sign in sign up
<<
Home , Anthophyllite, Gedrite, Tschermakite

AmericanMineralogist, Volume 65, pages 1103-lI //8, 1980

The - solvus and the composition limits of orthoamphibote from the Post Pond Volcanics,Vermont

FneNr S. Speln

Department of Earth and Planetary Sciences Massachusetts Institute of Technolo gy Cambridge, Massachusetts 02 I 39

Abstract

Coexisting orthoalnphiboles (high Al-Na gedrites and low Al-Na anthophyllites) from the Post Pond Volcanics, Vermont, display a variety of textures that are interpreted to be equilib- rium growth textures. These textures include: (l) discrete grains; (2) bladed intergrowths; (3) patchy intergrowths; (4) overgrofihs; and (5) lamellar intergrowths with lamellae parallel to (120). Coexistingorthoamphiboles display exsolution parallel to (010); exsolution is much more pronounced in gedrite than in anthophyllite, which suggeststhat the solvus is steeperon the anthophyllite limb. The solvus is defaed by discontinuities in the amount of Si, AlIv, Alvr, and Na(A) in the orthoamphiboles, which can be describedas gapsin the edenite (NaAl?2 n Si) and tscherma- kite (AlvrAllvaMgSi) substitutions.The width of the solvusis a function of Fel(Fe+Mg), with Fe-rich samplesshowing the widest gap. Distribution of elementsbetween anthophyllite and gedrite is systematic,with gedrite enriched in Alrv, AlvI, Fe2*, Na, Ti Ca, and Mn and anthophyllite enriched in Si and Mg. The composition timits of orthoamphiboles at this metamorphic grade are controlled by coexistencewith phasesmore enriched in certain ele- ments: the most Si-rich orthoamphiboles coexist with , the most Al-rich with or , the most Fe-rich with , the most Mg-rich with , chlorite or cordierite, the most Ca-rich with , and the most Ti-rich with ilmenite or rutile. Evaluation of published orthoamphibole analysesreveals that the crest of the solvus prob- ably lies at approximately6fi)+25"C.

Introduction This paper describesthe textures, chemis- The existence of a solvus in the orthoamphiboles try, and phase relations of orthoamphiboles and co- between low-Al anthophyllite and hiCh-Al gedrite existing from the Post Pond Volcanics with has been noted by Robinson et al. (1969, 1970, the purpose of (l) outlining the compositionalfield l91la), Ross et al. (1969), Stout (1969, t970, 1971, defined by these minerals, including the limits of the 1972),Spear (1977), and Sroddard(1979). miscibility gap between anthophyllite and gedrite; (2) In the southwesterncorner of the Vermont section investigating substitution mechanisms in ortho- (3) of the Mt. Cube Quadrangle, New Hampshire and ; and estimating the temperature of Vermont, the PostPond Volcanic Member of the Or- the crest of the anthophyllite-gedrite solvus.' fordville Formation consists of a variety of rock types Geologic setting that contain coexistingorthoamphiboles. The locality is particularly well suited for a chemographic study The samples were collected from the Post Pond of these minerals, becausewithin this one area a Member of the Orfordville Formation (Hadley,1942) highly diverse range of bulk compositions is found, from the Vermont portion of the southwestcorner of and the entire composition range of orthoamphiboles the Mt. Cube Quadrangle, New flampshire and Ver- at this metamorphic grade can be delineated. More- mont (seelocation map, Fig. l). This unit strikes to over, the solvus in the orthoamphiboles from the Mt. I In tlis paper the names anthophyllite and gedrite are used to Cube Quadrangle shows the widest compositional describe varieties of orthoamphiboles with Na(A)+Alrv<1.0 and gap yet reported for theseminerals. Na(A)+Alrv>1.0, following the nomenclaturcof Leake (1978). 0003-o04x/80/lI 12-l 103$02.00 I 103 I 104 SPEAR GEDRITE-ANTHOPHYLLITE SOLVUS

EXPLANATION [-G-l uowgroderock, undifferenlioled l-Dl-l Littteton Formotion lEl ctougn Formolion OrfordvilleFormolion BlockSchist Member 43" PoslPond Volconics Member 47' 5s,>, Strike ond dip of composifionolloy€ring 9 Strir" ond dip of schistositY 461 Trend ond plunge of minor folds . ?z-37 Somple locolion

43" Locotronof Mt.Cube Quodrongle

Fig. l. Geologic map of the southwesterncorner of the Mt. Cube Quadrangle, New Hampshire and Vermon! showing the location of the study area. Geology is after Hadley (1942) arrd Rumble (1969, personal communication). the northeastand dips to the northwestat 39-73o.To higtrly aluminous layers are found, which probably the west, and structurally above the Post Pond Vol- represent volcanic material with an admixture of de- canics, are the pelitic schistsof the Brick Hill Mem- tritus. To the west, at the upper part of the section' ber of the Orfordville Formation (Hadley, 1942). just below the Siluro-Devonian unconformity, there Rumble (1969) reported quartzite boudins from are cordierite and orthoamphibole gneisseshigh in along the contact between the Post Pond Volcanics MgO and low in CaO. The rocks are massive to and Brick Hill Member, which he interpreted as Silu- finely laminated and possibly represent a fossil rian Clough Formation, and therefore correlated the leached zone or soil horizon at the top ofthe volcanic Post Pond Volcanics with the middle Ordovician pile, or a volcanic sequence that has been hydro- Ammonoosuc Volcanics, which occupy the same thermally altered, possibly by seawater. stratigraphic position to the east (ThomFson et aI., Metamorphic grade is in the staurolite- 1968). zone, as shown by abundant staurolite and kyanite in In the southwest corner of the Mt. Cube Quad- the Brick Hill Member and staurolite in the Post rangle, the Post Pond Volcanics are composed of Pond Volcanics. Metamorphic temperatureswere es- massive to finely bedded mafic to aluminous am- timated with two geothermometers.Oxygen isoto,pe phibolites, with minor calc-sitcates, , fractionation between quartz and magnetite (1000 orthoamphibole gneiss, and cordierite gneiss. There tna : 9.8+0.11)yields a temperatureof 530+10"C are mafic dikes and agglomeratetextures in the rocks when the calibration of Downs and Deines (1978) is exposedin two quarries near InterstateHighway 91, used and 490+10"C when using the calibration of attesting to the original volcanic nature of at least Bottinga and Javoy (1973). The partitioning of Fe part of this unit. In the roadcut along U.S. Route 5, and Mg betweengarnet and biotite yielded a temper- SPEAK GEDRITE-ANT:HOPHYLLITE SOLVUS ature of 535+30"C when comparedto the calibration Mg-rich samples [Fel(Fe+Mg) in anthophyllite of Ferry and Spear(1978). If one assumesa temper- =0.301, gedrite and anthophyllite are di-fficult to dis- ature of 530oC, the presenceof kyanite implies a tinguish on the basis of ; distinction was minimgn pressure of 4.0 kbar, from the ALSiO, made in these sampleswith the electron microprobe. triple point of Holdaway (1971). The sampleswere Anthophyllite and gedrite coexist in a variety of collected from an area approximately 2 km by I km textures. They occur as discrete grains in some sam- 'ples, and are believedto have been equibratedunder ap- and sharp optical contacts can be seen where proximately isothermal, isobaric conditions. the grains touch. They also occur as oriented inter- growths that are optically continuous and exhibit Textural relations betweenanthophyllite and gedrite several habits including bladed intergrowths (Fig. Orthoamphibolesform large bladed crystalstypi- 2A), patchy intergrowths, and overgrowths (Fig. 2B). cally 0.1-5.0 mm long, but up to severalcm in length In most occurrenoesthe overgrowth is anthophyllite in rocks of diverse bulk comFosition and mineral as- on gedrite, but in a few samplesgedrite overgrowths semblages(see Table l). They occur in the foliation appear on anthophyllite. plane but also cut acrossit, indicating that thesemin- In several samples a lamellar intergrowth of an- erals may have formed late in the crystallization se- thophyllite and gedrite is observed(see Figs. 2B, C, quence. For comparison, hornblende and cum- D). Lamellae are oriented approximately parallel to mingtonite from this area almost invariably lie in the (120) and are readily observedin sectionscut normal foliation. to c. In Figures 28 and C the in the core The orthoamphiboles are distinguished from the of the grain is gedrite, whereas anthophyllite makes clinoamphibolesby their parallel extinction. Anthoph- up the rim. This is the case in most of the samples yllite is colorless and can be distinguished from Fe- observed with this intergrowth texture, but in some rich gedrite by the moderate pleochroism in gedrite. samplesthe textural juxtaposition is reversed.Figure There is a correlation between the degree of ab- 2D shows the distribution of Al between the an- sorption in gedriteand its composition.The strongest thophyllite rim (ow Al) and gedrite core (high Al). pleochroism is observed in A1-, Na-, and Fe-rich The (120) lamellae are readily observedin this figure gedrites; the weakest in gedrites with high Mg and becauseof the sharp contrast in Al content between low Al and Na. In Fe-rich samplesthe absorptionis the lamellae and the anthophyllite rim. Electron mi- pale-gray to yellow-gray parallel to a and gray to croprobe analysis revealed that the lamellae have ap- blue-gray parallel to B and y (absorptionis y > B > proximately the same composition as the core of the c). In more Mg-rich samplesthe pleochroism scheme grain (re., gedrite composition). in gedrite is similar, but much weaker. In the most The origin of this textural intergrowth is not

Table l. Representativenineral assemblagesof orthoamphibole-bearing rocks from the Post Pond Volcanics

Sample /l Ged** Anth Hbld Cum Gar Stau Cord Plag O.Ez Biot Naph Talc Ch Iln Rut

7 3-20D +* + ff ^3, x* x 73-30r + ++ An^- x x J) 77-56c + ++ An^^ x x JJ 77-37N + + h29-40 * xx 68-43211 An^^ x x x JJ 68-432U h15-29 * x x 68-432D + + Atro-30 * xxx x 7 3-29D + ++ At3o x 68-432A + ++ ^27 x

* The symbol + indicates that an analysis is presented in Tables 2-5. The symbol x indicates the mlneral is present but no analysls is presented. ** Abbrevlations: ced=gedrite; Anth=anthophyllite; Hbld=hornblende; Cum=cumingtonltei Gar=garnet; Stau=staurolite; Cord=cordierlte; Plag=plagloclase; Qtz=quartz; Biot=biot lt e; Naph=Na-phlogop 1t e; Ch=chlorite; Iln=ilmenlte; Rut=rutl1e. I 106 SPEAR: GEDRITE-ANTHOPHYLLITE SOLVUS

Fig. 2. Photomicrographs and X-ray photograph ofanthophyllite-gedrite intergrowths. (A) Bladed intergrowth ofanthophyllite and gedrite. @) Overgrowth of anthophyllite showing (120) lamellar intergrowth texture. (C) Intergrowth of anthophyllite and gedrite showing (120) lamellar intergrowth. Note exsolution lamellae (trending NW-SE) cutting (120) lamellae of gedrite. (D) X-ray photograph (AlKc radiation) for lamellar intergrowth shown in (C). Symbols are G : gedrite, A: anthophyllite, P: plagioclase,H : hornblende, B : biotite, C : chlorite. known, but it is apparently structurally controlled always coarser in gedrite than in coexisting anthoph- and related to the growth of the , inasmuch as yllite. In some samples,exsolution is present in ged- the lamellae grow nonnal to the faces and ex- rite whereas the coexisting anthophyllite displays no solution lamellae cut across the (120) lamellae (see exsolution(Figs. 2C and 3A). In Figure 38 it can be below). It should be noted that not all anthophyllite- seen that the exsolution lamellae cut directly pcross gedrite overgrowths exhibit the lamellar texture, but the larnellar texture and are not at all affected by its no systematic relation between the presence or ab- presence.This observation suggeststhat the lamellar seneeof lamellar intergrowths and minglsl composi- intergrowth formed before exsolution took place, tion could be detected. A sinilar lamellar inter- which is consistent with it having formed during growth between anthophyllite and gedrite on a growth of the crystals. submicroscopic scale was observed by Gittos er a/. (1976,their Fig. 3) with transmissionelectron micros- Mineral chemistrv copy. In the samplestudied by Gittos et al. the (l2O) lamellae resulted from processesrelated to ex- Analytical procedures solution. Data on representativemineral assemblagesare All grains of coexisting anthophyllite and gedrite tabulated in Table l. All chemical analyses were exhibit exsolutionparallel to (010) (seeFigs. 2C and made with €rnautomated MAC electron nicroprobe 3); the exsolution is thus most readily observedwhen either at the Geophysical Laboratory, Carnegie Insti- viewing sectionscut normal to c, but can also be ob- tution o1rya5hington,or in the Department of Earth served when viewing sections cut normal to a. Ex- and Planetary Sciences, MassachusettsInstitute of solution lamellae are less than I pm wide and are Technology. Natural and synthetic silicates were SPEAR: GEDRITE-ANTHOPHYLLITE SOLVUS I 107

Fig. 3. Photomicrographs showing exsolution features in gedrite and anthophyllite. (A) Exsolution lamellae (the light and dark lines trending E-W) are visible in gedrite but not in anthophyllite. (B) Exsolution lamellae (rending N-S) cut (120) lamellae of gedrite but are not present in anthophyllite. Fig. 3B is the same grain as shown in Fig. 28. Symbols are G : gedrite, A : anthophyllite. usedas standards;matrix correctionswere performed blende, where Fe'* was calculated from stoichiomet- after the proceduresof Benceand Albee (1968)using ric constraints following the procedure in Spear theoreticalcorrection factors (Albee and Ray, 1970). (1980).Cations were assignedto structural positions Analysesare estimatedto be accurateto within +3 as follows: all Si and sufficientAl to sum to 8.0 in the relative weight Vofor major elementsand i5 relative tetrahedral site; remaining Al, all Ti, Mg, Fe, Mn, weight Vofor mir.or elements.Several hundred analy- Ca, and sufficientNa to sum to 7.0 in the M,, Mr, Mr, sesfrom over 40 sampleswere made; representative and Mo octahedral sites;remaining Na in the alkali analysesare presentedin Tables 2-5. No KrO was site. detectedin any analysisof orthoamphibolefrom the study area (detectability limit =0.01 wt.Vo).Cations Results were calculated on the anhydrous basis of 23 oxy- Sample 73-29D contains gedritesthat are unusu- gens.All Fe was assumedto be Fe2*except in horn- ally rich in AlrO, and NarO, and are characterizedby

Table 2. Representativeelectron microprobe analysesof orthoamphiboles

73-20D 73-30r 71-56C 77-31N 68-432M (ta-43211 6a-432D 73-29D 6A-432A GAGA GA G

Weight Z oxides

si02 42.80 54.r2 46.4I 54.43 46.07 55.22 44.83 54.r0 47.07 55.42 48,20 55.18 49,L7 57,05 40.91 43.44 [zo: 15.52 t,75 L4.29 2.35 14.04 2.33 16. 81 2 .88 14.30 3.45 t4,4L 4.t2 14.11 2.'10 19.00 18,08 Ti02 0. 18 0.02 0.16 0.04 0.t7 0.04 0.18 0.05 0.17 0.05 0. 13 0.05 0.13 0.03 0.18 0,16 MCo 11.73 17.60 15.21 19.85 15. 38 20.11 15.53 20.Ut 16.93 21.08 18,14 21.97 19,90 24.40 11.4.3 16,15 Ieo 25,89 25.36 20.7r 2t.L3 20.09 20,34 18.74 19.82 17.85 18.48 76,t4 16,45 t4.L4 15.22 24.93 17.30 Mno 0.35 0.28 0.36 0.31 0.43 0.42 0.22 0.20 0.25 0.22 0.30 0,30 0.55 0,52 0.17 0.25 Cao 0.A7 0,29 0.47 0.35 0.68 0.31 0.45 0.36 0.56 0.39 0.35 0.31 0.24 0,11 0.40 0.29 Na^o 7.94 0.12 7.73 0.20 1.51 0.13 1.38 0.20 r.74 0.39 1,68 0.38 1.40 0.17 2.38 2.O5 98.88 99.54 99.40 98.67 98.37 98.96 98.24 97.93 98.87 99.49 99.35 98,76 99.64 LOO.20 99.40 97.72

Cations pe! 23 oxygens Si 6.346 1.806 6.634 7.767 6.643 7,817 6.4L7 1.730 6.670 7,725 6.724 7,684 6.161 7.778 6,014 6.239 ^,rv 1.654 0.194 1.366 o.233 1.357 0.183 1.583 0.270 I.330 0.275 L.276 0.316 L,233 0.222 1.986 t.76I ,. VI r.059 0.104 r.042 o.162 1.030 0.206 L.254 0.2t5 1.057 0.293 1.094 0, 350 1.054 0,210 1.307 1.300 T1 0.020 0.002 0.017 0.004 0. 018 0.004 0.019 0.005 0.017 0.004 0.014 0.005 0,012 0.001 0.020 0.0r7 Mg 2.592 3.783 3.253 4.222 3.305 4,242 3.334 4.288 3.516 4.382 3,171 4,559 4.084 4.958 2.504 3.457

Ie 3.2IO 3.059 2.t476 2.522 2.423 2 .408 2.243 2.368 2.tr2 2.t57 1.883 1.916 I.624 r.732 3.065 2.078 Mn 0.044 0.034 0.044 0.037 0.053 0.050 0.021 0.024 0.029 0.024 0.035 0,035 0.061 0.056 0.02r 0,030 Ca 0.075 0.045 0,072 0.055 0.105 0,056 0.059 0.055 0.083 0.057 0,052 0.046 0.032 0.016 0.063 0.045 Na(M4) 0.0 0.0 0.096 0.0 0.005 0.034 0,054 0.025 0.126 0.088 0.151 0.079 0,133 0,027 0.020 0.073 Na(A) 0.558 0.034 0, 384 0.055 0. 356 0.002 0,329 0.030 0.352 0.014 0. 303 0,024 0.239 0.017 0.658 0.498

Fe/Fe+Mg 0.553 o.441 0.432 0.374 0.423 0,362 0.402 0.356 0.371 0.329 0.333 0.296 0.285 0.259 0,550 0,373 I 108 SPEAK GEDRITE-ANTHOPHYLLITE SOLVUS

Table 3. Representative electron microprobe analyses of nisms: Fe S Mg (Fe-Mg exchange),AlvrAlrv lT hornblende MgSi (tschermak exchange) and NaAl'' S ! Si (edeniteexchange). Other substitutions,such as those 73- 73- 77- 68- 73- 20D 30r 56c 4321'r 29D involving Mn, Ti, Fe'*, or Ca are of minor impor- tance. Weight % oxides Gedrite differs from anthophyllite primarily in the si02 42.66 45.04 43.63 45.22 41.08 amount of edenite and tschermak substitutions: the A1203 16,30 16.82 16,29 L5,43 L1.98 miscibility gap betweenthese two minerals represents Tio2 0.45 0.53 0.42 0.42 0.37 a discontinuity in the amount of these two sub- Mgo 8.45 10.88 10.89 r2.L6 7.46 (1963,p.2la) FeO 18.86 13.60 14.64 13.07 18.85 stitutions.Deer, Howie and Zussman MnO o.L2 0.15 0.18 0.11 0.11 report an end-member formula for gedrite of Cao r0 .23 10, 50 10.30 10.39 10.15 Naro L.75 1.85 I.13 z.LO 1.98 Mg,AU'A$"SLO",(OH)", but Robinson et al. (l97la) K2o 0. 19 0 .2t 0. 15 0, 14 o.24 have emphasizedthe importance of Na in the A site 99.01 99. 58 98.24 99.O4 98.22 of gedrite and have postulated an ideal end-member

Cations per 23 oxygens gedrite composition of 5a 6 .200 6. 355' 6.262 6. 399 6. 040 Nao.r(Mg,Fe2*)r(Mg,Fe2*)rrAl, rSLAlrOrr(OH), etrv 1.800 L.64s 1.738 1.601 1.960 .nrvr 0.993 1.153 1.018 0.973 1.157 based on evaluation of all available wet-chemical Ti 0.049 0.056 0.045 0.04s 0.041 analyses. This formula represents a combination of Fe' 0.63s 0.422 0.608 0,s03 0.583 Mg 1.830 2,288 2.329 2.564 1.635 the edenite and tschermak substitution in a ratio of -2+ !e r.65'1 1.182 1.150 r.O44 L.735 one to three. Figure 4 is a plot of A-site occupancy vJ. Mn 0. 015 0. 018 0.022 0.013 0. 014 gedrites and anthophyllites from the study 1.593 r..587 1.584 1.575 1, 599 Al'' for Na (!14) 0.228 0.294 0.244 0.283 0,236 area. Only selectedcoexisting pairs have been shown Na(A) O:265 0.2\2 0.237 0.293 0.329 for the sake of clarity. The trend from anthophyllite 0.035 0.038 0.029 0.025 0.045 TA 0.300 0.2s0 0.266 0.318 0.374 to gedrite is very pronouncedand is, in general,to- "ideal" gedrite as defined r"3+/t"3++Mg0.277 0.263 0.346 0.325 0.252 wards the composition of r"2+/r"2++ttg0.475 0,341 0.331 o.2gg 0.515

* Fe3* detemlned from stoichiometry Table4. RepresentativeffiT;l*-*roprobeanalyses of

73- 73- I t- b6- a very strong blue-gray pleochroism. They are 20D 30r 55C 37lr 4321J slightly lessaluminous (Al,O' : 19.00to l9.ll wt.Vo) than a gedrite reportedby Robinson and Jaffe (1969, Weight Z oxides sampleI34I; AlrO, : l9.8l7o). The NaoO content of sio2 53.84 56.04 55.32 55.51 57.15 thesegedrites, however, is greater(2.27 to 2.38 wt.Vo) dz03 1. 61 0. 90 1.o0 1.03 1.18 than that of sampleI34I (Na,O : l.92%o\.The alumi- Tio2 0.01 0.01 o.o2 0.0 o.o2 MCO L7.44 L9..56 20.04 20.24 2t.99 nian ferroanthophyllite reported by Seki and Yam- Feo 25.63 2L.9L 20.75 20.82 18.09 asaki (1957)has Al,O, and NarO contentsof 19.47- Iho 0.19 0.37 0.40 0.23 0.29 CaO 0.41 0.51 0,54 0.38 0.38 19.72trtt.Vo and 1.08-1.16wt.Vo, respectively. To my Na20 0.15 O.OZ O.02 0.04 0.07 knowledge, the gedrites from sample 73-29D have 99.28 99.32 98.09 98.25 99.16 the highest NarO content of any orthoamphiboles re- Cations per 23 oxygens ported in the literaturc (cf., Rabbitt, 1948; Robinson si 7,802 7.954 7.923 7.927 7.964 and Jafe, 1969; Robinson et al., l97la; Lal and eLrv 0.198 0.046 o.oi\ 0.073 0. 036 Moorhouse, 1969; Gable and Sims, 1970). These Alvr 0.077 o.ro5 o.og2 o.1oo 0. 158 highly aluminqus gedrites coexist with staurolite, Ti 0. 001 0 . 001 0. 002 0. 0 0. 002 which is the most Al-rich phase to coexist with ged- Ms 3.761 4.138 4.277 4.308 4.568 Fe 3.106 2.601 2.485 2.487 2.t09 rite at these metamorphic conditions (staurolite- }'tn O.O23 0.044 0.049 0.028 0.034 kyanite zone). Ca 0.064 0.078 0.083 0.058 0.057 Na 0.042 0.006 0.006 0.011 0. 019 The major chemical variations in orthoamphiboles x 15.080 14.973 L4.994 74.992 L4.947 can be describedin terms of end-member anthophyl- Fe/Fe+Mg0.452 0.386 0.367 0.365 0.316 lite [Mg,SirO,, (OH)r] plus three substitution mecha- SPEAR: GEDRITE-ANTHOPHYLLITE SOLVUS I 109

Table 5. Representative €lectron microprobe analyses of garnet, by Robinson et al. There is considerable fanning of staurolite. and cordierite the tie lines, however;this representsreal differences in the A-site occupancy between di-fferent mineral Gatnet SEeurolite Cordlerite 73- 77- t3- 68- 13- 68- 68- pairs. In general, the more Mg-rich pairs have the 20D 37N 29D 432A 29D 432D 432A lower A-site occupancies,as can be seenin Figure 5. Thus the notion of a linear trend from anthophyllite Weight Z oxides

s i02 37.r1 39.16 38.55 27.88 27.5I 51.20 49.68 to "ideal" gedrite must be modified to include varia- A1203 21.46 21.38 21.51 52.a7 52.I5 33.43 32.84 tions in FelMg. The apparent tendency of horn- Ti02 0.08 0.01 0.0 0,67 0.50 0.0 0.0 blende-absent assemblagesto have relatively low A- Mgo 4.11 5.55 4,15 2.79 2.33 11.77 11.14 !eO 34.39 30.57 34.57 12.74 14.45 2.70 3.52 site occupanciesreflects the low Fe/Mg content of Mn0 o.82 1.13 0.58 0.14 0.06 0.05 0.05 Z^O n.d.* n.d. n.d. 0.43 1.25 n,d. n,d. these samples(see Fig. 5) rather than a crystal-chem- CaO 2.72 2,86 2.19 0.0 0.0 0.02 0.0 Na2O 0.05 0.0 n.d. 0.0 0.0 0.12 0.16 ical effect causedby low CaO contents. 100.73 100.76 r02.18 97.46 98.25 99.30 97.39 16s miscibility gap between anthophyllite and

Cations gedrite can be clearly seenin Figures4,5,6, and 7. It -----1?-erycgs:------19-g=yeel:-- -19-eseesq__- is manifested by a large discontinuity in the A-site 2.953 3.050 3.009 7.753 7,692 5.074 5.047 occupancy and in the Alvr and Al'' contents; that is, 2.013 1.963 r.979 17.315 17.189 3.905 3.936 Ti 0.005 0.001 0.0 0.140 0.105 0.0 0.0 along the edenite and tschermakite substitution di- o.481 0,644 0.483 r.157 0.97t r.736 1.687 Fe 2,289 1.998 2.257 2.964 3.379 0.224 0.299 rections. Mo 0.055 0.075 0.038 0.033 0.015 0.002 0.004 A-site occupancy in anthophyllite rangesfrom ap- Zn n.d. n.d. n.d, 0,088 0.258 n,d. n.d. 0.232 0.239 0.233 0.0 0.0 0.001 0.0 proximately 0.0 to 0.1 cations; in coexistinggedrite Na 0.002 0.0 n.d. 0.0 0.0 0.023 0.032 I 8.037 7.910 8.004 29.450 29.608 10.964 11.005 thesevalues range from 0.25 to 0.55. Values of Al'I

FelFe+Ug O.A24 0,756 O A24 0.719 0.711 0.114 0.151 range from 0.36 to 0.10 in anthophyllite and 1.24to

* n.d. means noc determined 0.92 n gedrite; Al'" is 0.2+0.1 in anthophyllite and l.4t0.2 in gedrite. Some of the scatterin maximum

o.7

o.6

zbo. *< oEo. o

CnHo. & o.2

v o.2 0.4 0.6 0.8 LO t.2 t.4 t.6 1.8 20 Allv ANTHOPHYLLITE "GEDRlrE"

Fig. 4. Plot of A-site occupancy vs. Alrv in gedrite (squares) and anthophyllite (circles). Tielines connect coexisting phases. Filled symbolsareforsamplescontaininghomblende;opensymbolsareforsampleswithouthornblende.Thesymbo|s plotting positions of gedrite and anthophyllite (connected by dashed line) with Fe3*/(Fe3* + Fe2+) calculated at lWo arrd 25%, respectively. 1ilo SPEAR GEDRITE-ANTHOPHYLLITE SOLYUS

o z

II l o o o 1! F a I

Fe/ Ge+Mg) Fig. 5. A-site occupancy vs. Fe,/(Fe+Mg) for coexisting anthophyllite and gedrite. Symbols as in Fig. 4.

and minimum Al content of gedrite and anthophyl- all as Fe2*. Coexisting anthophyllite and gedrite lite is brought about because of the pervasive ex- from an iron-rich bulk composition and a magne- solution and the difficulty in measuring the bulk Al sium-rich bulk composition have been recalculated, content of phases containing exsolution lamellae. assuming l$Voand25Vo feric iron in each phase.The The above values of A-site occupancy, Al"t, and Alt' results are indicated in Figures 4-7 by the (+) and encompassthe range in composition for coexisting (x) symbols connectedby dashed lines to the respec- gedrite and anthophyllite only. Orthoamphiboles not tive orthoamphibole. If comparable amounts of Fe'* constrained to lie on the solvus can have composi- are present in Fe- and Mg-rich gedrites, then the in- tions outside of this range. There is a slight tendency crease in A-site oocupancy with increasing Fe'*/ for Mg-rich orthoamphiboles to be enriched in Al'' (Fe2* + Mg) observed in Figures 4 and 5 is still pres- (Fig. 6); there is a strong tendency for Fe-rich ortho- ent. However, if gedrites from Fe-rich bulk composi- amphiboles to be enriched in A-site occupancy (Fig. tion are preferentially enriched in Fe3* over gedrites 5). Moreover, the size of the miscibility gap, as ex- from Mg-rich bulk compositions, then some or all of pressedby A-site occupancy, increaseswith increas- this variation will disappear. In fact, it is possible to ing Fel(Fe+Mg). choose ferric iron contents so that all recalculated The effect of ferric iron content on recalculated or- gedrites will have the same A-site occupancy, or al- thoamphibole formulae warrants discussion. Wet- ternatively so that all gedrites and anthophyllites plot chemical analyses of orthoamphiboles from similar on a linear trend, as has been noted by Robinson et rocks (Robinson and Jaffe, 1969;Rabbitt, 1948)con- al. (197h).It is unlikely that all gedriteshave identi- tain lessthan 3.0 wt.VoFerO, or Fe3*/(Fe'* + Fe"*) cal A-site occupancies;this would require the most less than 0.15, although Lal and Moorhouse (1969) iron-rich gedritesto have l5-25%oFe3* and the most report gedrite analyseswith Fe3*/(Fe'* + Fe2*) as -rich gedrites to have 07o.If this were the high as 0.25. lt some of the iron in the ortho- case, however, the relationship between Al'' and amphiboles is recalculated as Fe3*, the resulting Fe2*/(Fe'* + Mg) and Al'' and Fe'z*/(Fe'z*+ Mg) structural formula shows higher values for Al'' and would still be present and even accentuated(see Figs. lower values for Al"t, A-site occupancy, znd Fe2'/ 6 and 7).If all gedritesand anthophylliteswere re- (Fe2* + Mg) than the same analysis calculated with calculated to fall on a single line in Figure 4, system- SPEAK GEDRITE-ANTHOPHYLLITE SOLVUS llll

o.2 0.3 0A o.5 0.6 6. Alvr vs. Fel(Fe+Mg) for coexisting anthophyllite and gedrite. Symbols as in Fig. 4.

t.4

z a

o.2 o o.2 0.3 o.4 0.5 0.6 Fe/Fe+Mg)

Fig. ?. Alrv vs. Fel(Fe+Mg) for coexisting anthophyllite and gedrite. Symbols as in Fig. 4. ll12 SPEAK GEDRITE-ANTHOPHYLLITE SOLYUS

atic oxidation of Fe-rich samples would still be re- a TI quired. Moreover, the relationship between Fe2*/ In K9Ti (Fe'* + Mg) and A-site (Fig. D. occupancy 5), Al"' (Fig. I 6), and A1'" (Fig. 7) would still be present. I No petrographic or chemical evidence for the sys- tn KXn!no tematic oxidation of Fe-rich samples, such as the U .i"t presenceof hematite or of considerable FerO, com- ponent in ilmenite, has been observed.Although the Col question of ferric iron in the orthoamphibole must InKi 1.. remain unresolved, it is believed that much of the variation of A-site occupancy, Al"r, and AlIv with t T Fe2*/(Fe2* + Mg) observedin Figures 4-7 represents real crystal chemical differences between iron-rich In KRdtnD and magnesium-rich samples.' The miscibility gap depicted in Figures 4-7 for the orthoamphiboles is the widest gap reported for these t' ln KTtch ^ mine14ls,as wifl be discussedlater. l.{ote that the so- D1 called miscibility gap between actinolite and horn- t blende, which are the monoclinic calcic amphibole ,,{. equivalents of anthophyllite and gedrite, is in the lnKD same composition direction and spansapproximately t the same Bap(e.9., Klein, 1969;Brady, 1974). The distribution of the edenite and tschermakite o o4 oq components between anthophyllite and gedrite can FelFe+Mg in GEDRITE be seenin Figure 8. The definitions of Ko are as fol- Fig. 8. laKp vs. Fel(Fe+Mg) for coexisting anthophyllite- lows: gedrite. KD for cach plot is diferent and is defined in text. The synbols "+" and "x" indicate change in lnK$+vg if gedrite is recalculatcd assuming Fe3+/(Fe3+ + Fe2*) : 0.10 and 0.25, zEdcn_ [Na(A)l /[Na(e)] KBo'"= The symbol the arrow n ** *u respoctively. indicatcs that a I 6l "."/ [ 1n;:l point plots at infinity. Alvr LDr.r$h_f : Al"' I /t I ftvtgG\aa],* / | t"tgG\,Izy_l-" where E refers to a vacancy in the A site. The values A-site occupancy, Altt, and Al'". Gedrite is always of lnKS'o are approxim.ately2.5t0.75 and, except for enriched in Fe relative to coexisting anthophyllite. some erratic points, those for lnKff* cluster around Moreover, the partitioning is a function of FelMg ra- 2.8+0.5.The points on the lnKff* plot with valuesof tio: gedrite is more enriched in Fe relative to antho- infinily are for anthophyllites with 0.0 calculated A- phyllite in the Fe-rich than in the Fe-poor bulk com- : site occupancy, although these amphiboles do con- positions. Values of lnKB*Mt [Kl*"' (Fe/Mg)"J tain measurable NarO. Becauseof the great uncer- (FelMg)-ol range from 0.38+0.05 in the most Fe- tainties in calculating A-site occupancies,the values rich sample to 0.13*0.05 in the most Fe-poor sam- of lnKff'" should be considered only approximate. ples. The trend of increasing K' with increasing Fel 146s1importantly, however, there is no systematic Mg is still apparent if Fe-rich gedrites are recalcu- variation in Kf* or Kf,*h with Fel(Fe+Me). lated assuming lOVoFet*;the trend only disappearsif The distribution of Fe and Mg between coexisting Fe-rich gedrites contain 257oFe3* and Fe-poor ged- anthophyllite and gedrite can be seenin Figures 5, 6, rites contain Vo (see Fig. 8). and 7, where Fel(Fe+Mg) has been plotted against Figure 9 showsplots of total Fe vs.Al'I and Mg vs. Al'' for coexisting gedrite and anthophyllite. As- 2 sumptions about Fe3* in the onhoamphiboles affect Mdssbauer spectra of an Fe-rich gedrite [Fel(Fe+Mg) = 0.4E1 this diagram very little. The plots show that total Fe and an Mg-rich gedrite [Fel(Fe+Mg) : 0.28] reveal that both or- thoamphiboles have Fe3+/(Fer* + FC*) ratios of less than 0.05, varies very little between anthophyllite and coexist- thus substantiating the observcd trcnds with respect to Fel ing gedrite, whereas anthophyllite is always enriched (Fe+Mg). in Mg relative to gedrite. Furthermore, there is an SPEAR GEDRITE-ANTHOPHYLLITE SOLVUS n13

(1973), and Kamineni (1975), and the implications for exsolution mechanisms are discussedby Robin- son et al. (l97la, p. 1030-1031).Octahedral Al is highly concentrated in the M2 site of aluminous or- thoamphiboles and, relative to Fe, Mg partitions E strongly into this site (Finger, 1970;Papike and Ross, 0) L 1970).' The relationship depicted in Figure 9 is con- sistent with these site-occupancy assignments and with the notion that the tschermak exchange in or- thoamphiboles involves primarily the M2 octahedral site (in addition to tetrahedral sites to maintain charge balance). The distribution of Ti, Mn, and Ca between antho- phyltite and gedrite can be seenin Figures 8, 10, ll, and 12. Ti is always enriched in gedrite relative to .J anthophyllite: typical values for anthophyllite range g gedrite from 0'12 to o from 0.02 to 0.07 wt.Voand for 0.20 'xl.Vo. There is a tendency for Fe-rich ortho- CD Ti = amphiboles (especially gedrite) to contain more than Mg-rich orthoamphiboles (see Fig. l0). All as- semblagescontain a phase enriched in TiOr: rutile in Mg-rich samples and ilnenite in Fe-rich samples (Fig. 10). ,tl : 1l Values of lnK" {K* [TilMe (M2)1"^1/lTi/Mg Anth Alvl Ged (M2)l-") clusteraround 2.2!0.08 (exceptfor two er- Fig. 9. Total Fe vs. AlvI and (B) total Mg vs. Alvr for coexisting ratic points representing anthophyllite with no mea- (circles) and gedrite (squares). anthophyllite sured TiOr; seeFig. 8). There is no systematicvaria- tion in lnK, with Fel(Fe+Mg). approximately l: I inverse relation between the Mg The distribution of Ca between anthophyllite and and Al'I contents of coexisting gedrite and antho- gedrite (FrC. ll) is systematic:gedrite is always en- phyllite. ffue implication of this relation is that the tschermak exchange reaction (Al"'AlI"cMgSl) in- 3 of Fe* - Fe2+ + volves only the Mg-end member component in or- Papike and Ross show similar distributions Fe3* + Mn + Ti and Mg among the Ml, M2, and M3 sites of thoamphiboles. A similar relation in the ortho- gedrite. If it is assumedtlat Fe3+ and Ti are located exclusively in amphiboles has been described by Stout (1972, the M2 site, then calculations reveal that Mg is greatly enriched in personal comnunication), Abraham and Schreyer the M2 site relative to Fe2*.

+ RUTILE

o.3 o.4 0.5 o.6 Fe,/Ferftg

Fig. 10. Ti ys. Fel(Fe+Mg) for coexisting anthophyllite and gedrite. Symbols as in Fig. 4. Dashed line separatesthe assenblages containing rutile from assemblagescontaining ilmenite. ll14 SPEAK GEDRITE-ANTHOPHYLLITE SOLVUS o.l

o c)

o.3 o.4 o.5 0.6 Fe/{Fe+Mg) Fig. ll. Ca vs. Fel(Fe+Mg) for coexisting anthophyllite (circles) and gedrite (squares).Symbols as in Fig. 4.

\o nl

Ca Fe+Mg+Mn Fig. 12. Portion of the compositional triangle At-Ca-(Fe+Mg+Mn) showing compositions of tnalyzgd anthophyllites (Anth), gedrites (Ged)' and (Hbld). Tielines @nnect coexisting phases. Hornblende is the most Ca-rich phase in these rocks; the plot displays the maximum solubility of Ca in orthoamphiboles at this metamorphic grade. The most Al-rich gedrites and hornblendes coexist with staurolite (St), which is the most aluminous phase for Fe-rich assemblagesin tlese rocks. SPEAR: GEDRITE-ANTHOPHYLLIT:E SOLVUS ll15 riched in Ca relative to anthophyllite. The amount of in all but three samples. Values of lnKf [KX' = Ca in the orthoamphibole is a function of buft CaO [Mn/(Mn+Fe+Mg)]"-/[Mn/(Mn+Fe+Mg)-*] in- content. The hornblende-containing assemblages creasewith increasingFel(Fe+Mg) from 0.1t0.15 at (closed synbols) show relatively systematic Ca con- Fel(Fe+Mg) : 0.3to 0.7*0.3 at Fel(Fe+MB) : 0.55 tents in both gedrite and anthophyllite. Ortho- (seeFig. 8). amphiboles in hornblende-absentassemblages The composition of orthoamphiboles and coexist- (open symbols) show erratic CaO contents and val- ing minerals are displayed in a plot of Al-Fe-Mg in ues lower than those in hornblends-ssilxining as- Figure 13. This diagram is not a projection of the semblages.The distribution of Ca as well as Al be- phase relations, and crossing tielines should not be tween hornblende and orthoamphiboles can also be construed as indicative of reaction relations. The plot seen in Frgure 12, which illustrates that Ca contents is presented to show the compositional limits of or- in hornblende, gedrite, and anthophyllite are very thoamphiboles with respect to Al, Fe, and Mg. Or- consistent for all assemblages.The crossing tielines thoamphiboles with the highest Fel(Fe+Mg) are result from the condensationof Fe*Mg*Mn into found in associationwith garnet. Gedrite and antho- one component. phyllite coexisting with garnet have Fel(Fe+Mg) ra- Distribution of Ca between orthoamphibole pairs tios of 0.53-0.55 and 0.214-0.46,respectively, and Kff = [Cal(Fe,M4 + Mn,M4)]s^/lCa/(Fe,M4 + gedrite coexisting with garnet and staurolite has a Mn,M)l*,h) shows no dependencyon Fel(Fe+Mg) Fel(Fe+Mg) ratio of 0.55. (seeFig. 8). Valuesof tntr(S"cluster around 0.6+0.4. The most Mg-rich orthoamphiboles from this The Mn contents of gedrite and anthophyllite are grade are found coexisting with talc, chlorite, atd/or small (0.09-0.55 vrt.Vo)and depend largely on bulk cordierite, all of which are present in sample 68- MnO content. Mn favors gedrite over anthophyllite 432D.

Mg Fe Fig. 13. Al-Fe-Mg compositional plane showing compositions of analyzed gedrite (squares), anthophyllite (open circles), (triangles), garnet (Ga), sraurolite (St), cordierite (Cd), chlorite (Ch), and talc (Tc). This figure reveals the limits of the solubility of Al, Fe, and Mg in orthoamphiboles at the metamorphic grade of the Post Pond Volcanics. llt6 SPEAR: GEDRITE-ANTHOPHYLLITE SOLVUS

The most aluminsus anthophyllites coexist with the compositional data and estimates of metamor- gedrite and the most aluminous gedrites coexist with phic temperatures in these studies combined with staurolite (in Fe-rich samples,see also Fig. 12) or data from the present study, it is possible to construct cordierite (in Mg-rich sanples). The Al-poor limit a hypothetical T-X solvus for the orthoamphiboles. for gedrite is defined by coexistencewith anthophyl- In Figure 14 the solvus is drawn as a function of lite and for anthophyllite by coexistencewith cum- A-site occupancy, AlvI and AlrY. The lowest temper- mingtonite. ature data. and the data that record the widest mis- cibility arc from this study. The data from Stout Discussion Eap, (1970, 1972) are from the Telemark region, Norway, The miscibility gap between anthophyllite and where metamorphic conditions are in the kyanite-sil- gedrite has been discussedby Robinson et al. (1969, limanite zone. Spear(1977,1978) has estirnatedthat 1970, l97la), and Ross et al. (1969),who analyzed the Telemark have crystallized at a exsolution textures in super-solvusorthoamphiboles, highsr 7or lower Pr,o than the PostPond Volcanics. and Stout (1969, 197O, 197l, 1972),Spear (1977) and A single pair of coexisting garnet + cordierite from Stoddard (1979,personal comnunication), who ana- Stout (1972)yields temperaturesof 553o and 570"C lyzed coexisting sub-solvus orthoamphiboles. From using the calibrationsof Thompson(1976) and Hold- away and Lee (1977), respectively. Individual pairs of coexisting anthophyllite + gedrite from Stout (1972) have been plotted on Figure 14, assuminga temperature of 560o with an estimated error of () o^ooa o Boa +,25"C. On the average,these data show a smaller F miscibility gap than data from the Post Pond Volcan- ics, although there are some notably discrepant points. James et al. (1978) and Robinson et al. (l97la) o.5 LO r5 2.O present orthoamphibole'analysesthat completely Allv bridge the solvus. James et al. estimated metamor- phic conditions at 625"C, 3-6 kbar; Tracy et al. (1976)estimated that the samplesstudied by Robin- () a aoo to a son et al. crystallizedat 625450oC, -6 kbar. Note o F that the most A1- and Na-rich gedrites studied by Robinson et al. do not display exsolution (Fig. la). Based on the above temperature estimates, the crest of the anthophyllite-gedrite miscibility gap o.5 r.o t.5 should be below 625" and above570" and it is tenta- Alvl tively placed at 600+25oC, as shown in Figure 14. The solvus is asymmslrisal, with the anthophyllite- rich limb being steeperthan the gedrite-rich limb. (J o oao a o aoo a Conclusions Orthoamphiboles are found in a wide variety of geologic environments. Careful delineation of the solvus in T-X space will enable metamorphic tem- peratures to be estimated in samples where two or- A-SITE OCCUPANCY thoamphiboles coexist. Additional data on natural samples coupled with experimental determination Fig. 14. Hypothetical 7-X solvus for ortho,mFhiboles drawn at will serve to refne ttLe T-X solvus presented in Fig- Fe2+/(Fe2++Mg) = 0.4, based on data from this study (large ure 14. Exsolution features in sub-solvus boxes),Stout (1970,1972)(circles), James et al. (1978)(triangles), and super- and Robinson et al. (l91la) (diamonds). Filled di,monds are solvus orthoamphiboles may also reveal information eftoamphiboles from Robinson et al. showing no exsolution. e1 ss6ling rates (Ross et al., 1969;Robinson el a/., Temperature estimatesare discussedin text. r971b). SPEAK GEDRITE-ANTHOPHYLLITE SOLVUS lllT

Acknowledgments Klein, C., Jr. (1969)Two-amphibole assemblages in the systemac- tinolite-hornblende-glaucophane. Am- Mineral., 54, 212-237. Most of the analytical work for this study was conducted at the Lal, R. K. and W. W. Moorhouse (1969) Cordierite-gedrite rocks Geophysical Laboratory, Camegie Institution of Washington, and associatedgneisses of Fishtail Lake, Harcourt Township, where the author was a Postdoctoral Fellow. Support from this In- Ontario. Can. J. Earlh Scr'.,6, 145-165. stitution is gratefully acknowledged.Also acknowledgedare many Leake, B. E. (1978)Nomenclature of amphiboles.Can. Mineral., fruitful discussionswith J. Ferry, J. Stout,L. Finger, E. Stoddard, 16,501-s20. and H. S. Yoder, Jr. The manuscript was improved considerably Papike, J. J. and M. Ross (1970)Gedrites: crystal structuresand by constructive reviews by P. Robinsoq M. Ross, and C. Klein. intercrystalline cation distribution.s. Am. Mineral., Jt 1945- Special thanks go to D. Rumble, III, who contributed much time t9't2. and energy both in the field and in the laboratory and who con- Rabbitt, J. C. (1948) A new study of the anthophyllite series.lrz. tributed several samplesused in this study. Support of NSF grant Mineral., 33,263-323. EAR-79- I I 166 is also gratefully acknowledged. Robinson, P. and H. W. Jaffe (1969) Chemographic exploration of amphibole assemblagesfrom central Massachusettsand south- westernNew Hampshire.Mineral. Soc.Am., Spec.Pap. 2,251- 274. References of three amphiboles. Contrib. Mineral. Petrol., Abraham, K. and W. Schreyer(1973) Petrology of a fem.rginous coexistence hornfels from Rieckensgluck, Harz Mountains, Germany. Con- 22,248-258. trib. Mineral. Petrol., 40, 275-292. anthophyllite miscibility (abstr). Am. Albee, A. L. and L. Ray (1970) Correction factors for electron thophyllite and the BLp probe microanalysis of silicates, oxides, carbonates,phosphates Mineral., J5, 307-309. _ - (l97la) of the antho- and sulfates.Anal. Chem.,42, 1408-1414. and Composition gedrite and hornblende, Bence,A. E. and A. L. Albee (1968)Empirical correctionfactors phyllite-gedrite series,comparisons of Am. Mineral., J6, 1005- for the electron microanalysis of silicates and oxides. f. Geol., and the anthophyllite-gedrite solvus. 76,382403. 1041. (l97lb) Orientationof Bottinga, Y. and M. Javoy (1973) Commentson oxygen isotope H. Jaffe,M. Rossand C. Klein, Jr. and clinoamphiboles: geothermometry.Earth Planet. Sci.Lelt., 20,250-265. exsolution lamellae in clinopyroxenes phase Am. Mineral., 56, Brady, J. B. (1974) Coexisting actitrolite and hornblende from consideration of optimal boundaries. west-centralNew Hampshire.Am. Mineral., 59,529-535. 909-939. (1969)Exsolution tex- Deer, W. A., R. A. Howie and J. Zussman(1963) Rock-Forming Ross,M., J. J. Papike and K. Wier Shaw of sub-solidus thermal his- Minerals, v. 2: Chain Silicates.Longmans, London. tures in amphiboles as indicators 2,275-299. Downs, W. F. and P. Deines (1978)Experimental calibratiotr of toies. Mineral. Soc.Am., Spec.Pap. and PetrologicStudies the quartz-magnetite oxygen isotope geothermometer (abstr.). Rumble, D. (1969)Stratigraphic, Structural Hampshire. Ph.D. Dissertation, Har- Geol. Soc. Am. Abstracts with Programq 10, t92. in the Ml. Cube area, New Massachusetts. Ferry, J. M. and F. S. Spear(1978) Experimental calibration of the vard University, Cambridge, (1957) partitioning of Fe and Mg between biotite and Elrrl.et. Contrtb. Seki, Y. and M. Yamasaki Aluminian ferroanthophyllite Mineral. Petrol., 66, I l3-l 17. from the Kitakami Mountainlan4 northeastem Japat Am- Finger, L. W. (1970)Refinement of the crystal structureof an an- Mineral., 42, 56-520. thophyllite. CarnegieInst. Wash.Year Book,68,283-288. Spear,F. S. (1977)Phase equilibria of amphibolitesfrom the Post Gable, D. J. and P. K. Sims (1970)Geology and regional meta- Pond Volcanics, Vermont. CarnegieInst. Wash. Year Book, 76, morphism of some high grade cordierite ,Front Range, 613-619. Colorado. Geol. Soc.Am. Spec.Pap. 128. - (1978) Petrogeneticgrid for amphibolitesfrom the Post Gittos, M. F., G. W. Lorimer and P. E. Champness(1976) The Pond and Ammonoosuc Volcanics. CarnegieInst. Wash. Year phase distributions in some exsolved amphiboles. In H.-R. Book, 77,805-808. Wenk, Ed., Electron Microscopy in Mineraloglt, p. 238-247. - (1980) NaSiACaAl exchangeequilibrium between plagio- Berlin. Springer-Verlag, clase and amphibole: an empirical model. Contrib. Mineral- Pet- Hadley, J. B. (1942) Stratigraphy, structure and petrology of the rol., 72,3341. Mt. Cube area,New Hampshire.Geol. Soc.Am. Bull., 53, ll3- Stoddard, E. F. (1979) Phase petrology of cordierite-amphibolc 176. rocks, Old Worhan-Piute Range, Southeastern California Holdaway, M. J. (1971)Stability of andalusiteand the aluminum (abstr.). Geol. Soc. Am. Abstracts reith PTogam; Cordilleran silicate phasediagram. Am. f. Sci.,271,97-131. Section,11,130. - and Sang Man Lee (1977) Fe-Mg cordierite stability in (1969) microprobe study of coexisting or- high-grade pelitic rocks based on experimental, theoretical and Stout, J. H. An electron (abstr.). Am. Geophys.Union, 50, natural observations. Contrib. Mineral. Petrol., 63, 175- 198. thorhombic amphiboles Trans. J"mes,R. S., R. A. F. Grieve and L. Pauk (1978)The petrologyof 359. cordierite-anthophyllite gneissesand associatedmafic and pe- - (1970) Three-amphibole assemblagesand their bearing on litic gneissesat Manitouwadge,Ontaio. Am. I. Sci.,278,4l-63. the anthophyllite-gedrit€ miscibility gap (abstr.). Am. Mineral-' Kamineni, D. C. (1975) Chemical of some cordierite- 55,312-313. bearing rocks near Yellowknife, Northwest Territories, Canada. - (1971) Four coexistingamphiboles from Telemark, Nor- Contrib. Mineral. Petrol., 53, 293-310. way. Am. Mineral., 56,212-224. lllS SPEA K GEDRI TE-A NTH OPH YLLITE SO LVU S

- (1972) Phase petrology and nineral chemistry of coexist- Tracy, R. J,, P. Robinson and A. B. Thompson (1976) Gamet ing amphiboles from Telemark, Norway. J. Petrcl., 13,99-145. mmposition 41d zoning in the deternination of temperature Thonpson, A. B. (19?6) Mineral reactions in pelitic rocks: II. Cal- and pressure of , central Massachusetts.ln. culation of someP-T-X(Fe-Mg) phaserelations. Am. J. Sci, Mineral.,61,762-775. 276,425454. Thompson, J. B., Jr,, P. Robinson, T. N. Cli.fford and N. J. Trask (1968) Nappes and gneiss domes in west-c€ntral New England. lnB-an 7-el et al.,F/rs., Studiesof Appalachian Geologt, p.203- Manuscript received,November 9, 1979; 218. Wiley, New York. acceptedfor publication, May 7, 1980.