Composition and Cation Order Variations in a Sector-Zoned

American Mineralogist, Volume 65, pages 313-320, 1980

Compositionand cation order variationsin a sector-zonedblueschist pyroxene

MTcHAEL A. ClnpeNrnn Hoffman Laboratory, Department of Geological Sciences Harv ard University, Cambridge, M assachusetts 02 I 38

Abstract

Pyroxenecrystals in aknost monomineralicveins cutting a blueschistrock from California show sectorzoning. Portions of each crystal which grew at facescutting the c axis are sodic augite(e.9. Jd,,AcroAugur) whereas omphacite (e.g. JdrrAc,rAugaT) formed at facesparallel to the c axis (probably at {100}, {010}, {1l0}). Examination by transmissionelectron microscopy revealsthat the omphacitehas a primitive lattice with antiphasedomains on a scaleof lessthan -50A and, in a few areas,exsolution lamellae approximatelyparallel to (100) of a disorderedpyroxene. The sodicaugite contains fine lamellaeof amphibole(?) approximately parallelto (010). I suggestthat the activitiesof ions in the solutionsfrom which crystallizationoccurred dic- tated a pyroxenecomposition between omphacite and augitebut that this lay within a broad, low-temperaturemiscibility gap and two pyroxenesgrew simultaneously.The sectoralar- rangementdeveloped during rapid crystal growth; local chargebalance etrectsin each new layer of material controlled the compositiondistribution. Short-rangeordering occurreddur- ing growth in directionsnormal to the pyroxenechains (giving omphacite)but was restricted during growth parallel to the chains (giving sodic augite).

Introduction tion to differing extents (Leung, 1974). If the most important constraint is the structure of the exposed Sector zoning is well known in minerals. The phe- crystal faces, then a number of factors can influence nomenon is of special interest in that variations in the observed composition distribution. These in- composition between different sectors of individual clude: (a) the nature of partially formed cation sites crystals reflect local variations in crystal growth con- on the surface, "protosites" (Nakamura, 1973; ditions and therefore provide an insight into the fac- Dowty, 1976), at which cations can be adsorbed and tors controlling the formation of minerals. Although then incorporated in non-equilibrium proportions there appear to be as many theories to explain the into the crystal; (b) the need to maintain charge bal- origin of sector zoning as there are examples of it, ance within each layer or between successive layers most authors agree that it develops under metastable of deposited material, as for Al"' + 2Al* = H* + 2Si conditions where growth rates are too rapid for bulk substitutions in staurolite (Hollister, 1970); (c) the equilibration between crystals and their growth me- presence of a solvus and the orientation of best-fit- dium. The best documented cases are reviewed by ting crystal planes for epitaxial overgrowths, as in Dowty (1976) and include titanaugite, staurolite, and augite/pigeonite (Bence and Papike, 1971; Hollister augite/pigeonite. et al.,l97l; Boyd and Smith, l97l). Differences be- Possible causes of sector zoning can be divided tween growth sectors need not be restricted simply to into two categories: control by the supply of ions to composition. The degree of cation or anion order can the growth surface and control by the structure of the also vary, as in adularia (Akizuki and Sunagawa, crystal faces. If the rate of arrival of ions to the sur- 1978), topaz (Akizuki et al., 1979), and some stauro- face is inadequate to keep up with the crystal growth lites (Dollase and Hollister, 1969). rate, then extraneous ions will be metastably incor- In this paper a new example of sector zoning, in- porated. Because crystal faces not related by symme- volving sectors of sodic augite and omphacite, is de- try tend to grow at different rates, growth zones will scribed. The crystals appear to have grown rapidly in develop which depart from the equilibrium composi- a low-temperature (blueschist facies) metamorphic 0003-004X/80,/030,+-03I 3$02.00 313 314 CARPENTER: BLUESCHIST PYROXENE rock and have both compositional and order differ- Due to the interlocking of the crystals, growth ences between their sectors. A complete explanation faces have only rarely been preserved. There are oc- of the crystal growh mechanisms cannot be offered casional pockets of quartz into which protruding py- but several of the constraints discussed above may be roxene crystals have straight or slightly curved faces involved. For example, charge-balancing problems (Fig. la). These crystals appear to have grown into are thought to affect the distribution of R*, R'*, R'* cavities which were later filled by quartz, and their cations in ordered omphacite (Clark et al.,1969), and growth faces lie parallel to the c axis and at steep an- there is the possibility of a miscibility gap between gles to it. The most common growth habit of pyrox- omphacite and sodic augite (Brown et al.,1978). ene crystals includes {ll0}, {100}, {010}, which are Cation ordering in omphacite leads to a symmetry almost certainly developed in this case, and {l lT} as reduction from a C-face-centered to a primitive lat- the faces cutting the c axis. {l lT} faces would give tice (Clark and Papike, 1968) and the appearance of more sharply pointed terminations to the crystals antiphase domains (APD's) observable by transmis- than are observed, but other known possibilities, sion electron microscopy (Champness, 1973; Phakey such as {0ll}, {T0U, and {001}, do not appear to and Ghose, 1973). Champness argues that natural give the observed interfacial angles either and posi- omphacites grow metastably in a disordered (or par- tive identification of the terminal faces has not been tially ordered) state below their equilibrium ordering made. Growth normal to the c axis (presumably at temperature (7",o) and then order, whereas Yo- [00], {010}, {ll0} faces) was of omphacite, koyama et al. (1976) favor a low value of [.0 and whereas in the c direction sodic augite formed. The APD formation during cooling. Fleet et al. (1978) sodic augite portions have fine lamellar features par- demonstrated that 7",o may be as high as 725+20"C, allel to the cleavage (Fig. lc), which makes a useful well above the crystallization temperature of most contrast with the omphacite for optical identification omphacites, but suggested that the APD's are a purposes. growth feature. There is still disagreement therefore probe analysis about whether natural omphacites grow with meta- Electron stable limited order (Champness, 19731,Carpenter, The methods of analysis and recalculation are the 1978) or with a high degree of long-range order same as discussed by Carpenter (1979a). For the esti- (Fleet et al., 1978). mation of Fe'*/Fe'* proportions it was assumed that Si : 2.00 because the analyses show a slight excessof Specimen description and observations silica (possibly due to a systematic error; see Carpen- The sodic pyroxenes occur in emerald green veins ter, 1979a). Fe'* was then taken as Na - Al, and pro- up to I cm wide in a glaucophane-epidote-mus- portions ofjadeite, acmite and augite molecules were covite-sphene schist. The hand specinen was col- recalculated to a total of lODVofrom Jd : Al, Ac : lected by Dr. S. O. Agrell as a loose block in a land- Fe'*, Aug : Ca. Representative analyses are shown slip of low-grade Franciscan rocks at Russian River, in Table I and all the data are plotted in Figure 2. seven miles north of Cloverdale, California (Harker There is a sharp compositional break between the collection no. 102729, Dept. of Mineralogy and Pe- sectors (Fig. 2). The c-axis sectors have 8-157o jadeite trology, Cambridge, England). It closely resembles while the central omphacite portions have2S4lVo ja' the samples figured by Esseneand Fyfe (1967) except deite. No systematic zoning was detected within the that the pyroxene crystals have no obvious preferred sectors. Small pyroxene crystals in the host glauco- orientations. The veins consist of 99Vopyroxene with phane schist have compositions in the same range as tare quartz and muscovite. Individual crystals are the vein omphacite (Fig. 2). composite and have roughly tabular shape with di- microscopy mensions -lmm x lmm x 0.5mm with the short di- Electron mension parallel to the c axis. In most orientations, Specimens were prepared for transmission electron with the notable exception of two cleavage sections, microscopy by ion beam thinning and were exam- they show a distinctive hour-glass texture (Fig. l). ined in an AEI EM6G microscope operating at l00kV. They do not extinguish uniformly between crossed The crystals contained a high density of defects, such polars but appear to be slightly curved. The pyroxene as dislocations, which may account for their non-uni- cleavages show similar curvature. There are some form optical extinction. variations of the typical hour-glass texture, particu- Selected area diffraction (SAD) patterns from ho- larly at the sector interfaces where one sector may mogeneous omphacite regions contained diffuse h + branch into the other. k : odd reflections, indicating a P lattice and cation CARPENTER: BLUESCHIST PYROXENE 315

b d

Fig. l. Optical micrographs of sector-zoned vein pyroxene crystals (partially crossed polars). (a) Hour-glass t€xture of omphacite and sodic augite. Slightly curved crystal faces and some surface ledges (arrowed) can be seen against quartz (Q). (b) Sketch ofzoned crystal in (a) showing the branching sector interfaces. A : augite, O : omphacite. (c) Sector zoned crystal with fine lamellae in the sodic augite (arrowed). (d) Sketch of (c). ordering. The maximum size of equiaxed APD's sections from exsolved and honogeneous areas of which could be imaged using these reflections was omphacite typically have weak violations of an n -50A (Fig. 3a), but in general they were below the glide parallel to (010), suggestingthat the space resolution limits of the electron microscope.A few group of the orderedstructure may be P2 as opposed areaswith coarser,highly irregular domAinswere ob- to F2/n.' served.SAD patternsfrom sodicaugite had absences Sodic augite regions frequently contain fine la- appropriate for a C-face-centeredlattice (h + k : mellae approximately parallel to (010) (Fig. 3c) odd reflectionsabsent). which may correspondto the lamellar features ob- Exsolution, in the form of fine (-1004) corhposi- servedoptically. SAD patternsfrom theseareas have tion modulations approximately parallel to (100), considerablestreaked intensity along the D* direction was found in patches (Fig. 3b). Dark-field images with h + k: odd reflectionsshow one componentin rClark and Papike (1968)and Clark et al. (1969)reported a P2 contrastand the other out, suggestingthat one is or- spacegroup for orderedomphacite, but subsequ€ntauthors (Mat- dered and the other is not. This microstructureis dis- sumotoe, al., 1975; Curtis et al., I 975)have argued in favor of P2/ violations have, however, been observedin se- cussedin more detail elsewhere n. Weak n-glide in relation to the lected area electron diffraction patterns from severalspecimens, general problem of exsolution in sodic pyroxenes particularly where the APD sizeis small (Carpenter,1978, l9'l9a; (Carpenter, 1979b). SAD patterns showing a*-c* Carpenterand Okay, 1978). 316 CARPENTER: BLU ESCHI 57, PYROXENE

Table l. Representativepyroxene analyses.Augl and Oml afe epidote, as in the specimendescribed here, rather from sodic augite and omphacitesectors of a singlecrystal. Aug2 than lawsonite, does not necessarilyindicate higher and Om2 are from a second crystal and Aug3, Om3 from a third metamorphictemperatures and that the temperature

Analysis no. Augl on1 Aug2 on2 Aug3 om3 of vein formation may also have been in the vicinity of 300"c. si02 53,97 55.26 54.03 55.55 54.09 55.Ar AI2O3 2.34 8.19 L.52 9.00 2,42 7.72 The curvature of pyroxene crystals and their high

FeO * 11.05 9.73 l0, 81 9.64 10.54 LO.43 density of defectsare probably growth features be- MnO o.29 0,18 0.33 0.2r 0.19 0.00 causethe veins show no obvious signs of deforma- Mco 9.61 5.78 9.77 5.02 9.56 s.73 tion. This imptes rapid crystal growth. The sectoral CaO r7.22 ).I.76 18.63 10.64 18.36 11.82 arrangement of sodic augite and omphacite cannot Na2O 4.36 7,51 3.4L 8.30 3.72 7.47 easily be explained as overgrowth or replacement 'fotal 98.41 98.50 98.36 98.88 98. 58 during crystallizationof successivephases, as for ex- Catlons per 2,00 Si ample can the omphacite rims on impure jadeite L+ si 2.00 2.00 2. 00 2.00 2.00 2.oo from the Urals (Dobretsov and Ponomareva,1968), Al- 0.11 0.35 0.07 0. 38 0. 11 0. 33 and from Guatemala and Syros (Carpenter, 1978, F"+ ** o.2o o.r8 0.17 0.20 0,16 0.20 1979a).Simultaneous crystallization of two pyroxene F"2* o. 14 o.12 0. 16 0.09 0.r7 0,11 compositions is also compatible with the inter- l'n2* o. 01 o. ol 0. 01 0.01 0.01 0.00 interfacesbetween some sectors(Fig. la). ug2* o,53 o.3r o ,54 o.27 0.53 0.31 fingering parallel c^21 0.69 0.46 o.74 0.41 0.73 0,46 Omphacite appearsto have grown on faces Na+ 0.31 0.53 0.24 0.58 0.27 0.53 to the c axis whereassodic augite grew on facescutting the c axis. In placessmall surfaceledges can be 'M 3,99 3.96 3.93 3.94 3.98 3.94 seenwhich exposeboth typesof face (Fig. la). Depo- Jd 11 35 7 38 11 33 sition of successivelayers of the preferred phase at Ac 20 r8 I7 20 L6 20 branching sec- Aug 69 47 76 42 73 47 each face would lead to the observed tor boundaries(Fig. 4). * Total Fe glven a6 FeO ** Fer calculated as Na - Al The hiatus in compositionbetween sectors is most Trace amounts CrrO3, V203, TiO2 detected in some crystals .of NtO, K^O not detect6d- readily explainedif there is a miscibility gap between sodic augite and omphacite at low temperatures,as in the caseof augite/pigeonitesectoral intergrowths but apparently with positionsof maximum intensity where the solvus is between Ca-rich and Ca-poor (Fig. 3d). The diffraction patternsare similar to those compositions. Such a miscibility gap has already (1978),who described arising from exsolutionlamellae of amphiboleparal- beensuggested by Brown et al. slightly man- lel to (010) in augite (Smith, 1977).ln the present an apparentequilibrium coexistenceof As these specimen the lamellae are tentatively identified as ganese-enrichedomphacite and diopside. is frequently diffi- amphibole,but it is not clearwhether they formed by authors pointed out, however, it and metas- exsolutionor alteration.2 cult to discriminatebetween equilibrium table assemblagesin low-temperaturemetamorphic Discussion rocks. There is more substantialevidence for limited jadeite at low Esseneand Fyfe (1967) and Coleman and Clark miscibility between and omphacite l97l; (1968) estimated T - 2OO-300.Cand P - 6-9kbar temperatures(Dobretsov et al., Carpenter, ionic radii of the M- for the formation of the California blueschists, but 1979a),and considerationof the that non-ideal higher temperatures have been proposed for the site cations supports the argument whole high-grade tectonic blocks, some of which contain solid solution behavior should occur in the jadeite (Na*, l.o2A; Al'*, 0.54A)-diopside blueschistassemblages (Taylor and Coleman, 1968). range Deviations from ideality Esseneet al. (1965)suggested that the occurrenceof (Ca'*, 1.00A;MB'*, 0.724). will be enhancedby the substitution of Fe2* (0.784) for Mg'* but reduced by the substitution of Fe'* (0.65A) for Al1* (ionic radii from Shannon,1976). In 2 Fine lamellae of what appear to be amphibole have also been the absenceof cation ordering at intermediatecom- observed in aegirine-jadeite from a Syros blueschist (Carpenter, unpublisheddata). Colemanand Clark (1968)reported X-ray evi- positions(i.e. primitive omphacite),a solvusmight be dence for amphibole intergrown with a jadeite from California. anticipated betweenjadeite and augite at low tem- CARPENTER: BLUESCHIST PYROXENE 3t7

Jodeite Augite

Fig. 2. Electronmicroprobe analyses ofpyroxenes fuom 102729.Each point is one analySis.Crosses are from c-axissectors and filled circles from other sectors of vein crystals. Open circles are groundmass pyroxenes. Thc continuous and broken lines represent the composition limits of cation ordering observed in blueschist omphacites (from Carpenteg 1979a\. peratures.3If so, then omphacitemay only appear in at {100} faces.The cationswhich should be adsorbed low-temperaturerocks if it grows with some degree preferentially at these sites are those which can form of cation order to stabilizeit relativeto the two-phase the strongestbonds, ae. thosewith the largestcharge end-memberassemblage. The TEM observationsin- and smallestionic radii (Dowty, 1976).Of the cations dicatethat the maximum scaleof long-rangeorder in available.Al3* on Ml and Ca'* on M2 should be in- the omphacitedescribed here is -504, but is gener- corporated in preferenceto Mg2* and Na*, respec- ally much less,so that cation ordering achieveddur- tively. To maintain chargebalance these would have ing growth can only have been limited and short to be accompaniedby equivalent proportions of Na range in character. and Mg, giving omphacite(Fe'z* is expectedto follow Mg, and Fe3l to follow Al3*). Protositeson Constraintson sectorzoning {010}, {l l0}, {l I l}, (and {00U) faceshave more complete The activities of ions in the migrating solutions oxygen coordination but are similar to eachother. A from which the vein crystals grew apparently dic- protosite model would therefore predict the largest tated that a pyroxene intermediate in composition diference in sector compositions to be between between omphacite and sodic augite should form. {100} and the rest.This is not the casein the present However, this composition lay within a miscibility specimen. gap and the simultaneouserystallization of two (b) Charge balancing. The charge di-fferencesbe- phasesseparated by l1-2o%ojadeitecontent resulted. tween cationsin omphaciteplay a critical role in de- The questionremains as to why the two phasesgrew termining the ordered configuration (Clark et al., as different sectorsof single crystals.The constraints 1969).Ml sitesin chainsparallel to c are occupiedal- which apply to other examplesof sector zoning, as ternately by Mg and Al and M2 sites are filled pre- discussedin the introduction, can each be consid- dominantly by Ca or Na, according to whether two ered. Firstly the growh rate was rapid but the supply of the three Ml cations with which they share oxy- ofions to the growth surfacesdoes not seemto have gens are Mg or Al respectivelyo(Clark et al., 1969; been a problem becausethere is no systematiczoning Curtis et al., 1975;Matsumoto et al., 1975).Because within the sectors.The important factorsmust relate the low crystallizationtemperatures were almost cer- to differencesin crystal structuresbetween faces par- tainly below Ii.o, there must have been a tendency allel to or cutting the c axis. for ordering to occur at the earliestpossible opportu- (a) Adsorption at protosites.According to the nity, and it has already been suggestedthat limited modelsof Nakamura (1973)and Dowty (1976),the short-rangeorder was incorporatedinto the growing most flexible Ml and M2 protositesin pyroxenesare omphacite sectors.Omphacite may therefore have

3 A similar solvus was proposed by Bell and Davis (1965) but at a It is assumed that the ordering follows essentially F2/n high temperatures. Their experimental results, however, conflict symmetry and that the weak n-glide violations observed in with those of Kushiro (1964) obtained under approximately electron diffraction patterns in this case represent only minor similar conditions and have not been substantiated. departures from this. 318 CARPENTER: BLUESCHIST PYROXENE

Fig. 3. Electron micrographs of vein pyroxenes. (a) Antiphase domains in an omphacite sector. Dark field using an h + k : odd reflection. (b) Fine composition modulations approximately parallel to (100) in omphacite; bright field. (c) Fine lamellae approximately parallel to (010) of (?) amphibole in sodic augite; bright field. (d) Electron diffraction pattem from an area sirnilar to (c). The streaked reflections are from the lameflae. (a), (b), (c) ar€ at the same magnification.

formed at the facesmost favorablefor ordering in the 4l'-/. surfacelayer, and converselysodic augite may have grown where ordering in the layer being deposited I'r c was most difficult. At {100} facesalternate Ml sites in the octahedralchains are simultaneouslyexposed, allowing Mg'* to be adsorbedpreferentially next to (o) Al'*. Almost the same situation arisesat {l l0} and {010} exceptthat the Ml sitesare not quite coplanar. aug Adjacent M2 siteswithin the layersand in successive ',, oug layers can be filled accordingto the charge balance om requirements.The situation at {00U, however,forms (d) (c) a sharp contrast.The cation occupyingan Ml site in a new (001)layer would be influencedby an M2 cat- Fig. 4. Schematic two-dimensional cross section through a ion immediately next to it. In the orderedomphacite crystal to illustrate the effect of a ledge (shown in a) at the growth structurethese sites must be filled by Mg and, on av- faces. Successive layers (with thickness t) on faces growing perpendicular to the c axis are ofomphacite (shaded) but are sodic erage,l/3Ca 2/3Na (or Al and l/3Na 2/3Ca). Thus augite on faces growing in the c direction (unshaded) (b,c). The in most casesMg2* is next to Na* and Al3* next to result is a branched sector interface (dotted line) (d). Ca'*, giving a local charge deficiency or excess. CARPENTER: BLUESCHIST PYROXENE 319

These charge imbalances within the new (001) layer Acknowledgments would act in opposition to cation ordering during I am grateful to Dr. S. O. Agrell for providing the specimen and growth. The same applies to other faces cutting the c to Dr. D. L. Bish, Dr. C. Klein, Dr. L. S. Hollister, and Dr. E. J. axis, but the effect decreasesas they make a steeper Essene for criticisms of and improvements to the manuscript. I ac- and steeper angle to it. A requirement of both short- knowledge the receipt, from The Natural Environment Research Council of Great Britain, first of a Research Studentship held at range ordering and local charge balance in the the Department of Mineralogy and Petrology, Cambridge Univer- growth layers therefore can only be met at faces par- sity, England, and lately of a Research Fellowship. allel to c, where omphacite develops. Ordering and charge balancing are less compatible at the faces cut- References ting c, so that the growth of augite is preferred. Akizuki, M. and L Sunagawa (1978) Study of the sector structur€ (c) Nucleation and epitaxial growth. If original in adularia by means of optical microscopy, infra-red absorp- tion, and electron microscopy. Mineral. Mag., 42,453462. omphacite nuclei acted as nucleation sites for sodic augite, or vice versa, the shared plane should be the anomalous optical properties of topaz. Mineral. Mag., 43,237- one with best structural match. From the exsolution 241. relations of the ordered and disordered components Bell, P. M. and B. T. C. Davis (1965) Temperature-composition jadeite-diopside. (Fig. 3b), this appears to be (100) and not a plane at section for Carnegie Inst. lMash. Year Book, 64, 120-123. angle to the c axis. A further an reason for rejecting Bence, A. E. and J. J. Papike (1971) A martini-glass clinopyroxene this explanation is that the choice of growth planes from the moon. Earth Planet. Sci. Leu., 10,245-251. was not restricted to the earliest stages of crystalliza- Boyd, F. R. and D. Smith (1971) Compositional zoning in pyrox- tion. The two phases grew on their preferred faces at enes from lunar rock 12021, Oceanus Procellarum. I. Petrol., 12, surface ledges even during the later stages of crystal 439464. Brown, P., E. J Essene and D. R. Peacor (1978) The mineralogy growth, leading to the branching texture of some of and petrology of manganese rocks from St. Marcel, Piedmont, the sector interfaces. Italy. Cont rib. M ineral. P et r ol., 67, 22'l -232. The most satisfactory explanation for the sector Carpenter, M. A. (1978) Kinetic control of ordering and ex- zoning involves local charge balancing within each solution in omphacite. Contrib Mineral. Petrol., 67, 17-24. - (19'l9a) from Greece, Turkey and Guatemala: new layer of material being deposited combined with Omphacites composition limits of cation ordering. Am. Mineral., 64, 102- a tendency for cation ordering to occur at the earliest 108. possible stages of crystal growth. This mechanism is - (19'l9b) Mechanisms of exsolution in sodic pyroxenes. broadly similar to that proposed for sector zoning in Contrib. Mineral. Petrol. (accepted for publication)' staurolite by Hollister (1970), though for the present - and A. Okay (1978) Topotactic replacement of augite by in a blueschist rock from north-west Tvkey. Min- case there is probably the additional effect of a mis- omphacite eral. Mag.,42,435438. gap cibility between omphacite and sodic augite. Champness, P. E. (1973) Speculation on an order-disorder transformation in omphacite. Am. Mineral., 58,54U542. Clark, J. R. and J J. Papike (1968) Crystal-chemical character- Conclusions ization of omphacites. Am. Mineral.,5J, 840-868. D. E. Appleman and J. J. Papike (1969) Crystal-chemical Sector zoning in sodic pyroxenes from Cloverdale, characterization of clinopyroxenes based on eight new structur€ California developed during rapid crystal growth and refinements. Mineral. Soc. Am. Spec. Pap., 2, ll-50. involved differences of both composition and cation Coleman, R. G. and J. R. Clark (1968) Pyroxenes in the blueschist facies of Califomia. Am. J. Sci., 266, 43-59. order. The main constraint on sector formation ap- Curtis, L W., J. Gittins, V. Kocman, J. C. Rucklidge, F. C. Haw- pears to have been the need to maintain local charge thome and R. B. Ferguson (1975) Two crystal structure refine- balance within the growth layers so that cation order- ments of a P2/n titarian ferro-omphacite. Can. Mineral., 13, 62- ing during growth was possible only at faces parallel 67. to the c axis. The intergrown crystals provide sup- Dobretsov, N. L., Yu. G. Lavrent'yev and L. N. Pospelova (1971) Immiscibility in Na-Ca pyroxenes. Dokl. Akad. Naak. SSSR, porting evidence for a low-temperature miscibility 201, t52-t55. gap between omphacite and sodic augite. The om- - and L. G. Ponomareva (1968) Comparative characteristics phacite represents the first case where it can be posi- ofjadeite and associated rocks from Polar Urals and Prebalkash tively stated that reasonable-sized antiphase domains rcgion. I nternational Geol. Rev., 10, 22I-242, 24'l -2'19 - (larger than a few unit cells) did not form during Dollase, W. A. and L. S. Hollister (1969) X-ray evidence of ordering differences between sectors of a single staurolite crystal. growth below 2".o. Rather, metastable growth with Geol. Soc. Am. Abstracts with Programg 7,268-270. limited short-range cation ordering seemsto have oc- Dowty, E. (1976) Crystal structure and growth: II. Sector zoning curred. in minerals. Am. Mineral., 61,460469. CARPENTER: BLUESCHIST PYROXENE

Essene,E. J. and W. S. Fyfe (196?) Omphacite in Californian Nakamura, Y. (1973) Origin of sector zoning of igneous clinopy- metamorphic rocks.Contib. Mineral Petrct., tS, l-23. roxenes.Am. Mineral., 58, 986-990. --, - and F. J. Turner (1965) Petrogenesisof Franciscan Phakey,P. P. and S. Ghose(1973) Direct obsewationofantiphase glaucophane schistsand associatedmetamorphic rocks. Contrib. domain structure in omphacite. Contfib. Mineral. Parol., 39, Mineral. Petrol., I I, 695-7A. 239-245. Fleet, M. E., C. T. Herzberg,G. M. Bancroft and L. p. Aldridge Shannon, R. D. (1976) Revised effective ionic radii and systematic (1978)Omphacire srudies, l:The P2/n+ A/c transformation. studies of interatomic distances in halides and chalcogenides. Am. MineraL, 63, l 100-1106. Acta Crystallogr.,A 32, 751-7 67. Hollister, L. S. (1970) Origin, mechanismsand consequencesof Smith, P. P. K. (1977)An electronmicroscope study of arnphibolc compositional sector-zoningin staurolite. Am. Mineral., 55, lamellaein augite. Contrib.M ineral. Petrol., 59, 317-322. 742-766. Taylor, H. P. and R. G. Coleman(1968) Ot8/Ot5 ratios of coexist- --, W. E. Trzcienski Jr., R. B. Hargravesand C. G. Kulick ing minerals in glaucophanc-bearing metamorphic rocks. Geol. (1971)Petrogenetic significance ofpyroxenes in two Apollo 12 Soc.Am. Bull.,79, 1727-1756. samples.Proc.2nd" Lunar Sci. Conf,259-557. Yokoyama,K., S. Banno and T. Matsumoto(1976) Compositional Kushiro, I. (1965)Clinopyroxene solid solutionsat high pressures. range of P2/n omphacite from the eclogitic rocks of Crntral CarnegieInst. llash Year Book, 64, ll2-117. Shikoku, Iapan. Mineral. Mag., 40,773-779. Leung, I. S. (1974)Sector zoned titanaugites: morphology, crystal- chemistryand growth. Am. Mineral., 59, 127-138. Matsumoto,T., M. Tokonami and N. Morimoto (1975)The crys- Manuscript received, August 24, 1979; tal structureof omphacite.Am. Mineral., 60, 634-&1. acceptedfor publication, Novenber 30, 1979.