Contact Metamorphismby Ophiolite Peridotite from Neyriz,Iran · in Orogenic Andesites and Related Rocks, R. Thorpe, Ed. (Wiley, NewYork, in press). 14. The first sample was collected

history of Mount St. Helens, it was pre-dicted that the volcano was likely toerupt, possibly before the end of thiscentury (10). The early activity observedat Mount St. Helens is similar in charac-ter to the eruption of Lassen Peak Volca-no in Califqrnia during the period 1914 to1921 (11j. During the first year of Las-sen's activity, approximately 150 phreat-ic eruptions sent cold ash and steam col-umns to heights as great as 3000 m abovethe crater. No report was made of sulfuror acid fumes during that period. Mag-matic activity began on 19 May 1915 witha violent pyroclastic eruption of daciticmagma. Nuee ardente-type eruptionsfollowed on 20 and 22 May. Then theeruptive activity subsided with minor pe-riods of heightened activity, until at leastFebruary 1921. Our observations ofMount St. Helens are certainly consist-ent with the possibility that the currenteruptive style may evolve into eruptionswith a larger magmatic component.

Volcanic gas geochemistry can pro-vide indications of increasing magmaticcontributions to eruptions. At Mount Et-na in Sicily, the rate of release of SO2 hasbeen shown to increase before magmaticeruptions (12). The S/Cl ratio in leach-ates is related to SO2/HCl ratios in vol-canic gas (13). Monitoring the SO2 emis-sion and measuring the S/Cl ratio inleachates could provide important pre-monitory indicators of a magmatic erup-tion. Therefore, continued geochemicalmonitoring of Mount St. Helens is ad-vised.

RICHARD E. STOIBERSTANLEY N. WILLIAMS

LAWRENCE L. MALINCONICODepartment ofEarth Sciences,Dartmouth College,Hanover, New Hampshire 03755

References and Notes

1. D. Mullineaux et al., Sci. Event Alert Net-work Bull. 5 (No. 3), 2 (1980).

2. Oregon Journal (29 March 1980), p. 1.3. R. E. Stoiber and A. Jepsen, Science 182, 577

(1973).4. We use the term "rate" to indicate the mass that

would be emitted if the volcano were eruptingcontitulbusly. In fact, the eruptive periods atMount St. Helens probably totaled less than 1/10of a day.

5. R. E..Stoiber, G. B. Malone, G. P. Bratton,Ge6: Soc. Am. Program Abstr. 10, 480 (1978).

6. P. Taylor and R. E. Stoiber, Geol. Soc. Am.Bull. 84, 1031 (1973).

7. W. I. Rose, Jr., Geology 5, 621 (1977).8. Data in Table I are from volcanic eruptions that

represent a variety of these variables, althoughall combinations of all variables are not includedin any one sample set.

9. S. Thorarinsson, in Volcanic Activity and Hu-man Ecology, P. D. Sheets and D. K. Grayson,Eds. (Academic Press, New York, 1979), pp.125-156.

10. D. R. Crandell, D. R. Mullineaux, M. Rubin,Science 187, 438 (1975); D. R. Crandell and D.R. Mullineaux, U.S. Geol. Surv. Bull. 1383-C(1978).

11. A. L. Day and E. T. Allen, The Volcanic Activi-ty and Hot Springs ofLassen Peak (Publication360, Carnegie Institution of Washington, Wash-ington, D.C., 1925).

SCIENCE, VOL. 208, 13 JUNE 1980

12. L. Malinconico, Nature (London) 276, 112(1979).

13. W. I. Rose, Jr., R. E. Stoiber, L. Malinconico,in Orogenic Andesites and Related Rocks, R.Thorpe, Ed. (Wiley, New York, in press).

14. The first sample was collected by D. Swanson(USGS) and made available to us by T. Casade-vall (USGS). The second sample was collectedby L.L.M. with logistical support provided by a

USGS helicopter and D. A. Johnston (USGS).15. The assistance of D. Burrington of NBC News

and T. Casadevall and D. A. Johnston of theUSGS and the cooperation of the USGS person-nel in Vancouver, Wash., are gratefully ac-knowledged. This research was supported underNASA Cooperative Agreement 5-22.

21 April 1980; revised 14 May 1980

Contact Metamorphism by an Ophiolite Peridotitefrom Neyriz, Iran

Abstract. Ophiolites are conventionally regarded as fragments offormer oceaniclithosphere. Mineralogical andfield evidence indicates that peridotite of the Neyrizophiolite was intruded at high temperature into folded crystalline limestones, form-ing skarns. This excludes the formation of the ophiolite at a mid-ocean ridge but isconsistent with its origin by intrusion during continental rifting.

Basic and ultrabasic rocks of an ophi-olite complex are exposed in the Neyrizregion of southern Iran where they wereemplaced by thrusting during the lateCretaceous (1). One unusual aspect ofthis ophiolite is the presence of skarns atthe contact belween crystalline lime-stones and the ophiolite peridotite.These skarns have been interpreted (i) asremnants of an intrusive contact of pre-thrusting (1, 2) or postthrusting (3) age,(ii) as the product of metamorphism atthe base of an obducted ophiolite (4), or(iii) as the result of fortuitous associationwith the ophiolite during subduction (5).The mineral assemblages reported (1)suggest other possible origins, such ascalcium metasomatiSm associated withserpentinization or regional metamor-phism of impure limestones. The originof these skarns is critical to the inter-pretation of the Neyriz ophiolite and theevolution of the Zagros mountain chainand they are of more general interest asophiolite-related metamorphic rocks, yettheir detailed mineralogy is unknown. Ireport mineralogical data from newskarn localities that indicate contactmetamorphism of limestone by the ophi-olite peridotite.

Mountain-size marble blocks are ex-posed just east of Tang e Hana, a smallvillage 30 km northwest of Neyriz. Theycontain large-scale close folds plungingsteeply southwestward and rest on perid-otites with a subhorizontal tectonizedcontact offset locally by small steepfaults. All specimens of peridotite that Ihave examined from this area are whollyserpentinized harzburgite, often with acompositional banding of originally or-thopyroxene-poor and orthopyroxene-rich layers. Ricou (1) also concluded thatthe peridotite is harzburgite, although re-porting a few specimens with up to 10percent diallage. About 2 km from thewesternmost skarn locality is an isolated

hill (about 1 km2 in area) of partly ser-pentinized, mylonitized, harzburgitesand lherzolites.The marble-peridotite contact trun-

cates the folds in the marble blocks, andtwo groups of skarns are found within 30m of it. Type I skarns occur within theserpentinized harzburgite, whereas type2 skarns occur within the marbles. TypeI skarns are clinopyroxenites and clino-pyroxene-marbles. The pyroxenites re-semble rodingite "dikes"; they are sub-parallel to the marble-serpentinite con-tact and are composed of coarse,equigranular, pale-green pyroxene with afew scattered grains of chrome spinel.The pyroxene-marbles occur as angularfragments up to 0.5 m in diameter con-taining only calcite and idioblastic, ran-domly oriented grains of pyroxene.Some fragments contain areas of unde-formed, serpentinized harzburgite inwhich the contact between harzburgiteand marble is marked by a zone (2 to 3mm) of clinopyroxene partly altered tochlorite and calcite and completely un-disturbed except where crosscut by laterveins of calcite and chlorite. Type 2skatns are subcircular wollastonite-py-roxene-calcite bodies up to 10 m in diam-eter. They are exposed on flat surfaces,and their three-dimensional form is un-known; they are very coarse-grained andhave a granoblastic polygonal texturetypical of high-temperature metamorphicrocks. Their contacts with the marblesare sharp, completely undisturbed, andmarked locally by pegmatitic calcite.Garnet replaces wollastonite along somecleavage traces and grain boundaries,and adjacent dark-green pyroxene ispatchily altered to a paler variety. Irreg-ular subplanar zones crosscut the rocksin which plagioclase, garnet, and calciteare associated with poikilitic, altered py-roxene and wollastonite.

I interpret the field relations as in-

0036-8075/80t0613-1259$00.50/0 Copyright X 1980 AAAS 1259

on Decem

ber 4, 2020

http://science.sciencemag.org/

Dow

nloaded from

Fig. 1. A diagram of temper-ature versus fluid composition C02(T-Xco) showing selected ex- 1200- AlTs 2Woperimentally determined and Grcalculated equilibria for the FeTs + 2WoCaO-Al203-SiO2-CO2-H20 and Gr50And50CaO-Fe2O3-SiO2-CO2-H20 1000-

tems (11). Fluid pressure equals @ Type 3total pressure (2 kbar). The - Type 2position of the isobaric in- -variant point I ia experimental- ,_ 800- xly determined (11), and the po-sition of point II has been cal- C02culated from data in (10). The /lines labeled type 2 and type 3 600-

indicate the minimum temper-atures for the reaction fassaite+ wollastonite = garnet in the Om 0.2 0.4 Q6 Q8 tOtype 2 and type 3 skarns, esti- C02

mated from the positions of the equilibria AlTs + 2Wo = Gr, and FeTs + 2Wo = Gr5OAnd5O(11). The curve labeled Gr (0.47) indicates the shift of the curve Gr + CO2 = An + Wo + Cc forgrossular activity = 0.47, which is the maximum value for type 2 garnets; decreasing grossula-rite content of garnet will shift this curve to higher Xco, at constant temperature. The inter-section of this curve with the type 2 line thus locates an isobaric invariant point (Ila), indicatinga CO2-rich fluid accompanying retrogression. Abbreviations: AI Ts, aluminum Tschermak's py-roxene; An, anorthite; And, andradite; Cc, calcite; FeTs, ferric Tschermak's molecule (6); Gr,grossular; Qu, quartz; and Wo, wollastonite.

dicating that in most places little move-ment has occurred along an original in-trusive contact. Type 2 skarns, whichshow no signs of tectonism at all, ap-pear to represent originally intrusive"tongues" and are restricted to a zonewithin a few tens of meters of the presentcontact. The marbles often contain py-

roxene close to the contact but not else-where. Type 1 skarns appear to repre-sent part of a discontinuous zone at themarble-peridotite contact. Tectonismhas resulted in the disruption of this rigidzone and produced a few centimeters ofschistose serpentinite in places and hassmoothed out most of the original irregu-

larities of the contact and incorporatedthe type 1 skarns as blocks in serpentin-ite, although they remain close to theiroriginal position. In at least one place, anentire contact zone is apparently pre-served undisturbed (1).Another group of skarns (type 3) oc-

curs discontinuously at the contacts be-tween smaller marble blocks (less than100 m in diameter) and serpentiniteabout 10 km east of Tang e Hana. Theseskarns have a coarse mylonitic foliationand contain fassaite, garnet, wollaston-ite, and anorthite porphyroclasts in a re-

crystallized matrix chiefly composed ofwollastonite. They differ from types 1

and 2 in the much greater abundance ofgarnet and plagioclase and in the occur-rence of sphene and apatite. Theseskarns are clearly not in their position offormation. Their contact with highlyschistose serpentinite is sharp, and, al-though the mylonitic foliation is normal-ly parallel to the serpentinite schistosityand the marble-serpentinite contact, inone place it is highly oblique to both. Themylonitization obviously predates theserpentinization since the fabric is over-

printed by hydrous Ca-Al silicates typi-cal of low-temperature calcium meta-somatism associated with serpentiniza-tion. Intergrowth of fassaite and garnet in

Table 1. Representative analysis (oxide percent by weight), formulas, and end-members (mole percent) of coexisting minerals from skarns. Skarntypes are indicated parenthetically. All samples were analyzed with an energy-dispersive electron microprobe with estimates of Fe2O3 forpyroxenes and garnets by assuming cation sums of 4 and 16 normalized to 6 and 24 oxygens, respectively. Wollastonite and plagioclase wererecalculated to 6 and 32 oxygens, respectively. Analyses NY59-B6 and NY59-B9 are diopside and garnet formed by reaction of fassaite (NY59-B7) with wollastonite. For pyroxene end-members, see (6); for garnets, Gross = grossular, Andr = andradite, and Schorl = schorlomite; forplagioclase, An = anorthite and Ab = albite; N.D., not detected.

NY59-1 NY40-13 NY40-22NY5-2 NY59-B7 NY59-B6 NY59-B9 wollas- NY40-15 NY40-18 plagio- wollas-Sample fassaite fassaite diopside garnet tonite fassaite garnet clase tonite(type 1) (type 2) (type 2) (type 2) (type 2) (type 3) (type 3) (type 3) (type 3)

SiO2 47.37 45.68 52.64 38.02 51.54 41.77 37.69 44.41 51.31TiO2 0.73 0.93 0.02 1.09 N.D. 2.23 1.53 N.D. N.D.A1203 7.27 9.12 0.32 13.58 N.D. 12.27 13.05 35.78 N.D.Fe2O3 4.34 4.59 0.79 11.64 0.00 3.86 11.73 N.D. 0.00Cr2O3 0.30 0.04 N.D. N.D. N.D. N.D. N.D. N.D. N.D.FeO 0.00 3.09 8.29 0.00 0.31 7.40 0.61 N.D. 0.35MnO N.D. 0.10 0.28 0.39 N.D. N.D. 0.24 N.D. N.D.MgO 13.59 10.70 12.18 0.29 N.D. 7.72 0.44 N.D. N.D.CaO 25.42 25.28 25.46 35.82 48.32 24.04 34.95 19.36 47.87Na2O 0.24 0.18 0.13 N.D. N.D. N.D. N.D. 0.48 N.D.Total 99.26 99.71 100.11 100.83 100.17 99.29 100.25 100.03 99.53

Si 1.758 1.713 1.980 5.926 1.994 1.602 5.919 8.200 1.996Aliv 0.242 0.287 0.014 0.074 0.000 0.398 0.081 7.789 0.000AlVi 0.076 0.116 0.000 2.422 0.000 0.157 2.335 0.000 0.000Ti 0.020 0.026 0.001 0.128 0.000 0.064 0.181 0.000 0.000Fe3+ 0.121 0.130 0.)22 1.365 0.000 0.111 1.386 0.000 0.000Cr 0.009 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000Fe2+ 0.000 0.097 0.261 0.000 0.010 0.237 0.081 0.000 0.011Mn 0.000 0.003 0.009 0.051 0.000 0.000 0.032 0.000 0.000Mg 0.752 0.598 0.683 0.067 0.000 0.441 0.103 0.000 0.000Ca 1.011 1.016 1.026 5.983 2.003 0.988 5.881 3.830 1.996Na 0.017 0.013 0.009 0.000 0.000 0.000 0.000 0.172 0.000

AlTs8 AlTs12 AlTso Gross62 AlTsI6 Gross59 An96FeTs11 FeTs12 FeTs, Andr34 FeTs11 Andr34 Ab4TiTs2 TiTs3 TiTso Schorl2 TiTs6 Schorl3Others79 Others73 Others,g Others2 Others67 Others4 120SCECE OL 0t260 SCIENCE, VOL. 208

on Decem

ber 4, 2020

http://science.sciencemag.org/

Dow

nloaded from

porphyroclasts and development of gar-net rims on some porphyroclasts indicatethat some of the mylonitization mayhave occurred during cooling and retro-gression of the skarn assemblage soonafter its formation.

Representative analyses of skarn min-erals are listed in Table 1. Unaltered py-roxenes are characterized by high Ca,Al, and Fe3+ and low Na and are fas-saites in which charge balance is main-tained principally by the coupled sub-stitutions

(Mg,Fe2+)Si = (Al,Fe3+)Al

and

(Mg,Fe2+)Si2 = TiAl2in Tschermak's molecule [Ts, expressedas end-members AlTs, FeTs, and TiTs(6)]. Fassaites in the type 1 skarns(± calcite) have lower Ts contents thanthose in type 2 skarns (+ wollaston-ite + calcite), whereas type 3 fassaites(+ garnet + wollastonite + plagioclase)have the highest Ts contents. Garnetsare grossular-andradite solid solutionswith less than 10 mole percent of otherend-members; those in type 2 skarnshave a higher grossular content (57 to 70mole percent) than those in type 3 skarns(42 to 59 mole percent). Plagioclase intype 2 skarns has altered to albite, epi-dote, and calcite but in type 3 skarns isanorthite-rich (89 to 97 mole percent).Wollastonite and calcite in all skarns arevirtually pure.

Fassaite is found only in silica-poor,high-temperature environments (7-9) buthas a relatively large stability field. How-ever, the presence of wollastonite limitsfassaite stability by reactions including:

Ts + wollastonite -> garnet (1)

Some constraint on formation temper-atures for the skarn rocks is provided bythe appearance of garnet, which is form-ing retrogressively by two reactions:

Fassaite + wollastonite --

garnet + Ts-poor diopside (2)

Fassaite + wollastonite + CO2garnet + plagioclase +diopside + calcite (3)

Temperatures for these reactions (Fig. 1)have been obtained from thermochemi-cal data (10) and experimentally deter-mined equilibria in the systems CaO-A1203-SiO2-H20 (11) and CaO-AI203-Fe2O3-SiO2 (9), and the effect of othercomponents has been estimated (11). Noequilibria directly fix pressure in theskarns, and therefore a total pressure of2 kbar has been assumed by comparisonwith the stratigraphy of the nearby Oman13 JUNE 1980

ophiolite (12); fluid pressure has been as-sumed equal to total pressure. Reaction2 is only slightly dependent on pressure(approximately 50°C/kbar, dependingupon composition) and is independent offluid composition; garnet-fassaite com-positions indicate minimum temper-atures at 2 kbar of 880NC for type 2skarns and 920°C for type 3 skarns. Theassemblage garnet-fassaite-wollastonite-plagioclase-calcite found in type 2 skarnsas a result of reaction 3 fixes an isobaricinvariant point (Fig. 1) in T-Xco2 space.Its position will vary slightly with vary-ing Fe3+/Al in fassaite and garnet, but, ifit is assumed that plagioclase is pureanorthite similar to that in type 3 skarns(Table 1), the assemblage indicates aCO2-rich fluid accompanying retro-gression (Fig. 1).These temperatures are minimum esti-

mates for skarn formation (since theyapply to retrograde reactions) and thevery high estimated temperatures aresupported by the type of fluid inclusionsin fassaites and wollastonites of type 2skarns. These contain primary and pseu-do-secondary inclusions interpreted asmagmatic inclusions, an indication thatfassaite and wollastonite crystallizedinitially from a melt (13). Contact tem-peratures of at least 900°C thereforeseem reasonable. Similar assemblages tothose of the Neyriz skarns have been re-ported from dolomitic limestone xeno-liths at the margin of the Bushveld lopo-lith for which intrusion temperatures ofthe order of 1 100°C have been suggested(9, 14); differences in mineral assem-blages are a reflection of the initial Mg-poor composition of the Neyriz marbles.Fassaite-garnet assemblages producedfrom limestones metamorphosed by gab-bros of the Boulder batholith (8) appearto differ not only in bulk composition andfluid composition but were metamor-phosed at rather lower temperatures,as indicated by the absence of theassemblage garnet-fassaite-wollastonite(15).The only reasonable explanation for

the origin of these high-temperatureskarns is the intrusion of the ophioliteperidotite. Metamorphism of the perido-tite can be excluded because of the ab-sence of any signs of metamorphism,other than the later episode of serpentini-zation, and the large amounts of in-troduced Ca and CO2 required. Calciummetasomatism has occurred but has pro-duced a characteristic set of minerals in-cluding pectolite, prehnite, grossular,hydrogrossular, Ts-poor diopside, andchlorite. These minerals clearly postdateskarn formation and retrogression andare typical of low-temperature Ca-meta-

somatism associated with serpentiniza-tion (16). Regional metamorphism of im-pure limestones fails to explain the con-sistent and intimate spatial relationshipsbetween peridotite, skarn, and marble;the very high temperatures requiredwould also be extremely unusual. Thepresent structural position of the skarnson top of the peridotite and their mineral-ogy and textures are very different tosubophiolite metamorphic rocks (17)produced by obduction. The type 3skarns are not in their position of forma-tion and could be explained as remnantsof a contact with the ophiolite gabbrosnow exposed within a few kilometers ofthe easternmost skarn localities. Thiswould explain their relatively high Al,Ti, and P contents. However, skarns arenever found associated with the gabbros,and the mineralogical and chemical simi-larities to type 1 and type 2 skarns,which are clearly related to the perido-tite, suggest a similar origin for the type 3skarns. The zonation reported from acomplete contact (1) suggests that thetype 3 skarns represent the remnant of askarn zone closest to the peridotite.There are no other intrusive rocks in theNeyriz region which could have been re-sponsible for the skarn formation.The evidence therefore indicates that

the skarns formed partly by crystalliza-tion of a melt formed at the contact andpartly by diffusion metasomatism, as aresult of the intrusion of peridotite intocontinental basement limestones. Thisorigin of the crystalline limestones issuggested by the large-amplitude foldscrosscut by the peridotite contact. Theperidotite may have been intruded as acrystal mush with a small amount of par-tial melt produced by adiabatic meltinginduced by sudden continental rifting, orit may have been hot and solid but plasticand locally molten at the contact as a re-sult of the introduction of CO2. In a C02-rich fluid phase SiO2, A1203, Fe2O3, CaO,and MgO were probably the chief com-ponents (18), but the skarn zone formedwould probably have been narrow sincethe energy required for the assimilationof marble is supplied largely by the en-thalpy of crystallization of phases fromthe magma (18, 19). Formation of skarnminerals would therefore have acceler-ated the cooling and crystallization of theintrusion. In such a case, the composi-tion of the pyroxene would be a functionof the rates of diffusion and chemical po-tential gradients across the contact dur-ing multicomponent diffusion. The in-creasing Ts content of the fassaite fromtype 1 to type 3 skarns is a consequenceof the higher diffusion rates and higherinitial concentrations in the peridotite of

1261

on Decem

ber 4, 2020

http://science.sciencemag.org/

Dow

nloaded from

SiO2 and MgO compared to A1203 andFe2O3. Zoning of the pyroxenes is notobserved and is unlikely to be preservedsince intracrystalline diffusion is muchmore effective at high temperatures.The preservation of an intrusive con-

tact is therefore more consistent withmodels of the Neyriz ophiolite, that sug-gest an origin during continental rifting(20) (perhaps resulting in an ocean simi-lar to the present-day Red Sea) ratherthan at the spreading center of a wideocean destroyed through subduction (5)or partially obducted as a hot slab (4).Other ophiolites of the Tethyan belt,where evidence for subduction and ob-duction is often weak, may have origi-nated in a similar manner.

ROBERT HALLDepartment ofGeological SciencesQueen Mary College, University ofLondon, London El 4NS, England

References and Notes

1. L.-E. Ricou,Bull. Soc. Geol. Fr. 13, 146(1971).2. K. W. Gray, Q. J. Geol. Soc. London 105, 189

(1949).3. A. J. Wells, Geol. Mag. 106, 385 (1969).4. N. H. Woodcock and A. H. F. Robertson, Geol-

ogy 5, 373 (1977).5. S. J. Haynes and H. McQuillan, Geol. Soc. Am.

Bull. 85, 739 (1974).6. Pyroxene end-members have been calculated in

the order suggested by H. S. Yoder and C. E.Tilley [J. Petrol. 3, 342 (1962)]. Sodium has beenassigned to acmite (NaFe3+Si2O6), and Tscher-mak's molecule (Ts) has been calculated asthree end-members, aluminum Ts (AlTs,CaAl,SiO6), ferric Ts (FeTs, CaFe3+AISiO,),and titanium Ts (TiTs, CaTiAI206).

7. W. A. Deer, R. A. Howie, J. Zussman, Rock-Forming Minerals (Longmans, London, 1978);C. E. Tilley, Geol. Mag. 75, 81 (1938).

8. A. Knopf and D. E. Lee, Am Mineral. 42, 73(1957).

9. H. G. Huckenholz, W. Lindhuber, J. Springer,Neues Jahrb. Mineral. Abh. 121, 160 (1974).

10. H. C. Helgeson, J. M. Delany, H. W. Nesbitt,D. K. Bird, Am. J. Sci. 278A, 1 (199S8).

11. Experimentally determined equilibria are as fol-lows: (i) calcite + quartz = wollastonite + CO2[H. J. Greenwood, Am. Mineral. 52, 1669(1967)1; (ii) calcite + anorthite + wollaston-ite = grossular + CO2 [T. M. Gordon and H. J.Greenwood, ibid. 56, 1674 (1971)]; (iii) anor-thite + 2 wollastonite = grossular + quartz [A.L. Boettcher, J. Petrol. 11, 337 (1970)]; (iv)grossular5,andradite5o = FeTs + 2 wollastonite(9). The position of the invariant point I (Fig. 1)has been taken as 600°C and Xco2 = 0.15. Thepositions of other equilibria have been calcu-lated from data in (10). The effect of other com-ponents has been estimated: for garnet, by as-suming an ideal solution model [J. Ganguly,Contrib. Mineral. Petrol. 55, 81 (1976)]; for fas-saite, by assuming that AlTs activity = XAITS,which is consistent with experimentally deter-mined activity-composition relations for Di-Hd-AlTs (Di is diopside, and Hd is hedenbergite)solid solutions with XAITJ < 0.3 [B. J. Wood,Am. Mineral. 61, 599 (1976)] and assuming thatFeTs activity = XFeTS. The activities of plagio-clase, wollastonite, and calcite have been ikenas unity.

12. B. M. Reinhardt, Schweiz, Mineral. Petrogr.Mitt. 49, 1(1969).

13. Complex inclusions ranging in size from a fewmicrons to 20 ;L are common in the pyroxenesas planar groupings traversing the host crystalbut do not continue into the adjacent wollaston-ite, and they are therefore classed as pseudo-secondary inclusions. Similar, but isolated, in-clusions occur and are interpreted as primary.The inclusions commonly contain more than 50percent crystalline material and a low-densityvapor phase which rarely has a rim of liquid.Some of the crystalline material appears to becalcite. Similar, though less common, inclusionsoccur in wollastonite. All these inclusions are in-terpreted as magmatic inclusions, an indication

that fassaite and wollastonite crystallized from amelt.

14. J. A. Willemse and J. J. Bensch, Trans. Geol.Soc. S. Afr. 67, 1(1964).

15. R. J. Shedlock and E. J. Essene, J. Petrol. 20,71 (1979).

16. R. G. Coleman, Bull. U.S. Geol. Surv. 1247(1967).

17. , Ophiolites: Ancient Oceanic Lithosphere(Springer Verlag, Berlin, 1977); H. Williamsand W. R. Smyth, Am. J. Sci. 273, 594 (1973);J. Zimmerman, Geol. Soc. Am. Mem. 132(1972), p. 225.

18. R. Joesten, Geol. Soc. Am. Bull. 88, 1515 (1977).19. N. L. Bowen, The Evolution of the Igneous

Rocks (Princeton Univ. Press, Princeton, N.J.,1928); D. H. Watkinson and P. J. Wyllie, Geol.Soc. Am. Bull. 80, 1565 (1969).

20. R. Stoneley, Tectonophysics 25, 303 (1975).21. I thank M. Arvin for the opportunity to visit

Neyriz and H. S. Edgell for help with fieldwork.I thank P. Suddaby, N. Wilkinson, and K.Brodie, who provided assistance with micro-probe analyses carried out in the Geology De-partment, Imperial College, London. I thank A.H. Rankin for examination and interpretation ofthe fluid inclusions and M. G. Audley-Charlesand W. J. French for discussions.

3 December 1979; revised 14 March 1980

Aragonite Twinning in the Molluscan Bivalve Hinge Ligament

Abstract. Molluscan bivalve hinge ligaments are composed of long needle-shapedaragonite crystals embedded in a protein matrix. These crystals are twinned and, ingeneral, the twin forms a thin lamella through the center of the crystal.

By electron microscopy and electrondiffraction of single crystals in thin sec-tions, we have demonstrated that thearagonite (CaCO3) crystals in the mollus-can bivalve hinge ligament are twinned.Aragonite is orthorhombic belonging tospace group Pmcn. Inorganic sources ofthe mineral are commonly observed tobe twinned about the (110) morphologicplane (1). Thin twin lamellae have beenobserved in electron micrographs ofaragonite crystals from limestone depos-its (2). Our experiments are apparentlythe first demonstration of twinning inbiogenic aragonite, although Mutvei (3)has proposed that, on the basis of crystalmorphology alone, the aragonite crystalsof bivalve shell nacre are twinned.The bivalve hinge ligament is com-

posed of long needle-shaped aragonitecrystals embedded in an elastic, pre-dominantly protein matrix (4). The crys-tals are pseudohexagonal in cross sec-tion (Fig. 1), and the morphologicallylong (crystallographic c) axis is orientedperpendicular to the growing margin ofthe ligament. The crystals are about 100nm in cross section. For electron micros-

copy, 0.5-mm slices of the ligament werefixed in 5 percent glutaraldehyde in 0.1Mcacodylate buffer, pH 7.6, for 4 hours,then dehydrated with ethanol and em-bedded in Epon. With a dry diamondknife, 100-nm sections were cut per-pendicular to the c axis of the crystals.The sections were transferred to a watersurface with a fine hair and picked up im-mediately on copper grids to prevent dis-solution of the crystals. Images and dif-fraction data were obtained with a Phil-ips 200 electron microscope. Imageswere recorded at 60 kV and diffractiondata at 100 kV.The electron micrographs of Mya are-

naria and Spisula solidissima (Fig. I)show hinge ligament cross sections. Theelectron-opaque structures are thearagonite crystals; the electron lucentarea is the organic matrix. Since the caxes of the individual crystals (which arenearly perpendicular to the plane of themicrographs) are not perfectly parallel,not all of the crystals in a section are ina reflecting orientation. Those crystalsthat are in a strongly reflecting orienta-tion have opaque images as a result of

Fig. 1. Electron micrographs of hinge ligaments showing the aragonite crystals in cross section.(A) The Mya arenaria ligament. The thin bands through the center of the crystals are thin twinlamellae. (B) The Spisula solidissima ligament. Crystal a has multiple twin lamellae and crystalb has twin lamellae on both the (110) and (110) planes. Crystal c has a thick twin lamella and thelower third of crystal a is in the twin orientation. Bar equals 100 nm.

0036-8075/80/0613-1262$00.50/0 Copyright X 1980 AAAS1262 SCIENCE, VOL. 208, 13 JUNE 1980

on Decem

ber 4, 2020

http://science.sciencemag.org/

Dow

nloaded from

Contact Metamorphism by an Ophiolite Peridotite from Neyriz, IranROBERT HALL

DOI: 10.1126/science.208.4449.1259 (4449), 1259-1262.208Science 

ARTICLE TOOLS http://science.sciencemag.org/content/208/4449/1259

REFERENCES

http://science.sciencemag.org/content/208/4449/1259#BIBLThis article cites 22 articles, 8 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

trademark of AAAS. is a registeredScienceAdvancement of Science, 1200 New York Avenue NW, Washington, DC 20005. The title

(print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for theScience

Science. No claim to original U.S. Government Works.Copyright © 1980 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of

on Decem

ber 4, 2020

http://science.sciencemag.org/

Dow

nloaded from