(Hypogene) Sulphate Minerals in Ore Deposits2

WITH WHICH IS INCORPORATED

THE AMERICAN GEOLOGIST

VOL. XIV DECEMBER, I9I 9 No. 8

PRIMARY (HYPOGENE) SULPHATE MINERALS IN ORE DEPOSITS2

B. S. BUTLER.

CONTENTS. PAOE. Introduction ...... Sulphate Minerals ...... Character of Ore Solutions as Indicated by Volcanic Emanations ...... 590 Conditions of Format'ion of Sulphur Trioxide ...... 593 Formation of Sulphuric Acid in Nature ...... 594 Formation of Primary Sulphate Minerals ...... 596 Summary ...... 6o8 Acknowledgment ...... -...... 609

INTRODUCTION.

In studyingore depositsthe writer has severaltimes been con.- fronted with the problemof accountingfor the formation of sulphateminerals that are apparentlyprimary. These minerals occur at placeswhere no evidencecan be found of their formation by surface oxidation. Either they were formed directly 'by igneousemanations or, if they were formed by surfaceagencies, those agenciesmust have effectedthe depositionof other associatedminerals that would ordinarily •beregarded as of hypogene origin. An attemptwill be madeto analyzethis problemby start- ing with the belief, 'basedon deductionsfrom field observations, that the sulphatesin certaindeposits have •een formed directlyby x Published by permission of the Director, United State Geological Survey. 58x 5.82 B. S. BUTLER. igneousemanations. To avoidcomplications the occurrenceof the mineralsin depositsthat are not generallyregarded as associated with igneousrock is discussedonly incidentally. It is recognized that nearly all the sulphatesthat occur as apparently primary mineralsin ore depositsassociated with igneousrock have also been formed where they are not associatedwith igneous rocks. It is likewise recognizedthat in the formations of at least some of the primary sulphatesin ore depositsthat are associatedwith igneousrocks surface agenciesmay have played a part. The general conclusionreached is that some of the metals which at high temperatureare combinedwith oxygen,as the temperature is reduced,give up oxygen,and 'boththe metalsand the oxygen combinewith sulphur, producingsulphides of the metals and oxides of sulphur. At suitabletemperatures the sulphitesand sulphatesare doubtlessformed and the lesssoluble sulphates are deposited. Moreover, it is believedthat certain conditionslead to the formation of free sulphuric acid which, on reaction with potassium-aluminumrocks, forms alunite.

SULPHATE MINERALS.

Brief summaries.of the occurrenceof the principal primary sulphateminerals in igneousrocks or of the sulphateminerals that are apparentlyprimary in veins which are believedto be associatedwith igneousrocks are given below. No attempt is made to includein this summaryoccurrence of theseminerals in other relations,or to give a completelist of localities. Hauynite and Noselite.--tlauynite, Na2Ca(NaSO4' A1) A12Sia- O• and noselite,Na4(NaSO4'A1)'Al•SiaO1• are primary minerals in certain rather uncommontypes of eruptive rocks such as thoseof the CrippleCreek volcano. It is noteworthythat these minerals are characteristiccomponents of volcanic rocks,whereas the closelyrelated chlorideminerals are componentsof plutonic or deep-seatedrocks. Barite.--Barite, BaSO4_,is perhapsthe most commonprimary sulphatevein mineral. It occursusually if not invariably in PRIM.4RI / SULPH.4TE MINER.4LS IN ORE DEPOSITS. 583

'depositsformed at moderateand low temperature2 or amongthe laterminerals of depositsthat containminerals deposited at high temperature. Barite also occurswhere it is not known to be relatedto igneousrocks. It is difficultto determineeven approximately the temperature at which any mineral depositwas formed, but Lindgrena has grouped the minerals in such deposits.in three classes--those formedat low temperature(200 ø C. or less,perhaps much less), at moderatetemperatures (t75 ø to 3ooøC.), and at high temperatures(300 ø to 575ø C.). Contact depositsmay have been formed at much higher temperaturesnear that of magmas,which range from 8ooø to •4ooø C. ,'Inhydrite.--Anhydrite,CaSO4, has only recentlybeen recognized as a primary vein mineral and its associationsindicate deposition at moderate temperature. Anhydrite has the unusual propertyof decreasingin solubilitywith increaseof temperature, beingbut slightlysoluble at 2ooø C.4 It might be supposedthat anhydritewould form most abundantly in depositsthat replacedlimestone, but the recordedoccur- rencesdo not justify this supposition. Probably the ,abundant ca•'bondioxide that must necessarilybe presentin solutionsthat are replacing limestone inhibits the precipitation of calcium sulphate. Lindgren5 first observedanhydrite as a vein-formingmineral in the Cactus mine, Utah, where it occurs in a vein in monzonite associatedwith tourmaline,hematite, pyrite, chalcopyrite,barite and siderite. Anhydrite, barite and siderite were among the latestminerals to form. Lindgrengave the followingsuggestion' as to the origin of anhydrite: •-Emmons, W. I-I., "A Genetic Classification of Minerals," Ecoa. Gvox.., Vol. 3, p. 6•8, x9o8. a Lindgren, Waldemar, "Mineral Deposits," McGraw Hill Publishing Co., New York, x9x3. 4 Melcher, A. C.,/lm. Chem. Soc. Jour., Vol. 32, pp. 50-66, x9•o. 5Lindgren, Wald.emar, "New Occurrence of Willemite and Anhydrite," Science,.new ser., Vol. 28, p. 933, x9o8. 584 B. S. BUTLER.

It is suggestedas a possibilitythat during the later part of mineralization the anhydrite was precipitated by a reaction between ascending solutionsof sodiumsulphate and descendingsolutions containing calcium carbonate. The writerø acceptedthis interpretation,but with reservations and doubtsthat are amongthe motiveswhich led to the prepara- tion of this paper. Anhydriteoccurs in the Bully Hill district,Calif., in lodesin alaskite porphyry and is associatedwith pyrite, chalcopyrite, sphaleriteand barite. Graton7 apparentlyregards it as having beendeposited directly from ascendingsolutions. Boyle8 alsoregards the gypsum and anhydriteof the Bully Hill district as of deep-seatedorigin, supposingthat the calciumwas derived from the limestonethrough which the solutionspassed. Anhydriteis reportedfrom the Cobredistrict, Santiago, Cuba. The depositsare in andesitetuff cut by dikesof andesite. The associatedminerals are pyrite, chalcop.yriteand quartz? Geijer1ø describes anhydrite in depositsin Swedenassociated

with tremolite, galena,chalcopyrite and pyrite. ß Hewett,TM and Miller and Singewaldx2 have describedthe re- markabledeposits of the Minasragra vanadiummine of Peru. Hewett is quotedby Miller and Singewaldas stating that the Veta Madre, which is a mixture of earthy material, disseminated sulphideof vanadium,and anhydrite,the last largely alteredto gypsumfor I2o feet below the surface,represents shale that 'has 6 Butler, B. S., "Geologyand Ore Depositsof the San Franciscoand Adja- cent Districts, Utah," U.S. Geol. Survey Prof. Paper 80, p. 124, I913. 7 Graton, L. C., "The Occurrenceof Copper in Shasta County, Calif.," U.S. Geol. Survey Bull. 43o, p. IOO, I9m. s Boyle, A. C., Jr., "Geology and Ore Depositsof the Bully Hill Mining District, Calif.," Am. Inst. Min. Eng. Trans., Vol. 48, p. III, I915. • Emerson, E. H., "Geologia de las minas," Bol. de Minas, Cuba, No. 4, PP. 47-52, I918. •o Geyer, Per, Falutraktens berggrund ack malmfyndighet'er, Geol. Survey of Sweden. Aorsbok, I916, p. I56. • Am. Inst. Min. Eng. Trans., Vol. 40, pp. 274-299, 19o9. • Miller, B. L., and Singewald, J. T., "The Mineral Deposits of South America," pp. 487-49x, McGraw-Hill Book Co., Inc., New York, I919. PRIM•IRY SULPH•ITE MINERSILS IN ORE DEPOSITS. 585

'beenmore or lesssaturated by sulphideof vanadiumand replaced by anhydrite. Bastin•a has describedprimary anhydriteand gypsumin the Bradencopper deposits of Chile. The rocksand ores of the dis- . trict recorda complicatedseries of igneousevents, including ex- trusions,intrusions and three distinctperiods of mineralization. The first mineralization producedextensive tourmalinization with the depositionof relativelysmall amountso.f pyrite and chalcopyrite.The mineralsof the secondperiod were mainly quartz, pyrite and chalcopyritewith small amountsof black tourmaline and a little biotite.

ß The metallic minera!sof the third period includepyrite, chalcopyrite, bornire, galena, sphalerite, molybdenite,tennantite, enargiteand hfibnerite,and the gangueminerals. siderite, rhodochrosite,calcite, anhydrite, gypsum and barite. The mineralization of the third periodis thoughtby Bastin to have taken place at lower temperaturethan the earlier periodsand was character- ized by solutionof tourmalineas contrastedwith depositionof that mineral in the first two periods. Anhydrite is also presentin the copperdeposits of Cuka- Dulkan at Bor, Serbia24 It is regardedby Lazarevic,however, as secondary. Anhydrite occursmost abundantlywhere it is not associated with either igneousrocks or veins. Gypsum.--Gypsumis a commonmineral in ore depositsbut has doubtlessusually been formed'by the alteration of anhydrite or by reaction'between solutions of sulphuricacid (proc[ucedby the oxidationof sulphides)and calcium-bearingminerals. As already noted,Bastin considers gypsum in the Bradenmine as primary. Adolph Knopf has kindly furnishedthe following note on the occurrenceof gypsumat the Utica mine,Calif. The ore on the 2,•00-foot level of the Utica mine, on the Mother Lode at Angels, Calif.,--a low-grade ore averaging $2 in gold--consists xaBastin, E. S., private report based on detailed study. X4Lazarevic, M., "Die enargit-covellin-lagerst•itten yon Cuka-Dulkan bei Bor in Ost-Serbien," Zeitschr. prakt. Geolo•7ie,Vol. 20, p. 337, •9•2. 586 B. S. BUTLER. of quartz and subordinatedolomite, gypsum, and, as shown under the microscope,albite. The only sulphidepresent i.s someextremely fine- grained galena disseminatedin small patches. The vein, which is ver- tical, is inclosedin amphi'boliteschist, and the vein and countryrock are so imperviousand dry tha.t the mine workings are dusty, althoughthe mine above the 9oo-foot level makes large quantities of water. The gypsum is intimately intergrown with the quartz, and this fact together with its occurrence so far below the zone of oxidation and the obvious imperviousnessof the vein to descendingwaters, suggestthat the gypsum is a primary (hypogene) constituentof the ore. Under the micro- scopethe gypsumis seen to be intergrown with quartz in patterns some- what like micrographicintergrowths, and this feature possibly corrob- orates the evidenceof its primary origin. Celestite.--Celestite,like barite and anhydrite,occurs in deposits formed at intermediate to low temperature and also at many placeswhere it is not associatedwith igneousrocks. •llunite.--Alunite has the chemicalformula K20' 3A12Oa'4SOa ß6H20 in which Na may replace K in varying proportions. Alunite is perhapsthe most abundantand widely distributedsul- phate mineral that is associatedwith altered volcanic rocks. It occursalso as a secondary(supergene) .mineral in the oxidized zoneof ore deposits. Its genesishas been variously interpreted by differentgeologists, doubtless because it hasbeen formed in vari- ousways. The occurrencesof the mineral havebeen summarized by RansomeTM and later ,byButler and Gale?ø Perhaps the best known deposit worked for alunite is that at Tolfa, Italy, where the alunite occursin trachyte and is said to giv.e placein depthto pyritic trachy'te. Concerningthe formation of the alunite De Launay says Alunite is, in my opinion,a product of the decompositionof feldspar similar to kaolin, which is worked in the same region, and often from the sameveins, and, like this kaolin, is boundto disappearin depth. The • Ransome, F. L., "The Geology and Ore Deposits of Goldfield, Nev.," U S. Geol. Survey Prof. Paper 66, pp. x89-•95, •9o9. x6Butler, B. S., and Gale, H. S., "Alunite, a Newly Discovered Deposit near Marysvale, Utah," U.S. Geol. Survey Bull. 5xx, •9•2. x7Translation by Butler, B. S., and Gale, H. S., op. cit., p. 52, from De Launay, L., "La metallogeniede l'Italie," Cornpt.Rend. Tenth Internat. Geol. Con#., Mexico, pt. x, x9o7. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 587 theory which was formerly held is somewhatdifferent. It was thought that the sulphur vapors of solfataric kinds circulatedin the fissuresof trachyteand attackeddirectly in depththe feldsparsof {he latter, and a relation was supposedto exist betweenthese different phenomena and the trachyteitself. I believe,on the contrary,.that there are two entirely distinctphases in •thephenomenon--first, a vein deposit clearly delimited, of pyritic trachyte corresponding,perhaps, to the veins of a trachyte par.ticularly feldspathicand at the sametime pyritic like the granulites of Berzowsk(Oural); second,penetration by superficialwaters of the feldspathicrocks, producing, where pyrite fails, the'ordinary forms of altered feldspars•that is to say, kaolin,--but where, on the other hand, pyrite furnished sulphuric acid, crystalline alunite. Lindgren•s and othershave described the occurrenceof alunite formedby the actionof solutionsthat containedsulphuric acid, which were derived from oxidizing sulphides on potassium aluminum silicates. In the United States there are numerousdeposits of alunite which are believedto have been formed and others that are per- Imps now forming 'by the action of hot sulphurouswaters on potassium-alumniumrocks. Largebodies of volcanicrocks so altered are composedchiefly of quartz, alunite and pyrite. The calcium,magnesium, and sod•iumwere largely removedfrom the original rock, but iron appearsto have been convertedto pyrite at the same time that alunite was formed. In discussingthe origin of the gold depositsat Goldfield, Nev., RansomeTM postulates: that the ore constituentswere brought up in hot solutionscharged with hydrogen sulphide, a little carbon dioxide, and probably also with some alkali sulphides; that the hydrogen sulphide was oxidized at and near the surface to sulphuric acid, which percolateddown through the warm rocks to mingle with the uprising currents carrying sulphydric acid. For the depositsnear Marysvale, Utah, which occur as veins in effusiverocks and have been developedfor •he alunite, Butler and Gale say:20 •8 Lindgren, Waldemar, "The Copper Deposits of the Clifton-lVforenci Dis- trict, Ariz.," U.S. Geol. Survey Prof. Paper 43, pp. I•9, I2O, I9O5. •9 Ransome, F. L., op. cit., p. I93. ,0 Butler, B. S., and Gale, I-I. S., op. cit., p. 36. 588 B. S. BUTLER.

The evidencein the Marysvale district, however, indicatesthat the materialsconstituting ,the veinswere depositedby ascendingsolutions and that thesesolutions brought in the constituentsof the alunite. At just what stage,the sulphuric acid may have been formed can not now be positivelystated, but it seemsmost natural to supposethat it was a part of the original solutionsand that ..thepotassium and aluminumwere in part original in the solutionand in part dissolvedfrom the walls of the fissureat greater depth. The veinsin this districtare nearlypure alunitebut the altered wall rockis mainlyalunite quartz and pyrite. Concerningalunit- ized rocksat Rico Mountains,Colo., Crosswrites The alteration of the porphyry of Calico Peak into a rock consisting largely of aluni•te,a hydroussulphate of alumina and the alkalies.... can be explainedonly as the result of the attack of sulphurousagents, and from the circumstances of occurrence there can be no doubt that the action is to be attributed to solfa•taricemanations of the Rico eruptive center in the period of waning igneousactivity. Larsen has dfiscribed several areas of alunitized rock in Colo- rado.22 The altered rocks consistessentially of quartz, alunite and pyrite, one analysisgiving: quartz, 69 per cent.; alunite, 29 per cent.,and pyrite, 2 per cent. Concerningthe genesisof these depositsLarsen writes The evidence suggestshot ascending solutions as the cause of the alunitization. The field relationspoint stronglyto deep-seatedhot sulphuric acid solutionswithout the aid of the surface agents. However, in view of the fact that geologistsdo not generally admit the presence of such solutions,the evidence in the present case is not sufficient to justify the assumptionof sucha coursefor the alunitizationin the San Cristobal quadrangle. The alternative source is the mingling of hot ascending solutions or gases carrying H_•S and of surface oxidizing waters. Clap.p•4 regardsthe depositof Kynquot Sound,British Colum- 2xCross, Whitman, and Spencer, A. C., "Geology of the Rico Mountains, Colo.," U.S. Geol. Survey Twenty-first Ann. Rept., pt. 2, pp. 92-94, I9OO. 2• Larsen, S., "Alunite in the San Cristobal Quadrangle, Colo.," U.S. Geol. Survey Bull. 53o, pp. I79-I83. xaLarsen, E. S., op. cit., p. I83. 24Clapp, C. H., "Alunite and Pyrophyllite in Triassic and Jurassic Voi- canics at Kynquot Sound, British Columbia," Ecoa. Gv.ox.., Vol. IO, pp. 7o-88, 1915. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 589 bia, as formedby "hot ascendingsolutions of volcanicorigin and that at least these solutionscausing alunitization carried free sulphuricacid." A. L. Day and E. T. Allen have studiedthe hot springsof Mount Lassen,Calif., where alunite and pyrite are being depositedin the acid waters of the springs? Descriptionsof occurrencesof alunitecould be multiplied,but the aboveseem sufficient to give the generalmodes of occurrence. Hinsdalite.--Hinsdalite,2PbO. 3A12Oa-2 SOa' P20,' 6H20, has beendescribed by Larsen26 and by Irving and Bancroft27 from the GoldenFleece mine near Lake City, Colo. It is a primarymineral in a vein in volcanicrocks associatedwith pyrite, tetrahedrite, galena, pyrargyrite, quartz, rhodochrositeand barite. Barite is abundantalso in neighboringdeposits. The depositsare regarded by Irving and Bancroft as havingbeen formed at shallowto moderate depth. Creedite.--Larsen28 describes creedire, CaSO4-2CaF2'2A1 (F, OH)a.2H20 , as occurringwith barite and fluorite in a vein in Tertiary lavas near Wagon Wheel Gap, Colo. Larsen does not statewhether he regardsit as a primary or secondarymineral in the vein. Thaumaxite.--Theunusual mineral thaumasite,3CaO-SiOo•- SOa'CO2' •5 1-t20, occursin veinswhich cut contactaltered lime- stonein the Old Hickory mine, BeaverCounty, Utah. It is regarded by Butler2ø as forming under conditionssimilar to those favorable to the formation of zeolites. The best known localities of this. mineral in the United Sta'tes are at West Paterson and 25Allen, E. T., personal communication. 26Larsen, E. S., and Schaller, W. T., "Hinsdalite, a New Mineral," Am. ]our. Sc•., 4th ser., Vol. 3•, pp. 251-25.5,I9•I. 2, Irving, J. D., and Bancroft, Howland, "Geology and Ore Depositsnear Lake City, Colo.," U.S. Geol. Survey Bull. 478, pp. 54-55, •8 Larsen, E. S., and Wells, R. C., "Some Minerals from the Fluorite-barite Vein near Wagon Wheel Gap, Colo.,"Proc. Nat. Acad. Sci., Vol. 2, pp. 362- 364, I9•6. 2• Butler, B. S., "Geologyand Ore Depositsof the San Franciscoand Adja- cent Districts,Utah," U.S. Geol. Survey Prof. Paper 8o, p. •o4, •9•3. 590 B. S. BUTLER. other localitiesin New Jersey,where it occursin trap associated with zeolites. Wilkeite.NWilkei'te,•ø 3Caa ( P204)' 3Ca2SiO4' 3 CaSO•-CaCOa ßCaO, occurs in contact altered limestone associated with wollas- tonite and garnet near Riverside,Calif. Scapolite.--Scapolitesoccur as productsof contactmOtamor- phisinand as alterationproducts of igneousrocks. Someof the scapolitescontain the sulphateradicle. a• Svonber#ite.--Svonbergiteis apparentlya rather rare primary mineral in ore deposits.•

CHARACTER OF ORE SOLUTIONS AS INDICATED BY VOLCANIC

EMANATIONS.

The data concerningthe compositionof volcanic emanations have beensummarized by many writers. A rather completeout- line andbibliography has been prepared by Clarke,aa who briefly summarized the results as follows: That the volcanic gases appear in a certain regular order has been shownby the various researchesupon their composition,and especially by the labors of Deville and Leblanc. What, now, in the light of all the evidence,is that order, and what do the chemicalchanges mean? First. The gasesissue from an active crater at so high a temperature that they are practicallydry. They contain superheatedsteam, hydrogen, carbon monoxide, methane, the vapor of metallic chlorides, and other substancesof minor importance. Oxygen may be present in them, with some nitrogen, argon, sulphur vapor, and gaseouscompounds of fluorine. 3oEakle, A. S., and Rogers,A. F., "Wilkeite, a New Mineral of the Apatite Group, and Okenite, its Alt'eration Product from Southern California," •/rn. Your. Sci., 4th ser., Vol. 37, p. 262, x914. axBorgstrSm, L. H., "Die skapolithlagersfiitteyon Laurinkari," Corn.g•ol. Finland, Bullß 4I, P. :•3, I913. Brauns,R., "SkapolithfiihrendeAuswfirflinge aus dem Laacher Seigebiet," Neuss Yahrb., Beilage Band 39, P. •9, x9•4. a2Lacroix, A., "Pyritiferous Depositsat the Contactof Graniteat Chezeuil, ßSaone-et-Loire and its MetamorphicRocks," Bull. Soc.Franc. Min., 4•, •4-2•, •9•8. aa Clarke, F. W., "The Data of Geochemistry," U.S. Geol. Survey Bull. 6•6, pp. 260-290, x9x6. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 59 1

Second. The hydrogenburns to form morewater vapor,and the car- bongases oxidize to carbondioxide. From the sulphur,sulphur dioxide is produced.The steamreacts upon a part of the metallicchlorides, generateshydrochloric acid, and so acid fumarolesmake their ap- pearance. Third. The acidgases of the secondp.hase force their way through crevicesin the lava and the adjacent rocks, and their acid contentsare consumedin effectingvarious pneumatolytic reactions. The rocksare corroded,and where sulphides occur hydrogen sulphide is set free. If carbonaterocks are encountered,carbon dioxide is also liberated. Fourth. Only steamwi,th carbon dioxide remains, and even the latter compound soon disappears. Thisseems to bethe general course of ever•ts,although it is modified in detailsby localpeculiarities. All of thesubstances' enumera,ted in the listsof gasesand sublimates given in the earlierportions of this chapter maytake part in the reactions,but theydo not seriouslyaffect the larger processeswhich have just beendescribed. The orderis essentiallythat laid downby Deville and.Le'blanc, except that the early evolutionof hydrogenand carbonicoxide is taken into account.The currentof eventsmay be disturbed,so to speak,by ripplesand eddies--thatis, by subsidiaryand reversedreactions--bug its main courseseems to be clearly indicated? Harker states :35 Different types of solfataric action might be distinguished,and these are in somedegree characteristic of differentkinds of lavas. But it is also to be remarked that different volatile constituents,may figure promi- nentlyat a givenvolcanic center at differentstages in declineof aotivity. This is in part a matter of direct observation;for it has frequentlybeen remarked that only the hoetestfumaroles emi.t hydrochloricand hydro- fluoric acids, while sulphurousand hydrosulphuricacids are connected with lower temperaturesof emission,and water and carbonicacid with the lowest temperatures. Some indicationsof a like sequencehave alreadybeen noted in pneumatolysisunder platonic conditions. a4For a summary of our knowledge concerningthe magmatic gases pre- vious to the work of Brun and Chamberlin, see Lincoln, F. C., Eco•v. GEo,.., Vol. 2, p. 258, I9O7. Lincoln gives a good table of analysesand proposesa classification of the volcanic exhalations. For a theoretical discussion relative to "gas mineralizers"in magmassee Niggli, P., Zeitschr.anorg. Chemic, Vol. 75, P. 161, 1912,and Vol. 77, P. 321, 1912. Also Centralbl.Mineralogle, 1912, p. 321; and Geol. Rundschau,Vol. 3, P. 472, 1912. a*Harker, Alfred, "The Natural History of Igneous Rocks," p. 307, Methuen & Co., London, I9O9. 592 B. S. BUTLER. The volcanicemandtions contain c&rbon, hydrogen, sulphur, oxygen, nitrogen and minor constituentssuch as chlorine, fluorine and metals in varying proportionsand combinationsdepending proba'blyboth on the original characterand the temperatureof the gases. Day and Shepherd3ø collectedgases from Kilauea with great care to avoid contamination with air, and state that the absence of argon "affords a most desirableconfirmation of our 'belief that the volcano gaseswere successfullycollected before they had come in contactwith atmosphericair at all and.were therefore entirely uncontaminatedeither by reactionor admixture with it." Thesegases were analyzedwith the followingresults:

GASESFROM HALEMAUMAU (KILAUEA), MAY, 1912. (Percentages by volume.)

Tube x. Tube Tube 8. Tube xo, Tube

CO2 ...... 23.8 58.0 62.3 59.2 73.9 5.6 3.9 3.5 4.6 4.0 H2 ...... 7.2 6.7 7-5 7.0 lO.2 N• 63 -3 29.8 x3.8 29.2 lX.8 None I.$ 12.8 None None so•...;...... I None None None None None HydrocarbonsRaregates ...... I None None None None None

Thesegases contained abundant water and before coolingmore sulphurdioxide than is shownin the analysis. The authorssay: The SO_,/or example,has gone over in part or altogetherto SOs and gone into solution, and only two of the five tubes analyzed now show SO2 as such. Moreover the resulting acid solutionsmay have reacted to a limited extent on the glass tube, and accordinglybe responsible/or all or a part of the alkalies,lime, and alumina shownin the analysesof water.

The writer understandsthat the authors do not supposethat any sulphuricacid was formed from original constituentson the cooling of the gases. The analysesof the materials containedin the water are given in the following table: 80Day, A. L., and Shepherd, E. S., "Water and Volcanic Activity," Geol. $oc. America Bull., Vol. 24, p. 588, 1913. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 593

ANALYSES OF MATERIAL CONTAINED IN THE WATER COLLECTEDIN THE TUBES.

Tube x, Grams Tube•'2, Grams Na.•O ...... o.o214' o.o3I* K20 ...... o.oicrz* o.oiI* CaO ...... o.012o* o.•4' Fe20•'1, 0.08o* o.oIo* AI•O• J ...... " ...... C1 ...... 0.220 0.206 F ...... 0.565 0.492 NH• ...... o.ooi8 None TiO ...... 0.0o5(?) None Total S as SO3 ...... 0.480 0.508

ß *The major portion of these may have come from the glass or from Pele's hair.

The analysesrecorded in the literature indicatethat sulphuris present in different volcanicemanations as H2S, SO2, SOa and as sulphurvapor, and many of the analysesshow free oxygen.

CONDITIONS OF FORMATION OF SULPHUR TRIOXIDE. It is obviousthat the first stepin the depositionof the sulphate mineralsis the formationof the sulphateradicle and the conditions under which this forms are therefore of interest. The com- mercial importanceof sulphuricacid has led to careful investi- gation of the modesof formation of sulphurtrioxide. Much in- formation on this subjecthas beenbrought togetherby Lunge.a* Sulphuricacid can be' formed in manyways, but oneway in which it may be formed in igneousemanations is that known as the contactprocess. This processconsists essentially in bringing about the combinationof sulphur dioxide and oxygen by the aid of a catalyzer. The accompanyingdiagram, taken from Knietsch,as shows the reaction under different conditions at atmosphericpressure. Lungeaøsays: The most important result of Knietsch'sexperiments was that a line of stable equili,briaexists which divides the range of .temperatureinto a, Lunge, George, "The Manufacture of Sulphuric Acid and Alkali," 3 vols., New York, D. Van Nostrand Co., I913. asLunge, George, op. cit., p. I3O7. a9Idem, p. I3II. 594 B. S. BUTLER.

two parts [with the catalyzersused]. The range below 200ø and above 900ø or •,oooø may be called devoid of reaction in a technical sense; between200 ø and 45øo the reactionof formationprevails; above 45 øo the dissociationof SOs comesinto play very rapidly. For the purposeof thispaper we may startat a temperatureof IOOOø at which SOa will be dissociated,ir•to SO2 and O, and note the changethat will take placeon cooling. It is apparentthat the rangeof temperaturein which sulphur trioxideforms at all rapidlyand is stableunder atmospheric pres- sure and in a systemcontaining only sulphur and oxygen is not large. In the manufactureof sulphuricacid the catalyzer commonly employedis platinum, but ferric oxide and many other sub- stancesact as catalyzers,so that many catalyzersmay be present in mineral veins. Anothermethod of producingsulphuric acid, which is employed at theplant of theNew CorneliaCopper Co., at Ajo, Ariz.,should be mentionedhere. In the treatment of copperores the copper is leachedby sulphuricacid and electrolytically precipitated. For successfulprecipitation it is necessaryto keepthe ferric iron of the solutionslow. When the solutionsbecome polluted with ferric sulphateihat substanceis reducedto ferroussulphate by sprayingthe solutionthrough a chambercontaining sulphur dioxide.4ø The gas is cooledfrom 6ooø to I5 oø F. in passing throughthe chamber.The ferric iron in the solutionis practi- ocally all reduced,according to the followingreactions :4• Fe2(SO4)a -5 SO• -5 2H:•O: 2FeSO4n t- 2H:•SO4.

FORMATION OF SULPHURIC ACID IN NATURE. The possibleconditions under which sulphuric acid is formed in nature have beenconsidered in detail 'by Ransome? In dis- •0 Tobelman,H. A., and Potter, J. A., "First Year of Leachingby the New CorneliaCopper Co.," Am. Inst. Min. Eng. Bull. •46, p. 475,•9•9. •t• Idem, p. 478. 4zRansome, F. L., "The Geologyand Ore Depositsof Goldfield,Nev.," lJ. S. Geol. Survey Prof. Paper 66, pp. •89--I95, I9o9.

596 B. S. BUTLER. cussingthe origin of the ore depositsof Goldfield,Nev., Ransome considersthree hypotheses: ( • ) "The directvolcanic hypothesis," which postulatesthat the solutionscome from deep-seatedsources chargedwith sulphuricacid, a hypothesissupported by the presence of the sulphate-bearingminerals hauynite and noseliteas original constituentsof volcanicrocks and the presenceof barite and celestitein mineral depositsthat are generally believed to have been formed independentlyof surface agencies;(2) "the hypothesisof the derivationof sulphuricacid from the oxidation of pyrite," a processso well known as to require no discussion; (3) "The hypothesisof simultaneoussolfatarism and oxidation," which postulatesthe rising of solutionscontaining hydrogen sul- phideto or nearly to the surface,their oxidation to sulphuricacid by atmosphericoxygen and the descentof the 'acid solutionthus formedinto the veinsagain, where they react with the ascending solutionsand causethe precipitationof metals, sulphides,etc.

FORMATION OF PRIMARY SULPHATE MINERALS. In consideringthe formationof the primary sulphateminerals, it must of coursebe recognizedthat sulphates,either as minerals or in solution, are prevalent at the surface and that, if waters from the surface have had a large part in the formation of a mineral deposit,the insolublesulphates would most naturally be formed. On the other hand, sulphatesare often found where there is good reason to believe that surface •vaters were not involved in their formation, and we may inquire whether sulp.hates can form without the aid of surface agencies. First let us see what consequencesfollow the assumptionthat free oxygen is presentin a given magneticemanation, even thoughmany investi- gatorsregard the presenceof free oxygenin a magmaas improb- able. With such an emanationthe formation of sulphur trioxide would not only be possiblebut under certain conditionsit would be inevitable. For instance,a gas that containedonly sulphur or sulphur dioxide and free oxygen, at •,oooø or higher, on cooling would pass through the interval favorable to the forma- PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 597 tion o.f sulphurtrioxide and that compoundwould form. At the higher temperatures,where sulphur trioxide is unstable, it wouldbe in lowconcentration, andthe formation and precipitation of abundantsulphates would not be.expected,but where temperaturesfavorable to formationand stabilitywere reached a high concentrationmight result,and there wouldbe a propor- tionate tendencyto the formationof sulphates. It is natural to supposethat bariumpresent in the form of the more solublecom- poundswould be the first substanceto be precipitatedas sulphate on accountof the very slight solubilityof barite in ore solutions as indicatedby its commonoccurrence. The samereasoning ap- plies to someextent to the formation of anhydrite (CaSO4) at high temperatureand p•:obablyalso to the formationof celestite. Barite is 'by far the most insolubleof the sulphatesin pure water, and apparentlyalso in mineralizingsolutions, and, if it were not for this mineral, there would be little trace of sulphatesas primary mineralsexcept where near-surfaceconditions. are reached. Sucha developmentof sulphatewould seemto be a more natural method for the formation of anhydrite, for example, than that suggested•orthe deposits of the Cactus mine, Utah, which did not account for 'the formation of the sulphate radicle. If the sulphate radicle actually forms in t'hecooling solution, it is easy to postulatea moderatesupply of calciumfrom the alteration of the monzonite walls of the fissure from which it is known that much calcium has been removed. It is possiblethat under favorable conditionsthe sulphateradicle may form to an extent that would result in a solutioncontaining sulphuric acid,; which in rocks.high in alkali and aluminum'would at favorabletemperatures give the conditionsfor the formation of alunite. As Ransomehas pointed out, it is notablethat solutionswhich form aluniteapparently are not goodcarriers of metal. The .Goldfielddeposits are a distinct exceptionin their associationwith metallic sulphides,althou.gh cinnibaris associatedwith alunitein depositseast of Beatty, Nev., which have been describedby Knopf.43 The cinnabaroccurs in 4a Knopf, Adolph, "Some Cinnabar in West'ern Nevada," U.S. Geol. Sur- vey Bull. 620, p. 64, I916. 59 8 B. S. BUTLER. silicifiedand alunitizedrhyolite. In the great alunitizedareas there doesnot appearto have been extensiveremoval of iron, much of whichi whetheroriginally present as oxide or in silicates,seems to have been altered to pyrite and remained. In someof the depositsof the Marysvale,Utah region,hemalite is associatedwith alunite, but the relation of the minerals has not yet beencarefully studied. Moreoverthe alunitein the vein is remarkablyfree from other minerals. When sulphustrioxide haddeveloped in the cooling solution tothe extent of yielding sulphuricacid the solutionprobably ceased to be a carrierof most of t'he metals,44 and the formation of sulphur trioxide may pos- siblybe a factor in the precipitationof the metals. The 'behaviorof smelter 'gases'gives some ind,ication of the temperature'at which certain sulphateswill form abundantly undergiven conditions. Fu'lton 45 says: The smokestream also carries water vapor, .theorigin of which is the moisturein the ore chargesfed to the furnaces. As long as the smoke stream has a temperatureabove 440ø C., no combinationof sulphur trioxide with water vapor to form sulphuricacid is possible,but as the temperaturefalls in ,theflues as the stackis approached,sulphuric acid vapor forms by the combinationof water vapor and sulphur trioxide vapor, until at about35 øo C. one half of •chesulphur trioxide presentin the smokestream is in the form of sulphuricacid vapor. At ordinary atmosphericpressure (760 mm.) sulphuricacid, or rather a mixture of 98.54per cem.of sulphuricacid and 1.46per cent.of water, hasa boiling pointof 338ø C. From this it followsthat above338o C. all sulphuric acid.present must be in a vaporform or dissociated,but that below this temperatureit may be presentin the smokestream in the liquid form as a fine mist or small liquid particles.... When the ore charge smeltedor roastedcontains a considerablepercentage of volatile metals, suchas lead, zinc and cadmium,,these partly passto the fume and, under oxidizing conditions,form oxides. The fine fume particlescombine readilywith the sulphuricacid formed,giving rise to sulphatesand the consequentneutralization of the acid. The neutralizationor the formation of sulphatesprobably does not readily,take place until ,temperatures below 200ø C. are reached,meaning practically that before sulphatesare formedthe sulphuricacid mustbe belowits condensationpoint. 4• Ransome,F. L., op. clt., p. I95, suggestssulphuric acid as a precipitant. 45Fulton, C. H., '" MetallurgicalSmoke," U.S. Bur. Mines Bull. 84, pp. 25-26, I9•5. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 599

There seemslittle doubt th.atsome sulphates will form at tem- peraturesabove 200 ø. Certain theoretical considerations and facts of observation supportto somedegree the methodof formationof sulphateshere suggested. Sulphateswould not be expectedin deep-seatedrocks, and apparentlythey are absent. They might form in low tem- peraturemagmas, but their abundantformation in suchmagrnas would not be expected,and this seemsto agreewith observations. On the coolingof certain mixtures of gases,the formation of sulphur trioxide would b.eexpected, and under certain conditions the formation of sulphates,and this seemsto accord with the facts of observation. The fact must not be overlooked,however, that the sulphates under discussionare associatedwith sulphidesand were apparently depositedat the.same time. The methodwould appear to make it necessaryto supposethat sulphideswere depositedin the presenceof free oxygen. We may next assumethat free oxygen does not occur in the assumedmagmatic emanation,and see if there is an available supplyof combinedoxygen for oxidation of sulphurto form sulph.uric.compounds. There seems ample evidence to indicatethat the oxygenpresent in the magmaand in the surroundingmineral- ized zone is not sufficient to insure that all the elements will be in their higheststate of oxidation. It is sufficient,however, to insure that much of the iron in igneousrocks may be presentas ferric compounds,notably magnetite and hematite. In fact, ferric iron is an importantconstituent of most igneousrocks, and in the contactdeposits and veins formed at high temperatures,. ferric iron is abundantin the minerals,magnetite, hematite, garnet (andradite) and other ferric compounds. In depositsformed at lower temperaturesferric mineralsare far lessabundant and are often entirely absent,the iron present being in ferrous mineralsor in sulph.idesand allied compounds. Of veins formed at intermediatetemperature Lindgren 46 says: 46Lindgren, Waldemar, "Mineral Deposits,"p. 5•4, New York, •9•3. 600 B. S. BUTLER.

Scarcely ever do we find the oxides, such as magnetite, specularire, and ilmenite. The predominatinggangue mineral is quartz, but carbonates are also common, such as calcite, dolomite, and ankerite, more rarely siderite; fluorite and barite are occasionallyof importance. ßIn veins that contain abundant sulphate the ferric minerals seemto beat leastunusual, though ferrous minerals--siderite or ankeriteor manganousminerals--as rhodoch.rosite or rhodonite may be abundant. Thus in the Cceur d'Alene district, Idaho, Ransome4' findsthat magnetitewith garnet is largely confinedto the "contact" type of deposits. In most of the veins of the region ferric minerals are absent,though siderite is abundant. Barite is locally abundant. Spencer4s recognizesa transition in the so-called"contact" depositsof the Santa Rita d,istrict from a zone near the intrusive rock characterizedby garnet (andradite) to.one characterizedby manganiferoussiderite and hedenbergite. That is an inner zone where ferric iron predominatesand an outer zone where ferrous iron predominates. No sulphatesare recognizedin these zones. Primary ferrous and manganousminerals in abundanceare associated with some of the larg• ore depositsthat are believed to have formed,at intermediatetemperatures, such as thoseof Lead- ville, Gilpin County, Rico, Lake City, and Creede, Colo., Butte and Philipsburg,Mont., and many others. Barite is presentin some of the deposits;but it has not been reported from others. It is abundant in some deposits,as those of the Tintic district, Utah, a9where ferrous and manganousminerals are not abundant, thoughiron sulphideis plentiful. The Cactusmine, Utah,5ø appears to offer a good exampleof the relation of ferric, ferrous and sulphate minerals. In that deposithematite was abundantly formed. After its deposition 47 Ransome, F. L., "Geology and Ore Deposits of the C•eur d'Alene Dis- trict, Idaho," U.S. Geol. Survey Prof. Paper 62, p. 94, x9o8. 48Spencer, A. C., personal communication. 49 Lindgren, Waldemar, and Loughlin, G. F., "Geology and Ore Deposits of the Tintic District, Utah," U.S. Geol. Survey Prof. Paper xo7, p. •53, •9•9. 50 Butler, B. S., "Geology and Ore Deposits of t'he San Francisco and Adja- cent Districts, Utah," U.S. Geol. Survey Prof. Paper 80, p. x2•, •9x3. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 6Ot had ceasedsiderite, anhydrite, and barite were deposited(see Fig. 40). Pyrite and chalcopyritewere depositedthroughout the period. From a considerationof igneousrocks. and. different types of ore deposits,it would appearthat ferric iron is commonand ferric mineralsform in nearly all magmasand their emanationsat high temperature. Sulphur, on the other hand, doesnot tend to form combinationsthat crystallizefrom magmas. Under somecon-

Beg/nningof End of minendiza•on rnineraliza•on

Sericite

Quartz

Chalcopyrite

Pyrite

Hemat/re

l}faznetite(?)

Tourma.line Siderite

Anhyddte

4o. Diagram showing relative period of formation of the principal ore and gangue minerals of the Cactus ore zone. ditions sulphurhowever combines with the metals as sulphides in magmasand may form large depositsor it may be oxidizedto sulphate,which may then produce complex silicate-sulphate minerals. In the contactzone and the deepvein zonethe higheroxide of iron continues to be stable and ferric minerals as oxide and sili- cateare abundantlydeposited. Sulphurcombines with the metals to form sulphidesbut they are in large part later than the ferric oxidesan.d silicates, but the sulphateradicle is rarely presentand only in a/licateminerals. As no insolublesulphites are formed there is no recordas to the presenceor absenceof the sutphite radicle in the solutions that formed the contact minerals. In veinsof moderateand shallowdepth ferric mineralsare very 6O2 B. S. BUTLER. scarceor absent,whereas ferrous minerals in some depositsare abundant. Sulphidesare abundantand the presenceof barite, whichis by far the mostinsoluble of the sulphates,indicates that sulphatesprobably formed relatively abundan,tlybut for the most part •verecarried away in solution. This changefrom ferric to ferrous minerals,together with the appearanceof sulphate minerals associatedwith the ferrous minerals,may be interpretedas indicatingthat at the higher temperatures the conditions were favorable to the oxidation of iron and the reduction of sulphur, whereas at the lower temperatures the conditions were favorable to the reduction of ferric com- poundsand the oxidation of the sulphur. This interpretation agrees with the experimental data so far as the oxidation of sulphur dioxide and the reduction of ferric compoundsare con- cerned. Possibly this reductionof the ferric compoundstakes place to some extent even af.ter the ferric minerals have been depositedfrom 'the ore-bearing solutions. The hematite in the Cactus mine, which has already been mentioned,is distinctly magnetic and may have become so by the partial reduction of ferric oxide when sulphur or its compoundswere oxidized to sulphuriccompounds. This record of the ores, then, revealsone possiblesource of oxygen to form sulphatesapart from free oxygen in magmatic emanatio.n. If a large part of the iron in the magmais in ferric and ferrousoxide, 'the combination of part of the iron with sulphurto form pyrite would greatly reducethe amountof sulphurand at the sametime would furnish oxygento oxidize the excessof sulphur. It has already been pointedout that the iron in the wall rock of many veinsas ferrous and ferric oxide has combined,in part at least, with sulphurand thus freed the oxygen to go into someother combination. The depositsof nativecopper at Corocoro,Bolivia, are worthy of note in this connection. The depositsare in a seriesof red sediments. The ore solutionshave apparentlyreduced the ferric iron in the beds containingthe ores, and sulphateshave been deposited. Miller and Singewaldsay ,x Miller, B. L., and Singewald, J. T., "The Mineral Deposits of South America," p. 92, McGraw-Hill Book Co., Inc., New York, I9x9. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 603

The mineral solutions that circulated through the ore beds have bleached'themto a whiteor light greencolor, but the imperviousshales betweenhave not been affectedby the solutions. Patchesof red sand- stone within the ore 'bodiesthat have been protectedfrom the bleaching action of the mineralizers are barren of ore.... Gypsum and, less .abundantly,barite and celestite occur as gangueminerals. Steinmann,according to Miller 'andSingewald, 5•' believes that: The mineralizing solutionswere. analogousto those tha.t formed the other copper depositsof the Andes; that is, characterizedby the presenceof sulphurand arsenicto the subordinationof oxygen. On coming in contactwith the ferric oxide of the red beds,they reducedit and bleached.those strata, and the sulphur was oxidized to sulphuric acid. On accountof the greater affinity of sulphuricacid for lime, magnesium, and iron, the sulphatesof those metals were formed and copper was set free. ß There is evidencethat many of the elementsin the magmas besidesiron were combinedwith oxygen and that with decrease in temperaturethey tended to combinewith sulphur. The re- markabledeposits at FranklinFurnace, N.J.? are.instructivein this connection. There was probablya lack of sulphurin the emanationsthat formed the deposits,and under thoseconditions manganicminerals were deposited. Zinc was depositedabun- dantly as oxide and silicate. In depositscontaining sulphides, manganicminerals are certainlyrare, though manganous minerals may be abundantand undersome conditions manganese sulphide has beendeposited. Likewise if sulphuris presentin the emanations the zinc oxide and silicate apparentlyare never deposited, but zinc at the temperature'at which it will depositcombines with the sulphur. It may alsobe notedthat tungstenis mostcommonly deposited as tungstatebut it may be depositedas sulphide. The only known occurrenceof the sulphide,that in the Emma mine, Utah,'4 is a ,2 0p. cit., p. 94- •aSpencer, A. C., U.S. Geol.'Survey Geol. Atlas, Franklin Furnace folio (NO. 10I), 1908. ,4 Wells, R. C., and. Butler, B. S., Washington Acad. Sci. ]our., Vol. 7, PP. 596-599, 1917. 604 B. $. BUTLER. deposit that is believed to have been formed at moderate temperature. Here the sulphideof tungstenwas one of the latest mineralsto form and apparentlyin part at least replacedearlier sulphides: Tin is commonlydeposited as the oxide, but in some of the Bolivian depositswhich seemto have been formed at only mod- iratetemperatures and in whichsulphides are relatively abundant the sulphideof tin was deposited. Some of thesedeposits contain barite. Lunge55 statesthat "on boiling sulphurwith water, hydrogen sulphideis evolvedand sulphuricacid is found in the residue." The formation of sulphuricacid by heating sulphurwith water has beendemonstrated by Allen?ø Thus hydrogenunder certain conditionsgives up its oxygen,and the oxygen so releasedgoes to oxidizesulphur. The productionof sulphuricacid by the reaction of sulphurdioxide and water at about x5oø C. will also take place•7 accordingto the equation3S02 -3-2H20:2H2804 q- S. The resulting sulphur would of coursebe available for reaction with water under proper conditions for producing sulphuric acid (H=S04). If it be granted that sulphat•sform in the zoneof intermediate temperatureswithout the influenceof surfaceagencies, the ques- tion then arises whether a similar origin can be attri,butedto alunite deposits. There seemsto be universalagreement among geologiststhat alunite depositshave formed near the surfaceand alsothat someof them,at least,are formedby acid solutionsthat result from surfaceoxidation. It is manifestlydifficult to prove the part that surfaceand deep-seatedagencies have played under suchconditions, but if barite, anhydrite, and celestitecan and do 5• Lunge, George, op. cit., Vol. x, pt. x, p. x7. 50Allen, E. T., personal communication. 57Lewis, G. N., Randall, M., Bichowsky,R. V., "A Preliminary Study of ReversibleReactions of Sulphur Compounds,"Am. Chem. Soc. ]ou•;., Vol. 40, p. 356, x9•8. Randall, M., and Bichowsky, R. V., "Equilibrium in Reaction between Water and Sulphur at High Temperatures; The Dissociationof Hydrogen Sulphide,",4m. Chem. Soc. ]our., Vol. 4o, p. 368, x9•8. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 605

form from deep-seatedsolutions without surfaceoxidation, there seemsto be no goodreason why alunitemight not alsoand why this mode of origin may not be consideredif it seemsto accord best with the observed facts. The occurrenceof hinsdaliteas a primarymineral in deposits that are believedby the geologistswho havedescribed them to be of deep-seated(hypogene) origin is of interest. This mineral is chemicallyclosely allied to alunite•in fact is a memberof the alunite group of minerals--and if it can form from th.e same solutionsas sulphides,rhodochrosite, and barite, there seemsto be no inherentreason why alunitemight not also form from ascendingsolutions under the properphysical conditions. The geologicalevidence indicates that alunite can form only at low temperatures,probably considerably below the initialtem- peratureat whichsulphuric acid and the sulphatesof potassium and aluminummay appearin the solutions. The mineralogicallyallied minerals of the jarositegroup and otherallied basic ferric sulphates are apparentlyformed only by surface(supergene) solutions, and undernear surfaceconditions favorablefor oxidixingiron, where they may form abundantly. It is obviousthat theycan form onlyunder conditions favorable to thestability of thehigher oxides of bothiron and sulphur, and it appears that alunite forms under conditions that favor the reductio.nof the higheroxide of iron to ferrouscompounds or sulphide,on the one hand, and that favor the formation of the higheroxide of sulphur,on the other hand. The factthat ore solutions may change their character during thedeposition of theore has been long recognized. Thus Lind- gren5s in describingthe alterationof the rocksadjacent to the veinsin the De Lamar mine,Idaho, states: Thisconfirms ,the view set forth that two different processes have been active;first, an ordinaryprocess of sericitiza.tion, accompanied by a veinfilling of bariteand calcite, effected by waterscontaining alkaline 58Lindgren, Waldemar, "The Gold and Silver Veins of Silver City, De Lamar,and Other Mining Districts in Idaho,"U.S. Geol.Survey Twentieth Ann. Rept., pt. 3, P. x82, x9oo. 606 B. S. BUTLER. carbonates;second, pseudomorphic replacement of the filling by quartz and leachingof A1208from the sericitizedcountry rock by siliceous (probably acid) waters. Lindgren59 also states that "the lossof somuch A12Oa can be explainedon the suppositionthat the waterscontained sulphuric acid,as only such thermal waters are knownto dissolvealumina in largequantities." In anotherwork he says Veins formednear the surfacein volcanicregions are sometimessub- ject to peculiarchanges, which are rarelyobserved in depositsof more deep-seatedorigin. An earliergangue mineral, such as calcite or barite, maybe whollywiped out and replaced by a newgangue of quartzand adularia. This alterationhas nothingto do with surfacewaters; it is plainlycaused by a changein the compositionof ascendingcurrents. In the San Franciscodistrict, Utah, the writer has describeda notabledifference in the alteration of the wall rock and in gangue mineralsin neighboringdeposits in Tertiary lavaswhich it is believedrepresent different stages in the processof ore'deposition. Thus, regardingthe Horn 'Silver and Beaver Carbonate mines, it is stated6• that- Although the principalore mineralsof the .two depositsare the same there is a notable difference in the gangue minerals. Carbonatesare importantin the Beaver Carbonatemine and sulphatesin the Horn Silvermine.. This differencepoints to a differencein the characterof the solu.tions that depositedthe ores,and a similardifference is indicated in the alteration of the rock adjacent to the deposits. The extensive removal of alumina from the rock of the Horn Silver deposit and the presenceof abundantsulphates is contrastedwith .thefailure to remove aluminaand the presenceof calcitein the BeaverCarbonate deposit. The presenceof sulphuricacid in solutionthat depositedsul- phid.esis suggestedby Spurr.ø2 Concerningthe alterationof the wall rocksat Tonopah,Nevada, says: •o Idem, p. I8I. 6oLindgren, Waldemar, "Mineral Deposits,"p. 436, New York, I913. 6xButler, B. S., "Geology and Ore Depositsof the San Franciscoand Adjacent Districts, Utah," U.S. Geol. Survey Prof. Paper 8o, p. I33, 1913. 62Spurr, J. E., "Geology of the Tonopah Mining District, Nev.," U.S. Geol. Survey Prof. Paper 42, p. 234, •9o5. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 607

However,they [the mineralizingsolutions] attack the rock vigorously by virtue of the carbonicacid, probably also sulphuric acid, and perhaps to a less extent by acids of chlorine and fluorine. The many knownreversible reactions that go on with changing conditionsmake the problemsof changesin ore solutionsmost complex. Theseproblems really belongin the field of chemistry, but as the reactionsthat go on in ore solutionsas a result of changesin temperatureand in concentrationmay exert a large influenceon the deposition of theores, ;the geologist may properly point out someof the factsthat needexplanation and indicatethe evidenceof the changesthat have taken place as they are pre- servedin the rocksand ores. There are undoubtedlychanges of which no record is preserved; indeed, were it not for the pres- encein ore deposit.sof a few sulph'atesthat are relativelyinsol- uble, there would be little record of the existenceof the sulphate radicle in ore solutionsfrom .deepsources. In this properatten- tion has been directed to some possiblerelation of sulphur and ox.ygento otherconstituents of <)re solutions, 'but it is obvious that other elementsshould be considered,of which carbon is an important one. The possibilityof the formation of sulphuricacid and sulphates in magmaticemanations may have a bearingon many problems, but it is not the purposeto pursuetheir study in this paper. It may, however, 'bepointed out that closeobservation of the relations of minerals is essentialto a-clear understanding. It seems certainthat althoughbarite, for instance,does not form at high tem.perature, it may be associatedwith mineralsformed earlier at high temperature,or evenlater, as whenonce formed and per- hapscovered with otherminerals it mightpersist, even though the temperaturemight be raised. A studentof copperdeposits, when consideringthe possibility of a deep-seatedsource of sulphuric.acid, will read with added interestthe paper by Zies, Allen, and Merwin6• on reactionsbe- 6aZies, E. S., Allen, E. T., and Merwin, H. E., "Some Reactions Involved in Secondary.Cqpper SulphideEnrichment," Ecoa. G•.., Vol. H, pp. 407- 503, •9•6. 608 B. S. BUTLER.

tweencopper sulphate and sulphuricacid and varioussulphides at moderatelyhigh temperature. It seemsthat there is an almost unlimitedfield for similarinvestigation which will contribute directlyto thesolution of someproblems of ore.deposition.

SUMMARY. Sulphatesin igneousrocks and in depositsformed at high temperatureare confinedto a few complexsilicate minerals that containthe sulphateradicle. In depositsformed at intermediate temperaturebarite is commonand anhydriteand celestiteare not uncommon.Under favorableconditions and probablyat com- parativelylow temperaturealunite forms abundantly. In some depositsat leastthe sulphateradicle of the alunitewas probably derived from deep-seatedsolutions. A studyof volcanicemanations has shown that theyexhibit changesin characterand that in the later stageof fumarolic activitythey may containsulphurous and sulphuriccompounds. Sulphuricacid can readily be formedby the reducingaction of sulphurdioxide on ferric solutions.If igneousemanations contain free oxygenand sulphuror sulphurdioxide it would be expectedthat, as they becomecool, sulphur trioxide wouldbe formed,and that at suitabletemperature the sulphateswould be formed. Sulphurtrioxide is unstableat hightemperatures and the temperaturerange in whichit formsrapidly and is stableis narrow. If emanationscontain no freeoxygen that combine with themetals or withhydrogen at hightemperatures, they may at lowertemperatures comt)ine with sulphurto form the oxidesof sulphurand sulphuriccompounds. Thisinterchange of oxygen from certain elements at hightem- pera'tureto sulphurat lower temperatureis believedto be an im- portantfactor not only in theformation of sulphatesin solutions of deep-seatedorigin but also in theprecipitation of primary (hypogene) ore minerals. PRIMARY SULPHATE MINERALS IN ORE DEPOSITS. 6o9

ACKNOWLEDGMENT. The writer wishesto gratefully,acknowledge the criticismsand helpful suggestionsof co-workersin the subjectof ore deposits, especiallythose of E. $. Bastin,Adolph Knopf, E. $. Larsen,G. F. Loughlin, ChasePahner, F. L. Ransome,Max Roeslet,A. C. Spencerand R. C. Wells, of the United StatesGeological Surv. ey; Drs. E. T. Allen, J. B. Fergusonand R. B. Sosman,of the Car- negie Institution of Washington, and Prof. John Johnston,of Yale University, without implying, however, th,at all these stu- dents subscribeto all the ideasput forward.