Document nm4QxdrB97N9E1jproaVevbN2

'RS. GARD, SP1LIKMAX AND TAYLOR " cefitedfor publication, September l/, 1065. THE AMERICAN MINERALOGIST, VOL. St, JANUARY-FEBRUARY. l6u BRUCITE IN ALPINE SERPENTINITES P. B. Hostktleb, R. G. Coleman, U. S. Geological Survey, Menlo Park, Calforuia, F. A. Mumptox, Nuclear Division, Union Carbide Corporation, Tuxedo, New York, AND B. W. Evans, Department of Geology and Geophysics, University of California, Berkeley, California. Abstract ip~s ,Y-ray diffraction tracings of 102 serpentinizeri and partially scqieutinized Alpinc-typc ultramafics from the circura-Pacific orogenic belts show that brucite commonly forms durf*r;' Jng pervasive serpentinization of dunites and high-olivine peridotites. Characteristically. |\1 such brucite is fine-grained and intimately intergrown with lizardite. Micrnprobe analyses indicate that the lizardites contain less iron, and the brucites more iron, than the parent 1% w olivines and pyroxenes, although the brucites show broad variations in iron content. Extensive brucite formation during serpentinization indicates maximum temperatures i .'of about 400" C. and. unless massive amounts of bol h magnesia and silica are removed from the ultramafic mass, volume increases of about 40 per cent. Field evidence supporting such migration is generally lacking. The ellipsoidal, fault-bounded configurations typical of Alpine ultramafics suggest that large volume changes do not result from complete serpen- tinization in situ; rather, serpentinization probably results front incremental additions of w"' water derived from country rocks during the tectonic ascent of the ultramafic mass through ^ the earth's ccust. E " Reaction of brucite with COj-bearing groundwater at depth is probably responsible for much of the magnesite associated with serpentinites. In a surface oxidation environment, '0 - iron-bearing brucite is converted directly to pvroaurite or coalingite, or reacts with per ''.'W colating groundwater to form h-vdrom_ag-nesite and artinite. * Introduction 13"" Ar-rav diffraction studies by the authors of a large number of serpentinized and partial)}- serpentinized dunites and peridotites have revealed ' that pervasive serpentinization commonly produces significant quantities of brucite, in addition to serpentine-group minerals1 and magnetite. In ' this paper sonie typical brucile-';serpentine" assemblages are described is: in detail and. corollary implications of serpentinization are discussed concerning: So. (1) the transfer of magnesia and silica, (21 volume increase, and (3) maximum temperalures. s: Ij. pfl fy H. 1 Usage of the terms serpentinite, "serpentine", and serpentine-group minerals follows that given by I'aust and Fahey (1962). Where thephases in a specimen have been identified, they are named, as dinochn-sotile, lizardite, or antigorite. Where the phases were not identified, or ate referred to collectively, they are called serpentine-group minerals. Appropriate rocks are called serpentinites. There still remain occasions in which it is desirable to use the term "serpentine" (in quotes), but only where the meaning cannot be misconstrued. 75 PLAINTIFF'S EXHIBIT SA-5I3 76 HOSTETLER, COLEMAN, MUMPTON AND EVANS In addition, the conversion of brucite to hydrous magnesium carbonates during the oxidation and weathering of serpenlinite bodies is discussed. As long as serpentine-group minerals of nearly constant composition, together with accessory magnetite, are regarded as the sole products of serpentinization of Alpine ultramafics1 which vary significantly in their original proportions of magnesia and silica, serpentinization must imply substantial changes to the initial proportions of magnesia and silica. Serpentinization is also commonly regarded as a process occurring at constant volume. According to Turner and Ycrhoogen (1960, p. 318), "microscopic fabric and field relations of undeformed serpentinites show . . . that serpentinization is commonly accompanied by little or no increase in volume." Because of the difference in densities between ser pentine and the primary anhydrous silicates, such a concept can be ac counted for only by a substantial subtraction of material. In their pioneering experimental study of the MgO-SiO-HaO system, Bowen and Tuttle (1949) showed that the maximum stability limit of chrysotile serpentine is very close to 500 C. The specific reaction studied was that between clinochrvsotile and forsterite, talc, and water vapor. For talc, however, one may substitute enstatite, quartz, or a. vapor phase preferentially enriched in silica and thereby modify somewhat the stabil ity limit of chrysotile serpentine. This reaction, in what we might call its "expanded form", can thus account for the exclusive formation of the serpentine-group minerals from a variety of starting materials, and the reaction may be accompanied by a variety of assumed volume changes including no volume change at all. It is not surprising then, that such a reaction has been commonly adapted by petrologists who have also accepted a temperature of 500 C. as characteristic of the reaction (the pressure-temperature curve shown by Bowen and Tuttle is very nearly independent of pressure, passing through 500 C. at. 2000 psi water vapor pressure, and about 510 C. at 40,000 psi). Bowen and Tuttle (1949) also showed that the minimum temperature for the breakdown of forsterite to yield clinochrysotile plus brucite in the presence of 15,000 psi water vapor pressure is 365 C. This reaction, however, has been thought to be of little petrological significance because brucite has not been generally recognized to be a product of serpentini zation. Bowen and Tuttle themselves minimized its importance, for they stated (p. 452): "Olivine does appear to suffer a change in rocks whereby only serpentine remains in the rock." The formation of brucite during serpentinization implies both a sizeable drop in the maximum tempera- 1 The term Alpine ultramafics is used herein the sense propounded by Benson (1926) and Thayer (I960), that is, those ultramafics occurring astride the axes of mobile orogcnic zones. peuti' earliest) gcriptioi large g! Fcnogli cerned1 more e pvroaui {nemal 1945), - Koski, brucit cordin contai. pentini (1961) Shtelnf serpen mode tinite Th that petroi (Hah fc'.l/.LV, MUUPTON AND FXANS f brutite to hydrous magnesium carbonates it hermg of serpentinite bodies is discussed, p minerals of nearly constant composition, tetite, are regarded as the sole products of '.ramafics1 which vary significantly in their :sia and silica, serpentinization must imply nitial proportions of magnesia and silica, uonly regarded as a process occurring at to Turner and Yerhoogen (I960, p. 318), d relations of undeformed serpentinites n is commonly accompanied by little or no of the difference in densities between ser,-drous silicates, such a concept can. be aeitial subtraction of material. . cntal study of the MgO-SiOrHsO system, wed that the maximum stability limit of jse l.o 500 C. The specific reaction studied tile and forsterite, talc, and water vapor, istitute enstatite, ciuarta, or a vapor phase and thereby modify somewhat the stabilne. This reaction, in what we might call account for the exclusive formation of the > a variety of starting materials, and the by a variety of assumed volume changes all. It is not surprising then, that such a adapted by petrologists who have also 1 C. as characteristic of the reaction (the mm by Bowen and Tuttle is very nearly r, through 500 C. at 2000 psi water vapor 40,000 psi). showed that the minimum temperature 0 yield clinochrysotile plus brucite in the apor pressure is 365 C. This reaction. of little petrological'significance because ecognized to be a product of serpentinielves minimized its importance, for they pear to suffer a change in rocks whereby rock." The formation of brucite during uzeable drop in the maximum tempera- 1 here in the sense propounded by Benson (1926) afics occurring astride the axes of mobile orogenic BRUCITE IN SERPENTINITES 77 liure of serpentinization and a larger volume increase than would result Ilf excess magnesia had been removed in solution from the area of ser- :|;jjentinizatiom. ;f We have found that brucite not only is a common constituent of ser:!'pentinites but that it accompanies serpentine-group minerals at the Pearliest stages of alteration. Actually, there have been a number of de I fferiptions in the literature of brucite occurrences in serpentinites. One /large group of descriptions (Ferrari el al. 1950; Meixner, 1938, 1956; y Fenoglio, 1930, 1933; Allakhverdiev, 1958; Serdyuckenko, 1949) is conIpcerned with the mineralogiral occurrence of brucite with some of the rYipore exotic magnesium hvdroxycarbonate minerals (hydromagnesite, ;i;r pvroaurite, artinite) in serpentinites. The association of fibrous brucite :L: (nemalite) with chrysotile fibers has been reported from Alaska (Fackler, 1945), Switzerland (Eckhardt, 19S6), and Quebec (Allen, Gill, and Koski, 1957). Van Biljon (I960) describes thin veinletsand grains of if. brucite in a massive serpentinite sample from Swaziland, S. Africa. Ac>; cording to Sakomato (1959) the serpentinites around Kochi City, Japan, ;;; contain brucite in an intimate mixture with "serpentine." Massive serpentinite at Leopoldsgriin, W. Germany, (Hahn-Weinheimer and Rost, A. (1961) contains numerous small veinlets of brucite and magnetite, f Shteinberg (i960) has shown that brucite is an important constituent in serpentinized dunites in the Urals. According to Hess and Otalora (1964) modest percentages of brucite are scattered through much of the serpen ; tinite taken from the AMSOC core hole near Mayaguez, Puerto Rico. The cumulative weight of these references supports our observation ih that brucite is not uncommon in serpentinites, but their cumulative petrologic importance is diminished by the fact that in only two cases / (Hahn-Weinheimer and Rost, 1961; and Shteinberg, 1960) has any at tempt been made to relate genetically the formation of brucite with the. overall process of serpentinization. Hahn-Wienheimer and Rost (1961) do not believe that the brucite at Leopoldsgriin was formed during an initial, pervasive serpentinization stage, but rather they believe it was formed during a later stage of metamorphism and antigorit.ization. In our opinion the formation of fibrous brucite or brucite associated with ;. magnesium hvdroxycarbonate minerals is probably not related to initial : serpentinization processes, but rather, is the result of reprecipitation or later reworking of the initial serpentinite. Normative calculations from chemical analyses are an indication of the possible presence of brucite in serpentinites. Shteinberg (1960) pointed out that magnesium carbonates were not present in significant amounts in most of his samples, and therefore the excess magnesia, for rocks whose initial MgO/SiO; ratio was greater than 3/2, must be pres- 78 HOSTETLER, COLIMA If, MVMPTON AND fin .VS ent as brucite. Detailed sc-ray, optical, and thermogravimetric analysis (Shteinberg, 1960) of a serpentinized duniie, two partly serpentinized dunites, and a partly serpentinized peridotite confirmed the presence of significant brsudte in all but the peridotite. Thus it seems that serpentinites often contain brucite, but to our knowledge Shtemberg (1960) was the first to associate brucite formation with the process of serpentinization. Occurrences of brucite in serpentinite undoubtedly have been overlooked in the past, indicating that iden tification of such brucite in thin-section is difficult. Geologic Occurrences The bulk of this investigation rests on the evidence of .v-ray diffraction tracings of 102 serpentinite samples from a number of Alpine ullramafic localities in the circum-Pacific. orogenic belts. All the samples were col lected by the authors, although none was collected with the specific idea in mind of determining the distribution of brucite-bearing serpentinites. Eight serpentinite samples were collected over an 80-mile interval of the Great Serpentine Belt of New South Wales, as described by Benson (1913-1918). These samples were selected to represent the most intense areas of shearing and serpentinization. Eight more samples were taken from several of the isolated, partially serpentinized, periodotite masses in Mew Caledonia. The ten selected Mew Zealand samples are part of a larger study by Coleman on petrologic relationships in the Great Ultramafic Belt of New Zealand. Eighty of the samples were collected at numerous localities in Washington, Oregon, and California and are part of a continuing study on problems associated with serpentinization within the Pacific Coast ultramafic belts. Preliminary mineralogical determinations were made on the basis of a-rav diffraction charts. Most of the samples (eighty-seven) were found to be completely serpentinized and generally contain a mixture of the polymorphs lizardite and clinochrysotile. Brucite was detected in thirtyone samples, and, as would be expected, the proportion of brucite to serpentine-group minerals varied widely, but in no case was a sample found in which the molar percentage of brucite exceeded that of the "serpentine." The fifteen samples of partially serpentinized ultramafic rocks contained olivine, or olivine plus orthopyroxene. Brucite was found in ten of these samples. The initial olivine/orthopyroxene ratio obviously controlled the brucite formation in these samples, for it was greater than 1/1 (mol) in each of the ten samples containing brucite, but near 1/1, or less, in three of the five samples devoid of brucite. The other- two con tained only olivine and serpentine-group minerals, but were taken from Mine*. mme| tionsl graYf f'"' I T.E3TAX, MUMPTOX AXD /-T.t.VS BRUCITE IX .SERFEXTISITES ay, optical, arid thermogravimetric analysis penlinized dunite. two partly- serpentinized u'nized peridotite confirmed the presence of the peridotite. 'ntinites often contain brucile, hut to our ) was the first to associate brucite formation lization. Occurrences of brucite in serpentioverlooked in the past, indicating thatidenlin-section is difficult. OGIC OCCTOEKNCES ion rests on the evidence of x-ray diffraction ainples from a number of Alpine ultraniafic ic orogenic belts. All the samples were colgh none was collected with the specific idea istribution of brucite-bearing serpenlinites. were collected over an 80-mile interval of New South Wales, as described by Benson were selected to represent the most intense Linizatkm. Eight more samples were taken artially serpentinized, periodotite masses in ;cted New Zealand samples are part of a >etrologic relationships in the Great Ultra Eighty of the samples were collected at igton, Oregon, and California and are part, roblems associated with serpentinizatiou mafic belts. ieterminations were made on the basis of of the samples (eighty-seven) were found d and generally contain a mixture of the >chrysotile. Brucite was detected in thirty>e expected, the proportion of brucite to ied widely, but in no case was a sample rcenlage of brucite.exceeded that of the pies of partially serpentinized ultramafic ine plus orthopyroxene. Brucile was found tial olivine/orthopyroxene ratio obviously n in these samples, for it was greater than raples containing brucite, hut near 1/1, or es devoid of brucite. The other- two con iine-group minerals, but were taken from /weathering surfaces where, as will be discussed subsequently, brucite is demonstrably unstable. Thus the data from partially serpentinized /dunites and-pyroxene-poor periodotiles implies that formation of brucite ;-5s to be expected during initial serpentinizat ion of these rocks. .'Mineralogy and Petrology. . 1) Petrography. Detailed study to determine the characteristics and : habitat of the brucite was restricted to five samples selected from the original one hundred and two (Table 1). These represent dunites in vari ous stages of alteration to serpentine. In three of the samples (Dun Min., .Twin Sisters, and Burro Mtn.) more than 40 volume per cent of the .sample still consists of olivine and, in the case of the Burro Mountain sample, minor pyroxene. On the basis of any rock classification, the rocks represented by these samples would be considered dunites; in fact, the Dun Mountain material is representative of the type dunite described by : Hochstetler (1859). The remaining two samples are presumed originally to have been dunites also, but now consist primarily of serpentine-group minerals. In hand specimen, unweathered surfaces of serpentinized por tions of these ultramafics are black, whereas unaltered dunite is a light r gray Microscopic examination of these samples reveals textures character istically encountered in partially serpentinized ultramafic rocks. The olivines are in an advanced stage of replacement by serpentine-group ; minerals with individual grains separated one from the other by "serpen tine'-' mantles and veinlets. Fracture patterns and crystal boundaries appear to have controlled the initial alteration of the olivine, and these boundaries are ofLen preserved by trains of fine-grained, secondary mag netite. Progressive replacement of olivine grains from the rim inwards < contributes to the typical mesh structure characterizing serpentinized . dunites. Along these same original crystal boundaries brucite is com monly present as fine-grained veinlets. In other instances, particularly in ; the Dun Mountain and Burro Mountain specimens, the brucite may form a central core within a completely serpentinized olivine grain, or as a rim around those grains not yet completely altered. More typically, brucite \. occurs as an extremely fine-grained intergrowl.h with ''serpentine.'' Optically, brucite may be distinguished from "serpentine" if individual grains exceed 0.05 nun due to its higher birefringence and indices of re fraction. In the case of the New Idria material, brucite has formed large rectangular plates (up to 0.5 mm square) that are more easily recognized in thin section. The coarse-grained brucite, however, may be confused with chlorite, because, like chlorite, the brucite has anomalous reddish-  80 HOSTETLER. COLEMAN. Ml'.UPTON AND EVANS brown interference colors resulting from dispersion, and because the uniaxial post live character of the brucite could easily be mistaken for a chlorite with,a small 2V. In order to characterize more fully the brucite and coexisting serpen tine-group minerals by optical, x-ray and chemical techniques, each of the live samples was crushed and sized to obtain a 200-325 mesh fraction suitable for centrifuging in heavy liquids. The density of brucite is given as 2.39 g/cc (Deer, et al. 1962, v. 5, p. 89) and Iizardite and clinochrvsotile as 2.55 (Deer el al. v. 3, p. 170); therefore liquids with densities of 2.40 2.50 g/cc were used to attempt a separation of brucite from serpentinegroup minerals. This was not successful because the serpentine-group minerals and brucite were found to have overlapping densities, due prob ably to substitution of iron in the brucite. The brucite-``serpentine'' mixtures were carefully purified so that the primary silicates and spinels were completely removed. These purified separates were used to obtain the x-ray and chemical information given in this paper. The refractive index, e, for each brucite sample was measured as 1.580 + .004, except for the Burro Mtn. sample (1.585 .004). 2) X-ray. Y-ray diffraction studies were made using a Norelco High Angle Diffractometer and Xi-filtered CuX radiation. Powdered sam ples were sedimented on glass slides and x-raved from 2 to 6820 at a scanning rate of l.'minute. The x-ray diffraction data for the brucite in the five separates described above was sensibly identical to the AST.M standard brucite (card 7- 239) for the spacings (001), (101) and (102); d (110) ranged from 1.580 to 1.583 as compared with 1.573 for standard brucite. Figure 1 is an overlay comparison of six diffraction charts of; (1) pure hrudtc, (2) whole rock from the Dun Mountain sample. (3) `'serpentine"-brucite fraction of the Dun Mountain sample, (4) Iizardite plus minor clinochrvsotile, (5) clinochrysotile, and (6) antigoritc. The whole rock from Dun Mountain (Fig. 1, line 2) contains approxi mately 20 mol per cent (5 weight per cent.) brucite, and the (001), (101). (102) and (110) reflections of brucite are all plainly visible on the dif fractometer chart. If the molar percentage of brucite in a whole rock sample is approximately 10 per cent or less, the (001) reflection at 18.6 20 may be the only observed peak. Density separation of a "serpentine"-brucite fraction from the primary silicates will yield a positive identification for brucite by eliminating interfering olivine peaks and by increasing the relative amount of brucite in the sample. The normally difficult x-ray diffraction identification of specific ser- pentine-g olivine, s| fizardite | pie, the i| and 36.8*1 reflect ion| sample, if brucite-1'! ... line 3), if indicated! ferentiatii visible inf nificant but nonef The - minor cli| - in wliicEf olivine, In most. 3) CI4 Micropr SiiAvSi . olivines, identity | of the " tine" m< Fe/Mg-j . in olivim the parti An A-l neous di Iizardite; cite, oliij chrysotil analyses! microprc (Smith analysis brous h; electric gates ari brucite I 1 )LFAIAX, UCMPTOX AXO AT.4.YS BRUCITE IX SERPF.XTIX1TFS resulting from dispersion, and because the of the brucite could easily be mistaken for a nore fully the brucite and coexisting sevpen. ical, x-ray and chemical techniques, each of 1 and sized to obtain a 200-325 mesh fraction icavy liquids. The density of brucite is given 2, v. 5, p. 80) and lizardite and clinochrysotile 70); therefore liquids with densities of 2.-Ut ilpt a separation of brucite from serpent ineiot successful because the serpentine-group und to have overlapping densities, due probri in the brucite. The brucite-"serpentine" fied so that the primary silicates and spinels 'hese purified separates were used to obtain rmation given in this paper. The refractive tuple was measured as 1.580+ .004, except (1.585+ .004). m studies were made using a Norelco High ,'i-filtered CuKa radiation. Powdered samss slides and .r-rayed from 2 to 6826 at a The x-ray diffraction data for the brucite in above was sensibly identical to the ASTM 9) for the spacings (001), (101) and (102); 1.583 as compared with 1.573 for standard parison of six diffraction charts of; i the Dun Mountain sample, <3) "serpentine"-brucite fie, (4) lizardite plus minor clinochrysotile, (5) clino- fountain (Fig. 1, line 2) contains approxi ght per cent) brucite, and the (001), (101), f brucite are all plainly visible on the diflar percentage of brucite in a whole rock per cent or less, the (001) reflection at ved peak. Density separation of a "serpenhe primary silicates will yield a positive liminating interfering olivine peaks and by t of brucite in the sample, v diffraction identification of specific scr- pentine-group minerals is made even more complex by the presence of olivine, since strong reflections of olivine overlap the critical lines of lizardite and chrysotile. As seen in the whole rock Dun Mountain sam ple, the indicative (202) reflections of lizardite and chrysotile at 36.0 and 36.8 26, respectively, are masked by the strong (131) and (112) reflections of olivine. When the olivine fraction is removed from the I sample identification is made much easier, and, from the pattern of the I)rucite-"'serpentine" fraction of the Dun Mountain, sample (Fig. 1, fine 3), the presence of lizardite and a small quantity of chrysotile is indicated. The position of the (204) reflections is also diagnostic in dif ferentiating lizardite and chrysotile, but often these reflections are not :m yisible in the diffractometer tracings. It may be possible to delect sig F'- nificant amounts of anligorite without separation of an olivine fraction, ti*. but none of our samples contained antigorile in association with olivine. The Dun Mountain assemblage of olivine, brucite, lizardite, and minor clinochrysotile is a typical one; in fact, for each of the ten samples in which brucite and "serpentine" were found associated with primary olivine, lizardite was noted as the dominant serpentine-group mineral. S'-' -In most cases, minor clinochrysotile was also observed. 3) Chemical Composition. Two analytical methods were employed. Microprobe analyses were used to determine Fe/Mg+Fe ratios for olivines, brueites, and serpentine-group minerals, and to establish the identity of accessory minerals. Chemical analysis of an acetic acid leach of the "serpentine"-brucite density fraction yielded brucite '"serpen tine" molecular ratios and Fe/Mg+Fe ratios for brueites. The molar Fe/Mg+Fe ratios thus represent the molecular percentage of fayalite in olivine, Fe(OH)2 in brucite, and "Fe-serpeiitine" (FevSi205(0H) i) in the particular serpentine-group mineral. An A.R.L. electron probe microanalyzer was used to make simulta neous determinations of Mg, Fe, and Si on individual grains of olivine, lizardite, and brucite in polished rock sections (Table 1). Pure Mg-brucitc, olivine (Fan.,), hypersthene, and an analysed lizardite and clino chrysotile (U. S. Geological Survey, Menlo Park, Calif.,' unpublished analyses) were used for standardization. Analysis of olivines with the microprobe has already been shown to be simple, precise, and accurate (Smith and Slenstrom, 1965; Keil and Fredriksson, 1964). Microprobe analysis of serpentine-group minerals is more difficult because their fi brous habit makes a good surface polish more difficult to obtain, because electrical and thermal conductivities for fine-grained crystalline aggre gates are lower than for the single crystals, and because intergrowths of brucite interfere. Although analyzed lizardite and clinochrysotile sam- 82 HOSTETLER, COLEMAN, MUSIETON AND EVANS 1 Brucite Whole rock, Dun Mountain "'Serpentine"brucite fraction. Dun Mountain 4 Lizardite plus minor clinochrysotile Clinochrysotile pics were u ' wediary du| $te made, f' , *jf serpenfi: ^,tio meth Jjrucite lose: fe(OH); in | `"constant a amounts of accounted .1 In pure lizan ' subtract! analyses w` The oliv: rock; all fel tjltramafic I be regarded| K | mol per rified, non (IT. S. Geol| Indicate an; any one roc? data also sj per cent fa| lizardite. ' f Whereas! olivine, the lizardite, t| widely witl the five sau from 10 to tit Pi 1 .i- 1 50 40 -- T. 20 Degrees 2 9 Antigorite JWjt.Vll V" g|:!: Fig. 1. Di 11:1:,: traction from Lizardite ]. J*Kew Idria. C ^|?:"B-brucite. C gil-lminerals. TMAX, MCMPTOX AND EVANS l Brucite Whole rock, '-4-- Dun Mountain ''Serpentine"brucite fraction,, 4-------- c__ Dun Mountain J4 Lizard its plus minor clinochrysotile -MJ Clinochrysotile ; BRUCITE IN SERPENTINITES )ingles '"'ere used as standards, an olivine reference was used as an inter- j^^Hjdiavy during routine work. Provided that mass absorption corrections ffeaf* made, forsterite olivine serves as a reliable standard for the analysis ttLrnf serpentine-group minerals. Analysis of brucite is possible only if a method is used, for even under the gentlest excitation conditions ^'"' brucite loses water during electron probe analysis. The percentage of the y j?e(OH)-> molecule was determined from the Fe/Mg ratio which remains Ir"'constant and is independent of the extent of volatilization. Small jb" founts of lizardite in the analyzed volume of brucite can be accurately T'. recounted for by using the ratio of Si-counts in "brucite" to Si-counts pure lizardite. Since the brucite is not homogeneous, it is not possible Isubtract brucite from a lizardite analysis, and only those lizardite i."Vpialysf:s which give the correct maximum Si-counts were considered. The olivines were found to be very uniform in composition within one jfpck; all fell within the range already described for olivines from Alpine (jltrainalic bodies (Green, 1964). The serpentine-group minerals cannot be regarded as virtually iron-free. Typically the lizardiles contain about |F~ A mol per cent ;`Fe-serpentine". Conventional chemical analysis of pu- fcj.j'ified, non-brucile bearing lizardite and dinochrysolile from New Idria (0. S. Geological Survey, Menlo Park, Calif., unpublished analyses) also indicate an `T'c-serpentine" component of about 4 mol per cent. Within gnv one rock there is a limited range in composition (Table 1). The probe |C data also suggest that there may exist a positive correlation between ' per cent fayalite in the olivine and per cent "Fe-serpentine" in the ibliizardite. Whereas lizardites possess lower .Fe/Mg+Fe ratios than the parent olivine, the coexisting brucite is richer in iron. In contrast' to olivine and ^ fizardite, the Fe/Mg-f-Fe ratio of different grains of brucite varies b(:: widely within any one rock sample, but the range of ratios for each of |r;' the five samples is similar. The mol per cent of Fe(OH) in brucite varies L; from 10 to 20 per cent and averages around 15 per cent. Concerning ac- Ip #4 - r~ JO 2e 10 Antigorite Fic.. 1. Diffraction tracings of: 1) Pure brucite from Liming. Nevada. 2) Whole rock In fraction from Dun Mountain, 3) "Scrpentine"-brucile fraction from Dun Mountain, |l;4) Lizardite plus minor clinochrysotile from New fdria, California, 5) Clinochrysotile from Il'New Idria, California, 6) Antigorite from Milford Sound, New Zealand. Symbols as follows: Si-B-brucite, C-dinochrysotile, L-lizardite. O-olivine, S-undifferentiated serpentine-group ~;tninerals. I 84 HOSTETLER, COLEMAN, MUUPTON AND EVANS a Sa I i J gje 2 = '<C S r* ST 3 ? Jj sc ^ Is i S.S " Ss * 5 -"a`C r* -- *-* fc^(2 3 fc.3 3 55 a?1 | u S S1I 3^ W 3 3- C5 '! I o'B| t - * j*3; = *3 o M 5 rt 3 Nt 11 1 I c :i 2ss^ |Ae-- < I-- usn',?; i| oI 'f.i*.52 ' =Ly' f=ty*grS-. . Kac~ ,i <> Or <>&3.--- wS *oS , --i-}* i , ~^ ' lit |p' til 1<>aS3'-~> <>cs2 C> fStl'C? IS C3 3S X 2*1 --3 o`5 _S - 3SCic = s*t5 3S ^g*5= s&g 2c te*T3 li c 1 tegs "r "f 5 I * -e m I 0 v g y v- 3 2. c b - a "_. =S .^2.=y* <3 =ei 3, : S $ -~N --' >3 3 n 5 t- '"4 ,| ji I a J O C =C S- ; - ^ ai ,t i; 3I ,5- ?ii i*Ti Ii> 1 42=')fns4>.i*sa3 i,--6..2c*-c>->..^sci?" . i i -:: ^g-3 tfg-g > i = M 5o;1 iHiMjJf =5 st; lqtiJlS 4-g l i ll 1 i j 1 I l S Si - stg.-'g .-gS3CeJ'?l? o a,a =+c_ | 1 2^| * 1- H sr.S OJj t si^5 33 8 303 3& *WgT2, *-3*S 'Jy 'z, 2 jq?r*S t . d-5 3 * e^SiS ^ JSc2 s? 55 ,/S* 1 " * C "s1Z*? *rt*t3 <n 5 *5 ^5 aa= * S 2 2 ! i r? I T& cc 5 ~ y 4 "ISo / 35 3 f.s 2 * "i i' + i| 1" 82 3=5 S 2 *!'S=c . .S72sa * E ff i-i.Sl3=" es 2-^^ =-IsS g* alii 5* 4 3 X"i^s"1 " E*TaC 3C-O33C -- i '; J< I i |x -! Ip.lili " 5 ss ;'|S 8 8 55-91 s g r tx rt< e -f < , Cc//jaC < JIB 4*44 frc.tfSitjcv*: a j go: '2 !3 l2 1 S. & u c S ' * 3 Ow C^' 'eS-fc'5 Ml jS S-Jgri |r.d *,*5S*3W3 k^'S* C3 ^yTIos" 3 -- 2*?s^S^f*S2 $1*% I.52 MgS .Jja:s!i2sr=5.< = z e- S i3 % cessory mmj ' (Table 1), | , ' gpecimen, a| % Brucite ';. mined by le|, J a solution, c * water. The , ""' at room r.en| _ "serpentine ' ' (approxima was analyze j ' Idria "serpi }: been removi - also run to < ,, ' being leach ; \;, Results indl k .a lower, but; | Jeachate by ^Itetwcen th ; bearing Xe\| The aridf y-'-- amounts of] TJ- obtained b| t'T the probe aj 5'TM^ soluble mag y" tine-group i; | cr;-samples is | '^ - mined with] for the Twii |;y- samples ran - \ the ratio is i ^ ot both brui ;j; the brucite I ,-jr brucite crvs| %' ing sample ;.y. ing clinochr] Serpentinizi cerning the Jrr-MProy AXD EVAXS <f s3" 3 iar 2r/ A. . --tf *5 .A S 8 a *2 g .= w- 5 j** S 3 i,'*s--113C - a 4, ,I SP5|wf-e'i1l ,. <S^Sp755/.fee M2r=~^ 5*5 aj "if.2 Cl = si | idr-'s9 ; ? ti np v:*"* y3.S I1 . E*1-" a ' e>'c'S ! k~. g ' I1 ?sSc gu,.2S<O ii Sic53. .aSC |' OefjCj.t*s.25c 2BrSoUw 'l^PSSS |S 'H 4? 1c? -- 0 Ii> O g 3 J ^ 2 I ^.s Z h`J 1 <120.,23~ Iff S3.Sfc*e*2 'c 1 ^SSi r*, Ea" *s ; S0 I r!s vSi ai^: 0a3 to 3 -0*3 *:s-5{ d.i;; . a<r. 8u zL- _!l |3S| Sdsl .s' k I Es 1* -+3 V~ i ^ I fi S 5 - tss I "-2 l.d! ^ 3 =J 'c CTJ 3^3 3 ! 11*| ! --3rO=CJ5... s e IP* ---- ?:!?; aSe; J ijl . rvs: ^oSwS"!-.5 S'CVSC- I5jl;.fS!- 1 lig 5S^S= e83^^ Jv hS'E. 5s? 111 ifsfir S~2?: .5$ ^i2wx 30 ! iI J g I g I w<u5 .* l-f~-t_i*Ss?oSn , -CRw.SiJ_^pi jjjj. (* vI: II & ! s'- " I .if ifsfeid gg2-3.a>- I lfl -2S.^^5c i|3 ." tS?ec 1C *1H*1^*25*a*^*& 2* BRLXITE XV SERPEXTIXITES 85 l,..,.^jgssory minerals, awaruite (FeXLi) was detected in three of the rocks j ^Xable 1), iiiagnelite in all five, chromite in all save the New Idria specimen. arfrl heazlewooditc (NLA) in the New Tdria specimen only. .; .. . Urucite. "serpentine'' ratios for the five selected samples were deterri'- ^jned by leaching an approximately 500 mg portion of the sample with ^ solution composed of one part glacial acetic acid and four parts distilled i - water. The samples were weighed before and after leaching for 3 hours vV 3j room temperature on a Fisher oscillator. Diffraction diagrams of the r, 7,'."serpentine" residue after leaching showed no brucite. The leachate ' " {approximately 50 ml), considered to represent the dissolved brucite, l, ' was analyzed for magnesium and total iron. Portions of both the New ^ ..Jdria i!serpentine"-brucile fraction from which the magnetite had not been removed, and a "blank" of non-brucite-bearing dinoehrysotiie were |r also run to see whether significant amounts of magnesium and iron were , being leached from clinochrvsotile and minor residual magnetite. The results indicated that perhaps 5 per cent of the soluble magnesium and 3lower, but variable, percentage of soluble iron were contributed to the F leachate by clinochrvsotile. No difference in iron content was detected between the leachates from the magnetite-bearing and non-inagnetite^>s bearing New Idria brucite-``serpentine" fractions. The acid leaching results are also shown in Table 1. Although the amounts of Fe(OH)> in the brucites agree moderately well with those 7 obtained by microprobe analysis, the results are considered to support 1< the probe analyses only in a general way because of the variations in ^7,; soluble magnesium and iron contributed to the leachate from serpen17'' tine-group minerals. The average molar Fe(OH)* content for the five samples is If per cent: as compared with the 15 per cent figure detersi.'!;; mined with the electron probe. The molar brucite/"serpentine" ratios 17 ' for the Twin Sisters, Dun Mountain, Hat Island, and Burro Mountain 7"'- samples range between 0.4 and 0.6, whereas for the New Idria sample the ratio is 0.9. There is some evidence of reworking and reprecipiration of both brucite and "serpentine" for the New Idria sample in that: (1) b; the brucite crystals are about an order of magnitude larger than the TV' brucite crystals occurring in the other samples, and (2) the brucite-bear, mg sample was found adjacent to a band of splintery, non-brucite-bearing clinochrvsotile. Discussiox ||7 Serpenlinization -Temperature Control. Experimental information con cerning the pressure-temperature conditions for the univariant reaction: : fnrstcrite -t H,0 =; dinoehrysotiie -V brucite III 86 HOSTETLER, CO LEM A X, MCMPTOX AXD EVAXS is given by Bowen and Tuttle (1949) and Toiler (1952). in neither of these, studies was the curve fixed reversibly. Clinochrysotile and brucite are more stable than forsterite and water at least to 365 C. at 15,000 psi water vapoi*pressure according to Bowen and Tuttle, and to 430 0. at 15,000 psi according to Yoder. Olsen's (1963) calculated curve passes through 390 C. at 15,000 psi, with an uncertainty between 60 anil 100 C. Olivines in Alpine-type ultrainatics contain approximately 10 mol per cent fayalite (Olsen, 1963; Green, 1964). Bowen and Tuttle's (1949) estimate of an 80 C. lowering of olivine stability as a result of this amount of solid solution cannot be accepted since oxygen partial pressure was not controlled, and magnesioferrite (MgFeaOs), rather than mag netite. was produced. Olsen (1963), assuming that clinochrysotile and brucite are iron-free phases, calculated a 20-30 C. reduction in tem perature for the reaction: olivine f Faio) d H-0 f- O* rs clinochrysotile + brucite -f magnetite dl) as compared to (T). Combining this result with the experimental data, it seems likely that the stability limit of olivine of FosoFau composition is around 400 C. in the presence of 15,000 psi water vapor pressure. The stability limit of pure Mg-clinochrysotile, as given by the reac tion: ctinochrysorile ~ forsterite + talc + H;0 (HI) was shown by Bowen and Tuttle (1949) to be no higher than 505 C. at 15,000 psi. If Olsen's calculations on the influence of fayalite solid solu tion are again accepted, the equilibrium temperature for the reaction: clinochrysotile + magnetite == olivine <TaJ(i) + talc + H-.O + 02 IV) is no higher than 485 C. at 15,000 psi. The widespread occurrence of brucite in Alpine serpentinites implies that pressure-temperalure condi tions dturing serpenlinizat ion are commonly those characterizing reaction (II) rather than reaction (IV). The temperature difference between the two reactions may be assumed provisionally to be about 80 C., but it should be remembered that neither reaction considers the stability of lizardite or antigorite rather than clinochrysotile, or the influence of iron substitution in brucite and the serpentine-group minerals. Furthermore, talc is a relatively uncommon phase in serpentinized Alpine ult.ramatic rocks; olivine and pyroxene generally alter directly to serpentine min erals. Therefore, the upper stability limit for "serpentine" can be de scribed by a metastable reaction of the type: "serpentine'' ;=: olivine -f- enstatile { K;0 The inva since we js more water vaj Recogf of restrii tlnizatio^ have lakf pressur yield awi Idria, c| -silicate In thesej thomson| 360 C. kcratoph man, 19| the pods fempora| , generally Serpentii . to the ci - hoogeil, beremof .U ' : RemovE idunites .. : volume "V The adi| * ' (equatii (equatiper cenl ) VAN, MLWfPTON AND /-VANS BRVCITE IN SERPENTINITES S7 ; (1949) and Yoder (1952). In. neither 0{ ed reversibly. Clinochrysolile and brudte and water at least to 365 C. at 15,000 ding to Bowen and Tuitle, and to 430" c. er. Olsen's (1963) calculated curve passes . with an uncertain!}' between 60 and tmafics contain approximately 10 mol per reen, 1964). Bowen and Tuttle's (1919) g of olivine stability as a result of this be accepted since oxygen partial pressure esioferrite (MgFegOi), rather than mag1963), assuming that clinochrysotile and dculated a 20~30 C. reduction in tem- : * dinociirysotile -f hrucite -1- magnetite ai) g this result with the experimental data, y limit of olivine of FojuFain composition ice of 15,000 psi water vapor pressure, vfg-dinochrysotile, as given by the reac- forsterite + talc ; HsO (III) le (1949) to be no higher than 505 C. at ns on t.he influence of fayalite solid solu.nlibrium temperature for the reaction: e olivine ! PaTf.) + talc + H-0 + Ch i IV) 5,000 psi. The widespread occurrence of implies that pressure-temperature condie commonly those characterizing reaction The temperature difference between the provisionally to be about 80 C, but it uther reaction considers the stability of >n clinochrysotile, or the influence of iron serpentine-group minerals. Furthermore, phase in serpentinized Alpine ultramafic nerally alter directly to serpentine minbilitv limit for "serpentine" can be de n of the type: olivine cnslatile + HiO Tlie invariant for reaction (V) lies ai a higher temperature than for (III), Since we know from Bowen and Tuttle (.1949) thai. forsterite plus talc more stable'than ensiatite plus water in the 500 C.--one kilobar IL 'water vapor pressure region. Recognizing the widespread occurrence of awaruite, a nickel-iron alloy restricted composition (approximately Xi*Fc), as a product of serpen- tinization, Olsen (1963) suggested that serpentinization may frequently m Jiave taken place in the region near 400 C. and 24,000 psi water vapor pressure, since lower temperatures of hydration would be expected to yield awaruite compositions lower in nickel content. Within the Xew Jdria, California serpentinite mass, Coleman (1961) has described calcsilicate mantles between jadcite pods and the surrounding serpenlinite. these mantles the hydrogrossular composition and the presence of thomsonite suggest crystallization temperatures ranging from 240 C.-- ,360 C. The jadeite pods presumably originated from the desilicalion of keratophyres which were tectonically engulfed in the serpentinite (Cole man, 1961). Loss of SiOg and addition of Cat) to the mantle surrounding the pods favors the idea that alteration and crystallization were con temporaneous with serpentfnizarion. Within the Xew Idria mass brucile .'generally accompanies serpentine-group minerals. Serpentinization--Volume Changes and'Transfer of Material. According ;to the concept of serpentinization at constant volume (Turner and Verhoogen, 1960, p. 319), large amounts of both magnesia and silica- must be removed in solution from the ultramafic mass. if! 5 MgiSiOj - 4 HsO 2 Mg-Si;Os(OH) i -f 4 MgO - Si02 (VI) re.: olivine (704 gm, 220 cc) introduced (72 gm) "serpentine'' -- (161 gm) (60 gm) (554 gm, 221 cc) removed in solution ill Removal of magnesia alone will not. suffice; serpentinization of pure dunites with attendant, leaching of all excess magnesia will result in a I:; volume increase of 25 per cent. el : 2 Mg-SiO-t 4- 2 HsO -- Mg,Si20;OH), + MgO olivine introduced "serpentine" removed in solution (VII) : (281 gm, 88cc) (36 gml (277gm. llOcc) (40 gm) The addition of silica, either as pyroxene within the serpentinizing mass . (equation VIII), or from aqueous solutions derived from country rocks (equation IX), calls for larger volume increases of 46 per cent and 67 nj Mg-SiOi + MgSiOs - 2 IfjO MgjSijOsiOHi i (VIII) olivine pyroxene introduced "serpentine" (141 gm, 44 cc) (100 gm, 31.5 cc) (36 gm) (277 gm. 110 cc) t 88 HOSTETMill. COLEiTAX, MVMPTOX AXD UVAXS 3 Mg3SiO, + 4 ICO + SiO: -- 2 MgsSiiO;,'OHi olivine introduced "serpentine" IX) (422tgm, 132cc; (72 gm) (00 gm i (554 gm, 221 cc) The widespread leaching of magnesia and silica from a hypothetical ultramafic mass undergoing serpentinizalion at constant volume should result in notable metasomatic effects in the adjacent wall rocks. Evidence for such metasomatic effects has seldom been encountered in the field (Turner and Verhoogen, 1960, p. 319; Coleman, 1965), but it may be obscured by faulting along the contacts. The leaching of silica, however, from a serpentinizing ultramafic mass is particularly implausible on thermodynamic grounds because it would call for a chemical potential gradient for silica away from the uli ramafic mass, which is strongly un dersaturated with respect to quartz, and towards the country' rocks, commonly made up of graywackes, shales, and cherts, which contain free quartz. Preservation of olivine grain boundaries in mesh-structure serpentines and the presence of baslite-lype pseudomorphs after enslalite is often cited as evidence for "serpentine'' replacement at constant volume. This evidence is ambiguous. Interconnecting "serpentine" veins and the in numerable fine veinlets of 'serpentine" cutting through parent olivine grains which show optical continuity are, in the manner of septarian concretions, plausible evidence for expansion. Buse and Watson (I960), while describing the alteration of an ultramafic belt near Bryn Mawr, Penna., noted a volume increase due to a "spreading apart." of olivine grains along fractures during serpenlinization. Raleigh (1963), studying partially serpentinized enstatile and olivine grains in the ultramafic body on Cypress Island, Puget Sound area, Washington, noted that offsets in exsolution clinopyroxene lamellae and in extinction bands in olivine grains are caused by zones of "serpentine." According to Raleigh, "The offsets in every- case are in the sense which would be produced by expansion in the serpentine band normal to its boundaries." Measure ments from a number of grains indicated that volume expansions ranging from 20 to 58 per cent, were required to produce the observed offsets (Raleigh, 1963, p. 64). The geologic setting of serpentinized ultramafic bodies supports in a general way the concept of a large volume increase during serpentinization. Faulted contacts, slickensided and highly sheared border zones, and the lack of contact melamorphism suggest an upward diapiric movement in response to tectonic forces along paths of least resistance. Surface extrusions of serpentinites in California (Dickinson. 1963) and the occurrence of these serpentinites on topographic highs indicates upward and outward movements suggestive of volume increases. Inflow pf water fre and an incr| jpovement country roel Coleman (l| - Lastly-, tl jinization i$ volume alte, quantities dunite req he removed! 6.1 '865 If serpeni olivine -- ortlsi vi e may caU ratio of ortf but small, qj Calculation! K ;.0 to 1/1 an magnesia, account for) : mafic mass in magnet in distributed r FoffiFas, (OH)*, brui . analyzer], coexisting body are p . listed by roafics; in Assumed d p. 22), ortli lizardite ati Brucite of 5 .jV.I/.-1-V. MCMPTOX AXD f-A'.iXS Hjt) -1- SiO; 2 Mg-FhOyOIIq introduced 2 Kmj (60 gm) "serpentine" (554 gm. 221 cc) of magnesia and silica from a hypothetical serpentini^ation at constant volume should ; effects in the adjacent wall rocks. Evidence ; has seldom been encountered in the held >0, p. 319; Coleman, 1965). but it may be ie contacts. The leaching of silica, however, mafic mass is particularly implausible on ause it would call for a chemical potential the ultramafic mass, which is strongly unj quartz, and towards the country rocks, wackes, shales, and cherts, which contain n boundaries in mesh-structure serpentines ype pseudomorphs after enstatite is often ine" replacement at constant volume. This connecting ``serpentine" veins and the inerpenune" cutting through parent olivine ontinuity are, in the manner of sepiarian ~e for expansion. Buse and Watson (1960), >n of an ultramafic belt near Bryn Mawr, case due to a ``spreading apart" of olivine serpentinizafion. Raleigh (1963), studying tile and olivine grains in the ultramafic get Sound area, Washington, noted that oxene lamellae and in extinction bands in >nes of "serpentine." According r.o Raleigh, in the sense which would be produced hy >and normal to its boundaries." Measure s indicated that volume expansions ranging required to produce the observed offsets xntinized ultramafic bodies supports in a large volume increase during serpentinizasided and highly sheared border zones, and iism suggest an upward diapiric movement along paths of least resistance, tinites in California (Dickinson, 1963) and entiniles on topographic highs indicates -nts suggestive of volume increases. Inflow BRCCTTE IX SERPEXT/XITES 89 gf water from surrounding sediments with concomitant serpentinization "^d an increase in volume and decrease in density should enhance such piovenient and may result in the inclusion of numerous fragments of country rock as Observed by Dudoz and Yuagnat (1962) in Cuba, and Coleman (1961, 1963) in New Idria, California, and New Zealand. lastly, the formation of significant amounts of brucite during serpen^jnization is a powerful argument against the hypothesis of constant volume alteration, because it would necessitate the removal of enormous quantities of silica from the ultramafic mass. Such a removal for pure dunite requires (hat 35 weight per cent of the original silica in the olivine removed in solution from the ultramafic mass: 6.15 .XfeSiO: + 6 H?0 --* 2 Mg-.SiaOsfOHn + 2 Mg(OH)a . olivine introduced , "serpentine", ,, brucite <86o gm, 271 cc) (90 gm) <,o54 gm, 221 cc. (U6gm,49 cc) + 4.3 MgO -j- 2.15 SiOa 73 gm)__ (129 gm; removed in solution X) If serpenLinization is expressed by the general reaction: olivine + orthopyroxene + HUD + Oa -* "serpentine" -- brucile -f- magnetite MgO (migrated), (XI) we may calculate a range of volume increases depending upon the initial ratio of orthopyroxene to olivine and upon the migration of variable, but small, quantities of magnesia away from the considered rock volume. Calculations were made for molar orthopyroxene-'olivine ratios of from ;0 to 1/1 and for migrations of 0, 5, 10, 15, and 20 per cent of the original magnesia. Calculations performed according to reaction (XT) do not ^account for any silica that may be added to or removed from rhe ullratnafic mass. Several other assumptions should also be mentioned. Except ;in magnetite, iron was assumed present entirely as ferrous iron and was distributed according to the analytical results described earlier: olivine - Fo^Fas, orlhopyroxene--En<>.>Fss, "'serpentine" (Mg.hsFe.odjSisOs (OH)4, brucite--(Mg.ssFc.u)(OH)a. Although no orthopyroxenes were analyzed, Green (1964) has shown that Mg/Mg+Fe+Mn ratios for coexisting olivines and enstatites within the Lizard Alpine peridotite body are practically equal. Among the coexisting olivines and pyroxenes listed by Ramberg and Devore (1951) are two pairs from Alpine ultramafics; in these also, the Mg/Mg+Fc ratios are nearly identical. Assumed densities were: olivine (Fo'ojFas)--3.26 (Deer, el al. 1962, v. 1, p. 22), orthopyroxenc (EngjFss) 3.26 (Deer, ct al., 1963, v. 2, p. 28), lizardite and clinochrysolile--2.55 (Deer, el al., 1962, v. 3, p. 170). Brucite of composition (Mg.g5Fe.15) (OH)* was also assumed to have a 90 HOSTETLER. COLES!AX, ML'MPTOX AXV EA'AXS density of 2.55 because we were not able to effect any specific gravitv separation of brucitc from serpentine-group minerals. Lastly, possible differences in porosity between the initial dunite or peridotile and the final serpentinite were neglected.1 A sample calculation with an initial olivine-orthopyroxene ratio of 9 l and a migration of 5 mol per cent of t.he original magnesia, is shown in reaction (XII), below: 9(Mg.)Fe.os''2SiOj + (Mg.9iFe.w)SiOs -f- 12.595 ICO | 0.0885 0-. olivine orthopyroxene introduced introduced (1311.7gm, 401.4 cc) (102.9 gm. 31.6cc) (227.0 gm) (2.8 gm) of.Nfg.asFe.cdsSiaOatOrijj 2.595(Mg.iisFe isKOH's "serpentine" (1404.5 gm, 543.5 cc'i brucite (1(53.7 gm. 64.2 co -f 0.177 Fe.-.Oi -r 0.874 MgO (XII) magnctil c migrated (41.0 gm, 7.9 cc) (35.2 gm) Reaction (XU) was balanced as follows: (1) Moles of "serpentine" produced were calculated from total amount of Si present in olivine and orthopyroxene, (2) The moles of Mg originally present in the primary silicates, less the molar sum of Mg migrated and present in "serpentine." represents moles of brucite produced, (3) Moles of magnetite produced represents one-third of the moles of Ke origi nally present, less moles of Fe present in "serpentine" and brucite, (4) Moles of water and oxygen required were obtained by balancingfirst for hydrogen, then for oxygen. The volume increase of 182.6 cc is a 42.1 per cent increase over the original peridotite volume. The molar ratio of brucite to "serpentine'' is 0.52, the volume and weight ratio is 0.12, and the weight of migrated Mg (as Mg4-1-) is slightly more than one per cent of the rock mass. Calculation results are graphically depicied in Fig. 2. The vertical coordinate represents the percentage volume increase; the lower horizon tal coordinate shows molar ratios of brucite/"serpentine", and the upper horizontal coordinate shows volume or weight ratios.2 Each dashed line represents the serpentinization reaction from a constant olivine/ortho pyroxene ratio but with various molar amounts (0- 20 per cent) of mag nesia migration. Each solid line represents reaction from various original ratios of olivine/orthopyroxene but: with a fixed amount of migrating magnesia. Assuming no significant addition of silica from country rocks, a knowl- 1 Although we have no measured values as guide lines, the ability of "serpentine" cores to sorb several excess per cent of water during shelf storage (Hess and Otalora, 1964. p. 165; suggests that serpentinites may be more porous than primary dunite or peridotites. 1 Volume and weight ratios are identical because the density of the iron-bearing brucite is considered to he identical with lizardite and clinochrvsotile. edge of th| - ratio of thi "of volume -from four tain, Hat cles. The M 54 22C 50 in 42- avoo--> 2<2v 3S B JD c 34 >4 30 26 J1.0 Fits. 2. V< rounding couj - composition magnesium "serpentine" | pies shown! tinized du dunite, an dotite wit Brucite Calculated! tions of tl migration: of the T\v| pure olivi Olli.MAX, Ml'MPTOX A.XP F.I'.-l NS ' were not able to effect any specific gravity i serpentine-group minerals. Lastly, possible ween the initial dunite or peridotite and the ected.1 h an initial olivine-orthopyroxene ratio of 9 '1 >er cent of the original magnesia, is shown iu e.osiSiOs -- 12.595 H=0+ 0.0885 0-. /roxcne introduced introduced ` i.ol.6cc) (227.0 gni) (2.8 gm) - 301,.i(Fe..u);SiiOi(OH), + 2.595(Mg.aFe Isb.OH\. "serpentine" brucitc (1-104.5 gm, 543.5 cc.) (165.7 gm, 64.2 cc) + 0.177 FesO, -f- 0.874 MgO . (XU) magnetite migrated (41.0 gm, 7.9 ccj (35.2 gm] need as follows: ed were calculated from total amount of Si present in e moles of Mg originally present, in the primary silicates, and present in "serpentine," represents moles of hrucite produced represents one-lhird of the moles of Fc midsent in "serpentine" and brucile, (4) Moles of vat er and mlancing first for hydrogen, then for oxygen. i82.6 cc is a 42.1 per cent increase over the The molar ratio of brticite to "serpentine" ght ratio is 0.12, and the weight of migrated ore than one per cent, of the rock mass. ;raphically depicted in Fig. 2. The vertical :rcentage volume increase; the lower horizonatios of brucite/"serpentine", and the upper volume or weight ratios.2 Each dashed line ion reaction from a constant olivine/orthoious molar amounts (0-20 per cent) of mag line represents reaction from various original cene but with a fixed amount of migrating ddition of silica from country rocks, a knowl- d values as guide lines, the ability of "serpentine" coves ter during shelf storage (Hess and Otalora, 1964, p. 1657 norc porous than primary- dunite or peridotiles. dentical because the density of the iron -bearing brucire M'dite and dinochrysotile. BJiUCITF. IX SFJiPUXTIXITF.S 91 igdge of the hrucite - "serpentine" ratio and of the olivine 'orthopyroxene yatio of the original ultramafic is sufficient to determine the percentage jjf volume Increase and the amount of magnesia migration. Examples from four samples, described earlier, from Dun Mountain, Burro Moun tain, Hat Island, and Twin Sisters, are shown on Figure 2 by solid cirfcles. The Dun Mountain and Twin Sisters samples are partially serpen- (wekjhfj rut' hrucite/" serpentine" .220 .200 .180 .i60 .140 .120 ,100 .080 .060 040 .020 000 54------- A - A- no Mg migration B -B'- migration of 5% original Mg 50 c-c- io%- D- O'- ' " 15% - " E- E- " " 20% ' " 8 46 60%0I- $o 42-1i -- SSi 3 o 341 301 01 -Olivine Opx -Orthopyroxene Burro Mtn. Dun Mtn. Hat island Twin Sisters 26 i.O 0.9 o'.B o `:r o!6 o'.s 0(4 o!3 ~o'.2 molecular ratio brucite/ serpentine Fig. 2. Volume increase during serpentinization, assuming no addition of silica from sur , rounding country rocks. Volume increase plotted as a function of original mineralogic i composition and resultant brucite/"setpentine" ratios. Molecular amount of migrated magnesium also derived from original mineralogic composition and resultant brudte/ . "serpentine" ratios. Dun Mountain, Hat Island, Twin Sisters, and Burro Mountain sara - pies shown by solid circles. tinized dunites, the Hat Island sample is an almost totally serpentinized dunite, and the Burro Mountain sample is a partially serpentinized peridotitc with a molar olivine,-'orthopyroxene ratio of approximately 9/1. Brucite-'"serpentine" ratios from these samples are shown in Table 1. Calculated volume increases, applicable only to the serpentinized por tions of the rock samples, range from 37-41 per cent and the magnesia migrations from 7-10 per cent. If the molar olivine/orthopyroxene ratio of the Twin Sisters and Dun Mountain samples is 9/1 instead of being pure olivine as shown, the percentage of volume increase would be 41 92 HOSTETLER, COLEM (.V, MCifPTOX AXD HVAXS per cent instead of 37 per cent, and the amount of magnesia migration would he only 5 per cent instead of 10 per cent. Several chemical mechanisms may account for the 7-10 per cent mi gration of magnesia, but the observed lack of magnesia metasomatism in the country rocks associated with serpentinites indicates, that a largescale removal of magnesia from the ultramafic mass is probably not realistic. Incremental serpentinization may indicate that small incre ments of magnesia have dribbled away into adjacent wall rocks during the ascent of the ultramatk mass. Fixation with available CO as inanesite is sporadic, but may be of local importance. At least part of the magnesite that has apparently migrated may actually be incorporated in brucites with a higher Mg content than the (Mg.ssFe.wXOHja com position used in our calculations. Juxtaposed dunite and peridolite layers with imbalances of silica and magnesia are probably important in fixing migrating magnesia. .Vs shown on Figure 1, serpentinization of a primary mass containing forty or more molecular per cent (28 weight per cent) of pyroxene will not form brucile but will instead free excess silica. Movement of magnesia from dunite layers into more silica-rich peridotite layers would allow the serpentine-group minerals to form in the veins so commonly observed within serpentinite bodies. The weight per cent of brucite will, under the most favorable condi tions, appear so modest that one may be tempted to neglect its impor tance in determining volume relationships. For instance, the weight per cent of brucite in the partially serpentinized Dun Mtn. sample is only 5 per cent, yet. the observed brucite/<:serpentine" ratio enables us to calculate a volume increase of 37 per cent for the serpentinized portion of the sample. This figure for volume increase cannot be lowered signifi cantly unless it is supposed that massive amounts of silica have been removed from the sample; a supposition which, as mentioned earlier, lacks field verification and is thermodynamically implausible. Other factors'may modify somewhat the calculated estimates of vol ume increase. First, the Mg/Fe ratios in the primary silicates may be somewhat different than the 92/8 figure used for our calculations. For example, in the Glen Urquharl, Scotland serpentinite (Francis, 1956) optical data indicate an olivine of 96 per cent forsterite and an ortho pyroxene of 94 per cent enstatite. Although we do not have a good idea of the fluctuations of iron content in associated serpentine-group min erals and brucites, differences in calculated volume increases would be negligible. Secondly, orthopyroxene is commonly more resistant to ser pentinization than is olivine (Bowen and Tuttle, 1949, p. 457). Prefer ential preservation of orthopyroxene in parti all}' serpentinized rocks would probably result in too low an estimate of the original olivine-' T/'/atthopyroxc r*" percentage < Injuring serpe k'.r Is aomewhat ".. LastK, tl Silica to the l" silica, rathe, :s-s%ne increas / .culated vo(u ss/ylountain a. "silent of the l.,.; Calculated oi is 52 per cn Enormous ^'additions of /'Iperature of ^/volume of v " '-serpentine' i...-original rods ks ultramafic r for water vo ;--. pvobably mi water rec |s/sample is 7 Immigration o |:-/samplc. If, < ssy.from quartz ) /observed M -Si total volum s#: ihnes the v i):- Kennedy Hi One can sj ||Z as a means he effective. |i|7 of diffusion tested mechi rocks and th important : within the number of grations of; MgO'SiOj /.I.V. MUMPTOX AXD I-.I'AXS , and the amount of magnesia migration .d of 10 per cent. s may account for the 7-10 per cent minerved lack of magnesia metasomatism jn ith serpentinites indicates, that a large st the ultramafic mass is probably not lization may indicate that small incre ed away into adjacent wall rocks during sss. Fixation with available COo as tnag of local importance. At least part of the migrated mat' actually be incorporated ontent than the (Mg.ssFe.i5)(OH)2 com . Juxtaposed dunite and peridotite layers agnesia are probably important in fixing n Figure 2, serpentinization of a primary molecular per cent (28 weight per cenr) cite but. will instead free excess silica, lunite layers into more silica-rich periU'pentine-group minerals to form in the thin serpentinite bodies, ie will, under the most favorable condie may be tempted to neglect, its imporadonships. For instance, the weight per serpentinized Dun Mtn. sample is only riicite.'"serpentine'* ratio enables tts to 7 per cent for the serpentinized portion lume increase cannot be lowered signifit.t massive amounts of silica have been ipposhion which, as mentioned earlier, trmodynamically implausible, newhat the calculated estimates of volratios in the primary silicates may be , 8 figure used for our calculations. For , Scotland serpentinite (Francis, 1956) of 96 per cent forsterite and an ortho ;. Although we do not have a good idea :nt in associated serpentine-group min i calculated volume increases would be ene is commonly more resistant to ser>wen and Tuttle, 19-49, p. -457). Preferoxene in. partially serpentinized rocks w an estimate of the original olivine,' BRVCfTK ix si-:rp/-:xtixiti:s 93 arthopyroxene ratio, and from Fig. 2, somewhat too high a figure for the percentage of volume increase. Thirdly, a significant increase in porosity during serpentinization would mean Lhat the calculated volume increase jk somewhat too ltfw. ' Lastly, there remains the possibility of Important contributions of 'silica to the ultramafic mass from surrounding sediments. Addition of , silica, rather than migration of magnesia, indicates that calculated vol ' tune increases are lovi by a considerable margin. For example, the cal culated volume increases shown in Fig. 2 for serpentinization of the Dun Mountain and Twin Sisters samples (assuming migration of 10 mol per cent of the original magnesia) are 37 per cent but if serpentinization is calculated on the basis of addition of silica, the resultant volume increase is 52 per cent. Enormous volumes of water would probably be required for even small additions of silica to the ultramafic mass. Depending on depth and tem perature of the ultramafic mass at the time of serpentinization, the volume of water required for the conversion of the primary silicates to "serpentine"-brucite is 50 to 80 per cent of the total volume of the original rock. The introduction, of reasonable amounts of silica into the : ultramafic rock mass by solution transport would require this estimate :for water volume to be increased by at least an order of magnitude and iprobably much more. For example, at 400 C. and 1500 bars the volume of water required for complete serpentinization of the Dun Mountain sample is 75 per cent of the original volume of the sample, assuming ;rnigration of 10 mol per cent of the original magnesia away from the sample. If, on the other hand, it is assumed that additional silica, derived :from quartz-saturated water in the country rocks, is responsible for Ihe observed MgO SiCh ratio in the serpentiue-brucite fraction, then the total volume of water required to provide this additional silica is 140 times the volume of the original sample (quartz solubility data from Kennedy (1950)). One can speculate on the feasibility of rapid rates of diffusion for silica as a means of drastically reducing these water volume requirements. To be effective,- the chemical potential gradient for silica must lead to a rate of diffusion markedly more rapid than the flow of water. This is an un tested mechanism to account for the movement of silica between country rocks and the interior portions of large ultramafic masses, but it may be important in accounting for local migrations of magnesia and silica within the ultramafic mass. Chemical analyses of a sufficiently large number of samples from an ultramafic mass will average out local mi grations of magnesia and silica and show, for the bulk rock, whether the MgO/SiOj ratio has changed during serpentinization. Constant bulk 1 94 HOSTETLER, COLEMAN. Sll MELON AND EVANS MgO ;SiOs ratios during serpenlinizafion were noted by Sbteinberg (1960) in his study of the serpentinized dunites of the Urals. Regardless of {.he applicability of the speculation in the several pre ceding paragraphs, ishe initial volume of water ultimately incorporated as (OH) groups within the crystalline brucitc and serpentine-group minerals must have been nearly as large as the combined volume of the primary anhydrous silicates; a fact which of itself makes implausible the concept of serpentinization as an autometasomatic process. Further more, the large volume increases required for complete serpentinization suggest that this process does not occur in situ or as the result of one enormous injection of water, because of the problem of accommodating such a volume increase within the surrounding country rocks. We prefer the concept that pervasive serpentinization is a gradual, possibly con tinuous, possibly intermittent, process that feeds on water derived from adjacent country rocks and that accompanies the rise of the ultrainaik mass during tectonism. Volume increases resulting from serpentinization serve as an impetus to the continued diapiric ascent of the mass. This concept of serpentinization is harmonious with structural and geomet rical features typical of Alpine ultramafics; namely, their ellipsoidal, fault-bounded configurations, the numerous inclusions of country rock near the margins and especially the distal ends of the ellipsoidal mass, and the occurrence of the masses along topographic highs, even where the density of the ultramafic mass is considerably greater than the sur rounding rocks. COa is at p0o- values a {MgaiFe-CO; ^finite znass, > JiTihe authors < "' ";: piagnesium c .. /stable (30-60 "[vjshort fibre 01 ///contained wi proach a 1 / ~~ Surface and , r.U ,,cottonballs: .... dialing artir s/k:the near esc.' /'/pf massive sc y - brown coalir "V ently result 1 5; the tmweart where air an ///bearing serf ,5- carbonates. to groundwa Krucite Stability and Carbonate Formation. Shtelnberg (1960J has noled the lack of brucite in aphanitic serpentinites which occur as small veins along diagonal fractures in the serpentinized, brucite-bearing dunites of the Urals. Lack of brucite may be due to variations in compositional type, or to local, but intensive, removal of magnesia and/or introduction of silica. Thus, brucite is probably not stable along the sheared borders of ultramafic masses, or in the immediate viciniiv of fresh fractures opened in brucite-bearing serpentinites, or in weathering zones where COa may be present or low pH values could develop. The dark, brucitebearing core of serpentinized dunite from Hat Island, describecl earlier, is surrounded by a greenish weathering rim of identical serpentine-group minerals buL devoid of brucite. Brucite formed during pervasive serpentinization must be regarded as a prime candidate for reaction with available CO-beaving solutions at depth to form the magnesite commonly associated with ultramafic rocks. In a hydrous surface or near surface environment brucite is con verted to hydromagnesite [Mgi(CO~)3(OH)2-3H20] if the partial pressure The logic; study in de; conclusions attention sb fresh, prirna gation and zones of pai : should indie to which sili of volume i tween the \ man- silicat explored. 0 serpentiniza group mine isotopic rati erations of: i.mr.ix. .ixnmcmptox evaxs erpentinization were noted by Shtelnberg erpentinized dunites of the L'rals. biiity of the speculation in the several preal volume of water ultimately incorporated e crystalline brucite and serpentine-group arlv as large as the combined volume of the a fact which of itself makes implausible the as an autometasoinatic process. Furtherlases required for complete serpentinization es not occur in situ or as the result of one because of the problem of accommodating n the surrounding country rocks. We prefer serpentinization is a gradual, possibly cont, process that feeds on water derived from that accompanies the rise of the ultramatic ae increases resulting from serpentinization ontinued diapiric ascent of the mass. This s harmonious with structural and geometiine ultramafics; namely, their ellipsoidal, , the numerous inclusions of country rock dly the distal ends of the ellipsoidal mass, asses along topographic highs, even where mass is considerably greater than the sur- te Formation. Shteinberg (1960) has noted ic s.erpeminites which occur as small veins e serpentinized, brucite-bearing dunites of ay be clue to variations in compositional , removal of magnesia and or introduction >ably not. stable along the sheared borders he immediate vicinity of fresh fractures penliniles, or in weathering zones where l values could develop. The dark, bruciteiunite from Hat Island, described earlier, alhering rim of identical serpentine-group rasive serpentinization must be regarded .ion with available CO>-bearing solutions ite commonly associated with ullramafic near surface environment brucite is con(CO3)3(OIi>JH20] if the parliai pressure BRUCITE IX SERPXXnxrTXS -of CO2 is at least 10-6 atm. (Hostetler, 1960). Probably even smaller pco2 values are required for conversion of brucite to artinire [Mgs(COj) (OHb 3KoO) or pyroaurite [Mg6Fe2(C03) (OH)i6 4H=0] and coalingite "{iVlgKiTe^CO.dOfel'.M 2H2O]. Within the New Idria, California, serpen tlnite mass, recent open-pit mining for short fibre asbestos has afforded the authors an opportunity to relate paragenetically brucite with these magnesium carbonate minerals. In those pits extending below the water table (30-60 feet) brucite is a common constituent of both the sheared '"$hort fibre ore and of the larger, massive, unsheared serpentinite blocks contained within the sheared material. The amount of brucite may ap proach a 1/1 molar ratio with the serpentine-group minerals. At the Surface and above the water table as exposed in these open pits, white "cottonballs': of hydromagnesitc (1-5 mm in diam.) and occasional ra diating artinite needles are ubiquitous in the sheared serpentinite to ft the near exclusion of brucite. Commonly the surfaces of residual blocks of massive serpentinite exposed in the same zone are coated with reddish"brown'coalingite and pyroaurite (.Wumpton, el al., 1965) which appar ently result from the oxidation of iron-bearing brucite in situ. However, the unweathered interiors of these blocks do contain brucite. Thus, ' where air and or COs-bearing groundwater have easy access to brucitebearing serpentinite, the brucite is readily converted to magnesium carbonates, whereas within those massive serpentinite blocks impervious ft to groundwater percolation, brucite is preserved. s: 1'trrcRK Probiems The logical sequel to a general and pragmatic paper such as this is a study in depth of a particular ultramafic body where assumptions and conclusions can be tested against critical field observations. Particular attention should be paid to zones of partial serpentinization between fresh, primary silicates and fully serpentinized areas. Detailed investi: gation and comparison, of MgO-'SiOi, FeSiO?, and Fe/Mg ratios in zones of partial serpentinization and in adjacent unserpentinized zones 1. should indicate the extent of magnesia and silica migration, the extent . to which silica has been added from country rocks, and reliable estimates : of volume increase during serpentinization. Possible relationships be : tween the various serpentine-group minerals and association with pri mary silicates or with successive generations of "serpentine" should be ; explored. Our data suggest that lizardite-brucite is the initial result of ; serpentinization of olivine. Knowledge of the variation in serpentine : group minerals of NiO and CoO content and of oxygen and hydrogen : isotopic ratios should prove useful in tracing initial and successive gen ; erations of serpentinization. 90 HOSTETLER, COUCMAX. ilU.UJPTOX AXD HVAXS Summary The apparently widespread occurrence of brucite in scrpentinized Alpine ultcamafics (up to 10 weight per cent), as related to the cxp.-riraental work of Bowen and Tuttle (19-19) indicates that the temperatures of serpentinization are probably some 100 C. lower lhan generally ac cepted, because the presence of brucite lowers the maximum value (500 C.) given for the reaction: olivine -) H;0 ~ silica -- '`serpentine" Moreocvcr, the occurrence of brucite as a product of pervasive serpenlinization demands a large volume increase during serpentinization unless it can be shown that truly massive quantities of both silica and magnesia have been removed from the ultramafic mass. Field observa tions yield no evidence in support of such massive migrations. Depending on the original ratio of pyroxene to olivine, our calculations show that volume increases of 35-40 per cent are common during serpentinization. The tectonic implications of such expansions are immediately apparent. The complete serpentinization of a large. Alpine ultramafic body in silu by water derived from adjacent country rocks is unrealistic because of the difficult}- in accommodating a great increase in volume within the surrounding country rocks. We assume that serpentinization of the large Alpine ultramafic belts is accomplished by moderate additions of water during the tectonic ascent of these bodies through the earth's crust. The widespread serpentinization of these Alpine ultramafics combined with their nearly universal occurrence in and along structural discontinuities suggests that much, or all, of the expansion is accommodated during tectonism. Further careful field and laboratory work is necessary before the process of serpentinization can be related to the origin of the ul tramafics. Acknowledgments -We would like to rhank the Union Carbide Corporation for access to their workings and for logistical support during our investigation of the New Idria ultramafic mass. We are also grateful to Don Donaldson and Don Freitag of the U. S. Geological Survey for chemical analyses and to Norman Page of the University of California (Berkeley) for critical review and discussion. Financial support from the National Science Foundation made (he microprobe analyses possible. References Allakhverdiev, Sir. I. (1958) Carbonates from the weathered crust of ultrabasic rocks of Azerbaidzhan. Ievest. Akad. Naitk, Azerbaidzhan S.S.R., Ser. Gcol.-Geograf. Xauk. 3, 79-87. ' Alias, C eba Yoa ) IAX. MCUPTOX AXD fiVAXS Summary ! occurrence of brucile in serpetuitiized veight per cent), as related to the c-xporitie (1949) indicates that the temperatures ly some 100 C. lower than generally ac- of brucite lowers the maximum value n.: 0 4- silica -- `'serpentine" of brucite as a product of pervasive servolurne increase during serpcnlinization uly massive quantities of both silica and from the tdlramafic mass. Field observa nt of such massive migrations. Depending ne to olivine, our calculations show that ent are common during serpentinization. ch expansions are immediately apparent, of a large Alpine uitramafic body in sihi it country rocks is unrealistic because of g a great increase in volume within the assume that serpentinization of the large nplished by moderate additions of water ese bodies through the earth's crust. The these Alpine ultramafics combined with '.e in and along structural discontinuities the expansion is accommodated during and laboratory work is necessary before can be related to the origin of the ui- 'OWLEDGMEXTS Jnion Carbide Corporation for access to support during our investigation of the are also grateful to'Don Donaldson and igical Survey for chemical analyses and si tv of California (Berkeley) for critical al support from the Xalional Science ie analyses possible. F.FEREXCES les from the weathered crust of ullrabasic rocks of Azerbaidzhan S.S.R.. Ser. Geol.-Geograf. Xatik. 3, - BRUCITE IX SERPEXTIXU'ES 97 I'ijAES, C. C,, J. C. Gii.l as J. S. KoSKI '1957) The Jeffrey Mine of Canadian Johnsi Manville Company. Limited. In, The Geology of Canadian Industrial Mineral Dc : posils. Sixlh Commonwealth Min. Metall. Cong., Vancouver. B. C. 27- 36. SJtENSOX, IV. X. 61913--1918.3 The geology and pcLrology of the Great Serpentine Belt of ! Xew South WalfS. Proc. Linn. Sop. XJS.W. 38, 490-517, 569-596, 662-724: 40, 121- bf 171,540-62-1; 42,223-283,693-700,- 43,320-393,593 -599. V____ (1926) The tectonic conditions accompanying the intrusion of basic and ultrabasic ' igneous rocks. C.S. .Vat. Acad. Sci. I, 'Bowks, X. I., and 0. l-`. Tuttle (1949) The system MgO-SiOs-HsO. Geol. Soc. Am. Bull. 60,439-460. ' Jjrsit. 11. L. .VXD E. H. Watson (19601 Alteration of ultrabasic rocks near Bryn Mawr, Pennsylvania. Pa. A cad. Sci. Proc. 34,117-123 :<3Cr.EMAX, R. G. i 1961) Jadc-ito deposits of the Clear Creek Area. New Idria District, San Benito County, California. Jour. Petrology 2,209-247. --------(1963.) Serpentinites, rodingites, and tectonic inclusions in Alpine-type mountain ; chains. Geol Soc. .4 m. Spe. Paper 73,130. iicc-------(1965) New Zealand serpentinites and associated metasomatic rocks. Geol. Bull. .Vea!ZealandSurrey (in press). : Peer, \V. A.. R. A. Howie and J. Zussvian 11962) Rock-forming Minerals,J. Ortho- and :v Ring-silicates. John Wiley and Sons. Inc., Xew York. .--------- (1962) Rock-forming Minerals, v. 3, Sheet Silicates. John Wiley and Sons, inc., Xew York. ---------(1962) Rock-forming Minerals, v. 5. Xon-silicaies. John Wiley and Sons, Inc., Now : York. --------- (1963) Rock-forming Minerals, p 2, Chain Silicates. John Wiley and Sons, Inc., :> New York. (Dickinson, W. R. (1963) Mrecciated serpentine extrusion, on Table Mountain in central . Caliiornia Coast Ranges.Geol. See. Am. Spec. Paper. 73, p. 34. :PuciOZ, C. and M. Vuagnat (1962) A propos de l'age dcs serpentinites de Cuba. Archives ties Sciences, Centre, Vol. 13,2,309-332. ; Eckh.vrdt. I-'. J. (1956) Rontgenographische Untersuchungen am Schwcizerit. Xettes Jahrb. Mineral. Monatsh., 32-43. F.vckler. W. C. (1945) The occurrence of nemalitcin Alaska. .1?k. Mineral. 30,640-641. : Faust, G. T. and J. J. Fahey (1962; The serpentine-group minerals. C.S. Geol. Survey Prof. Paper384 A, Fenooi.io, M. (1930) Sulla presenza della nesquehonite nella serpentina di Viv in Val di Lanzo. Aecad. Xas. dci Lincei, Send. 11,310-316. --------- (1933) Sulla lansfordite della minicre di Cogne in Val d'Aosta. Period Mineral. 4, 443-462. ' Ff.rkaki, A., G. Scanini and I.. Cavalca (1950) La brucite di Carro ed i minerali che l'accompagnano. Period. Mineral. 19,41--19. Feancis. G. H. (1956) The serpentinite mass in Glen Urquhart, Invemcsshire, Scotland. .4 m. Jour. Sci., 254,201-226. Gref.n. D. H. (1964) Petrogenesis of the high-temperature peridotite intrusion in the I-izard Area, Cornwall. Jour. Petrology, 5,134-1S8. Haiix-Weinheimer. P. and F. .Rost (1961) Akzessorische Mineralien und Elemcntc in Seipcntinit von Leopoldsgriin {Miinchberger Gneismasse). Ein Betrag zur Genchemie ultrabasischer Gesteine. Geochim. Cosmochim. Acta. 21,165-181. Hess. H. H. and G. Otalora (1964) Mineralogical and chemical composition of the Mayaguez serpentinite cores. In, .4 study of serpentinite--the AMSOC core hole near Mayaguez, Puerto Rico. Xal. Acad. Sciences-XRCPiibl. 1188,152-168. Sf>. . 98 HOSTETLER, COLFMAX, MCMPTOX AXD F.VA.XS HcarsTETiER. F, von (1859) Lecture on the geology of the Province of Nelson. Xei. Zealand Govt. Gazette {Prov. Kelson). 7, (20), 90-102. Hostetler, P. B. (1960) Low temperature relations in the system: MgO-SiO-COj ILQ Ph.D. diss.; Harvard Univ., Cambridge, Mass. Kf.il, K. and K?Fkbdricksson (196-1) The iron, magnesium, and calcium distribution j,, coexisting olivines and rhomhic pyroxenes of chondrites. Jour. Gcophys. Res., 09, 5487-3515. Kennedy, G C. (1950) A portion of the system silica-water. Econ. Gcol. 45,629-655. Meixner. H. (1938) KraubatherLagcrstatlenstudien. Zentralblatt Minera!.. Abt. A, 5-19. .........-- (1956) Seltene Magnesiumminerale aus dem Serpentingebiet von Krauhath, Obersteiermark. Dor Kartnthin, 33,166-171. Mumptox, F. A., H. W. Japfe and C. S. Thomson 11965) Coalingite, a new mineral from the New Tdria serpentmite, Fresno and San Benito Counties, California. .1 in. Mineral. 50, 1893-1913. Olsen, E. (1963) Equilibrium calculations in the system Mg, Fe. Si, 0. H. and Xi. Am. Jour. Sci. 261,943-956. Raleigh, O. B. (1963) Fabrics of naturally and experimentally deformed olivine. Ph.D. diss. California Univ.. Los Angeles, Calif. Bamberg, H. and G. DeVore (1951) The distribution of Fe-1- and Mg'f" in coexisting olivines and pyroxenes. Jour. Geol. 59,193--210. Sakamoto, C. (1959) Studies on serpentine containing brudlc. Kogyo Kagakti Zass/ti. 62, 1485-1488. Seroyuchenko, I). P. (1949) Hydromagnesit.e. Oohlady Akad. Xauk S.S.S.R. 68, 581-584. ShteInderg, T). S. (1960) New data concerning the serpentinization of dunites and peridotites of the Urals. Inter. Geol. Cong., XXI sess. Doil. Soeet. Geol. Felrograf. Proe. Izverzh. Metamnrji. Gornye Parody, IzdaleTstvo Akad. Xauk S.S.S.R.. Moskva. 250--260. Smith, J. V. and R. C. StenstrOM (1965) Chemical analysis of olivines by electron micro probe. Mineral. Mag., Tilley vol. .136-159. Thayer, T. P. (1960) Some critical differences between Alpine-type and stratiform peridotite-gabbro complexes. Inter. Geol. Cong., XXI Sees.. Copenhagen. Rcpt. Pi. 13, 247 259. Turner, F. and J. Verhoogen (1960) Igneous and MclamorpMc Petrology, 2nd cd. Mc Graw-Hill Book Co.. New York. Van Biljon. W. J. (1960) Chemical reactions in the thermal breakdown of the serpentine minerals. XeuesJaUrb. Mineral., Abba. 94,1223-1240. .Voder, H. S.. Jr (1952) The MgO-AMVSiOa-H-eO system and the related met amorphic facies. .4 in. Jour. Sei.. Bowen tol. 569-627. Manuscript received, June 11,1065; acceptedfor publicalioir, October 16,1065. waJ