THE AMERICAN MINERALOGIST, VOL. 43, SEPTEMBER_OCTOBER, 1958

SIGNIF-ICANCE OF AI,IPHIBOLE I,ARAGENESIS IN TH]' BIDWF]I,L BAR REGION, CALIFORNIA I{onntr R. ColrproN, Stanford Llnitersily, Stanford, California.

ABsrRAcr In the northwest Sierra Nevada, regionally extensive low-grade metabasaltic rccks containing abundant actinolite [(Na,K)o rsCar-s(Mg,Fe"), +A1or(OH)r eSizzA]o rOzzl are converted to hornblendic rocks in the contact aureole of a small batholith. Hornblende in the outer part of the aureole is associatedwith oligociase,, and minor chlorite, and has the composition (Na,K)6 rrCar z(Mg,Fe")r(Fe"',Al)r r(OH)t rSiuzAlr sos:, u'hile horn- blende in the inner part of the aureole is associated with andesine (and locally clinopyrox- ene), and has the composition (Na,K)o rCaz(Mg,Fe")a 3(AI,Fe"')r r(OH)r rSiurAlr rOn. The abrupt conversion of actinolite to aluminous hornblende in the outer part of the aure- ole contrasts with gradational conversions in sonre regionally metamorphosed schists, and this contrast is probably caused largely by difierences in the dehydration and reaction rates of the two environments. Ferrohastingsite [(Na,K)6 ECa2(Fe",Mg)i s(Fe"',Al)y- (OH,CI), rSieAlzOzzlformed in the inner part of the aureole from rocks (rnetatuffs) low in Si and high in Fe. Igneous hornblende that crystallized with biotite in the border zone of the batholith has the composition (Na,K)s rCar g(Mg,Fe")r +(Fe"',Al)6 s(OH)o sSizr- Alo sozr. fr.rtnooucrtoN Amphiboles are essentialconstituents of almost all of the abundant metavolcanic rocks of the northern Sierra Nevada, and also occur widely in granitic rocks that have been emplacedinto this terrane. In general, the metamorphic amphibolesare products of two plutonic episodes,the first being low temperature dynamothermal metamorphism on a regional scale,and the secondcontact metamorphismby the granitic intrusions. Much of the igneous amphibole formed directly from the more mafic granitic magmas, but some is doubtlessthe result of contamination of acidic magmasby amphibole-richmetamorphic rocks. Becausereactions involving amphiboles are very important in the evolution of the Sierran rocks, and becauselittle is known about the exact nature of thesemin- erals, pure amphibole samples were separated from five rocks that form a reasonably well known genetic series.The purposes of this paper are to present the chemical compositions and physical properties of these and to consider the reactions involved in their formation, esDeciallvthe conversionof actinolite to aluminoushornblende.

Pornorocrc Besrs or rHE SrUDY The general geologic setting of the Bidwell Bar region has been de- scribed elsewhere(Compton, 1955), and only the most necessaryrela- tions will be presented here. The region is an excellent one in which to explore the general amphibole evolution outlined above becausealmost all the metamorphic rocks have similar (basaltic) compositions,the

890 AMPHIBOLE PARAGENESIS 891 relations between these rocks and a small granitic batholith are well exposed,and the contact metamorphic aureolearound the batholith is distinct. \'Ioreover, the aureole shows progressive variations of texture and mineralogy that indicate variation in metamorphic intensity, and it was mapped in two zoneson the basisof textural criteria. The outer contact of the aureole was drawn where greenschistsand greenstones give way (in a few hundred feet) to fine-grained tough hornfelses,and the contact between the outer hornfels zone and the inner zone of the aureolewas drawn where metamorphic amphibole is clearly visible to the unaided eye. The contact between the batholith and the metamorphic rocks is generally sharp, but veins and nodes of occur widely in the inner zoneof the aureole,some being intrusive and somemetasomatic. Although the aureole zones were not mapped on the basis of thin sections,the cartographicunits can be correlatedwith grosspetrographic variations. The dominant assemblageoutside the aureole is actinolite--epidote-chlorite;in the outer zone of the aureole it is hornblende-oligoclase-epidote(locally with chlorite) ; and in the inner zoneoI the aureoleit is hornblende-andesine(locally with clinopyroxene). The metasomatic veins in the inner zone generally consist of andesineor labradorite,hornblende, and , while the intrusive veins consistof andesineor oligoclase,quartz, hornblende,and biotite. The five analyzed amphiboles were chosen so as to give a picture of amphibole variation correlative with the critical changesin this para- genesis.One is from the low grade regional terrane, one from the outer zoneoI the aureole, two from the inner zoneof the aureole,and one from an intrusive outlier of the bathoiith. The specimenfrom the regional terrane is a mafic greenstonethat lies 6,000feet outside the aureole.It is especiallyimportant to this study becausea Iarge part of the analyzed actinolite clearly pseudomorphs igneous clinopyroxene, while the same actinolite is abruptly overgrown here and there by thin rims of green metamorphic hornblende identical to that of the contact aureole. Thus two-stage amphibole reactions are well displayed, and the sharpnessof the overgrowths of aluminous hornblende is important in the interpreta- tion of the actinolite-hornblende reaction. The specimen is of further interest becauseit shows that partial contact metamorphism extended far beyond the mapped aureole. The second amphibole is from a hornfels that lies near the middle of the outer zone oI the aureole. The rock contains less epidote than most metabasalticrocks of the outer zone; however,it was selectedbecause the presenceof chlorite suggestedits amphibole might be intermediate between actinolite and the aluminous hornblendes of the inner zotre,a suggestionthat was not borne out by the chemicalanalysis. 892 ROBERTR. COMPTON

The first of the two amphibolesfrom the inner zone was separaled from an that lies only 200 feet from the batholith, and is veined by tonaiite that is probably metasomatic.The separationutilized only the amphibolite,although the hornblendesof amphiboliteand veins are physically identical. Like several other from the im- mediatecontact zone,this rock containssmall amounts of clinopyroxene, and it may vvellrepresent the most advanced stage of contact meta- morphism. The secondamphibole from the inner zone is from a ferrohastingsite schist that is interbedded with andesine-hornblendeamphibolites and biotite-quartz-hornblende-oligoclaseschists about 2,500 feet from the batholith. It is of interest in this study becausesimilar ferrohastingsites occur in several parts of the inner zone, and this mineral is the only unusual amphibole of the entire complex. Finally, the hornblendeseparated from the mafic outlier of the batho- lith occurs only a few feet from a sharp intrusive contact with typical amphibolite of the inner zone. The abundance of hornblende at such contactswith amphibolites,and the heterogeneityof the border suggestthat the hornblendeof the tonalite might well have formed by contaminationof a more acidic magma by amphibolite (Compton, 1955, p.33 37). The purposeof the analysiswas to determinethe composi- tional differencesbetween hornblende of the wall rock and hornblendeof the new biotite-rich igneousassemblage.

Arrpnreotos Seporation The fractionations of the difierent constituentsfrom the greenstone and the hornfels were complicated by the intricately interdigitated shapesof many of the minerals, yielding a large number of composite grains in all but the finest materials. This was overcomeby using very well-sortedfraclions of 0.01 to 0.03 mm. size,obtained by repeatedde- canting of - 250 mesh material for settling periods of between3 and 5 minutes. These well sorted samplesresponded well to centrifuging in heavy liquids, the principal separatory procedureused. The impurities were counted in all samplessubmitted for analysis, and in each case amounted to lessthan 0.5 per cent of the sample.

Compositions The chemicalanalyses are listed in Table 1. Analyses2 and 3 are each the averageof two analyseswhich difieredonly slightly from the amounts listed. The HrO+ determinations were made by the Penfield method, using an oxy-coal gas blast flame. The numbers of atoms in the formulae of the amphiboles (Table 2) Tesr,B 1. CouposrrroNs .qNl Pnvsrcel Pnoprnrrns ol Frvr Alrpnr- BorES lRoM rrln Bmwnr,l Ben RrcroN, C,c.LrronNrA,

Sio: 53 41 45.33 4.)-6/ 38 04 47.80 Alzo: 4.56 r0.12 t|.17 11.09 6.94 TiOz 037 0.54 057 087 0.80 FerOr 0.83 551 4.82 7.38 498 FeO 9.70 14.61 9.72 21.30 9.6s MnO o.25 059 o.25 036 0.38 Mgo 15.09 10.00 t1.69 4.00 14.58 CaO 12.52 10.53 t2.62 1t 99 12.r9 BaO 0.10 0.02 0.03 0.10 006 NazO 0.88 104 114 1.n 1.4+ KzO 0.63 0.36 0.29 1.69 0.63 HzO+ 163 1.89 173 185 077 H':O- 0.05 0. 1.5 032 0.19 0.05 F 0.01 0.01 0.01 0.o2 0.06 CI n.d. n.d. n.d. 0 .66 nd P:Ou 0.04 0.01 0.01 0.11 0.07

Total 100 07 100.71 t00.24 101.05 100.40 LessO for F and Cl 0.01 001 001 0.16 003

Total (corrected) r00 06 100.70 10023 100 89 10037

Specific gravity 3.07-312 3 25-3.29 3 22-3.24 3 39-3.40 3.20-3.21 d 1.626 1.65+ 1 660 1.692 1 651 a 1.645 1.674 1.680 1.71+ 1 670 a-q 0 019 0.020 0 020 0.022 0 019 2Vo 77-80" 65-69. 64-66. 30 31" 64 67" zAc 15-18' 16-18. 20-21" 15 17 ' 16 18' Orientation Y:b Y:b Y:b Y:b Y:b X Colorless Straw Straw Grayish Straw yellor. Y Very pale Olive Olive Deepolive Olive yeilow-green Z Very pale Blue green Slightly blu- Deepgreen- Slightly blu- green ish green ish blue ish green

1. Actinolite from mafic greenstone (88-1-16). Bidwell Bar (7t) quadrangle; lat 39"36'12"N., long 727o27'39'W.* Analyst, R. Klemen. 2. Hornblende from hornfelsed metadiabase (BB-4-171c). Bidwell Bar (7]') quad rangle; lat. 39"33'03'N., long. 721"25'50, W. Analyst, R. Klemen. 3. Hornblende from clinopyroxene bearing amphibolite (F-1-120b). Forbestolvn quacl- rangle; lat. 39"35'45' N., long. 121"19,43, W. Analyst, R. Klemen. 4. Ferrohastingsite from ferrohastingsite schist (CM-1-11). Clipper Mills quadrangle; lat. 39'35'38" N., long. 121'14'26'W. Analyst, R Kiemen (except that Cl determi- nation was made by writer). 5. Hornblende from mafic tonalite (RRC-9). Bidwell Bar (74) quadrangle; Iat. 39"32'54'N., long. 121"25'O6nW. Analyst, R. Klemen. * Note: to plot specimen localities on geologic map of area (Compton, 1955, plate 1), change all latitudes of map from 37" to 39". 894 ROBERT R. COMPTON

Taer-r 2. Nuunrns or Arorls rw Fonuulee oI AMPHTBoLES FRoMTHE Browrrl Ben Rncrow, CelrnonNra.

si 7.66 6.72 6.68 6.04 7.06 (z) R- oo-- R OO-' -',R OO" - R" oo"- 8 OO Al 0.34 1.28"' r.32 1.96 094-

Al o.43 0.48 061 0.10 0.25 (Y) Ti 0.04 0.56 0.061.11 0.07 1.20 0.091.07 0.090.89 Fe"' 0.09 0.s7 0.52 0.88 0.55

Mg 3.25 2.23 2.57 0.95 3.21 (x) Fe" l.16 4 44 1.804.10 1.183.78 283383 1.184.44 Mn 0.03 0.07 0.03 0.05 0 .05

Ca 1.92 1.69 1.97 2.02 t.92 Ba 0.01 ooo 0.00 ^ -- 0.0r^ -^ 0.00 a /< -, -",q 04 l'rr z t6 ''-" (w) Na o.24 o;;2 0.31 o.42 0.4t K 0.11 0 .07 005 0 .33 0.12

OH I. JO 1.89 |.69 1.96 076 CI n.d. 1 56 n.d 1 89 n.d. 1.70 0.r7 2 14 n.d 0.79 F 0.00 0.00 0.01 0.01 0.03

were calculatedon the basis of 24 (O,OH,F,Cl) (Warren, 1930,p. 504 516),and they are closeto the valuesdetermined structurally by Warren, except that number 5 is distinctly low in the OH group and high in the total of the X and Y Broups.The total positive chargesof number 5 also are too high, and can be balancedonly by assumingthat the "va- cant" OH positions are occupied by O ions; however, the rather low FezOadoes not support this possibility. If, on the other hand, it is as- sumed that the HzO+ determination is low, the composition can be recalculatedso as to agleereasonably well with the ideal slructural ratio o{ atoms. Since this possibility cannot be certain, however, the values in Table 2 are those calculatedfrom the originai analysis. The impurities were calculatedout of the analysesbefore the atomic compositionsof the amphiboleswere computed, but these corrections are so small that the only value improved by more than 0.01 of an atom is the Fe"' of amphibole 2, which is decreasedby 0.03 of an atom to correct for included magnetite.

Physical,properties The principal physicalproperties of the analyzedamphiboles are given in Table 1. The range of specificgravity listed is that of the two heavy Iiquids in which the final mineral sample was separatedfrom heavier AM PII I BOLE PA RAGENES I S

'I'he and lighter materials. amphiboles in rock specimens3, 4, and 5 were so homogeneousthat by {ar most o{ the monomineraiicamphibole 'Ihe grains lr,ereini:luded in the final san.rple. amphibolesof specimens1 and 2, on Lheother hand, u'ereso heterogeneousthat much monomineralic material had to be separatedfrom the analyzedsamples in order to free the samplesfrom composite grains. Judging from the composition of these impure fractions, amphibole 1 is somewhat more denseand pre- sumably a little richer in iron than the averageactinolite of the rock, and amphibole 2 is distinctly more dense, more deeply colored, and probably containsmore iron and aluminum than the averageamphibole of the rock. The indicesof refraction were determinedby the immersion method, using a light source; ail liquids were measuredat once with a refractometer. The very fine-grainedmaterial submitted for analysis rvas used, and becausethe material was somewhat variable (note the specihcgravity ranges)an error as large as 0.002is possiblefor samples 3, 4, and 5, and an error as large as 0.003for samplesI and 2. The values for 2\,'andZ/\c were determinedin sodium light on the universalstage. 'fhe rangeslisted include all valuesthat recurredfrequently in measure- ments of a number of grains,and this variation is likely to be causedat least in part by the compositionalvariations within each sample.

Pernocn,rpuy The generaipetrography of the Bidrvell Bar rocks has been described (Compton, 1955), and only the petrography of the rocks from which amphiboleswere separatedwill be presentedhere. These rocks are in many respectsrepresentative of the other rocks with which they are associated;the greenstone,however, is considerablymore mafic than the typical metabasalticrocks of the regionalterrane.

Maf,c greenstone Amphibole 1 is from a non-schistose,actinolite-rich greenstonethat originally contained abundant large of augite and caicic ,but now consistsentirely of metamorphic minerals. The following approximate mode, in weight per cent, was computed from a count of 1,000 points:

albite 8 0 chlorite... 4.0 epidote" 26.0 leucoxene 5.0 actinolite 53.0 2.O hornblende. 2.0 An approximatechemical composition calculated from this mode is given in Table 3. Actinolite occurs both as pseudomorphsafter augite and as acicular ROBDRT R COMPTOI|

T.ler,n 3. Celcur-areo Cnolrtcar, CouposrrroNs op Alrpnrsoln-Brltrmc Rocrs lRoM TrrE Brorvnr,r, Bln Rncrox, Car,rronNrl

t2 345 Mafic Hornfelsic l|errohasting- .\nruhilrolite Tonalite greenstone metadiabase ' stte scnlst

Sio: 46.3 49.3 492 37.9 59.5 AI:O: 126 143 16.5 109 18.0 TiOz 2.0 11 17 16 05 FeiOr 34 60 30 7.2 1.8 FeO 9.0 62 20.6 38 MnO 0.2 0.3 0.1 0.4 01 Mo() 8.9 67 a1 3.8 +.1 CaO 14.0 8.6 11.6 TZ.J 59 NauO 1..) 2.9 JI l.J K:O 0.4 0.5 0.4 1.6 t.4

HrO 2.7 l.J 1.1 1.8 0.5 C1 0.6 COz 0.9 prisms intergrown with albite, chlorite, and epidote.Here and there the faintly colored actinolite is abruptly overgrown by small patches of a more deeply coloredhornblende that is blue-greenin Z, with',/:1.674, 2V:62"-68", andZ\c:17o 18o.These data indicate that it is identical with the hornblendeof the contact aureole (seenumber 2, Table 1). The original plagioclasegrains are largely pseudomorphedby clusters of epidote and calcite. N{ost of the epidote is iron-rich, but some grains are zoned to iron-poor rims, and euhedral clinozoisitecrystals occur in clustersassociated with chlorite. The chlorite is optically positive, has a birefringenceof 0.002, and a refractive index of approximately 1.580. H ornfelsic metadiabose Amphibole 2 is from a fine-grainedhornfelsed metadiabase that has the following mode, in per cent by weight: qtartz 1 5 titaniferous ore 5.0 oligoclase(An22) 28.5 chlorite 0.4 hornblende 62.5 apatite 0.1 epidote 2.0 The point count runs used to compute this mode were unusually con- sistent, and therefore the caiculated chemical composition in Table 3 should be quite reliable. Nlost of the hornblendeis in small rather deeply coloredprisms that interfinger with feldsparin a randomly oriented hornfelsicfabric. Some larger (0.5 mm.) grainshave lighter coloredcores that gradeabruptly out AMPHIBOLE PARAGENESIS

to the deeperhornblende of the prisms. The cores probably represent large actinolitesfrom the original metadiabasethat were not completely reacted to aluminous hornblende;it is also possible,though lesslikely, that they are iron-poor aluminous hornblende. It is the more deeply colored hornblende that rvas analyzed. Chlorite occurs here and there with the aluminous hornblende,but its genetic relation to the amphibole could not be determined.Epidote is iron-poorand is commonly includedin oligoclase.Rather large skeletal ore grains are scatteredevenly through the rock. Clino pyr orene-b earin g amphib oli te Amphibole 3 is from a typical amphibolite of the inner zone, charac- teristically veined and spotted by mafic tonalite. Judging from the con- sistencyof point counts, the following mode (measuredon amphibolite only) should be quite reliable: andesine (Aq3) 360 magnetite 0 1 hornblende. . 60.0 apatite 0.3 clinopyroxene 0 3 zircon tr. sphene 3.3 alkali felclsparand zeolite, tr. The texture is granoblasticbut the hornblendeshows a moderale de- greeof crystallographicorientation, c and { 1OO}tending to lie in a plane. The hornblendeis evenly colored and free of inclusions.Clinopyroxene was detectedonly in crushedsamples of the rock; however,judging from its textural relations in other sectionedrocks, it is granular, as coarse- grainedas the hornblende,and showsno obviousreaction relation to the hornblende.The propertiesof the clinopyroxene(determined in part on the universalstage) aret t:1.7 19, a:I.689, Z\c:48",2V:60". Andesineis zoned from Anasout to about An35. F errohastingsite schist Amphibole 4 is from a heavy, very dark greenlineate schist that con- sistsof ferrohastingsitealong with only 2 per cent of sphene,0.5 per cent of oligoclaseor qvartz, and 0.5 per cent of epidote. The unusual rock compositionlisted in Table 3 should be very reliable.In thin sectionthe amphibole is seen to be evenly colored, medium-grained,and moder- ately prismatic; {100} tends to parallel the schistosil-yand the prism axesdefine the rock's strong lineation. The spheneis very fine-grained and is scatteredevenly throughout the amphiboleas though it had sep- arated from it.

Bi oti te- h or n blen ile lott alite Amphibole 5 is from one of the most mafic rocks of the batholith;how- ever, its propertiesare essentialiythe sameas thoseof hornblendesfrom 898 ROBERT R. COMPTON

'l'he other batholith rocks. calculated chemical analysis of the tonalite must be consideredas approximate only becausemany point counts showedthe rock to be heterogeneous,especially with respectto the ratio of biotite to hornblendeand of qvarLz to plagioclase.The approximate modeis: quartz... 14.0 magnetite 1.0 plagioclase (Ana) 58.0 epidote ., 0.2 hornblende.. - 14 5 apatite 0.1 biotite 12.O zlrcon tr. Plagioclaseis subhedralto euhedraland is zonedin an oscillatoryman- ner from coresof about Anaoto rims of about Anzz(as determinedby the Rittmann method on the universal stage). Quarlz is in fine-grained ag- gregatesbetween the plagioclasegrains, while clots of mafic grains tend to wrap around plagioclase.Hornblende prisms are subhedraland about 1-2 mm. long, and biotite forms raggedanhedra that are twice that size. Hornblendeis not rimmed by biotite, though it is locally replacedby it. Coarse-grainedpistachite occurshere and there with chloritized biotite. The minor chlorite and fine-grainedalteration products of plagioclase are included in the mode as the minerals thev replaced.

P.q.necBNBSrs The sequenceof metamorphic assemblagesin basaltic rocks of the Bidwell Bar area is outiined in Fig. 1. The arrows indicate only the principal sourcesof materialsfor each mineral of a new assemblage;ac- tually, almost all of the mineralsin adjacent columnsare linked in some degree. The derivation of minerals from basaltic rocks is based purely on textural evidence because igneous relics are very scarce. These

BASALT GREENSTONE HORNFELS AMPHIBOLITE

leucoxene ------> + spnene -.---= sphene lron oxides ore + ore

+ AGTINOLITE HORNBLENDE\ chtorite > HoRNBLENDE /"-z epidote ---.-.--=\ (clinopyroxene) OLIGOCLASE+ ANDESINE

Frc. 1 Sequenceof mineral assemblagesin metabasaltic rocks of the tsichvell Bar region. The principal minerals are in capital letters. AMPHIBOLE PARAGENESIS 899 reactionswere definitely earlier than the contact metamorphism,and all of the rocks of the aureolewere once similar to the rocksnow found in the regional terrane. The conditions of greenschistmetamorphism are not critical here,and have alreadybeen described (Compton, 1955,p. 17-1S). The progressivecontact metamorphic assemblagesare essentially iso- chemical,except that water was lost at each stage.The changesof leu- coxeneand hydrated iron oxidesto ore and sphene,and the conversion of actinolite to hornblendeapparently took place abruptly in the outer zone of the aureole.The amphibolereactions involved both the addition of Al to actinolite and the combination of chlorite and epidote to yield more hornblende,with magnelite and water as additional products.The principalpart ial reactionsare:

1) chlorite f epidote* Si+hornblende*Al *Fe//,+water 2) actinolite*Al+hornblende*Sif (Mg, Fe,,) 3) Fe//,f Fe,,+magnetite

The overall reaction formed by combining these partial reactionsis: actinolite+chloritef epidote+hornblende f magnetitef waterf NIg For partial reaction 2, the analysesindicate that Mg rather than Fe,, is the dominant product, and it is thereforenecessary to considerplagio- claseas a reactant with the excessMg, forming more (though minor) hornblende.rn the inner zone of the aureole,aluminous hornblende was increasedonly slightly by the further reactions of epidote and minor chlorite. rn the caseof the plagioclasereactions, however, albite reacted with epidote to produce oligoclasein the outer part of the aureoleand then continuedto react in a progressiveway so that plagioclasecomposi- tions tend to be more and more calcic as the bathorith is approached. The appearanceof clinopyroxenenear the batholith contact probably involves the destruction of some hornblende; however, there is insuffi- cient data on the compositionsand distribution of these pyroxenes to de- termine whether they were formed by local high temperatures,by the dehydrating effectsof the adjacentmagma, or simply by local concentra- tions of Ca and Mg. The paragenesisof the ferrohastingsiteschist involves rocks with un- usual compositions.Ferrohastingsite occurs at several localities in the inner zone of the aureole,in one place with abundant andesineand ac- cessorymagnetite and sphene,in another place with abundant labrador- ite, clinopyroxene,and titaniferousore. rn an important exposureon the cascade road 1,300 feet northeast of the Feather Falls village Forest service station, ferrohastingsiteschist is associatedwith metacherts, suggestingthat the schistswere once iron-rich basic tuffs, perhaps pala- gonite tufis enriched in iron hydroxides. That they were not impure 9OO IIOBEIIT R. COMPTON

Iimy beds that have undergoneiron metasomatismis indicated by the fact that they are not associatedwith tactites that occur in the same part of the aureole.In any case,the hastingsiticamphibole formed be- causeof the high Fe/Mg ratio oI the rocks and their low silica content; it cannot be correlatedwith a higher grade of metamorphism' The hornblende of the tonalite is notably less aluminous than the hornblendesof the aureole,and this fact, coupled with its large grain sizeand euhedral(110) faces, indicate that it crystallizedfrom a magma' rather than being xenocrysticafter amphibolite. The low alumina con- tent is doubtlessa result of partition of materialswith aluminousbiotite that was growing at the same time.

SrcNrncaNcE oF THE AcrrNor-rB-HonNer-BNoE REAcrroN Conversionof actinolite and other greenschistminerals to aluminous hornblendeis petrologicallysignificant at Bidwell Bar becausethe two minerals are abruptly superimposed,the rocks involved are extensive, and the reaction produced a textural change that affords a basis for map- ping the extent of a new metamorphic environment. How does this abrupt change correlate with what might be expected from the progres- sive metamorphismof actinolitesto hornblendes?Comparison of analy- ses of many actinolites, hornblendes,and hastingsites(from rocks of various compositions) indicates that these minerals form a continuous solid solution series,with the number of Al atoms in Z positionsranging from 0 to somewhatmore than 2 (seeesp. Hallimond, 1943,p' 80-87)' In metamorphic rocks containing appreciablealumina, the number of AI atoms in Z positionsof hornblendesis greaterin higher grade rocks than in lower grade rocks, and it seemsmost reasonableto expect a continu- ous increment of Al atoms in Z positionsof actinolitesand hornblendes during the progressivemetamorphism of basaltic rocks (Foslie, 1945, p. 98; Turner,!948, p. 90,98; Turner and Verhoogen,1951, p' 465)' The inconsistencybetween these views and the Bidwell Bar paragenesisre- quires a review of the mineral data that substantiate ideal gradation fiom actinolite to hornblendein metabasalts.When this is done, it will be possibleto considerwhy the Bidwell Bar metamorphismproduced a difterent paragenesis. Compositions of all chemically analyzed actinolites and hornblendes from ,greenstones, and epidote amphibolitesare presented in Table 4, in order of increasingamount of Al in Z positions.Number 4 was calculatedfrom an unpublished chemical analysis generouslyfur- nished by Prof. c. o. Hutton, and the complete analysis and optical measurements,all made by Hutton, are given in Table 5. Ali of the rocks concernedhave compositionsreasonably close to that of basalt,although A M P II I BOLE P 1 RAGEN DS I.] 901

Tlnr,n 4. NuMsans or PnrNcrpel Arous rw Fonuur,aB or AMpnrsolns rnol4 GnrnNscnrsts, GnnrNsronrs, aNo Eltnotr Aupnrgor,rrBs

t1

(z) si 7.73 7.74 I .66 7 64 7.61 7.s4 7 4i 716 706 672 6s8 648 Al 024 0 .26 0. .34 036 039 o16 053 0 84 0.9.1 128 142 152

At 000 0 19 0.43 027 002 o 21 0.13 0.56 0.s5 0 48 016 o.'il (Y) Ti o.o2 0 03 0.04 004 005 008 004 014 008 0.06 005 007 000 o12 009 0.2.5 0 28 042 052 0.33 0.32 0. 57 0 64 0..59

Ms 378 3.25 3.25 3.48 2.85 2.17 3.12 1 58 1.98 2 23 2 80 2.28 (x) Fe" I 11 1.39 1.16 0.97 1 87 1.82 1 28 2.22. 2 14 1_80 111 1.56 Mn O.O2 000 003 o02 003 004 000 0.03 0 03 007 003 00.5

Ca 193 1.90 I 92 1.68 1.73 1.60 1.57 1.68 1 63 169 199 t92 (W) Na 039 015 02.1 0.27 0 27 o 44 0.41 0.39 0.19 0 28 0.18 0 3.5 K 000 0.02 0.1l 002 009 o 02 0.00 0.15 0 26 0.07 008 010

OH 2.11 1 9,1 1 .56 2 00 2.02 192 206 206 202 1 89 2.60 2 1+

1 "Hornblende from fibrous hornblende clinozoisite-albite schist, Limebury point, Salcombe Estuarl.' South Devon." (Tilley, 1938,p. 505). 2 "Hornblende from hornblende-epidote-albite schist below Signal Station, PrawJe Point, Start District, South Devon " (Tilley, 1938,p. 504). 3 Actinolite from mafic greenstone (Tables I and 2, analysis 1). 4 Actinolite from albite-epidote chlorite-actinolite schist of Chl. 3 subzone; 32 chains E.N.E. of Jessie Peak, Mid-WaLatipu Survey District, Western Otago, New Zealand. Analyst, C. O. Hutton .5. "Hornblende out of chlorite-epidote-albite-amphibolite (26); 0.4 mile south-east of Loch Gair Hotel, Argyll." (Wiseman,19J4, p. 368). 6 "Actinolite from albite-stilpnomelane-actinolite-schist, No. 2646, .56chains east of Billy Goat Creek- Arrow River junction, westernOtago.,, (Hutton, 1940,p. 13-14). 7. "Amphibole from albite-epidoteactinolite-calcite-schist, No 2718,from the summit of Coronet peak, Wakatipu region." (Hutton, 19J8,p.209; 1940,p. 13). 8 "Hornblende out of garnet-biotite-eDidotealbite-amphibolite t68/4),o.2 mile N 51" E from northern end of Loch-na-Craige,near Achahoish,South Knapdale.,,(Wiseman, 1934, p. 382). 9. "Hornblende out oI biotite-epidote-albite-anphibolite(65/41." Same locality as 8. (Wiseman,193.1, p. 383). 10 Hornblende from hornfelsic metadiabase (Tables 1 and 2, anatysis 2). 11' Ilornblende lrom hornblendeschist (with quartz, oligoclase,epidote, and biotite); northwest of Lakc Pappilanselka, Orivesi (Seitsaari, 1953, p g9, table 5). 12. Hornblendefromstratiliedhornblendeschist(withquartz,oligoclase,epidoteandbiotite);northeastof Lake Vaavujarvi, Orivesi,Finland. (Seitsaari,1953, p. 89, table 4).

somewere originally clasticrocks. Numbers 1 and 2 are frbm the green- schistsof the Start region of Devonshire,England; numbers 4,6, and,7 are from the widespreadgreenschists of the Otago region,New Zealand; numbers 5, 8, and 9 are from the progressivelymetamorphosed epidio- rites (metadiabases)of the central and SouthwestHighiands of Scotland; numbers 11 and 12 arc from epidote-oligoclaseamphibolites (originally basicsediments) of the Tampereschist belt, Finland; and numbers3 and 10 are from greenstoneand epidote-oligociase-hornblendehorn{ers of the Bidwell Bar area. rt is notable that the actinolites from the lolver gradeparts of the variousareas are closelycomparable (numbers 1,2,3, 4, and 5). None, however,are pure "end member" actinolite;apparently 902 ROBERT R. COMPTON

Tesln 5. Couposrrrot aNo Pttvstcer, PnopEnttns or AcrtNor,rto lnolr Oteco, NBrv ZBar-eNo

Sio,: 54.07 o:7.629 AlrOr 3.81 p:r.643 TiOz 0.40 7:1 650 FerOs )L\ 7- r:0.021 Cr:O: 0.08 z/P:19" Nio 0.03 2Y":73"+2" NIgO 16 45 X: colorless FeO 8.25 lf :pale green MnO 0.24 7:pa!e green rvith bluish tinge CaO 11.13 Z>Y>X NazO 1.03 Gzo":3.17 K:O 0.17 -812U' 213 Analysis ancl physical measurementsby C. O. HzO- 0.11 Hutton. See Table 4 {or locality description.

Total 100.35

even at lowestmetamorphic grade a quarter or third of an atom of the 8 Z atoms is Al, a fact that must be kept in mind when comparing actino- lites of metabasaltic rocks with those of less aluminous metamorphic rocks. The actinoiites of the Otago region are of particular interest be- cause they show a variation within the greenschistrealm itself. The validity ol progressivemetamorphism in theserocks is undeniable,being basedon regional mapping of zonesof progressivemineral and textural changes,keyed by severalthousand thin sections(Hutton and Turner' 1936).The chlorite zone was divided into four subzoneson the basisof degreeof reconstitution of the original graywackebeds. In Chl. 3 sub- zone,from which amphibole number 4 was collected'the rocks are fine- grained schistswith a few relics of sedimentarygrains, while in Chl. 4 subzone,the sourceof amphiboles6 and 7, the rocks are exceptionally coarse-grainedand even show segregationbanding. In keepingwith this progressivechange, actinolites 6 and 7 have more Al in Z positionsthan does actinolite 4, and lhey are also more aluminous than any analyzed actinolites from greenschists.Still another amphibole from Chl. 4 sub- zone,not listed in Table 4 becauseit is from too mafic a rock, has 0.71 atom of Al in the 8 Z positions(Hutton, 1940,p. 15; 1956,p.231-232). The approachof theseamphiboles to aluminous hornblendemay be ex- plained by Turner's view (1948,p. 39) that "In Chlorite 4 the tempera- ture appearsto have been somewhathigher than in the other subzones' and was certainly maintained for a much longer period." Certainly, the compositionalrange of theseamphiboles supports the conceptof gradual 4 ],1P H I BOLE PARAGEN ESI S conversionof actinolite to hornblendeduring progressivemetamorphism. The amphibolesstudied by Wiseman (5, 8, and 9 of Table 4) supply an additional step in the progressiveincrease of Ai in Z positionsduring increasinggrade of metamorphism.Proof of progressivemetamorphism lies in the zonal pattern of mineral changes,based in this caseon tire classicalmineral zonesmapped in metashalesof the SouthernHighlands (Barrow,1912,p.274-279; Tilley, 1925).The compositionalchange from the greenschistactinolite (number 5) to the two hornblendesfrom epidote amphibolitesof the garnel zone (numbers8 and 9) is distinct; however, the garnet zoneamphiboles are not much more aluminousthan Hutton's Chl. 4 subzoneamphiboles, and they are lessaluminous than the horn- blendesof typical epidote-freeamphibolites of comparableareas. There- fore, thesegarnet zone hornblendesfit exactly into a serial changefrom greenschistthrough epidote amphibolite to amphibolite facies condi- tions, further supporting the view that reactionsof actinolite to horn- blende are continuous,not discontinuous. Just as surely as in the SouthernHighlands, the epidoteamphibolites at Bidwell Bar representan intermediatestep in the conversionof green- schists to epidote-freeamphibolites. Furthermore, the texture of the hornfels from which amphibole 10 (Table 4) was separatedindicates it was a massive metadiabase,directly comparablewith Wiseman's epi- ;nor was it any more aluminousthan the rocks from which Wise- man separatedhis two hornblendes.The one obviousdifference between the Bidwell Bar rocks and thoseof either the SouthernHighlands or the Otago region is that contact metamorphismproduced the rocks at Bid- well Bar while regional metamorphism produced the other two assem- blages.What are the physical difierencesbetween these particular cases of contact and regional metamorphism that account reasonablyfor the differencesin mineral parageneses?Probably, there are three principal factors that produced the differences,and perhaps the most important of these was the differencebetween rate of escapeof water and rate of increaseof temperatureduring inception and progressof the amphibole reactionsdescribed above. The theoreticalbasis for this {actor has been presentedin some detail by Thompson (1955) and in part by Yoder (1955,p.513-516) . On this basis,I interpretthe gradationalamphibole changesin the Otago and Highlands schistsas occurring in broad ter- raneswhich were heatedso slowly that r,vaternever accumulatedrapidly enoughto inhibit the progressof the reactionat eachtemperature incre- ment. Deformation and continuousrecrystallization may have been im- portant factors in moving water continuouslyout of the terrane.In the Bidwell Bar aureole,rapid increasein temperature relative to rate of escapeof water created a high water pressurealmost at the outset of 904 ROBERTR. COMPTON

metamorphism, and the amphibole reaction was inhibited in the low temperature range. By the time that escapeof fluids was sufficient to allow appreciablereaction, temperatureswere so high as to produce an aluminous hornblende.It is possiblethat a complicating abnormality may have been produced by fluids forced through the outer part of the aureoleby rapid dehydration nearer the batholith. Another {actor that probably contributed to the contrast in mineral paragenesesis the difierence between rates of reaction at any one Lem- perature and rates of heating. In the regional terranes, temperatures rose slowly enough so that even sluggish (low temperature) reactions were usually at or near equilibrium. The coarse-grainedOtago schistsof Chlorite subzone4 seem to emphasizethis factor strongly. In contrast to this true progressivemetamorphism, the contact rocks were heatedso rapidly that low temperature reaction of actinolite was volumetrically ineffective in the relatively short period of time between inception of heating and rapid reactionat intermediateto high temperature.Finally, the third factor is based on the degree of comminution and mixing of constituentsbefore and during recrystallization.Dynamic metamorph- ism of the rocks of the Southern Highlands and Otago produced a fine- grainedmixture of mineral units that could react readily as temperatures increased.Because of the lack of granulationin the outer zoneof the Bid- well Bar aureole,only the finest grained constituentsof the greenschists could react easily in the lower temperaturerange, a situation substanti ated by minor hornblendeovergrowths and relict actinolite coresin the larger Bidwell Bar amphiboles. There is insufficient data to determine which of these factors was dom- inant in controlling the amphiboleparageneses of the two environments; however,epidote amphibolites from other areasindicate the Bidwell Bar conditions were not unique. The epidote amphibolites of the Tampere schist belt of Finland are comparabiein severalways to the Bidwell Bar epidoteamphibolites, although there is apparently no chanceof tracing the progressionof changesfrom greenschiststo amphibolites in the Tampere terrane.The hornblendesof theserocks are aluminous (11 and 12 of Table 4) and in severalcases they are heterogeneous,like those in the outer part of the Bidwell Bar aureole(Seitsaari,1953, p.87 88). With regard to their origin, Seitsaari (1953, p.96) "suggeststhat the crystallization of the blue-greenhornblende in the Tampere schists is generallyrelated to a Iate increaseof temperature,and that the penetra- tive movements have been weak, if not terminated, during its forma- tion." A closeapproach to contact metamorphic conditionsis supported by thesefacts: 1) the hornblendeoccurs as porphyroblasticprisms that cut acrossthe schistosity,2) the hornblende occurs only in the finer AMPIIIBOLE PARAGENESIS 905 grainedrocks, and 3) the analyzedhornblendes were collectedless than a mile from granitic intrusions of batholithic size. A caseof contact metamorphic amphibolesshowing abrupt reaction relationsis reportedfrom the llisaka seriesof Japan by Sugi (1931).No analysesof theseamphiboles are available,but Sugi studied textural and refractiveindex changescarefully enough so that there is little doubt but that he detectedthe same amphibole relations as those at Bidwell Bar. The Nlisaka rocks are basic to intermediatelavas and greenschiststhat have been progressivelymetamorphosed to epidote amphibolites and amplribolitesnear a quartz intrusion. In the transition between greenschistsand amphibolites, actinolite and hornblende occur to- gether with epidote,sodic plagioclase, and chlorite.It is notable that the new blue-greenhornblende of this transition forms sharp rims on actino- lite, or elseclear-cutprismsthat lie on or nearlarge actinolite grains(Sugi, 1931,p. i15-116).In maficschists, hornblende typically hasformed first in areascontaining chlorite, indicating the importanceof this mineral in the formation of aluminous hornblende.Perhaps the greatest value of Sugi's work on theserocks was to demonstratethat abrupt reaction of actinolite to hornblendemay be detectedwithout laboriousseparations and analyses. Considering the general situation presented above, the actinolite- hornblendereaction must be consideredcarefully in assigning rocks of basalticcomposition to oneof the metamorphicfacies. Wiseman's epidote amphibolites fit exactly the albite-epidoteamphibolite facies of many authors (seeesp. Turner, 1948,p.88-89), but the compositionsof Hut- ton's actinolitesfrom the Chl. 4 subzoneindicate that somemetabasalts from the biotite zonemight well contain hornblenderather than actino- lite. According to the picture presentedabove, the temperature realm of albite-epidoteamphibolite facies in the Bidweli Bar region is repre- sented by metastablegreenschists and greenstonesthat lie outside the aureole;in the aureoleitself, conditions of the transitional facies have been essentiallyoverstepped, and the aureoleis entirely in amphibolite facies.The abrupt appearanceof cordieriteand andalusitein the outer- most parts of some pelitic aureolesmay be explainedby the same sort of overstepping of transitionai assemblages.Refore the hornblende analysesof the Bidwell Bar amphiboleswere available, I interpretedthe oligoclase-epidote-hornblendehornfelses as representing a stress-free variant of the albite-epidoteamphibolite facies(Compton, 1955,p. 12). On the basis of the hornblendereaction, however, I rvould reject this view, much as Thompson questionedthe "stress" and "antistress" con- trol of aluminum silicatesin regional and contact terranes (Thompson, 1955,p. 98). 906 ROBERTR. COMPTON

The eventual disappearanceof epidotefrom epidoteamphibolites pro- vides a useful physical basisfor defining the upper limit of the epidote amphibolitefacies as used by Foslie (1945), Seitsaari(1953, p.95), Barth (1952,p.338), and others.This limit is, of course,well beyondthe upper limit of the albite-epidoteamphibolite facies.The analysespre- sentedabove show clearly that hornblendesoccurring with epidote and oligoclaseare almost as aluminousas thoseof epidote-freeamphibolites, indicating that the actinolite-hornblendereaction takes place over a lower temperaturerange than the epidote-plagioclasereaction.

CoNcr-usroNs Through more data are wanting, it is likely that the seriesactinolite- hornblende-hastingsiteis controlledby temperature,and that a continu- ous reactionseries might be producedby ideally progressivemetamorph- ism of basalticrocks. Discontinuitiesin the seriesappear to be valuable in pointing up physical conditionsof metamorphism,especially those af- fecting reactionrates. It is suggestedthat seriaianalyses of any tempera- lure-variant solid solution serieswill be of unique value in interpreting metamorphismof a given suite of rocks,even lvhere the absolutetempera- ture-pressuredimensions of t.heseries are not known.

AcxNowr-eocMENTS I am grateful to Prof. C. O. Hutton for his generoushelp and sugges- tions regarding the separatory and analytical procedures,for his con- tribution of an unpublishedanalysis of an Otago actinolite, and for his reading the manuscript. A Supplementary ResearchGrant from Stan- ford University coveredthe cost of the analyses.

RrnnnoNcr,s

B,tnrn, Ton. F W (1952), Theoretical petrologl,: N. Y., John Wiley and Sons, Inc., 387p. BARRow, G. (1912), On the geology of lower Dee-side and the southern Highland borcler: Geol Assoc, Proc.,23r 274-290 Corrlrox, R. R. (1955), Tronclhjemite batholith near Bidn'ell Bar, Caiifornia: GeoI.Soc. Amey , BuIl., 66,944. FosLrE, SrnNen (1945), Hastingsites, and amphiboles Irom the epidote-amphibolite facies: lflorsft geol. T'id.sshr.,25, 7 4-98 llallruoNo, A. F. (1943), On the graphical representation of the calciferous amphiboles: Am Mi.neral.. 28. 65-89. IIurrox, C O. (1938), A chemical and optical study' of a lorv grade melamorphic actino litic anrlrhibole frorn Coronet Peak, I'cstern Otago, Nel. Zealand: Min. l[og.,25, 207-211 ---- (1940), Iletamorphism in the Lake Wakatipu region, western Otago, Nerv Zea- land: Nezr Zealand.Dept Sci. ond.Ind.u,str. Res , Geol. Mem.,5, l-84 - - (1956), The composition of an actinolite: Acto Cryst.,9, 231 232. A],1P H I BOLE PA RAGEN ES I S 907

.+ND'fuRNItR, l'. J. (1936), N{etamorphic zones in north west Otago: Roy. Soc' iY eu Zealand., I\rans, 65, 405-406. Sntr:-q,lau, JuHaNr (1953), A blue-green hornblende and its genesis from the Tampere schist belt, Finland: Comm. gEol.Finland.e, 8u11.,26,83-98. Sucr, KnN-rcru (1931), On the metamorphic facies of the n{isaka seriesin the vicinity of Nakagawa, Prov. Sagami: Jap. Iour. Geol.and, Geogr.,9187-142. Trroursox, J B, Jn.(1955), The thermodynamic basis for the mineral facies concept: Am. Jour. 9ci..253,65 103. Trttov, C. E. (1925), A preliminary survey of metamorphic zones in the southern High- lands of Scotland: Quart. Jour. Geol.Soc. London,8l, 100-112. (1938), The status of hornblende in low grade metamorphic zones of green schists: Ceol.M ag., 75, 497- 5l 1. TLTRNER,F. J. (1948), Mineralogical and structural evolution of the metamorphic rocks: Geol.Soc. Amer., Mem,.30,l 342. --' ANDVrnnoocnN, J. (1951), Igneous and metamorphic petrology: N. Y., McGraw- Hill Book Co., 602 p. WannnN, B E. (1930), The crystal structure and chemical composition of the monoclinic amphiboles:Zeit. K rist..72, 493-517. Wrsoulx, J.D.H. (1934),The central and south-westHighiands epidiorites:a study in progressivemetamorphism: Quart. Iour. Geol. Soc.London,90,354-417 . 'rvater Yonnn, H. S., Jn. (1955), Role of in metamorphisrn: GeoI.Soc. Amer, Spec. Paper 62, 5O5 524.

Monu,scrip! receited February 3, 1958