ECONOMIC

VOL. XXII

GEOLOGY
927 No. t

JANUARY-FEBRUARY,

TIN

ORES OF CHOCAYA,

BOLIVIAY

M. J. BUERGER AND JESSE L. MAURY.

INTRODUCTION.

TEE purpose of the present paperis to describe brieflya suiteof fin orescollected in the summer of i92I by Dr. WaldemarLindgren. In reference to the generalfeaturesof the occurrence of the ores,Dr. Lindgren hasbeen kind enough to furnishus with the followingstatement:
The tin and silver district of Chocaya is situated in southern Bolivia
in the Department of Nor Chichas,about IOO kilometers from the Ar-

gentineboundary.It is 45 kilometers southeast of Uyuni and connected by railroad with thiscenter. Topographically it is located in the Eastern Cordillerawhich here, however,is rather a tablelandthan a mountain range. The elevation of the minesis about4,22ometers, that of therailroad station 200meters less. The elevation of Uyuniis 3,70o
meters.

Fromthebroad playa andsaltflatsof Uyunitheroadrises gradually to a broad, low dividecovered with rhyolite flows;a sharp descent of

30ofeetcarries therailrfad down intothePilcomayo drainage andthe trackcontinues in a shallow, gradually deepening valley,bordered by rhyolite bluffs. Weathered slates of Paleozoic agebegin to appear in thevalley at theChocaya railroad station.Themines lie in anopen valley,thelowerpart of whichis deeply incised in therhyolite. At the camp, a couple of milesfrom the station, are concentrating works. The mines arehigher up among a landscape of yellow rhyolite

bluffs. There aretwoprincipal mines.Oneis mined bythe Compafiia

x Since the preparation of the manuscript,the general geologicrelations of

the deposit havereceived treatment at the hands of F. R. Koeberlln. See Eng.

& Min. four.-Press,April x7, 926, pp. 637 and 64x.

M. 1. BUERGER

AND

IESSE

L. MAURY.

Aramayo de Minas and is locally referred to as "Las Animas." The other, more northerly, is near the camp and operatedby the Compafiia Minera y Agricola de Oploca,and is referred to as the Oploca mine.

The two companies are operating on the same vein Which, however, differs materially in the two properties. The vein is containedin the rhyolite flows, is 2 to 3 kilometerslong, strikes N. 26 E., and dips steeply. The average width of the Las Animas vein is about 60 centimeters, that of the Oplocaperhaps2 meters. Both propertiesproducesilver and tin; Las Animas containingchiefly silver. The ore is rich, averaging about oo ouncesof silver and 7 per cent. of tin to the ton. The Oploca ore carries more cassiterite, but has
also one well-defined silver shoot.

The Las Animas mine is developed by a tunnel and a shaft 25 meters deep,probablydeepernow. The ore is beautifully banded,but the banding is not parallel to the walls, but rather surroundsfragmentsof rock containedin the vein. The richestlevel was probably25 metersbelow the tunnel. The cassiterite is recovered,but the tin in the stannite is

a complete loss.Thei2)ploca Mine isconnected withtheLasAnimas by

a I,OOO meter drift; the ore largely forms concentric bandsof pyrite and cassiterite aboutroundedfragmentsof partly replaced countryrock. The differencein mineralizationis explainedas probablyhaving 'beencaused by a split or branching of the vein between the two mines. North of this split there is very little silver in the ores.

The collectioncontainsspecimens taken from both the Las Animas and Oploca sections of the vein. While a somewhat more detaileddiscussion will be given presently,it will facilitate matters to include herea brief description of the suite. The specimens from the Oploca sectionof the vein include samples of the wall rock, of the main part of the vein and from an ore-shootcontaininghigh valuesin silver. Those from the main part of the vein exhibit bandingand are composed almost entirely of pyrite, quartz, and cassiterite. In two of these,the bav. ding is arrangedconcentrically aboutsmall cores,giving rise to an appearance (seeFig. ) similarto that of orbicularigneous rocks. Beyschlag, Vogt, and Krusch2 apply to this structure the terms "cockade" or "concentric." Another specimen(see
2 Beyschlag,F., Vogt, J. H. L., and Krusch, P.: ' The Deposits of the Useful
Minerals and Rocks." Translated by S. J. Truscott. lan & Co., London, x9x4. Vol. x, p. xx4. MacMil-

TIN

ORES OF CHOCAYA,

BOLIVIA.

Fig. 2) showsbandedstructurebut the arrangementis more layer-like than concentric. The specimens from the silver-oreshootare massivein structure.

Bandedstructureis alsodisplayed in the samples from the Las Animas sectionof the vein. In these, however, the bands are

coarser and somewhat lessregularthan thoseof the specimens from theOploca section (seeFig. 3), andsulphide minerals, such as stannite, and tetrahedrite are importantconstituents.
COUNTRY ROCK.

While unaltered samples of the countryrock were not available,it will be of interestto present briefly a description' of the
altered wall rock.

As seen in thehandspecimen, the rockis a grayporphyry with

numerous smallfeldspar phenocrysts and less numerous, but some-

what larger onesof quartz. Quartz seams cut across the rock in connection with disseminations of pyrite. In thin section, the rock appears highlyaltered and original minerals are almost lacking. The groundmass is quitesilicified. The feldspars are entirely replaced by a network of sericite foils.
Biotite has been bleachedto clear shredswith intercalatedbands

of shapeless rutile. In apparently advanced stages, it is replaced bya delicate, often radial, growth of a fibrous mineral (a serpentine ?). Zircon prisms, usually associated withthealtered biotite,
and quartz phenocrysts alone retain an unaltered condition.

Theserounded quartzphenocrysts contain, among otherinclusions, anisotropic mineral of lowindex, probably fluorite, which
occurs aslinearstrings of minutespecks andlargerindividuals of diamond-shaped crosssection. Pyrite crystalsare abundantly scattered throughoutthe rock, but appearsomewhat segregated about seamsof interlocking quartz crystals. Such seams containrare small crystals of cassiterite.

M. I. BUERGER
OPLOCA

.4ND

IESSE

L. M.4URY.

SPECIl{ENS.

The cockade structureis displayedin the form of irregularly oval shells,concentrically arrangedabout a core (Fig. I). The cockades vary in size from a quarterof an inchto two inches and consistof as many as ten distinctlydefinedshellsof alternating minerals. The outer shellis alwaysof quartz which, in the free cavitiesoften occurringbetweencockades, forms drusy coatings. The cores are, in general, of altered country rock in various stages of replacement by the vein minerals;the structureis, therefore, a concentric depositon a rubble of fragmentsof country rock. The vein minerals are pyrite, cassiterite,and quartz, in order of their relativequantities. They are distinctlycrystallized, the quartz and pyrite often exhibiting forms whichmay be identified. In size the crystalsvary from microscopic dimensions to
individuals one sixteenth of an inch across.

A thin section of part of a cockade is shownin Fig. 4- Some of the quartz was undoubtedlydepositedin the colloidal state. In suchgrains,extinctionof the crystallized aggregate is normal to regular stringsof inclusions, eachstring beingconcentric with the outer, usuallyrounded,boundaryof the grain. Cassiterite associated with this colloidalquartz would appear,on first ex-

amination,to be also a recrystallized gel. It has a fine texture with individual crystalsdeveloped normal to the bandedstructure. Alternating light and dark bandingsapparently are continuous acrosssuch crystals. Careful examinationwith high

magnifications, ontheother hand, reveals thatthedarker banding is dueto a moreintense pleochroism, whichordinarilyis from yellowish to colorless or somewhat greyish. These more pleochroiczones do not cut indiscriminately across individualcrystals, as in the caseof regularlyarrangedconcentric inclusions in colloidal quartz,but appear in slightlydifferentpositions in contiguous crystals. This would seem to suggestthat the cassiteritecrystals were deposited directly from solutions rather than recrystallizedfrom a gel. If this is the case,the variation in intensityof pleochroism may be attributedto a difference in iron contentof variousparts of the samecrystal,and in this we seea fluctuationin the characterof the depositingsolutions.

TIN

ORES OF CHOCAYA,

BOLIVIA.

}$

Fro. t. Portion of polished face of a specimenof cockadeore from the Oploca section of the vein. X .. c, cassiterite with quartz; p, pyrite; q, quartz druse; r, altered country rock cores. Fro. 2. Polished face of a banded specimenfrom the Oploca section of the vein. X .. Dark bands, cassiterite and quartz; light bands, pyrite; r, highly altered rock; s, thin stannite bands. Fro. 3- Polished face of a banded specimen from the Las Animas
section of the vein. Numerals refer to bands listed in text. Bands 6

and 7 are here represented only by the quartz druse q. Fro. 4. Part of a thin sectionof a cockade, showingcolloidalquartz.

Note centralcolloform masswith concentrically arrangedstringsof inclusions. The faint but distinct inclusions have been inked over.

Crossed nicols,X 67. Black, pyrite; dark, high relief, cassiterite;white and shades of gray, quartz.

M. I. BUERGER

AND IESSE

L. MAURY.

Thin sections show,amongthe rarer constituents of the cock-

ades, the presence' of irregular hematite flakes and bits of some

unidentified,colorless,isotropic mineral having an estimated refractive index of about 1.65. In determining the sequence of formation,the assumption was made that the growth of the cockades took placefrom the core
outward, and that, therefore, the earliestmineral in the inner band
was the earliest of the series.

The paragenesis of the vein mineralswas studiedon a large polished handspecimen (Fig. I) andseveral individual polished cockades.The cores display various stages of replacement by the vein minerals,somebeingalmostentirelyof pyrite, otherscontainingonlytracesof scattered pyriteand quartz. In mostcases
thesemineralsare arrangedin distinctveinlets,within which the pyrite occursas corrodedresidualstrings, while the evidently later quartz is continuous with the inner quartz-pyriteband. The first band of the cockades is essentially pyrite and quartz with subordinate amountsof cassiterite. The pyrite has been

extensively replaced by quartz,the latter often veiningit. The second bandis of cassiterite, whichmoldsitself uponcertainremaining euhedral faces of thepyriteandalso,to a minordegree, replaces both quartz andtpyrite. This established sequence: pyrite, quartz, cassiterite, is repeated in subsequent bandsand often three distinctcycles are represented.Specks of stannite occurin the second pyrite band. Fig. 2 shows another typeof specimen from the Oploca section of thevein,whichmaybe interpreted as a display of crustification on the vein wall. One-halfof the specimen consists of an intimatemixtureof fine-grained pyrite,quartz,and the remainsof the alteredcountryrock. The rest of the pieceis a series of comparatively narrowbands, varyingfrom threeeighths of an inchto a veryslightwidth. The minerals constituting these bands are pyrite, quartz, cassiterite,and stannite. The latter often carriesminuteblebsof chalcopyrite and tetrahedrite. The majority of the darkerbandsshownin the figureare composed of both quartz and cassiterite, either in separate, smallerbands,

TIN ORES OF CHOCAYA, BOLIVIA.

or intermixed. The lighter bands arethemetallic minerals. A noteworthy feature connecting thisdeposition withthatof the
cockade ore is foundin thealternation of two thin bands of stan-

nitewiththefirstthree of pyrite. In thecockades, stannite only appeared asoccasional pieces in thesecond pyrite band.
SPECIMENS FROM TI-IE SILVER-ORE SI-IOOT.

The specimens fromthesilver-ore shoot arecharacterized by

the absence of bandedstructureand the presence of numerous

smallrugs containing well crystallized minerals. Pyrite, stannite,tetmhedrite, quartz, galena, sphalerite, cassiterite, jamesonite, matildite(?) andthe unidentified mineral previously mentioned,
are represented in the ores.

In all specimens pyrite, quartz, and cassiterite are the oldest minerals. In polished sections, the other mineralsare seento be moldedaboutthe crystaloutlines of theseearlier ones,which usually showcorroded edges. Occasionally somereplacement is shown,particularlyof quartz by sphalerite. In specimens containing sphalerite and galena,quartz is most abundant, and'the later minerals display a curious concentric arrangement. Within rudely polygonal spacesin quartz are found sphalerite, stannite, and galena,the last forming the center. Examination with high magnifications shows that the stannite is replaced by the sphalerite and galena, the former sending minute anastomosing apophyses into it. In other parts of the same specimens galenareplaces sphalerite, apparently followingcleavage directions, leaving scallopedand veined residuals. The process indicated is oneof combined filling andreplacement.The stannitefirst deposited in interstices betweenquartz crystals. Subsequently sphalerite replaced both quartz and stannite along contacts, forming a ring betweenthesemineralscloselysimulating the outline of the original opening,and still later galena replaced the core of the stanniteforming the center. Tetrahedriteis very inconspicuous but high powers revealrare shreds
of it in galena.

In'most specimens, however,galenaand sphalerite are absent.

M. J. BUERGER

AND

JESSE

L. M./1URY.

In thesethe stannite contains moreabundant streaks of obviously

later tetrahedrite. Associated with these are somewhat lath-

like or streakedareasof a doubtful galena-graymineral of low hardness, which is negativeto all Davy-Farnham 3 reagents. Blowpipetests on piecesof stannitecontainingno other inclusions(except possiblysmall blebs of tetrahedrite) give indications of bismuth and silver. The presumptionis strong, therefore, that we are dealing with matildite, which would account, in part, at least, for the high silver values of the ore in this shoot. No native silver or other silver-bearingminerals
were found here.

The final phaseof mineral deposition is shownby tufts of acicularand bladedjamesonite supported on crystals of quartz, tetrahedrite, and corroded pyrite. Polished section studyproves that bothtetrahedrite and pyrite are well alongin the process of replacement by this late mineral.
THE LAS ANIMAS ORE.

The ore from this part of the vein exhibits a roughbanding

of minerals. The specimens will probablybear a somewhat detailed description. In two of these,the bandingfrom the wall

outward is (Fig. 3):

() A broad band.of cassiterite, quartz, and pyrite. The first two are intimately intermixedwhile the pyrite occursas scattered grains, often clearlyveinedand corroded by the cassiterite. In somecases the mineralsexhibit a very interesting structure. About apparently earlier pyrite grains are arranged delicate, alternating bandsof cassiterite and more pyrite in such a manneras to often reproduce in the entire structures, the outlinesof the originalnuclear grains. This effectof banding about a nucleus on a smallscale may be correlated with the banding in
the cockades. The identicalmineralogical, and similar structural

relations may be accepted as evidence of similarconditions, possiblycontemporaneity, of deposition of these phases.

a Davy, W. M., and Farnham, C. M.: "Microscopic Examination of the Ore
Minerals." McGraw-Hill, 92o.

TIN

ORES OF CHOCAYA,

BOLIVIA.

(2) A discontinuous bandof tetrahedrite with a little stannite,replacing some of the outerpyritegrains of the preceding
band.

(3) A thin bandof pyrite,with crystal outlines directed toward the following band.

(4) Minutelybanded quartzof almost chalcedonic appearance, containing darkerbandsof cassiterite whichconformwith the general colloform structure, molded uponthe euhedral pyrite
facesof band 3-

(5) A broad,r/therill-defined bandof practically purestannite, bordered by smallpyrite grains. This gradesinto: (6)An intimatemixtureof quartz,stannite, chalcopyrite and
tetrahedrite,grading into:

(7) A wide band of tetrahedrite, often terminatingin large crystals. In one specimen theselast two bandsare missingand
in their placeis drusy quartz.

While definiteage relationscan be established within an individual band, theseare complicated by the various generations of any particularmineraland by the indefinite boundaries of most
of the bands.

In polished sections, age'elations between various minerals

are, in general, the same, regardlessof the band studied. In tetrahedrite-stannite bands, quartz and pyrite are found to be the earliestminerals,displaying crystaloutlineswhich are often corroded next to the later minerals. Chalcopyrite sometimes replacespyrite. It is in turn veined on a very minute scaleby delicate threads of stannite andtetrahedrite. Somesmallpatches of chalcopyrite, however,seemto bear somerelationto irregular cracksand late quartz veinsin tetrahedrite and stannite, but

the greaterpart of it, at least,is earlier than theseminerals.

Although tetrahedrite and stannite usually occur in separate blockymasses, certain definitepinchingand swellingveinletsof tetrahedrite are found cutting stannite. These minerals are invariably associated with rich silver values which are to be ac-

counted for by numerous microscopic inclusions of nativesilver and matildite,which often occurtogether. The matildite has a

IO

M. J. BUERGER

AND

JESSE

L. M.4URY.

streaky,veinlike habit, apparentlyveining tetrahedriteand stannite, while the silver ordinarily forms shapeless blebs or more rarely isolatedlozenges. These are taken to be the latest minerals presentand appearto be hypogene. It is noteworthythat the presence of silver values is confinedto tetrahedrite-stannite bandsand is absentin pyrite-quartz-cassiterite bands. This is in

harmonywith, and explains,the absence of rich silver values in the cockadeore of Oploca. Viewed broadly, the general features suggestthat an overlapping deposition has ensued,without extensivereplacement of
one mineral by another. The sequence of minerals4 appearsto be: Pyrite, quartz and cassiterite--earliest; chalcopyrite; stannite;
tetrahedrite; matildite and silver.

In an early band of one Las Animas specimen, arsenopyrite occursin crystal clusters surroundingearlier corrodedmasses of pyrite. Later, somewhat massive bandsare rich in chalcopyrite, and the samesequence of completion of crystallization is observableas alreadynoted. The process suggested by examination of polishedsections is essentially a filling, by later minerals,of spaces existingbetweenearlier ones. This is borneout by the

fact that stannite,with a discontinuous tetrahedriteselvedge, forms a final, thin, regularcoatingcovering the entire massive band itself, apparently depositing there after havingfilled all the availableinterstitialspace.
CONCLUSIONS.

The deposit occurs in connection with rocksof acidcomposition and while it is not to be supposed that the rhyolite was directly connected with its formation, it seemshighly probablethat the vein was geneticallyconnected in depth with a parent chamber
of acid magma. In considering the mechanics of origin of this deposit,let us first eliminate the possibilityof its formation by meansof the

injectionof an "ore magma." Had suchan hypothetical "ore

4 Frankeite and cylindrite are also mentioned from Las Animas by A. W. Stelzner, "Die Silber-Zinnerzlagerst/itten Bolivas." Zeit. der det. leol. Gesell., 49, PP. 5-42, x897.

TIN

ORES OF CHOCAYA,

BOLIVIA.

II

magma" beeninjectedand congealed in situ, we shouldexpectto find evidence similar to that found in dike rocks. Doubly terminatedquartz and cassiterite crystals (early minerals) freely suspended in a matrix of other crystals,possiblyintergrowths, are 'to be looked for. These textures are notably absent. Instead,suchfeaturesas bandedand comb-structure, replacements

of bothore minerals and countryrock,quartzdeposited as a gel, and extensiveopen spaces containingwell-crystallized minerals,
are present, features which are inconsistent with the known or

supposed behavior of eithernaturalor "ore" magnaas. Indeed. thesefactsfind readyexplanation on no otherbasisthan that of deposition by mobile solutions, in part fillingthe openings along fissures or between a rubble of rock fragments. Granting that circulating solutions werethe mediaresponsible for deposition, it remains to offer a rationalexplanation for the outstanding featureof often-repeated cycles of similarmineralization in the various bands.

The occurrence of the vein in a Tertiary rhyoliteflow whose originalsurfacecouldnot have beenmuchhigherthan the present, 5andthepresence of quartz originally deposited asa gel, are indications of its near-surface originat moderate temperatures. Where the ore is deposited aboutcountryrock fragments the dontinuously interlocking natureof the bands and the

positions of the remaining, sometimes highly replaced andrelic, cores indicate that the fragments wereoriginally in contact and may,therefore, bereasonably interpreted as faultbreccia.
The generalized sequence of deposition seems, from minera-

graphic study,to have been:Pyrite earliest; arsenopyrite; quartz andcassiterite; chalcopyrite; stannite; sphalerite; galena; tetrahedrite; silver andmatildite; james0nite andlatequartz. It willbenoted thatthesequence agrees closely withthatgiven
by Davy. Chalcopyrite and stanniteappearhere in reversed
relation, however.
Lindgren, W.: Personal communication.

Davy, W. Myron; "Ore Deposition in the Bolivian Tin-silver Deposits," EcoN.

GEo,,., vol. 5, P. 49, 92o.

12

M. J. BUERGER

AND

JESSE

L. M.4URY.

It is interestingto note that the early positionof the fine-

grained cassiterite makes it an'undoubted primary mineral.

Arsenopyrite,while a mineral of somewhatpersistanttendencies,is most characteristically associated with depositsformed

at thehigher temperatures. This,whenpresent, assumes a very

early position in the bands together with the simple minerals quartzand pyrite. The morecomplicated tetrahedrite, matildite, and jamesonite are late minerals. In keepingwith the generalization that, other thingsbeingequal,chemistry at higher temperatures favors the formation of simpler compounds than at low temperatures, this relationseems to indicatea decrease in temperature of the depositing solutions. Acccrrding to this view, solutions, repeatedlyrefreshed,entered the fissure at intermediate temperatures and, due to closeproximity to the surface, were rapidly cooled, depositing,upon previously formed minerals, other minerals characteristic ot[ the changingconditions. The pressure,being essentiallyhydrostatic, was not a controlling factor in the changing conditions, and its smallmagnitudeurged no extensivereplacement, favoring, rather, open spacedeposition. Quartz, delicatelybandedand well crystallized,although
often of almost chalcedonicappearance,represents the final tages of magmatic emanations,coveringall free surfaces.
The variation in the nature of the ores of different sections of

the vein is a difference more of degree than of kind. The Oploca ore containsa little stannite,so characteristic of the Las Animas ore, while the latter contains, in early bands, pyrite, quartz, and cassiterite, which form the bulk of the Oplocamin-

erals. The higher temperatureindicationsof, the Oploca ore maybe assigned to a lateralcomponent of migrationof the depositingsolutions, whichwouldfavor the deposition of the leastsoluble, hightemperature, minerals at theOploca section of thevein; while the lower temperature minerals wouldbe deposited almost exclusively in the coolerLas Animasportion.

TIN

ORES OF CHOCAYA,
SUMMARY.

BOLIVIA.

I3

A few generalobservations may be made in summarizing the mineralization indicated by the suite. The vein occurs in a fault fissurein a rhyolite flow; it .containslarge quantitiesof two commonly persistent minerals,pyrite and quartz. The latter is here sometimes of colloidalorigin. Tourmaline and topaz are entirely absent. Bandedstructureand small vugs are the rule. The mineralization of any onepart of the vein appears to remain essentially constant throughout the entireperiodof deposition, but
between different sections of the vein there is a variance in the

natureof the ores. Cassiterite alwaysappears closely associated with quartz,although the converse is not necessarily true. The laststage of themlnerogenetic periodis,.in general, quartz,forming distinctly crystalline crusts on all free surfaces. The evidence points to the formation of the deposit by deposition from mobilesolutions. The occurrence of similar mineralogicalcharacter in later bandsis interpreted as meaning that freshsolutions had entered and, on cooling, deposited minerals

characteristic of thechanging conditions. The typeof mineralizationcontrasts strongly with that of the hypothermal, replacement tourmalinedeposits of Caracoles, described recentlyby Lindgren. 7
MASS.INST. TECHNOLOGY, CAMBRIDGE, MASS.

7Lindgren, W.,: "Replacement in the Tin-bearing Veinsof Caracoles, Bolivia,"

Ecoa. GEOL.,vol. z, pp. 35--I44, 9z6.