American Mineralogist,Volume63, pages 100A-1009, 1978 Newbiopyriboles from Chester, Vermont:I. Descriptive mineralogy Dnvn R. VnsLeNl lNo CHlnr-rs W. BunNHenr Department of Geological Sciences, Haruard Uniuersity Cambridge, M assachusetts 02 I 38 Abstract Four new magnesium-iron chain-silicate minerals have beenidentified from a metamor- phosed ultramafic body near Chester, Vermont. They occur with anthophyllite, cummingto- nite,and talc between the chlorite andactinolite zones at the boundary ofthe body. The cell parameters of the minerals are diagnostic: (l) jimthompsonite is orthorhombic, Pbca, a : 18.6,b:27.2,c:5.30A; (2)clinojimthompsoniteismonoclinic,C2/c,a:9.87,b--27.2,c : 5.32A,0: 109.5"; (3)chesteriteis orthorhombic,A2,ma,a:18.6,b:45.3,c:5.30A; (4) an unnamed mineral is monoclinic, A2/m, Am, or A2, a : 9.87, b : 45.3, c : 5.294, B : 109.7o. The physical and opticalproperties are close to those of low-Caamphiboles. The cleavage angles (37.8' and44.7") arelower thanthose of amphiboles, andintergrowths of the minerals with anthophyllite andcummingtonite arepetrographically distinctive. Theminerals are biopyriboles and are chemically intermediate between anthophyllite and talc.The ideal chemical composition for jimthompsonite andclinojimthompsonite is (Mg,Fe),oSi,roor(OH)n, and that of chesterite is (Mg,Fe),rSiroOEl(OH)6. The new minerals might easily be confused with amphiboles if theirelectron microprobe analyses were considered alone. Introduction Physical similaritiesamong pyroxenes, amphi- boles, and micas led Johannsen (1911) to call these mineral groups collectively the "biopyriboles." Solu- tion of the major biopyribole structure typeslater showed that the similarities werenot fortuitous, but rather werea direct resultof structural similarities. Thompson (1970) haspointed out that most amphi- boles can be thoughtof as alternating slabs of mica and pyroxene structure cut parallel to (010) along the ideal C2/m mica a-glide planes and pyroxene c- glides. These M (mica)and P (pyroxene) slabs, when assembled with the proper operations, form the MP double-chain amphibole structure. Thompson (1978) has also suggested that slab mixtureswith different mica-pyroxene ratios mightbe found. Our work con- firmsthis prediction andreveals the biopyriboles asa coherent mineralfamily,comprising several distinct but closely related structure types. This paperdescribes four new minerals and their occurrence in a talc quarry at Chester, Vermont (Veblen and Burnham, 1975, 1976; Veblen, 1976l' I Present address:Departments of Geology and Chemistry, Ari- zona State University, Tempe, Arizona 85281. Veblen et al., 1977). Physical and optical data,unit- cellparameters, electron microprobe chemical analy- ses, and X-ray powder diffraction patterns calculated from the refinedcrystal structures are reportedfor three of the new minerals. Nomenclature for these three well-characterized minerals is also presented. Only preliminary unit-cell dimensions and possible space groups are reportedfor the fourth mineral, which remains unnamed. All four minerals, whichare structurally andchem- ically intermediate between anthophyllite and talc, werefound duringexamination of single crystals by X-ray precession photography, and the crystal struc- turesof three of them havebeen solved and refined using diffractometer-measured X-ray diffraction data. Two of the minerals consist of triple silicate chains connected by octahedral cation strips, while the othertwo arethe first knownmixed-chain silicate structures, containing both double and triple silicate chains. The structural crystallography is described in a subsequent paper (Veblen and Burnham, 1978). Identification of the new minerals is not simple, because they occur together as fine intergrowths. Nevertheless, the information provided in this paper shouldenable mineralogists and petrologists to dis- 0003-0Mx / 7 8 /09l 0- I 000$02.00
Transcript
American Mineralogist, Volume63, pages 100A-1009, 1978
New biopyriboles from Chester, Vermont: I. Descriptive mineralogy
Dnvn R. VnsLeNl lNo CHlnr-rs W. BunNHenr
Department of Geological Sciences, Haruard UniuersityCambridge, M assachusetts 02 I 38
Abstract
Four new magnesium-iron chain-silicate minerals have been identified from a metamor-phosed ultramafic body near Chester, Vermont. They occur with anthophyllite, cummingto-nite, and talc between the chlorite and actinolite zones at the boundary ofthe body. The cellparameters of the minerals are diagnostic: (l) j imthompsonite is orthorhombic, Pbca, a :1 8 . 6 , b : 2 7 . 2 , c : 5 . 3 0 A ; ( 2 ) c l i n o j i m t h o m p s o n i t e i s m o n o c l i n i c , C 2 / c , a : 9 . 8 7 , b - - 2 7 . 2 , c: 5 .32A ,0 : 109 .5 " ; ( 3 ) ches te r i t e i s o r tho rhomb ic ,A2 ,ma ,a :18 .6 ,b :45 .3 , c :5 .30A ; (4 )an unnamed mineral is monoclinic, A2/m, Am, or A2, a : 9.87, b : 45.3, c : 5.294, B :109.7o. The physical and optical properties are close to those of low-Ca amphiboles. Thecleavage angles (37.8' and44.7") are lower than those of amphiboles, and intergrowths of theminerals with anthophyll ite and cummingtonite are petrographically distinctive. The mineralsare biopyriboles and are chemically intermediate between anthophyll ite and talc. The idealchemical composition for j imthompsonite and clinojimthompsonite is (Mg,Fe),oSi,roor(OH)n,and that of chesterite is (Mg,Fe),rSiroOEl(OH)6. The new minerals might easily be confusedwith amphiboles if their electron microprobe analyses were considered alone.
Introduction
Physical similarities among pyroxenes, amphi-boles, and micas led Johannsen (1911) to call thesemineral groups collectively the "biopyriboles." Solu-tion of the major biopyribole structure types latershowed that the similarities were not fortuitous, butrather were a direct result of structural similarities.Thompson (1970) has pointed out that most amphi-boles can be thought of as alternating slabs of micaand pyroxene structure cut parallel to (010) along theideal C2/m mica a-glide planes and pyroxene c-glides. These M (mica) and P (pyroxene) slabs, whenassembled with the proper operations, form the MPdouble-chain amphibole structure. Thompson (1978)has also suggested that slab mixtures with differentmica-pyroxene ratios might be found. Our work con-firms this prediction and reveals the biopyriboles as acoherent mineral family, comprising several distinctbut closely related structure types.
This paper describes four new minerals and theiroccurrence in a talc quarry at Chester, Vermont(Veblen and Burnham, 1975, 1976; Veblen, 1976l'
I Present address: Departments of Geology and Chemistry, Ari-zona State University, Tempe, Arizona 85281.
Veblen et al., 1977). Physical and optical data, unit-cell parameters, electron microprobe chemical analy-ses, and X-ray powder diffraction patterns calculatedfrom the refined crystal structures are reported forthree of the new minerals. Nomenclature for thesethree well-characterized minerals is also presented.Only preliminary unit-cell dimensions and possiblespace groups are reported for the fourth mineral,which remains unnamed.
All four minerals, which are structurally and chem-ically intermediate between anthophyllite and talc,were found during examination of single crystals byX-ray precession photography, and the crystal struc-tures of three of them have been solved and refinedusing diffractometer-measured X-ray diffractiondata. Two of the minerals consist of triple silicatechains connected by octahedral cation strips, whilethe other two are the first known mixed-chain silicatestructures, containing both double and triple silicatechains. The structural crystallography is described ina subsequent paper (Veblen and Burnham, 1978).
Identification of the new minerals is not simple,because they occur together as fine intergrowths.Nevertheless, the information provided in this papershould enable mineralogists and petrologists to dis-
0003-0Mx / 7 8 /09 l 0- I 000$02.00
VEBLEN AND BURNHAM: NEII BIOPYRIBOLES
t inguish the new minerals from other biopyr iboles,and also from each other.
Nomenclature
Names for three of the new minerals have beenapproved by the International Mineralogical Associ-ation Commission on New Minerals and MineralNames. The orthorhombic and monoclinic mineralswith b axes of -27 A that contain only triple silicatechains are named jimthompsonite and clinojim-thompsonite, after Professor James B. Thompson ofHarvard University. The orthorhombic mineral con-taining mixed double and triple silicate chains with b= 45A is named chesterite, for the locality near Ches-ter, Vermont. Although what is presumed to be themonoclinic analog of chesterite has been identified,its occurrence as very fine intergrowths in chesteritehas so far precluded measurement of its physicalproperties and confirmation of either its chemicalcomposition or crystal structure.
Fol lowing the usage of Johannsen ( l9l l ) andThompson (1970), biopyriboles are minerals that canbe represented as mixtures of (010) pyroxene andmica slabs; pyriboles are the subset of biopyribolesexcluding the micas. In this structural classification,talc is considered as a member of the mica group.
Cell parameters and space groups
The new minerals were first distinguished fromamphiboles by their b cell dimensions. Figure I com-pares 0-level a-axis precession photographs of ensta-tite, anthophyllite, jimthompsonite, and chesterite.The photographs show the effects of differing b-axislengths and exhibit the following extinction criteria: k: 2n 1 l missing for enstatite and jimthompsonite,and k * I : 2n * I missing for anthophyllite andchesterite.
Table I compares the cell parameters and spacegroups with those of anthophyllite from Chester. Thecell dimensions of anthophyllite, chesterite, and jim-thompsonite were determined by least-squares refine-ment with 84,95, and 98 measurements respectivelyfrom precision back-reflection Weissenberg films, us-ing the program LcI-sQ (Burnham, 1962) with sys-tematic correction terms for absorption, film shrink-age, and camera eccentricity errors. All data forchesterite and jimthompsonite were obtained fromthe same crystals used for X-ray intensity measure-ments by remounting the crystals to obtain a secondorientation. The clinojimthompsonite cell parameterswere refined by a least-squares method using twelvediffractions manually centered on a four-circle dif-
1001
Fig. l. 0-level a-axis precession photographs of enstatite (En),
anthophyl l i te (An), j imthompsoni te (Jt ) , and chester i te (Ch),
showing differences in D* length.
fractometer, while those of the unnamed mineralwere measured from precession photographs.
Space groups of the minerals listed in Table I weredetermined by examination of extinction criteria onlong-exposure precession films. The space group ofjimthompsonite (Pbca) is uniquely determined, butthe others are not; reasons for selecting one ofseveraldiffraction-equivalent space groups for each mineralare discussed in conjunction with the structural crys-tallography in a subsequent paper (Veblen and Burn-ham. 1978) .
Occurrence and associations
The new minerals all occur in the blackwall zonezof a metamorphosed ultramafic body that is exposedin the Carleton talc quarry near Chester, Vermont.The geology of the quarry has been described byGif lson (1927), Phillips and Hess (1936), and Chides-ter et al. (1951). The generalized zoning sequencefrom country rock to ultramafic body is: (l) mus-covite-quartz-garnet gneiss; (2) altered gneiss; (3)biotite and chlorite blackwall; (4) actinolite; (5) talc;(6) talc, magnesite, and serpentinite.
The material used in this study is from a singleblock of blackwall found on the quarry dump, al-though similar material has been found in place onthe northern wall of the flooded quarry (Richard
2 The term "blackwall zone" is commonly used to describe the
chlorite- or biotite-bearing metasomatic reaction zones that are
frequently found at the boundaries of metamorphosed ultramafic
bodies.
VEBLEN AND BURNHAM: NEW BIOPYRIBOLES
J inthonpsonlte
a = 1 8 . 6 2 6 3 ( 3 ) A
b = 27 .2303(6)A
c = 5 . 2 9 7 0 ( 3 ) A
v = 2686.6(2)A3
Pbca
z = 4 [ (Mg,Fe) r ' s112032(oH)4 ]
Ches terlte
a = 1 8 . 6 1 4 0 ( 3 ) A
b = 4 5 , 3 0 5 ( I ) A
c = 5 . 2 9 6 6 ( 3 ) A
v = 4 4 6 5 . 8 ( 3 ) A 3
A 2 , u
z = 4 [ (Mg,Fe) r rs120054(o t t )6 ]
Anthophyll lte
a = 1 8 . 5 8 6 3 ( 2 ) A
b = 1 8 . 0 6 4 9 ( 2 ) A
c = 5 . 2 8 9 5 ( 4 ) A
v = 1 7 7 6 . 0 ( 1 ) A 3
PM
z = 4 [ (Ms,Fe) 7S18O22(OH)2]
C11noj iEthoDpsonlte
a = 9 . 8 7 4 ( 4 ) A
b = 2 7 . 2 4 ( 3 ) L
c = 5 . 3 1 6 ( 3 ) A
B = 1 0 9 . 4 7 ( 3 ) '
v = 1 3 4 7 . ( 3 . ) A 3
c 2 l c
z = 2 [ (Mg,Fe)10s112032(oH)41
Unnaned mlneral
a = 9 . 8 6 7 A
b = 4 5 . 3 1 A
c = 5 .292A,
B = L09.7"
v = 2227.A3
A2ln, An, or 42
Table l. Cell parameters and space groups of low-calcium chainsilicates. Chester. Vermont
Orthorhomblc Monocllnlc
tremolite to anthophyllite. At Chester the anthophyl-lite is replaced by chesterite, the unnamed mineral,jimthompsonite, clinojimthompsonite, and fibroirstalc. The new minerals are thus part of a retrogradereaction sequence from anthophyllite to talc.
Habit, color, cleayages, and partings
The new minerals occur as intergrowths parallel to(010) in anthophyllite and cummingtonite, and formradiating sprays of prismatic crystals up to 5 cm long.The anthophyllite and the new minerals are trans-parent and have the same color, ranging from color-less to very light pinkish brown. All are colorless inthin section.
Cleavage angles were measured with a reflectiongoniometer on the single crystals used for X-ray in-tensity measurements. Chesterite possesses perfect{ l l0} c leavage intersect inBat44.To and 135.3o, whi lejimthompsonite has perfect {210} cleavage at 37.8oand 142.2". In addition, both minerals break along{100} and {010}, but these direct ions may be part ings,{100} resulting from breakage along fine monocliniclamellae and {010} from separation along lamellae ofanother orthorhombic pyribole. The cleavagesshould be valuable diagnostic properties in well-crys-tallized specimens, but some material from Chesterconsists of several minerals so finely intergrown thatthe cleavage is not apparent; these specimens areoften fibrous.
Structurally, the chesterite and jimthompsonitecleavages are analogous to the {210} cleavage of theorthopyroxenes and orthoamphiboles. Because themonoclinic polymorphs from Chester occur only aslamellae, their cleavages could not be observed, buton structural grounds one would predict a {120}cleavage for the unnamed mineral and a {110} cleav-age for clinojimthompsonite.
Optical properties
Optical data for jimthompsonite, chesterite, andanthophyllite from Chester were obtained by using aspindle stage (Wilcox, 1959). The jimthompsoniteand chesterite crystals were the same ones used for X-ray intensity measurement. The crystals were firstmounted with their b axes coincident with the spindleaxis, precise orientation being achieved by directlyremounting the crystals from their X-ray goniometer-head mounts. Optic axial angles (2V,) for sodiumlight were measured directly by spindle rotation, anddispersion was observed in blue-filtered light from anincandescent bulb. Principal indices of refractionwere then measured for sodium light by mixing index
Sanford, personal communication). The block maybe a section of a continuous blackwall zone, or it maybe part of an isolated pod. The simplified zoningsequence observed in this block is: (l) chlorite; (2)fibrous talc; (3) fibrous talc, jimthompsonite, clino-jimthompsonite, chesterite, the monoclinic analog ofchesterite, anthophyllite, and cummingtonite; (4)anthophyllite, cummingtonite, chesterite, the mono-clinic analog of chesterite, and actinolite; (5) acti-nolite and massive talc. Magnetite is found through-out as an accessory mineral, and cummingtonite hasbeen observed as lamellae in actinolite and an-thophyllite.
Actinolite and anthophyllite commonly occur asoriented intergrowths on planes near (100), attaininglengths of 5 cm. Three similar occurrences of tremo-lite and anthophyllite from the Gouverneur miningdistrict in New York have been interpreted by Ross etal. (1968) as representing retrograde alteration of
VEBLEN AND BURNHAM:
oils until a match was obtained (Becke line method);the index of the matched oil was measured with anAbb6 refractometer calibrated with oils matched ton(fluorite) and <.r(quartz). The crystals were then re-mounted with their c axes coincident with the spindleaxis, and the indices were checked. Optic axial anglescalculated from the indices of refraction agree withthe observed values of 2V, within the estimated preci-sion of the measurements, The consistency of thesemeasurements was further checked by measuring re-tardation ratios of the components of multiphasegrains in thin section with a Berek compensator.
Optical properties (Table 2) can be used to differ-entiate between anthophyllite, chesterite, and jim-thompsonite from the Chester wall zone, but com-par ison o f these da ta w i th the numerousdeterminations for anthophyllite given by Rabbitt(1948) indicates that optical methods are probablynot the best determinative means for distinguishingthese minerals at other localities. Anthophyllite andgedrite show a wide range of optical properties, andthe new minerals will probably show similar varia-tions with compositional changes.
Clinojimthompsonite occurs only as thin lamellaeand was not suitable for optical measurements. How-ever, the extinction angle NAc = 10o was measured.
The new minerals closely resemble amphiboles inthin section. In (001) sections the lower cleavage an-gles are revealed in some instances, but the cleavage isoften indistinct or fibrous. The minerals usually occurtogether in (010) intergrowths, however, leading to astriking appearance under crossed polars for crystalswith b near the plane of the section (Fig. 2a-c)s.Because pyribole exsolution features are usually re-stricted to planes near (100) and (001), there shouldbe little difficulty in recognizing these intergrowths;anthophyllite-gedrite exsolution on (010) could,however, cause confusion. When the intergrowthshave b inclined more steeply, but not normal, to thesection plane, they can take on a softly striped ap-pearance (like a multi-colored candy cane) in crossedpolars (Fig. 2d).
Portions of some grains possess interference colorsintermediate between those shown by anthophyllite,chesterite, and jimthompsonite, resulting in streaksparallel to (010) (Fig. 2a). These intermediate opticalproperties suggest that anthophyllite, chesterite, and
NEW BIOPYRIBOLES 1003
Table 2. Optical properties of anthophyllite, chesterite, andjimthompsonite from the Chester wall zone
Anthop[yll1te Chesterite Jirthonpsonite
c =a -
I -
2\I
L.6201 . 5301. 5l+r&
c
d = r . o f I
B = r .632Y = r.5l+o
Z = c
xNegatiYer>v, veak
d = f . o u )
B = r .626y = r .533X = a
Z = e2''I = 62o
xNegativer>v, veah
e A color photograph of such anScience, October 28, 197'l (Veblenalso shows the cleavage angles ofj imthompsonite.
intergrowth is on the cover ofet al., 1977). This photographanthophyllite, chesterite, and
Est i ra ted er ro rs : re f r . ind . : lO .OOO5, 2V: 12o
j imthompsonite are intergrown on a scale too fine tobe optically resolved, or that there are stil l morestructural ordering schemes or areas of disorderedchain sequence. Most grains, however, are opticallyuniform, and intergrowths usually exhibit optically-sharp boundaries between anthophyllite, chesterite,and jimthompsonite.
Chemical analysis
The results of electron microprobe analyses in-dicate that the major-element compositions of thenew minerals are intermediate between those of an-thophyllite and talc. Owing to the intergrown natureof the new minerals, pure samples could not be sepa-rated for water analysis.
Analyses (Table 3) were performed on an auto-mated MAC Model 5 electron microprobe. All speci-mens and standards were carbon-coated. Operatingvoltage was l5 kV, with a reference current of 300mA. A diopside-jadeite glass was used as the stan-dard for Si, Al, Mg, Ca, and Na; natural aenigmatite,ilmenite, forsterite, chromite, rutile, and orthoclasewere used for Fe, Mn, Ni, Cr, Ti, and K, respectively.Counting time was 20 seconds for both backgroundand peaks; counting was stopped when the number ofcounts exceeded 30,000. Compositions were cor-rected by the alpha-matrix method of Bence and Al-bee (1968), using alpha factors calculated as de-scribed by Albee and Ray (1970).
The microprobe analyses show that the Chesterpyriboles contain Mg, Fe, Mn, Ca, Al, Si, and Na;the elements Ni, Cr, Ti, and K are not present inmeasurable amounts. Energy scans performed withan Eonx 707A energy-dispersive analyzer on a Cam-eca MS46 microprobe confirmed the absence of sig-nificant amounts of unexpected elements. The jim-thompsonite analysis is the average of 1l analyzedpoints on the same crystal used for X-ray intensitymeasurement, plus three points from optically-identi-
1004 YEBLEN AND BIJRNHAM. NEW BIOPYRIBOLES
Fig. 2. Chester b iopyr iboles in th in sect ion. Scale bars represent 0.1 mm. (a) Sect ion paral le l to (001) o lgrain consist ing most ly ofanthophyl l i te. The anthophyl l i te is beginning to t ransform to chester i te, producing streaks paral le l to (010). The stra ight crack paral le l to(100) (arrowed) is probably a remanent "herr ingbone" twin p lane, indicat ing that th is grain was or ig inal ly a monocl in ic amphibolecrystal . Crossed polars. (b) A (010) intergrowth of anthophyl l i te and chester i te, wi th the b axis near the plane of the sect ion. Crossedpolars. (c) A sharp (010) intergrowth of anthophyl l i te, chester i te, and j imthompsoni te. The b axis is in the plane of the sect ion, but nocleavage is v is ib le. Crossed polars. (d) A (010) intergrowth of the pyr iboles wi th the b axis near the sect ion normal . Crossed polars.
f ied grains in polished thin section. The clinojim-thompsonite analysis is the average of four points onone of the lamellae in the crystal used for intensity-data collection; because the lamella width (2p) isnarrower than the volume activated by the electronbeam, this analysis is undoubtedly "diluted" by j im-thompsonite, which is the host mineral. The chester-ite analysis is the average of l4 points in thin section.In addition, anthophyll ite, foliated talc, and acti-nolite grains from the blackwall zone were analyzed,and the results given in Table 3 are averages of 17points from anthophyll ite, f ive from talc, and threefrom actinolite.
The average analyses of chesterite, j imthompson-ite, and clinojimthompsonite are consistent with therefined structures (Veblen and Burnham, 1978). The
ideal chemical formula for chesterite is M,rSi2oO,n(OH)6, and that for j imthompsonite and clinojim-thompsonite is MroSirrOrr(OH)n, where M refers toMg and Fe'+. The chesterite formula is thus the sumof the anthophyll ite formula, M?SisOrr(OH)r, and thejimthompsonite formula. Figure 3, a ternary plot ofoctahedral cations us. tetrahedral cations us. HOo.r,shows that the new pyriboles ideally are chemicallyintermediate between anthophyll ite and talc and arecoll inear with enstatite, anthophyll ite, and talc.
The analyses show that small amounts of Mn andCa substitute for Mg and Fe; the Mn may be respon-sible for the l ight pinkish color, observable in handspecimen, of the low-Ca pyriboles. Only clinojim-thompsonite appears to contain octahedral Al, butthe quality of the analyses may not be good enough
VEBLEN AND BURNHAM
to rule out the presence of some octahedral Al in theother minerals; small amounts of Al replace Si in thetetrahedral sites of all these pyriboles. If it is assumedthat the Na shown in the analyses is restricted to theA sites of the minerals (which are analogous to theamphibole A site), then these sites in the analyzedanthophyllite, chesterite, jimthompsonite, and clino-j imthompsonite are approximately 1,2/3,2 l /2, and2 percent filled respectively.
The recalculation program of Brady (1974), usedwith chemical assumptions analogous to those uti-lized by Stout (1972) for the amphiboles, shows thatFe3+ is not a major component of the pyriboles fromChester. When the averaged microprobe analyses ofTable 3 are normalized to the appropriate number ofoxygens, and when it is assumed that all Na is re-stricted to A sites, the resulting Fe3+/(Fe3++Fe2+)ratios are 0.04 in anthophyllite, 0.00 in chesterite,0.04 in jimthompsonite, and -0.05 in clinojim-thompsonite.
From the averaged analyses, there is no obviouschemical difference between jimthompsonite andclinojimthompsonite. The largest apparent differ-ences are in SiO, and AlrOr, but there is overlap inthe values of individual analysis points from the two
Table 3. Average electron microprobe analyses of biopyribolesfrom the Chester blackwall zone
Anth
NEW BIOPYRIBOLES 1005
Fig. 3. Chemistry of the Chester biopyriboles. The ideal
compositions, derived from crystal structures and averaged
electron microprobe analyses, are represented here on a molar
MO-SiOr-HOos ternary diagram, in which M represents divalent
octahedral cations (primarily Mg and Fe) and water is given as
HOo ".
In this representation, enstatite, anthophyllite, chesterite,jinithompsonite, clinojimthompsonite, and talc are composition-al ly col l inear.
minerals, and the clinojimthompsonite average wascalculated from only four analysis points; the SiO,difference and other differences are probably not sta-tistically meaningful. Because anthophyllite and cum-mingtonite can apparently overlap in composition(Rice el al., 1974), it is not surprising that jim-thompsonite and clinojimthompsonite can be compo-sitionally undifferentiable.
The individual analyses that yielded the averages ofTable 3 exhibit so,me scatter. The ratio (Si+Al)/(Si+Al*Mg*Fe*Mn*Ca) is the rat io of tetrahedral cat-ions to tetrahedral plus octahedral cations, assumingthat all Al is tetrahedral, that there is no tetrahedralFe3+, and that the Na is restricted to the A site. Aplot of this ratio for all the individual analyses (Fig.4) exhibits considerable overlap for anthophyllite,chesterite, jimthompsonite, and clinojimthompson-ite. It is not known how much of the scatter, if any, isdue to real compositional variations, and how muchis the result of analytical error. In spite of the scatter,however, it is clear that the new minerals are moresilica-rich than anthophyllite and less so than talc.Because the new pyriboles are intermediate reactionproducts between anthophyllite and talc, it is con-ceivable that some of the chemical scatter is real.Variations in the tetrahedral/(tetrahedral * octahe-dral) ratio could result from stuffing cations whichare normally octahedral into the A sites, or fromcreation ofoctahedral vacancies during the reactions.Submicroscopic lamellae of other biopyriboles couldalso contribute significantly to the analytical scatter.
It is well established (e.g. Papike and Cameron,1976) that Fe, Mn, and Ca are usually concentratedover Mg in the outer distorted M sites of pyriboles,
**Ihe uout of vater vas c€lculated. by assuing alf OH sites tobe f i l f ed { i th OHl- (Veblen @d Blmhu, 19?8 ).
1006 VEBLEN AND BURNHAM:
An rr l l l . r
Ch ., l f , . . .
J I
cj t 1 . . . .
Tc
At ' , '
low Co r, t l t .hl l . l . r l rr .r . . r '
t lo.58
Fig. 4. (Si+Al ) / (Si+Al+ Mg+ Fe* Mn*Ca) rat ios of indiv idualelectron microprobe analyses of anthophyllite, chesterite,jimthompsonite, clinojimthompsonite, talc, and actinolite fromChester. The number of analyses with a given ratio (rounded to0.001) is indicated by the height of the vertical line segments; theshortest lines represent one analysis. ldeal ratios, based onstructural formulas, are indicated by the vertical line segmentsbelow the horizontal lines. The line marked "low Ca" shows all theanalyses, excluding those of actinolite. The observed ratios are notevenly distributed about the ideal amphibole ratio.
while Mg is concentrated in the inner more regular Msites. It might be expected, then, that the ratio Mgl(Mg+Fe+Mn*Ca) would increase from anthophyl-lite to chesterite to jimthompsonite when these miner-als coexist, because the ratio ofregular to distorted Msites increases. Figure 5 shows the Mgl(Mg*FefMn*Ca) variations exhibited by the individual probeanalyses, and, even though there is significant scatter,the expected increase in Mg over other octahedralcations is clear.
In spite of radical structural differences, an-thophyllite, chesterite, and jimthompsonite haveideal chemical compositions that are very close toeach other. Given the scatter demonstrated in Figures4 and 5, whether it be from analytical errors or realcompositional variations, it would be very difficult todifferentiate these pyriboles unambiguously fromchemical methods alone. The bottom lines of Figures4 and 5 show the distribution of the tetrahedral/(tetrahedral * octahedral) and Mgl(Mgf Fe*Mn*Ca) ratios for all of the analyzed points, excludingthose for actinolite. These total distributions are simi-
NEW BIOPYRIBOLES
lar to the distribution that one would achieve byrandomly analyzing a mixture of these minerals, andbecause there are no obvious maxima in the distribu-tions an unwary analyst might easily explain them ascompositional variations in a single phase. The totaldistribution in Figure 4 is, however, clearly moresilica-rich than that to be expected from ideal an-thophyllite; that observation alone might alert thepetrologist to the presence of phases intermediatebetween the amphiboles and the micas.
On balance, chemical data should prove most valu-able when the minerals present have already beenidentified by other means. The ambiguity of the mi-croprobe data emphasizes that these new mineralsmust be defined primarily on structural rather thanchemical grounds. Similar phases with significantamounts of Al would be even harder to distinguishchemically, because the tetrahedral/octahedral ratiofor Al is determined from microprobe data by using astructural assumption.
Powder diffraction
Powder diffraction profiles for chesterite, jim-thompsonite, and clinojimthompsonite were calcu-lated with the computer program Powo5 (Clark etal., 1973), the same program used to calculate pat-terns for Geol. Soc. Am. Memoir 122 (Borg andSmith, 1969). Using refined atomic positions, iso-tropic temperature factors, and occupancies (Veblenand Burnham, 1978), the patterns were calculated forCuKa radiation, assuming peak half-widths of 0.1 l'20 at 40o 20 and a scan speed of 2" 23 per inch.Summaries of peaks with integrated intensitiesgreater than I percent of the largest intensity forchesterite, jimthompsonite, and clinojimthompsonite
An
chJI
cjl
Tc
low Co
Fig. 5. The Mgl(Mg+Fe+Mn*Ca) rat ios of individual electronmicroprobe analyses of biopyriboles from Chester. Presentation isthe same as for Fig. 4. There is a significant degree of scatter in theanalyses, but it can be seen that in going from anthophyllite tochesterite to jimthompsonite and clinojimthompsonite theminerals become increasingly magnesian.
t t t t lo .52 0.54 0.56
r t t t t t0.65 0.70 0?5 0 80 0.85 0 90
VEBLEN AND BURNHAM: NEW BIOPYRIBOLES 1007
dnthmpsonnc
Chcsr€ri'" Cilnojlmthompsonn.
Fig. 6. Calculated X-ray powder-diffraction profiles of anthophyllite (Finger, 1970), chesterite, jimthompsonite, andclinojimthompsonite. Patterns were calculated using refined atomic coordinates, isotropic temperature factors, and occupancies (Veblenand B urn ham, 1978 ). The patterns simulate random-orientation diffractometer traces with Cui(a radiation, a peak half-width of 0.1I' 20al 40o 20, and a scan speed of 2o per inch. The numbers beneath each trace are oN for CuKa.
are listed in Table 44. The powder diffraction patternsare plotted in Figure 6, which also shows anthophyl-lite for comparison. The summary tables are in thesame format as that used for Geol. Soc. Am. Memoir122: 2THETA and D give the 20 and interplanarspacing for reflection (HKL), and I(INT) and I(DS)give the uncorrected integrated intensity and in-tegrated intensity corrected for Debye-Scherrer ab-sorption, assuming that the peak can be resolved;PEAK and I(PK) refer to the peak positions andpeak heights of the pattern produced by summing allthe diffractions, after each has been given a Cauchyprofile with the input half-width.
a To obtain a copy of Table 4, order Document AM-78-078 fromthe Business Office, Mineralogical Society of America, 1909K Street, N.W., Washington, D.C. 20006. Please remit $1.00 inadvance for the microfiche.
Table 4 and Figure 6 suggest that X-ray powderdiffraction could be used to differentiate between thevarious biopyriboles from Chester. The peaks below15" 20 (CuKa) may be particularly useful, becausethey are very intense, but higher-angle peaks are re-quired to separate jimthompsonite from clinojim-thompsonite. The calculated powder-diffraction pro-files might be useful for identifying fine-grained runproducts of petrologic experiments if the phases arepresent in large enough amounts. It must be remem-bered that these calculated patterns assume randomorientation of the specimen, which will be difficult toachieve, given the prismatic or fibrous habit of all theminerals. Real intensities will differ from these calcu-lated ones because of preferred orientation.
Powder-diffraction photographs of a mixture ofthe Chester pyriboles estimated to contain 60 percentanthophyllite, 20 percent chesterite, and 20 percent
Mhillitc
l 008 VEBLEN AND BURNHAM: NEIII BIOPYRIBOLES
j imthompsonite are significantly less diagnostic thanthe powder calculations. A Debye-Scherrer f i lm wasunable to resolve the low-angle peaks, The only l ineson a Guinier photograph that could be unambig-uously assigned to minerals other than anthophyll itewere the (200), (220), (620), and (2. 13.l ) of chester-i te atd: 9.35,8.59,3.08, and2.78A respect ive ly andthe (210) and (610) of j imthompsoni te at 8.82 and3.104. Small lamellae of the new minerals in an-thophyll ite would be very diff icult to recognize bypower-diffraction techniques.
Summary: identification of the Chester pyriboles
Although differences among the low-calciumbiopyriboles from Chester have been demonstratedoptically, chemically, and by powder diffraction, it isl ikely that positive identif ication of the new mineralsat other localit ies wil l require single-crystal X-raytechniques. Optical study may prove diff icult, owingto the intergrown nature of the minerals, and theeffects of chemical substitutions on the optical prop-erties are not known. The intergrowths are distinctivein thin section, however, and the cleavages ofthe newminerals when observed in non-fibrous occurrencesmay be seen to intersect at angles lower than those ofamphiboles. Chemically, the new minerals are allvery close to anthophyll ite and cummingtonite, andelectron microprobe analyses show too much scatterto be definit ive. Powder diffraction is useful onlywhen the minerals are individually present in largeamounts.
The new minerals can easily be recognized by theirdistinctive X-ray precession photographs, which pro-vide both cell-parameter and symmetry information,and by electron diffraction patterns. Even the com-ponents of f ine multiphase intergrowths can be iden-tif ied by these techniques. Single-crystal methods are,in fact, the only means by which the lamellar mono-clinic phases can be recognized. Identif ication andstudy of these new minerals can best be accomplishedby using a combination of X-ray, optical, and chem-ical methods, and not restricting oneself to a singletechnique.
AcknowledgmentsWe wish to give special thanks to James B. Thompson for his
insight on many matters re lat ing to b iopyr iboles. Discussions wi thJames F. Hays, Cornel ius S. Hur lbut , Cl i f ford Frondel , John B.Brady, Richard Sanford, and Kenneth Shay were also helpful.Peter Buseck and John Armstrong of Arizona State Universityprovided energy-dispersive microprobe analysis. This paper repre-sents part of a Harvard Univers i ty Ph.D. Thesis (Veblen, 1976).Critical reviews by Cornelis Klein, James J. Papike, and Joseph V.
Smith are gratefully acknowledged. Financial support for this re-search was provided by NSF grant GA-41415 to Char les W. Burn-nam.
References
Albee, A. L. and L. Ray (1970) Correction factors for electronprobe microanalysis of silicates, oxides, carbonates, phosphates,and sulphates. Anal . Chem.,42, 1408-1414.
Bence, A. E. and A L. Albee (1968) Empirical correction factorsfor the electron microanalysis of silicates and oxides. J. Geol.,76, 382-403.
Borg, L Y. and D. K. Smith (1969) Calculated X-ray powderpatterns for silicate minerals. Geol. Soc. Am. Mem. 122.
Brady, J. B. (1974) Coexisting actinolite and hornblende fromwest-central New Hampshire. Am Mineral., 59, 529-535.
Burnham, C. W. (1962) Lattice constant refinement. Carnegie Inst.Wash Year Book, 61, 132-135.
Chidester , A. H. , M. P. Bi l l ings and W. M. Cady (1951) Talcinvestigations in Vermont preliminary report. U.S. Geol. Sun:.Circ. 95.
Clark, C. M., D. K Smith and G. G. Johnson (1973\ A Fortran IVProgram for Calculating X-ray Powder Pattents-Version 5.Rep. Dept. Geosciences, Pennsylvania State University.
Finger, L. W. (1970) Ref inement of the crystal s t ructure of ananthophyllite. Carnegie Inst. llash Year Book, 68, 283-288.
Gi l lson, J. L. (1921) Or ig in of the Vermont ta lc deposi ts, wi th adiscussion on the format ion of ta lc in general . Econ. Geol . ,22,246-28s.
Johannsen, A. ( l9 l l ) Petrographic terms for f ie ld use.J. Geol . , 19,3t7-322.
Papike, J. J . and M. Cameron (1976) Crystal chemistry of s i l icateminerals of geophysical interest. Reu. Geophys. Space Phys , 14,37-80.
Phi l l ips, A. H. and H. H. Hess (1936) Metamorphic d i f ferent iat ionat contacts between serpentinite and siliceous country rocks.Am. Mineral.. 21. 333-362.
Rabbi t t , J . C. (1948) A new study of the anthophyl l i te ser ies. lz .Mineral . . 33.263-323.
Rice, J. M., B. W. Evans and V. Trommsdorf f (1974) Widespreadoccurrence of magnesiocummingtonite in ultramafic schists,Cima di Gagnone, Ticino, Switzerland. Contrib. Mineral Pet-rol., 43, 245-251.
Ross, M., W. L. Smith and W. H. Ashton (1968) Tr ic l in ic ta lc andassociated amphiboles from Gouverneur mining district, NewYork. Am. Mineral . , 53,751-769.
Stout, J. H (1972) Phase petrology and mineral chemistry ofcoexisting amphiboles from Telemark, Norway. J. Petrol., 13,99 - t 45 .
Thompson, J. B. (1970) Geometr ical possib i l i t ies for amphibolestructures: model biopyriboles. Am. Mineral , 55,292-293.
- (1978) Biopyriboles and polysomatic series. Am. Mineral.,63,239-249.
Veblen, D. R. (1976) Triple- and Mixed-chain Biopyriboles fromChester, Vermont. Ph.D. Thesis, Harvard University, Cam-br idge, M assachusetts.
- and c. w. Burnham (1975) Tr ip le-chain biopyr iboles:newly discovered intermediate products of the retrograde an-thophyllite-talc transformation, Chester, Vermont (abstr.).
Trans. Am Geophys Union,56, 1076.- and - (1976) Biopyriboles from Chester, Vermont: the
first mixed-chain silicates (abstr.). Geol. Soc. Am. Absnacts with
VEBLEN AN,D BURNHAM: NEW BIOPYRIBOLES
Programs,8, 1153. Wilcox, R.E.(1959)Useof thespindlestagefordeterminationof- a1d - (1978) New biopyriboles from Chester, Ver- principal indices of refraction of crystal fragments. Am. Min-
monl II. The crystal chemistry of jimthompsonite, clinojim- eral.,44, 1272-1293.thompsonite, and chesterite, and the amphibole-mica reactioi.Am. Mineral.,63, in press.
-, P. R, Buseck and C. W. Burnham (1977) Asbestiformehain silicates: new minerals and structural groups. Science, /.98, Manuseript receiued, October /,3, 1977; accepted
