Petrography, Geothermobarometry and Metamorphic History of Metapelites from the Central Ruby Range Southwest Montana

Home , Geothermobarometry

Petrography, Geothermobarometry and Metamorphic History of Metapelites from the Central Ruby Range Southwest Montana HAMELIN, Clémentine1, BALDWIN, Julia A.2, HARMS, Tekla A.3, and BRADY, John B.1 1 Department of Geosciences, Smith College, Northampton, MA 01063, [email protected], 2 Department of Geosciences, University of Montana, Missoula, MT 59812, 3 Department of Geology, Amherst College, Amherst, MA 01002

RESULTS RESULTS ABSTRACT PETROGRAPHY GEOTHERMOBAROMETRY Figure 4. left column: CH1 results, right column, CH4 results. From top to Stable assemblage fields in mineral assemblage diagrams match petrography. The Ruby Range is one of several uplifted Precambrian blocks in southwest Montana, near the northwestern margin of the Archean • Schistose texture with well-developed foliation defined by bottom: a) T-XH2O diagrams a 7.5 kbar showing variations due to water con- • Conventional P-T (GTB) • Thermodynamic P-T modeling (Theriak-Domino) • QuiG Barometry (GeobaRamanTry) tent adjustments; b) Pseudosection using loss on ignition (LOI) water values,; Wyoming province. It was metamorphosed during the Big Sky orogeny ~ 1.72-1.79 Ga (Baldwin et al., 2014; Cramer et al., 2013). abundant biotite and sillimanite grain alignment. Two Modeled (Theriak-Domino) vs. observed modes (SEM phase map) also match well c) Pseudosection using water saturation at solidus determined using a T-XH2O This study seeks to characterize further the P-T metamorphic conditions of the Big Sky orogenic event as recorded in the central Ruby samples with a gneissic texture, with higher feldspar + quartz diagram for XRF bulk rock composition; d) Pseudosection with 30% added for both LOI water value vs. H2O saturation at solidus values (Table 1) Range. Metapelite samples were collected during July 2014 as part of a Keck Geology Consortium summer research project. Gar- content. xH2O at P = 7.5 Kbar xH2O at P = 7.5 Kbar melt from a ‘wet’ peraluminous granite ; e) Raman spectra with quartz peak, net-biotite-sillimanite-bearing metapelite rocks associated with marble, amphibolite and metamorphosed banded iron formation • Petrography: NKFMASH petrogenetic grid, preliminary constraints on P, T are • Garnets: anhedral, range in size from 1 mm to 10 cm 900 900 showing Raman shift between matrix and included quartz grains; f) Combined were sampled from the Christensen Ranch Metamorphic Suite (CRMS) at the highest structural levels of the central Ruby Range near diameter. Generally Fe-rich (~80% almandine), little to no geothermobarometry results (QuiG, conventional P-T, and 30%added melt 650 ºC - 900 ºC (muscovite-OUT, K-feldspar-IN), 2-11 kbar (cordierite and kyanite Stone Creek. Textures and mineral assemblages preserve evidence for partial melting. All samples contain the assemblage biotite + combined diagram) zoning, but with evidence for two garnet-growth episodes ksp + plag + plag + grt + ilm absent) (Figure 4, grey area). garnet + sillimanite +quartz, consistent with upper amphibolite to lower granulite facies metamorphism. Four samples were further grt + ilm + liq + liq + q + sill (inclusion free or different rims, Ca and Fe zoning in one 850 ksp + plag + 850 + q + sill investigated with petrographic observations, geochemical XRF and SEM/EDS analyses and modeling to determine P-T conditions grt + bt + liq Further P-T constraint from conventional P-T and pseudosection modeling: sample). + q + sill + • and history recorded in the samples. A peak-temperature value of ~800ºC at a pressure of 9kb and a second, possibly re-equilibrat- ilm PSEUDOSECTION ISOPLETHS • Biotite between dioctahedral and trioctahedral biotite, Fe-Mg ~700-800 ºC and 7-9 kb (Figure 4), ed T,P value of ~700ºC and ~ 7 kbar was determined from further investigation of five samples. Quartz-in-garnet (QuiG) barometry ksp + plag + Garnet Mg/(Mg+Fe) bt + grt + ilm ratios in biotite range from ~ 1:1 to 2:3 across samples but, 800 ksp + pl + grt 800 Biotite Mg/(Mg+Fe) was used to determine entrapment pressures of quartz grains includes in garnets, providing insight into the prograde metamorphic + bt + liq + q + liq + q + sill Peak-T at ~800 ºC, ~8.5 kbar (CH1, CH9b, CH11a) generally consistent within samples. + sill history of the metapelite samples from the CRMS, yielding peak-pressure results of ~12 kbar between ~700-800 ºC. These results ksp + plag + SOLIDUS Garnet grossular Ca/(Fe+Mg+Ca)

are consistent with the P-T conditions and metamorphic history for the Big Sky orogeny determined in the adjacent Tobacco Root • Generally muscovite-free, a few samples contain minor grt + bt + q AT Re-equilibration values at ~725 ºC, ~7 kbar (CH4, CH5a) + sill + ilm plag + grt + bt Garnet almandine Fe/(Fe+Mg+Ca) Mountains (Cheney et al., 2002; Brady et al., 2004, Harms et al., 2004). amounts of retrograde-muscovite. 750 ksp plag + grt + 750 + ilm + liq + q + ksp + plag + + sill Plagioclase anorthite Ca/(Ca+Na+K) plag bt + liq + q Overall, results from thermodynamic modeling consistent with mineral grt + bt + q

T [C] •

• Kyanite and cordierite-absent. + grt + sill T [C] + bt + sill • Most samples k-feldspar poor, only samples CH1, 9b and 11a + sill compositional analyses and determined modes, with some differences within P-T + q plag + bt + ilm + liq + q + sill contains abundant matrix k-feldspar. 700 + ru SATURATION 700 window (lower P, T vs. higher P, T ). Some differences with conventional P-T results, plag + bt + grt + plag + bt + grt + ARY DIAGRAM ARY INTRODUCTION VS. ilm + liq + q + ky ilm + liq + H2O+ q + ky PHASE MAPS - KEY plag + bt + grt + ilm + H2O+ q + ky but within the 700-80 ºC, 7-9 Kbar window. ksp + plag + • The Ruby Range is one of several Laramide uplifted blocks of exposed Precam- grt + bt + q 650 VALUE 650

BIN plag + grt + bt MINERAL ASSEMBLAGES + ky GARNET PLAG KSPAR MAGN QuiG barometry give insight into proprage and peak-P garnet growth pressures, + ilm + st + q • brian basement in the Wyoming Province. + H2O from ~8 kbar at ~600ºC to ~12 kbar between 700-800 ºC (peak-P). H2O

k - D k - D a o a o i i r m r m

WATER

e e

i i

n n

h h

grt + bt + sill + pl + qtz ± ap ± k-spar ± o o

T T

X pl + grt + bt + ilm + st + mu + q + H2O RUTILE BIOTITE ILM SILL QTZ APAT • Metamorphosed at 2.45 Ga and 1.72-1.79 Ga during the Big Sky Orogeny, 600 600 LOI chl ± mu ± ilm ± ru ± gr ± trm. 0.00 0.25 LOI value 0.50 sat. solidus 0.75 1.00 T- 0.00 LOI value 0.25 sat. solidus 0.50 0.75 1.00 PHASE GARNET BIOTITE SILL. QUARTZ PLAG. K-SPAR ILMENITE (Harms et al., 2004; Cheney et al., 2004). 0.01 H = ‘dry’ 5.89 H 12.29 H 18.03 H = ‘wet’ 0.01 H = ‘dry’ 2.10 H 6.32 H 18.03 H = ‘wet’ [Normalized] [Normalized] composition [Normalized] [Normalized] composition composition composition modes (vol%) LOI H2O+ ROCK LOI H2O+ ROCK LOI H2O+ ROCK LOI H2O+ ROCK LOI H2O+ ROCK LOI H2O+ ROCK LOI H2O+ ROCK • This project examines metapelites along the Stone Creek drainage, from CH1 17 16 19 28 33 31 3 5 3 23 24 21 10 8 7 20 16 17 0 0 0 the structurally highest unit of the Ruby Range, the Christensen Ranch Meta- Sample CH1 SI(49.69)TI(0.39)AL(19.63)FE(10.07)MG(5.41)CA(1.14)NA(1.19)K(6.50)F3(0.08)H(5.89)O(?) SI(57.00)TI(0.98)AL(19.58)FE(9.17)MG(3.22)CA(1.52)NA(3.70)K(2.66)F3(0.08)H(2.10)O(?) Sample CH4 CH4 18 12 13 10 19 23 5 6 5 31 33 30 28 26 24 7 0 0 1.4 1.2 0.7 morphic Suite (CRMS) overlying the Dillon Gneiss and Pre-Cherry Creek suite. Table 1. Phase mode comparison in samples for calculated LOI and H2O melt saturation at solidus modeling, and quantified phase • fine-grained, homogeneous 9 • medium to coarse grained, poikiloblastic garnet modes (SEM/EDS phase map for CH1, CH4, CH9b; estimated modes for CH11a). 9 plag • K-feldspar rich matrix + ksp plag + ksp + clusters around quartz + sillimanite melt pods. • CRMS only show Proterozoic monazite ages 1.72-1.79 Ga (Cramer et al., 2014) plag + + grt grt + bt + ilm ksp + ksp + • garnet (Fe, Ca) and tourmaline (Ca) zoning plag + + bt + plag + ky + q plag + • No K-feldspar, no muscovite in matrix q + grt + bt ksp + grt ksp + grt 8 grt + ilm sill + ilm + • The purpose of this study is to determine metamorphism conditions of pres- • Rare, single-grained garnet inclusions (bt, mu, qtz, apt) + bt + q + bt + q + liq + • Flat garnet profiles, no compositional zoning but clear 8 + ky + liq liq + sill + sill + sill + q • prismatic sillimanite in matrix plag + liq + ilm + q inclusion domains. DISCUSSION sure and temperature recorded in these metapelites to further constrain and ksp + grt + bt + q P [Kbar] • Kinked fibrous sillimanite bundles + sill VALUE 7 understand the Big Sky orogeny as recorded in the Ruby Range, complement- P [Kbar] plag + ksp Results plotted on NCKFMASH grid, within consistent range, but with some 7 grt + bt + ilm • ing our understanding of this orogenic event studied in the adjacent mountain + sill + q differences, which could be due to polymetamorphism (evidence for multiple

k - D

WATER a o i r m

o T

k - D a o i r m

o T episodes of garnet growth & consumption). ranges such as the Tobacco Root Mountains or the Gravelly Range. 6 LOI 650 700 750 800 850 650 700 750 800 850 All CRMS ages 1.72-1.79 Ga: P-T conditions recorded for Big Sky orogeny in T [C] • • Petrography, mineral chemistry and thermodynamic modeling. T [C] the Ruby Range, not from metamorphic events at 2.45 Ga. PSEUDOSECTION 21’15” R7WR6W 22’30” P ~ 8.7 kbar, T ~ 730 ºC P ~ 6-9 kbar, T ~ 695-820 ºC 14 Plagioclase anorthite isopleth did not plot in the P-T window None of the isopleths plotted in the P-T window in this diagram MODELING ISSUES & NaKFMASH SOURCES OF DISCREPANCIES IN RESULTS P2 ➊ 12 MONTANA peak-P SI(46.31)TI(0.37)AL(18.29) FE(9.39)MG(5.05)CA(1.07)NA(1.11)K(6.06)F3(0.07)H(12.29)O(?) SI(54.54)TI(0.93)AL(18.73) FE(8.78) MG(3.09)CA(1.46)NA(3.54)K(2.55)F3(0.08)H(6.32)O(?) Sources of error: a-x models, no Mn, no t • y 10 plag + ksp + k 10 bt gr E1 grt + bt + apatite correction, composition analy- sp liq sill mu ab liq + q 10 as k

+ ky plag + grt sta mu sta

+ bt + ilm ses, H2O content, melt loss, polymeta liq as bt t r + liq + ky g P3 9 9 + q morphism. P1 8 ➋ peak-T MONTANA METASEDIMENTARY 110˚ If assume H2O loss from dehydra- SOLIDUS • P (Kbar) HIGHLAND SUBPROVINCE plag + CH1 - Raman shift (min & max) between matrix quartz and quartz inclusions in garnet E2 8 plag + AT 8 P4 MTNS ksp + grt ksp + grt tion melting reactions only (Spear et al., ➌ 9 + bt + liq 9 plag + grt + 6 + bt + liq bt + ilm + liq TOBACCO + q + sill matrixksp + + q + sill 1100 + sill + q Re-eq ROOT N MADISON + ilm quartzplag + 1999), use LOI diagrams, no melt be plag + ksp + plag + ksp + A = 465.09 Raman Shift MTNS plag + plag + grt-1 + bt RANGE grt + bt + ilm ksp + grt + bt + ilm ksp + ksp + ksp + cm+ ilm + -1 ky 7 ky + q plag + plag + ky + q plag + incl 6 = + 0.29 cm grt + bt 7 grt + bt liq + sill low 750-800 ºC. 8 grt + bt grt + ilm 8 grt + ilm -1 4 and O + liq+ + ilm q + + ilm + + q incl9 = + 1.38 cm 2

P [Kbar] plag + plag + liq + + liq + RUBY + liqsill + sill liq + sill SCHEMATIC STRUCTURE bt +H2O + bt sill + q 900 sill + q Melt-loss not represented in the bulk sta mu SECTION - RUBY RANGE + q + q RANGE + q + sill + liq • t bt as H GRAVELLY BEARTOOTH RANGE P [Kbar] gr sill CHRISTENSEN + mu + q + and RANCH GROUP P [Kbar] 6 sill P [Kbar] 6 rock chemistry: Melt addition sill RANGE SATURATION WATER BLACKTAIL S MADISON 7 7 2 and 45˚ plag + ksp PSEUDOSECTION plag + ksp RANGE plag + grt RANGE P grt + bt + ilm grt + bt + ilm U + bt + ilm + incl6 500 600 700 800 900 O show theoretical pre-melt loss sta R plag + bt+ sill + q + sill + q bt T (ºC) G 700 A = 465.38

H2O + sill + q -1 + H2O +

H Intensity cm C DILLON GNEISS - - q + sill k D k D HENRYS LAKE P a o a o N i i plag + grt + r m r m

e e

i i

A n h h

o o T

R T bility fields, isopleths show matching BEARTOOTH-BIGHORN O 5 plag + R 5 bt + ilm + liq MTNS N G Figure 4. Proposed P-T bath for BSo on petrogenetic SE S 6 6 MONTANA N IS ksp + grt E E incl9 MAGMATIC ST N + sill +crd + q I G R + bt + liq H 650 700 750 800 850 650 700 750A = 466.47 800 850 grid in NaKFMASH system for pelites. Light grey area = C N -1 compositions for added melt vs. LOI. IDAHO K SUBPROVINCE LO EE + q + IL R cm sill sill D C k - D a o Y crd i k - D

r m a o R i maximum P-T constraint from petrographic observations, m e

i r

500

R n

h e i

E h

0 50 km o

CH PRE-CHERRY T [C] T [C] T E- R CREEK GROUP Proposed P-T path for BSo (Fig.4). P 4 4 • colored spots = thermodynamic modeling P-T constraints

WYOMING 0 0.1 0.5 1.0 mile 600 700 800 900 600 700 800 900 q q + conventional P-T, colored triangles = QuiG barometry T [C] T [C] results. (Modified from Spear et al.,1999). Figure 1. Sample distribution and location map along Stone Creek. Metapelite samples were generally sampled in the undifferentiated Garnets for QuiG barometry: 480 470 460 450 P ~ 9 kbar, T ~ 795 ºC P ~ 6.8Wavelength kbar, absorbance T ~ A 700(cm-1) ºC metasedimentary units (mapped in pink), a few samples were collected in the amphibolite unit (mapped in green). Other units mapped are • preserved garnet zoning in CH1 garnets: high-Ca and Garnet alm isopleths were not calculated for this diagram Garnet alm isopleths were not calculated for this diagram Garnets for QuiG barometry: calcitic and dolomitic marbles (blue), metamorphosed banded iron formation (purple) and quartzite (magenta). Note the contact with the Dillon low-Fe euhedral garnet cores (garnet gowth 1), low-Ca and CONCLUSION & SIGNIFICANCE gneisses (olive green) to the left of the map. (Modified from James, 1990). CH1 - Raman shift (min & max) between matrix quartz and quartz inclusions in garnet CH4 - Raman shift between open/matrix quartz and quartz inclusions in garnet • Inclusions patterns indicative of two garnet growth high-Fe garnet rims (garnet growth2) with quartz inclusions episodes: plagioclase inclusions in cores vs. ilmenite matrix in both domains. 1100 600 matrix quartz Implications for the Big Sky orogeny in the Ruby Range: constraint quartz A = 465.00 Raman Shift inclusions in rims or ears, strong inclusion trails. quartz • A = 465.09 Raman Shift -1 -1 cm -1 cm incl 6 = + 0.29 cm-1 incl 5 = + 0.48 cm inclusions in both subdomains on peak-T, peak-P, and re-equilibrium P-T conditions. -1 METHODS incl9 = + 1.38 cm-1 incl 4 = + 0.94 cm incl9bis = + 1.10 cm-1 • 21 metapelite samples collected in summer of 2014 for KECK project, along rim 900 500 Calculated reequilibrium temperature of ~700 ºC at pressure of ~7 kbar. incl4 • A = 465.94 -1 r Stone Creek drainage, structurally covering base to top of the CRMS (Figure 1). cm im

incl6 Intensity i Calculation results affected by bulk rock composition H2O content 700 A = 465.38 lm • -1 400 • All samples have mineral assemblage garnet + biotite + sillimanite + quartz. Outcrop Intensity cm . FOR QUIG BAROMETRY in estimations, mineral composition measurements, mineral a-x models, melt incl5 c incl9 A = 465.48 l and sample positions recorded in NAD 84 UTM coordinates using a GPS. Samples -1 A = 466.47 cm loss and polymetamorphism. grt -1 cm and field relationships with surrounding units were described in the field. 500 core 300 Result for peak-T and peak-P for the Big Sky orogeny comaprable to result in RAMAN SPECTRA incl9bis • RAMAN SHIFT A = 466.10 FOR PSEUDOSECTION GEOTHERMOBAROMETRY: cm-1 grt the Tobacco Root Mountains, recorded in pelitic schists from the Indian 480 470 460 450 475 470 465 460 455 (1) 25 sections were made, phases, modes and reaction textures identified, described Wavelength absorbance A (cm-1) Wavelength absorbance A (cm-1) Creek Metasedimentary Suite (Cheney et al., 2004) despite not being and quantified using a petrographic microscope. equivalent (ICMS 1st cycle vs. ICMS 2nd cycle metasediments). (2) Bulk rock chemistry for major and trace elements (X-Ray Fluorescence, XRF) and 14CH4 - Raman shift between open/matrix quartz and quartz inclusions in garnet 14 600 matrix quartz REE (Inductively Coupled Plasma Mass Spectrometry, ICPMS) was obtained through Sample CH1 A = 465.00 Raman Shift Sample CH4 ACKNOWLEDGEMENTS cm-1 incl 5 = + 0.48 cm-1 I would like to thank my advisor, John B. Brady, for putting the time and energy in this project, for letting me make Acme Labs, Vancouver Canada for 9 samples. incl 4 = + 0.94 cm-1 my own decisions about this research and giving me a sense of ownership in the work I am doing, making this qtz incl. -1 qtz incl. Ca-poor 12 grt rims incl9bis = + 1.10 cm 12 an intensive learning and growing collaborative research experience. I am also extremely thankful to Julia A. (3) 5 samples analyzed using and FEI Quanta 450 SEM with EDAX-EDS. Garnet, biotite 500 grt rims Baldwin, for helping me with the Theriak Domino modeling and for giving me precious advice on how to think rim incl4 ky ky about rocks in the Ruby Range, in the field and in the lab; Tekla A. Harms for her patience and understanding and plagioclase mineral chemistry was quantified, modal phase maps were A = 465.94 cm-1 sill sill with technical issues, eternal support and motivation to keep going and not get discouraged. I would also like to thank Jack Cheney at Amherst for K-feldspar staining, Frank Spear and his graduate students at RPI for inviting generated. Intensity ADDED MELT, me to use the Raman spectrometer for QuiG, Mike Vollinger for the cookies and Judith Whopereis for her help 40010 10 core and technical support with the SEM/EDS system. Finally, I would like to thank the KECK Geology Consortium and incl5 the Ruby Rangers, for making exciting research in the field and in the lab possible, and for fun times in Montana! • Pseudosection modeling for 5 samples in Na-Ca-K-Fe-Mg-Al-Si-H-Ti-O (NCKFMASHTO) A = 465.48 Ca-rich -1 cm plag incl. QUIG RESULTS system with TheriakDomino v. 03.08.2009 & pelites (Holland and Powell dataset 5.5), core • 300 REFERENCES P (kbar) incl9bis P (kbar) for LOI water value vs. H2O determined using a T-X H2O binary diagram calculated in A = 466.10 • Baldwin, J.A., Cramer, M.A., Voarino, B., 2014, Late Archean to Paleoproterozoic evolution • Harms, T.A., Brady, J.B., Burger, H.R, and Cheney, J.T. (2004) Advances in the geology of the 8 cm-1 8 qtz incl. GTB MODELING +30% of the Ruby Range, Montana, and its implications for the growth and modification Tobacco Root Mountains, Montana, and their implications for the history of the northern

qtz incl. • Theriak Domino vs. melt-addition modeling from ‘wet’ peraluminous granite 475 470 465 460 455 grt cores P-T of the Northwestern Wyoming Province: Geological Society of America Abstracts Wyoming Province, in Brady, J.B., Burger, H.R., Cheney, J.T., and Harms, T.A., eds. grt cores Wavelength absorbance A (cm-1) with Programs, v. 46, No 5, p. 20. Precambrian Geology of the Tobacco Root Mountains, Montana: Geological Society of plag +ksp + America Special Paper 377, p. 227-243. plag + grt + bt + (granite + 25% H2O) No apatite correction was done, Mn was left out and small ferric (CH4) DIAGRAM • Cheney, J.T., Brady, J.B., *Tierney, K.A., *DeGraff, K.A., *Mohlman, H.K., *Frisch, J.D., *Hatch, ksp + ilm + liq + grt + bt C.E., *Steiner, M.L., *Tuit, C., *Steffan, K.J., *Cady, P., *Lowell, J., *Archuleta, L.L., *Hirst, plag + grt + sill + q • Kohn, M. J., (2014) “Thermoba-Raman-try”: Calibration of spectroscopic barometers and + liq + bt + ilm + J., *Carmichael, S.K., *Fisher, R.G.M., *Wegmann, K.W., and Monteleone, B. (2004) q + sill thermometers for mineral inclusions: Earth and Planetary Science Letters 388, p 187- iron values were used. 6 GRANITE (CH1) OR SAT.SOL PERALUMINOUS 6 liq + sill + q Proterozoic metamorphism of the Tobacco Root Mountains, Montana, in Brady, J.B., COMPOSITE DIAGRAM COMPOSITE RESULTS GTB 196. plag + bt + Burger, H.R., Cheney, J.T., and Harms, T.A., eds., Precambrian Geology of the Tobacco • Stable mineral assemblages, actual modes vs. calculated modes, and mineral mu + q + sill plag + ksp Root Mountains, Montana: Geological Society of America Special Paper 377, 105- • Powell, R, & Holland, TJB, 1988 An internally consistent thermodynamic dataset with + bt + liq + 129. uncertainties and correlations: 3: application methods, worked examples and a q + sill computer program. Journal of Metamorphic Geology 6, 173–204. compositions compared for LOI vs. T-X H2O determined H2O value. • Cramer, M.A., Voarino, B., and Baldwin, J.A. (2013) Metamorphic evolution of Precambrian Figure 2. CH1 Sample Petrography: Phase maps, thin section maps and textural Figure 3. CH4 Sample Petrography: Phase maps, thin section maps and textural basement in the southern Ruby Range, SW Montana in Geological Society of America • James, H.J., (1990) Plate 1: Geologic map of the Southwestern Ruby Range, Montana in • Conventional geothermobarometry using Sprear & Kohn’s GTB program. 4 4 Abstracts with Programs, 45, 601. Precambrian Geology and Bedded Iron Deposits of the Southwestern Ruby Range, details - top: phase map and PPL thin section of CH1, prismatic sillimanite 600 700 800 900 600 700 800 900 details - top: phase map and PPL thin section of CH1, fibrous sillimanite texture, Montana: US Geological Survey Professional Paper 1495, 44. • De Capitani C. and Petrakakis K. (2010): The computation of equilibrium assemblage • Quartz-in-garnet (QuiG) barometry (Kohn, 2013), using GeobaRamanTry program. texture, PPL & XPL; bottom: two episodes of garnet growth, euhedral core T (ºC) T (ºC) PPL & XPL; bottom: two garnet growth episodes, plagioclase-free, ilmenite- diagrams with Theriak/Domino software. American Mineralogist 95:1006-1016 • Spear, Frank A., Kohn, Matthew, J., Cheney, John T. (1999) P-T paths from anatectic pelites. and crinkled rim (XPL and Ca-zoning EDS map), rich rim inclusions (PPL) and plagioclase-rich core inclusions (XPL), Contributions to Mineralogy and Petrology 134: 17-32