mg
+ Fe) relations. Note break in slope of data trend at the Mg-rich end of the diagram. (b) Ti – (Mg + Fe + Mn) relations. The arrows represent the
direction of the possible exchange vectors. (c) Ti-Si relations. (d) Ti-Altotal relations.
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY 325
rocks do not contain graphite, the maximal amounts of H2O in
the metamorphic ß uid could be higher than in graphite-bearing
rocks and any inß uence of the deprotonation exchange reaction
TiO2R–1(OH)–2 may be reduced so that biotite Ti contents and the
Ti-in-biotite temperatures will be lower. This feature may help to
explain the tendency for non-graphitic rocks to yield somewhat
lower temperatures.
Ti substitutions in low-Al biotite. Ti substitution mechanisms
can be extended beyond peraluminous biotites by considering an
additional extensive biotite data set from high-grade metaluminous
bulk compositions. Harlov and Forster (2002) include a detailed
biotite data set in their study that reported on a gneiss terrane that
has experienced a high-grade metamorphic overprint at ~800 °C
and 8 kbar under locally variable aH2O conditions. A hornblende-
bearing metatonalite from the Seward Peninsula, Alaska, locally
develops an 80 cm dehydration zone next to a marble layer, appar-
ently associated with a CO2-rich and low aH2O ß uid. The host rock
metatonalite has a uniform assemblage of hornblende + biotite +
plagioclase + quartz + pyrrhotite + graphite, but no Ti-saturating
mineral (ilmenite or rutile). The dehydration zone, lacking horn-
blende, develops in a zone within 50 cm of the marble unit and
has an assemblage of orthopyroxene + clinopyroxene + Ti-rich
biotite + plagioclase + quartz + minor K-feldspar + pyrrhotite.
There is a 30 cm wide transition zone that represents partial de-
hydration with some hornblende remaining. The factors that are
likely to inß uence biotite chemistry include: (1) the rocks do not
contain a Ti-saturating phase and, consequently, the biotite will
be unsaturated with respect to Ti and subject to variation in bulk
composition; (2) there is graphite present but no Fe3+-saturating
phase, hence, Fe3+ should be relatively low and constant; (3) the
rocks do not contain an Al-saturating phase so that Al levels will
be lower than peraluminous metapelites; and (4) the biotite will
be primarily inß uenced by local variation in aH2O because they
are derived from an isothermal and isobaric setting. A total of
470 biotite analyses from 12 samples are available with multiple
biotite analyses from the same grain and from several grains from
the same sample.
The Ti – Mg/(Mg + Fe) relations illustrate that there are two
clusters of data (Fig. 9a). (1) The lower-Ti biotite samples from
the amphibolite and transition zone exhibit a range of Mg/(Mg
FIGURE 9. Isothermal biotite data from low-Al biotites from metaluminous amphibolites to maÞ c granulites from Alaska (Harlov and Forster
2002) normalized on a 14 cation basis. The estimated metamorphic conditions are 800 °C and 8 kbar. (a) Ti – Mg/(Mg + Fe) relations. Each of the
12 samples has multiple biotite analyses (470 total analyses) with the distance (in cm) from the marble unit indicated next to the symbol. The g
next to the sample number indicates a sample with a granulite-facies assemblage, t represents a transitional sample and a represents a sample with
amphibolite-facies assemblage. (b) Ti – (Mg + Fe +Mn) relations. The arrows represent the direction of the possible exchange vectors. The dashed
line is a reference line with a slope of 1. (c) Ti-Si relations. (d) Ti-Altotal relations.
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY326
+ Fe) (0.52–0.68) and Ti (0.27–0.49 apfu) among the different
samples. Application of the Ti-in-biotite geothermometer to this
group of data yields a maximum apparent temperature of 745
°C, representing a minimum temperature estimate because of
the lack of a Ti-saturating phase. Biotite data from some of the
samples exhibit signiÞ cant dispersion, with a common trend be-
ing a decrease in Ti associated with an increase in Mg/(Mg + Fe)
ratio, a pattern consistent with partial resetting during cooling
(Fig. 6). (2) The high-Ti biotite samples are from the dehydrated
maÞ c granulite zone within 50 cm of the marble. The biotite
has a range of Mg/(Mg + Fe) (0.53–0.65) similar to the low-Ti
data, but with signiÞ cantly higher Ti (0.53–0.82 apfu). Apparent
Ti-in-biotite temperatures are higher than the lower-Ti samples,
with a range of 775–810 °C, close to the regional temperature
estimates, but likely related to the manner in which Ti substitutes
in lower-Al biotite under variable aH2O conditions.
A series of binary composition diagrams help deÞ ne the most
likely substitutions in the isothermal low-Al biotites with variable
aH2O. In addition to variable Mg/(Mg + Fe) ratios, the lower-Ti
biotites also exhibit a minor dispersion of data roughly parallel
to the Al2R-1Si–1 and AlOR–1(OH)–1 vectors (Fig. 9b, 9c, 9d). In
the granulite facies samples with higher-Ti biotite, the increase
in Ti in biotite has a slope of roughly 1.0 when plotted against
Mg + Fe + Mn (Fig. 9b) and a slightly positive slope plotted
against Si (Fig. 9c). The Ti exchange vector that best satisÞ es
both of these criteria is TiO2R–1(OH)–2. The separation between
the low-Ti biotite and the maximum values of the high-Ti groups
is approximately 0.4 Ti apfu, implying that the operation of the
TiO2R–1(OH)–2 vector by 0.4 units and a resulting deprotonation
of the highest-Ti biotites by about 0.8 H apfu. However, the Ti-
Altotal diagram (Fig. 9d) shows an additional inverse relationship
between Ti and Al, implying another exchange takes place that
reduces the amount of Al by about 1 apfu for each 0.4 Ti apfu.
The substitution that is most consistent with these relationships
is the RSi(Al–2) vector. Consequently, the major changes between
low- and high-Ti samples can be deÞ ned by a combination of
0.4 TiO2R–1(OH)–2 and 0.05 RSi(Al–2) vectors, and a resultant
cumulative exchange vector of Ti0.4O0.8Si0.05(R–0.35Al–0.1(OH)–0.8).
The reduction of Al in both the octahedral and tetrahedral layers
in conjunction with the introduction of the larger divalent cations
in the octahedral layer and smaller Si cation in the tetrahedral
layer probably represents the response of the biotite structure to
introduction of the smaller Ti into the octahedral layer.
SUMMARY STATEMENT
The Ti-saturation surface generated for biotites from peralu-
minous metapelites at low-to-medium pressures contains useful
information with petrologic and crystallographic implications. A
reformulation of the surface Þ t expression results in an empiri-
cal Ti-in-biotite geothermometer for biotites in peraluminous
metamorphic rocks. The geothermometer is valid over a range
of 480 to 800 °C and has an estimated precision of ±24 °C. Fur-
ther, the systematics of Ti in biotite provides a means to assess
the approach to chemical equilibrium of biotite in metamorphic
rocks. Finally, the Ti-saturation surface provides evidence
that Ti is incorporated into biotite by at least two substitution
mechanisms. In magnesian biotite, Ti primarily substitutes via a
Ti-tschermaks substitution controlled by octahedral-tetrahedral
layer misÞ t. However, at intermediate XMg, Ti-deprotonation
substitution becomes the dominant mechanism for Ti incorpo-
ration, particularly at metamorphic grade above the staurolite
zone where there is signiÞ cant reduction of activity of H2O in
the metamorphic ß uid from graphite-bearing metapelites. This
study highlights the petrologic mineralogy approach, as described
by Guidotti and Sassi (2002), in which natural solid-solution
minerals, constrained by phase-rule variance considerations
and equilibration, are used to provide mineralogic and petro-
logic insights not generally available via typical experimental
or crystallography studies.
ACKNOWLEDGMENTS
We are pleased to acknowledge all of our colleagues who have contributed
to the biotite data set over the years including Kathy Bussa, Jack Cheney, Paul
Conatore, Barb Dutrow, Angela LaGrange, Steve Lonker, David Pattison, and
Virginia Peterson. We are grateful to the many of the people cited in terms of our
sources of data as in many cases they provided us electronic Þ les of their data,
thereby making it much easier for us to process. Amber Hawkins at the University
of Maine did much of this processing. Finally, discussions about Ti in biotite with
Bernardo Cesare, Darby Dyar, and Ed Grew have been helpful in many ways.
Insightful reviews by Dexter Perkins, Tom Hoisch, and Ted Labotka greatly
improved the Þ nal manuscript.
REFERENCES CITED
Abrecht, J. and Hewitt, D.A. (1988) Experimental evidence on the substitution of
Ti in biotite. American Mineralogist, 73, 1275–1284.
Ahlin, S. (1988) Interpretation of biotite-garnet partitioning data from two Sveco-
karelian metamorphic rock units. Lithos, 22, 25–29.
Alias, G., Sandiford, M., Hand, M., and Worley, B. (2002) The P-T record of syn-
chronous magmatism, metamorphism and deformation at Petrel Cove, southern
Adelaide Fold Belt. Journal of Metamorphic Geology, 20, 351–363.
Aranovich, L.Y. and Newton, R.C. (1998) Reversed determination of the reaction:
Phlogopite + quartz = enstatite + potassium feldspar + H2O in the ranges
750–875 °C and 2–12 kbar at low H2O activity with concentrated KCl solu-
tions. American Mineralogist, 83, 193–204.
Arima, M. and Edgar, A.D. (1981) Substitution mechanisms and solubility of
titanium in phlogopites from rocks of probable mantle origin. Contributions
to Mineralogy and Petrology, 77, 288–295.
Berman, R.G. (1991) Thermobarometry using multiequilibrium calculations:
a new technique with petrologic applications. Canadian Mineralogist, 29,
833–855.
Bohlen, S.R., Peacor, D.R., and Essene, E.J. (1980) Crystal chemistry of a meta-
morphic biotite and its signiÞ cance in water barometry. American Mineralo-
gist, 65, 55–62.
Brigatti, M.F. and Guggenheim, S. (2002) Mica crystal chemistry and the inß uence
of pressure, temperature, and solid solution on atomistic models. In A. Mot-
tana, F.P. Sassi, J.B. Thompson Jr., and S. Guggenheim, Eds., Micas: Crystal
Chemistry and Metamorphic Petrology, 46, 1–98. Reviews in Mineralogy and
Geochemistry, Mineralogical Society of America, Washington, D.C.
Brigatti, M.F., Galli, E., and Poppi, L. (1991) Effect of Ti substitution in biotite-1M
crystal chemistry: American Mineralogist, 76, 1174–1183.
Brown, M. (2002) Prograde and retrograde processes in migmatites revisited.
Journal of Metamorphic Geology, 20, 25–40.
Burt, D.M. (1988) Vector representations of phyllosilicate compositions. In S.W.
Bailey, Ed., Hydrous Phyllosilicates (exclusive of micas), 19, 561–599. Re-
views in Mineralogy, Mineralogical Society of America, Washington, D.C.
Cesare, B., Cruciani, G., and Russo, U. (2003) Hydrogen deÞ ciency in Ti-rich
biotite from anatectic metapelites (El Joyazo, SE Spain): crystal-chemical
aspects and implications for high-temperature petrogenesis. American Min-
eralogist, 88, 583–595.
Connolly, J.A.D. and Cesare, B. (1993) C-O-H-S ß uid composition and oxygen
fugacity in graphitic metapelites. Journal of Metamorphic Geology, 11,
379–388.
Cruciani, G. and Zanazzi, P.F. (1994) Cation partitioning and substitution mecha-
nisms in 1M phlogopite: a crystal-chemical study. American Mineralogist,
79, 289–301.
Deb, M. and Saxena, S.K. (1976) A method of appraisal of chemical equilibrium
- application to terrestrial and lunar rocks. Lithos, 9, 225–234.
DeYoreo, J.J., Lux, D.R., Guidotti, C.V., Decker, E.R., and Osberg, P.H. (1989)
The Acadian thermal history of western Maine, Journal of Metamorphic
Geology, 7, 169–170.
Dutrow, B.L. (1985) A Staurolite Trilogy: III. Evidence for Multiple Metamorphic
Events in the Farmington Quadrangle, Maine. Ph.D. Thesis, Southern Method-
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY 327
ist University, Dallas, TX.
Dyar, M.D., Colucci, M.T., and Guidotti, C.V. (1991) Forgotten major elements:
Hydrogen and oxygen variation in biotite from metapelites. Geology, 19,
1029–1032.
Dymek, R.F. (1983) Titanium, aluminum and interlayer cation substitutions in
biotite from high-grade gneisses, West Green land. American Mineralogist,
68, 880–899.
Edwards, R.L. and Essene, E.J. (1988) Pressure, temperature and C-O-H ß uid
fugacities across the amphibolite-granulite facies transition, NW Adirondack
Mountains, NY. Journal of Petrology, 29, 39–73.
Engel, A.E.J. and Engel, C.G. (1960) Progressive metamorphism and granitization
of the major paragneiss, northwest Adirondack Mountains, New York, Part 2.
Mineralogy. Bulletin of the Geological Society of America, 71, 1–58.
Ferry, J.M. (1981) Petrology of graphitic sulÞ de-rich schists from south-central
Maine: an example of desulÞ dation during prograde regional metamorphism.
American Mineralogist, 66, 908–930.
Florence, F.P., Spear, F.S., and Kohn, M.J. (1993) P-T paths from northwestern New
Hampshire: metamorphic evidence from stacking in a thrust/nappe complex.
American Journal of Science, 293, 939–979.
Forbes, W.C. and Flower, M.F.J. (1974) Phase relations of titan-phlogopite,
K2Mg4TiAl2Si6O20(OH)4: A refractory phase in the upper mantle?: Earth
Planetary Science Letters, 22, 60–66.
García-Casco, A., Torres-Roldán, R.L., Millán, G., Monié, P., and Haissen, F. (2001)
High-grade metamorphism and hydrous melting of metapelites in the Pinos
terrane (W Cuba): Evidence for crustal thickening and extension in the northern
Caribbean collisional belt. Journal of Metamorphic Geology, 19, 699–715.
Graessner, T. and Schenk, V. (2001) An exposed Hercynian deep crustal section
in the Sila Massif of northern Calabria: Mineral Chemistry, petrology and a
P-T path of granulite-facies metapelitic migmatites and metabasites. Journal
of Petrology, 42, 931–961.
Guidotti, C.V. (1970a) The mineralogy and petrology of the transition from lower
to upper sillimanite zone in the Oquossoc area, Maine. Journal of Petrology,
11, 277–336.
——— (1970b) Metamorphic petrology, mineralogy, and polymetamorphism in
a portion of N.W. Maine. In Boone, G.M., editor, Guidebook for Þ eld trips
in the Rangeley Lakes—Dead Rivers Basin region, Western Maine. New
England Intercollegiate Geology Conference, 62nd Annual Meeting, Field
Trip B2, 1–23.
——— (1974) Transition from staurolite to sillimanite zone, Rangeley quadrangle,
Maine. Geological Society of America Bulletin, 85, 475–490.
——— (1984) Micas in metamorphic rocks. In S.W. Bailey, Ed., Micas, 13,
357–467. Reviews in Mineralogy, Mineralogical Society of America, Wash-
ington, D.C.
——— (1985) Metamorphic map of Maine. In P.H. Osberg, A.M. Hussey II,
and G.M. Boone, Eds., Geologic map of Maine. Maine Geological Survey,
Augusta, Maine.
——— (1989) Metamorphism in Maine: an overview: In R.D. Tucker and R.G.
Marvinney, Eds., Studies in Maine Geology, 3, 1–17. Maine Geological
Survey, Augusta, Maine.
Guidotti, C.V. and Dyar, M.D. (1991) Ferric iron in metamorphic biotite and its
petrologic and crystallographic implications. American Mineralogist, 76,
161–175.
Guidotti, C.V. and Sassi, F.P. (2002) Constraints on studies of metamorphic K-Na
white micas. In A. Mottana, F.P. Sassi, J.B. Thompson Jr., and S. Guggenheim,
Eds., Micas: Crystal Chemistry and Metamorphic Petrology, 46, 419–448.
Reviews in Mineralogy and Geochemistry, Mineralogical Society of America,
Washington, D.C.
Guidotti, C.V., Cheney, J.T., and Guggenheim, S. (1977) Distribution of titanium
between coexisting muscovite and biotite in pelitic schists from northwestern
Maine. American Mineralogist, 62, 438–448.
Guidotti, C.V., Cheney, J.T., and Henry, D.J. (1988) Compositional variation of
biotite as a function of metamorphic reactions and mineral assemblage in
the pelitic schists of western Maine: American Journal of Science - Wones
Memorial Volume, 288A, 270–292.
Guidotti, C.V., Teichmann, F., and Henry, D.J. (1991) Evidence for equilibrium
chlorite in the polymetamorphic metapelites of the Rangeley area, western
Maine. American Mineralogist, 76, 867–879.
Harlov, D.E. and Forster, H. (2002) High-grade ß uid metasomatism on the local and
regional scale: the Seward Peninsula, Alaska, and the Val Strona di Omegna,
Ivrea-Verbano Zone, Northern Italy. Part I. Petrography and silicate mineral
chemistry. Journal of Petrology, 43, 769–799.
Henry, D.J. (1981) SulÞ de-silicate relations of the staurolite grade pelitic schists,
Rangeley quadrangle, Maine. Ph.D. thesis, University of Wisconsin—Madison,
Madison, Wisconsin.
Henry, D.J. and Guidotti, C.V. (2002) Ti in biotite from metapelitic rocks: Tempera-
ture effects, crystallochemical controls and petrologic applications. American
Mineralogist, 87, 375–382.
Holdaway, M.J. (1980) Chemical formulae and activity models for biotite,
muscovite and chlorite applicable to pelitic metamorphic rocks. American
Mineralogist, 65, 711–719.
Holdaway, M.J., Guidotti, C.V., Novak, J.M., and Henry, W.E. (1982) Polymeta-
morphism in medium- to high-grade pelitic metamorphic rocks, west-central
Maine. Geological Society of America Bulletin, 93, 572–584.
Holdaway, M.J., Dutrow, B.L., and Hinton, R.W. (1988) Devonian and Car-
boniferous metamorphism in west-central Maine: the muscovite-almandine
geobarometer and the staurolite problems revisited. American Mineralogist,
73, 20–47.
Ibarguchi, G.J.I. and Martinez, F.J. (1982) Petrology of garnet-cordierite-sillimanite
gneisses from the El Tormes thermal dome, Iberian Hercynian foldbelt (W
Spain). Contributions to Mineralogy and Petrology, 80, 14–24.
Ikeda, T. (1998a) Phase equilibria and the pressure-temperature path of the highest-
grade Ryoke metamorphic rocks in the Yanai district, SW Japan. Contributions
to Mineralogy and Petrology, 132, 321–335.
——— (1998b) Progressive sequence of reactions of the Ryoke metamorphism
in the Yanai district, southwest Japan: the formation of cordierite. Journal of
Metamorphic Geology, 16, 39–52.
Johnson, T.E., Brown, M., and Solar, G.S. (2003) Low-pressure subsolidus and
suprasolidus phase equilibria in the MnNCKFMASH system. Constraints on
conditions of regional metamorphism in western Maine, northern Appalachians.
American Mineralogist, 88, 624–638.
Kwak, T.A.P. (1968) Ti in biotite and muscovite as an indication of metamorphic
grade in almandine amphibolite facies rocks from Sudbury, Ontario. Geochi-
mica et Cosmochimica Acta, 32, 1222–1229.
Labotka, T.C. (1983) Analysis of the compositional variations of biotite in pelitic
hornfels from northeastern Minnesota. American Mineralogist, 68, 900–914.
Moraes, R., Brown, M., Fuck, R.A., Camargo, M.A., and Lima, T.M. (2002)
Characterization and P-T evolution of melt-bearing ultrahigh-temperature
granulites: an example from the Anápolis–Atauçu Complex of the Brasília
Fold Belt, Brazil. Journal of Petrology, 43, 1673–1705.
Ohmoto, H. and Kerrick, D.M. (1977) Devolatilization equilibria in graphitic
schists. American Journal of Science, 277, 1013–1044.
Patiño Douce, A.E. (1993) Titanium substitution in biotite: an empirical model
with applications to thermometry, O2 and H2O barometries, and consequences
form biotite stability. Chemical Geology, 108, 133–162.
Patiño Douce, A.E. and Harris, N. (1998) Experimental constraints on Himalayan
anatexis. Journal of Petrology, 39, 689–710.
Patiño Douce, A.E. and Johnston, A.D. (1991) Phase equilibria and melt productiv-
ity in the pelitic system: implications for the origin of peraluminous granitoid
and aluminous granulites. Contributions to Mineralogy and Petrology, 107,
202–218.
Pattison, D.R.M. (1985) Petrogenesis of pelitic rocks in the Ballachulish thermal
aureole. Unpublished Ph.D. thesis, University of Edinburgh, 590 pp.
——— (1987) Variations in Mg/(Mg + Fe), F, and (Fe,Mg)Si = 2Al in pelitic
minerals in the Ballachulish thermal aureole, Scotland. American Mineralo-
gist, 72, 255–272.
——— (1989) P-T conditions and the inß uence of graphite on pelitic phase relations
in the Ballachulish aureole, Scotland. Journal of Petrology, 30, 1219–1244.
——— (1991) P-T-a(H2O) conditions in the thermal aureole. In G. Voll, J. Topel,
D.R.M. Pattison, and F. Seifert, Eds., Equilibrium and kinetics in contact
metamorphism: The Ballachulish Igneous Complex and its aureole. Springer
Verlag, Heidelberg
Pattison, D.R.M. and Harte, B. (1991) Petrography and mineral chemistry of pelites.
In G. Voll, J. Topel, D.R.M. Pattison, and F. Seifert, Eds., Equilibrium and
kinetics in contact metamorphism: The Ballachulish Igneous Complex and its
aureole. Springer Verlag, Heidelberg.
Pattison, D.R.M., Spear, F.S., DeBuhr, C.L., Cheney, J.T., and Guidotti, C.V.
(2002) Thermodynamic modeling of the reaction muscovite + cordierite →
Al2SiO5 + biotite + quartz + H2O: constraints from natural assemblages and
implications for the metapelitic petrogenetic grid. Journal of Metamorphic
Geology, 20, 99–118.
Peterson, V.L. (1992) Structure, petrology, and tectonic implications of highly
strained rocks along the west margin of the Acadian granulite-facies high,
south-central Massachusetts: Ph.D. thesis, University of Massachusetts,
Amherst, Massachusetts.
Pitra, P. and Guiraud, M. (1996) Probable anticlockwise P-T evolution in extending
crust: Hlinsko region, Bohemian Massif. Journal of Metamorphic Petrology,
14, 49–60.
Pressley, R.A. and Brown, M. (1999) The Phillips pluton, Maine, USA: evidence for
heterogeneous crustal sources for granitic ascent and emplacement mechanisms
in convergent orogens. Lithos, 49, 335–366.
Robert, J.L. (1976) Titanium solubility in synthetic phlogopite solid solutions.
Chemical Geology, 17, 213–227.
Shau, Y., Yang, H., and Peacor, D.R. (1991) On oriented titanite and rutile inclusions
in sagenitic biotite: American Mineralogist, 76, 1205–1217.
Shearer, C.K., Papike, J.J., Simmons, S.B., and Laul, J.C. (1986) Pegmatite-wall-
rock interactions, Black Hills, South Dakota: Interaction between pegmatite-
derived ß uids and quartz-mica schist wallrock. American Mineralogist, 71,
518–539.
Spear, F.S. and Cheney, J.T. (1989) A petrogenetic grid for pelitic schists in the
system SiO2-Al2O3-FeO-MgO-K2O-H2O. Contributions to Mineralogy and
HENRY ET AL.: TI-IN-BIOTITE GEOTHERMOMETRY328
Petrology, 101, 149–164.
Spear, F.S., Kohn, M.J. and Cheney, J.T. (1999) P-T paths from anatectic pelites.
Contributions to Mineralogy and Petrology, 134, 17–32.
Stephenson, N.C.N. (1979) Coexisting garnets and biotites from Precambrian
gneisses of the south coast of Western Australia. Lithos, 12, 73–87.
Symmes, G.H. and Ferry, J.M. (1995) Metamorphism, ß uid ß ow and partial melting
in pelitic rocks from the Onawa contact aureole, central Maine, USA. Journal
of Petrology, 36, 587–612.
Thomson, J.A. (1992) Petrology of high-grade schists, gneisses, and cordierite
pegmatites in south-central Massachusetts: Ph.D. thesis, University of Mas-
sachusetts, Amherst, Massachusetts.
——— (2001) A counterclockwise P-T path for anatectic pelites, south-central
Massachusetts. Contributions to Mineralogy and Petrology, 141, 623–641.
Tracy, R.J. (1978) High grade metamorphic reactions and partial melting in
pelitic schist, west-central Massachusetts. American Journal of Science,
278, 150–178.
Tracy, R.J. and Robinson, P. (1988) Silicate-sulÞ de-oxide-ß uid reactions in
granulite-grade pelitic rocks, central Massachusetts. American Journal of
Science, 288-A, 45–74.
Tronnes, R.G., Edgar, A.D., and Arima, M. (1985) A high pressure-high temperature
study of TiO2 solubility in Mg-rich phlogopite: Implications to phlogopite
chemistry. Geochimica et Cosmochimica Acta, 49, 2323–2329.
Veblen, D.R. and Ferry, J.M. (1983) A TEM study of the biotite-chlorite reaction
and comparison with petrologic observation: American Mineralogist, 68,
1160–1168.
Waters, D.J. and Charnley, N.R. (2002) Local equilibrium in polymetamorphic
gneiss and the titanium substitution in biotite. American Mineralogist, 87,
383–396.
Waters, D.J. and Whales, C.J. (1984) Dehydration melting and the granulite transi-
tion in metapelites from southern Namaqualand, S. Africa. Contributions to
Mineralogy and Petrology, 88, 269–275.
Zen, E. (1981) Metamorphic mineral assemblages of slightly calcic pelitic rocks
in and around the Taconic Allochthon, southwestern Massachusetts and
adjacent Connecticut and New York. Geological Survey Professional Paper
1113, 128 p.
MANUSCRIPT RECEIVED SEPTEMBER 30, 2004
MANUSCRIPT ACCEPTED OCTOBER 11, 2004
MANUSCRIPT HANDLED BY THEODORE LABOTKA
