2OO 14,'ILLI.4M F. FOSH.4G. mud. It crystallizesin clear glassymonoclinic crystals in a great variety of habits. It is easilydistinguished from the other borate mineralsby its eminentcleavage. When heatedin a flame cole- manite loseswater and falls to a powder. Sincethe other borates fuse easilybut do not decrepitatethis givesa simpletest for the detection of colemanite. Ulexite.--At one time ulexite constituted one of the main sourcesof the boratesof commerce. Today its productionis restricted to a few localities in South America. It is the double borate of calcium and sodium, NaCaB,O0.8H20. It is abundant in manyof the playasof North and SouthAmerica. It occursin aggregatesof looselycoherent, acicular crystals.. Theseaggre- gatesare popularlycalled "cotton-balls." They are foundin the surfacesalt crustsof the playa lakesas well as in the mud layers below. At the Lang depositsin California ulexite also occurs in lenticular masses with the colemanire at the foot wall of the beds. In the center of these masses the fibers are oriented in all directionswhile at their peripheriesthe fibers are parallel and normal to the surfaces. These masseshave the appearanceof "cotton-balls" that have been compacted,by' pressure. Ulexite fuseseasily in a flame,coloring the flameintensely yellow. Inyoite hasbeen found in the Mount BlancoDeposits in Death Valley. Recentlyit has been found in the gypsumdeposits of Nova Scotia. The fresh mineral is in clear glassy monoclinic crystalswith eminentcleavage but the mineral at Mount Blanco is almostcompletely altered to a fibrouscrystalline mass of meyer- hofferite. In compositioninyoite is similar to colemanitebut differs in its water content from that mineral. The iny'oitecon- tains thirteen molecules of water as shown by its formula, Ca2B00•. •3 H20. Meyerhofferiteis a third'member of the colemaniteseries with seven moleculesof water, Ca2B00•.7H20. It has been found only in the Mount Blancodeposits as a dehydrationproduct of the inyoite. Howlite is a borosilicateof calcium, H,'Ca2B,SiO•4. It occurs in all the importantborate deposits of California in considerable ORIGIN OF COLEMANITE DEPOSITS. 201 abundance. It is found in friable massesof monocliniccrys- tals or in porcelain-like,nodular masses. It is white in color but rarely brown. It fuses in a flame coloring the flame green. Bakerite is a similar if not identical mineral. Gypsumis ,presentin all the deposits.It is alwayssecondary with respectto the bedsin which it is found. It occursin clear glassy monocliniccrystals, as selenite,or as satin spar. No chemicallyprecipitated gypsum appears in any of the depositsas far as the writer has been able to observe. Calcite-ara#onite.--Calciumcarbonate appears in two forms in the deposits. As calciteit is presentin the clays,perhaps as an original chemicalprecipitate. As aragoniteit very rarely occurs as buff colored,acicular crystals perched upon and .perhaps derived from colemanite. Celestite.--Strontium sulphate, SrS04, is a.s far as known found only in the Calico beds. Here it occurs in colorlessor blue, long orthorhombiccrystals in the geodesof colemanite.

THE PLAYA LAKES. In order to understandwhat has taken place in the Tertiary lakes it will be well to observewhat is taking place in similar basinsof today. Bora:r Lake, in Lake County, California, was the first borax- producinglocality in the United States. It is situatedin a tri- angularvalley surrounded by hillsof basalt. A basaltridge sepa- ratesthe lake from the larger, fresh water Clear Lake. Formerly Borax Lake shrunk to a small volume in the summer but it is now fed by an artesianwell in the lake floor and is essentiallya permanentbody of water. The dissolvedsolids carried borax to the extent of t8 per cent.,with about62 per cent. of sodiumcar- bonateand 20 per cent. of sodiumchloride. In the mud of this lake were numerouscrystals of sodiumborate, a rather peculiar feature, since the lake waters were not saturated with salts. About the edge of the lake were numeroussmall spiresof cal- careoustufa ranging up to a foot in height. A small solfatara. 202 WILLIAM F. FOSH.4G. known as the Little Sulphur Banks, is not far distant from the shores..Across the ridge that separatesBorax Lake from Clear Lake is a solfataraof large dimensions,known as SulphurBanks. The springsand waters emanatingfrom this solfatara are highly borifer6us. The boron in Borax Lake is without doubt derived from the adjacent solfataras. Searles .Lake in San Bernardino County offers a somewhat' different type of playa lake. This is a concentrationbasin receiving its salines from the drainage of a considerablearea. The area now drainedis restrictedto the Argus and SlateRanges, but at someearlier periodOwens Lake overflowedinto the Seaties Basinby way of Indian Wells and Salt Wells Valleys and it was these waters that contributedmost of the salts now coveringthe floor of Searles Basin. In the winter a few inches of water covers the lake bed, but in summer the •vater level falls to a few inches below the salt crust. The crust is a mere efflorescence at the outer edgebut acquiresa thicknessof about50-75 feet in the center. Beneath this salt crust are beds o.f fine mud with em- beddedcrystals of halite,hanksite, gay-lussite and others. These salts, especiallythe hanksite and, pirssonite,are in sufficient abundanceto form definitesolid layers or strata. Many of these muds are rich in precipitatedcalcum carbonate, in the form of calcite. The minerals occurringhere are chiefly sulphatesand carbonates of calcium and sodium, chlorides of sodium and potassium,and borax. Ulexite, if presentat all, is very rare, and no colemanite occurs. Danby Lake coversa large area near the town of Areboyin San BernardinoCounty. Although it presentssome unusual features, it has never been studied in detail. Near the western edgeof the lake a small coneof basaltarises and someof the flows cover a portionof the lake bed. The surfaceis covered with gypsumand mud. Theemud is rich in small crystalsof glauberiteand the brines carry a largeamount of calciumchloride as well as sodium chloride. A stratum of salt underlies the mud but its extent and thickness are undetermined. No ulexite or colemanitehave been reported from here. ORIGIN OF COLEMANITE DEPOSITS. 2o3

RhodesMarsh is one of the types of playa lake that formerly producedconsiderable quantities of borax before the discovery of the colemanitebegs. It is situated in Esmeralda County, Nevada, According to the descriptionof LeConte,2 common salt occupiesthe center of the flat. Around this to the margin, the depositsdiffer from place to place. In someparts borates occur as borax or "tincal," in others the borate is utexite. Com- mon salt is found nearly everywhere,more or lessmingled with the other salts. Sodium carbonateand sulphateare present in considerable amounts. The well known ulexite "balls" occur in a semicirculararea surroundingthe central salt area on the north. They are embeddedin a stiff, wet clay. If the loose earth is removed to the depth of a foot or more, until the stiff clay is reached, the "cotton balls" are found. In places the beds of ulexite reacha thickness'ofsix to eight feet. This concentration of the ulexite into patchesseems to be a commonfeature of both westerndeposits and the South American lake.s. In someof the later depositsthe ulexite occursin considerablequantities and of great purity. GEOLOGY OF THE DEPOSITS. GeneralGeolo9y.--The colemanite deposits are foundinter- stratifiedwith lake bed depositsof Tertiary age. Theselake bed depositsare for the most part volcanic ash or material derived from volcanic rocks laid down under shallow water conditions in closed basins. The beds in direct association with the colemanite are fine-grained,thin-bedded shales of variouscolors. Lavas of rhyolitic, andesiticor basalticcharacter are often presentin the boron-bearingseries. For the most part they al•ear to be sur- face flows. The borate beds themselvesfollow the bed•dingof the series but are somewhatirregular in form. They do not form continu- ous,well-defined beds but rather representshale beds in which the boratesare more or lessirregularly distributed. They are, how- ever, restricted to a definite member of the series. Third Ann. Rept. Calif. State Min. Bur., p. 5I, I883. •o4 WILLIAM F. FOSHAG.

The borate-bearingbeds have all undergonesome movement. They range from bedsalmost perpendicular to bedsbut slightly tilted. There is also evidence of movement within the beds them- selves. Lan#.--The sedimentaryrocks in the regionabout Lang are coarselystratified sandstones of a light buff color. Thesesand- stonesgrade into conglomorates. In theimmediate vicinity of the colemanitedeposits they are fine,thin-bedded shales. The borate- bearingbeds have an east-westtrend and dip at an angleof about 7ø0 to the south. The commercialproduct is co.lemanite, in cleavagemasses and very often in columnarbands. There is evidenceof movement and pressurein the ore-bearingstrata, but hardly sufficientto accountfor the large massesof clearablecolemanire. Much of thecolemanire is of gray color,due to includedmud, but second- ary colemanireis of a purewhite color..Crystals are rare. How- lite is presentin cor•siderableabundance as botryoidal concretions which sometimes are several feet in d'iameter. The howlite nodulesare imbeddedin the colemanireand form "augen" in the strata. Rarely howliteand colemanireare intergrown,and still more rarely the howliteis later than the colemanire. The mineralof greatestsignificance is ulexite,which occurs in abundancein someof the workings. At the 25o-footlevel the footwall is composedlargely of ulexite. It is not the ordinary "cotton-ball"type but is massiveand fibrous. It hasthe appear- ance of cotton balls that have been consolidatedby pressure. That it is typicalulexite is shownby the followinganalysis:

CaO ...... I4. I4 H20 ...... 35.68 B2Os ...... 43 -12 Na20 ...... (7.05) (by difference)

The ulexiteoccurs in irregularmasses, more or lesslensqike and surroundedby thin layersof clay. In structurethese lenses are compact-fibrous,thefibers oriented in all directionstn thecenters, and parallelat the peripheries. ORIGIN OF COLEMANITE DEPOSITS. 205

Calcite is rare in the d'epositsand none was observedin direct associationwith the colemanite. Aragonite occurssparingly and is always secondary. Borate.--The borate depositsare situatedin the Calico Moun- tains not far from the old silver camp of Calico. The nearest town is Yermo on the Salt Lake Railroad, eight miles distantby wagon road. At the time of the writer's visit (May, 1920) a small amount of work in the nature of "gophering" was being done. Sincethe mineshave been idle for a numberof yearsthe extensiveold workingsare no longeraccessible, so that the only data obtainablewas from the dumpsand shallowworkings. The sediments here are sandstones and shales. The borate beds are in thin-beddedshales similar to thoseat Lang. The bedshave an east-westtrend and dip to the south. South of the depositsrises the main rhyolite massthat forms the core of the Calico Mountains. Storms a states that "the rocks have not suffered,in the region about the borax deposits,the slightest metamorphism." The surface exposuresare unaltered, loosely coherent,shales. The colemanitebeds outcrop along the hill for a distance of about one-half mile and show, beside colemanite. occasionalbands of howlite and numerousveins of satin spar gypsum. Storms describesthe bed,sas being much mixed with sandy sedimentsand gypsumat their westernend, giving the appearanceof having formed "near the shore line of a basin." A peculiarityof the colemaniteoccurring here is the large numberof geodalnodules lined with excellentcrystals of cole- manite with sometimescelestite and selenite. In someparts of the depositthey occur.in great numbersimbedded in the shales, much after the fashion of the ulexite nodulesdescribed by LeConteat RhodesMarsh. 4 The geodesoften show two genera- tions of colemanitecrystals. In certain other portionsof the depositscolemanite gives way to cavitiescontaining clear and brilliant selenitecrystals. The shale containingthese selenites has a decidedlyhoneycombed structure, and the cavitiesnow con- 8 Calif. State Min. Bur., Hth Ann. Rep., p. 346. 4 See above, under Rhodes Marsh. 2O6 WILLIAM F. FOSH•IG. tainingthe selenite crystals were undoubtedly formed by the removal of some previous mineral. Associatedwith both the colemanite and the selenite are small nodules of howlite. The colemanitecrystals of many of the geodes,especially those near the surface,are partially or wholly altered to calciumcarbonate, in many casesstill retainingthe external form of the colemanite crystals. Fentura.--The Ventura depositshake beendescribed' by Gale* and were not visited by the writer. They are situatedon the south flank of Mt. Pinos, near the San Emigdio Range, in Ven- tura' County. The general trend of the borate-bearingbeds is northeastand southwestand they dip to the southeast. These bedshave been extensivelyfolded and faulted. The colemanire occurs in shales and limestones intercalcated with flows of basalt. The colemanireis typical of the mineral as it occurselsewhere. It is crystallinewith excellentcleavage, white in color, but in placesgray from includedmud. The massesfollow the bedding of the sedimentarystrata with which they are associatedbut appearto be very irregular in form. The mineral is often radiated and in placescrystallized in open cavities. The cole- maniteappears to havebeen developed in immediateassociation with a bed of limestone included within the shales. The lime: stoneis massiveor of a roughlyporous character. The outcrops rarely showany colemanitebut are indicatedby an abundanceof gypsum. This gypsumis in the form of seleniteand is of vein character. Death Falley.--The geologyof the Death Valley depositsha.• beendescribed by Keyes? He characterizesthe borate-bearing bedsas fine olive greenclays that weatherpale greenor white. Numerous basalt sheets from ten to a hundred feet thick are inter- bedded. In the upperpart of the sequencemuch selenite gypsum, bedsof colemanireand thin layersof limestoneof probablechem- ical origin,occur. Underneaththese beds are strataof sand- stone and conglomerate. • U.S. Geol. Survey Prof. Paper 85, p. 5. • Trans. Am. Inst. Min. Eng. Bul. 34, P. 870, x9zo. ORIGIN OF COLEMANITE DEPOSITS. 9-07

The richer borate beds are from a few inchesto fifty feet thick. In the unweatheredportions they consistof bluish clays interspersedwith milk-white layers, nodular bands and nodules of colemanite. Through the stratacarrying the coarselycrystal- lized colemanitethe clays are more or less highly impregnated with fine particlesof the boratemineral, and yield uponleaching ten to twenty-five per cent. of anhydrousboric acid. Mingled with the coarsecolemanite are often found large amountsof crystallizedgypsum. In someplaces the gypsumis so abundant that the borate minerals are all but completelyobscured. Fre- quentlythere are presentlarge amountsof pure limestone. How- lite, althoughnot mentionedin any of the pu•blisheddescriptions, is a[undant in many of the colemanitespecimens examinec• from this' locality by the writer. In appearanceit is very similar to the Calico howlite.

CHEMISTRY OF THE MINERALS. Boric acid is one of the weakest of inorganic acids, dissociat- ing to a lessextent than carbonicor hydrosulphuricacids. The dissociating.constant for boric acid is i. 7 X IO-ø at i8 ø as compared/with carbonic,(H2COa), 3.04 X io -7 and hydrosul- phuric (H2S), o.9x X IO-7. Besidesorthoboric acid there exist in the form of salts metaboric acid, HBO2, tetraboric acid, H•B40 •, and a numberof poly acids. In respectto its formation of poly acids,boric acid is similar to silicicacid. The dissociationof metaboric acid takes place as follows: HBO• = H + q- BO-•. Theoretically orthoboriccan give three hydrogenions, but the dissociation HoBO a = H +q- H2BO-a takesplace 'to sucha veryslight extent that furtherionization is not measurable. The orthoboratesin nature are accordinglyof the type R'H2BOa, e.g., pinnolte,Mg(H•BOa)2.H•O; and such salts as MgaB•O0 do not form from solution. 208 WILLIAM F. FOSHAG.

Orthoboricacid uponheating loses water and passesover into metaboricacid. Further heating convertsit to the oxide, B20 a. The oxideis volatileonly at high temperatures.The orthoboric acid, however, is volatile in steam and this explainsits presence in fumaroles and volcanic exhalations. Although orthoboricacid is weaker than carbonicit may po•- si'blyreplace the latter to someextent in hot concentratedsolu- tions. This would be due to the greater volatility of the carbonic acid. The reaction, however, has never been studied'in any detail. The writer subjectedsections of calcite to the action of boric acid for a period of severalmonths without any changeof the carbonate to the borate. When boric acid dissolvesit tendsto form co.mplexmolecules with the formation of meta, ortho, tetra and poly acids, so that a solution of boric acid does not contain the orthoboric acid ion alone but the others in equilibrium with each other. Borax solutionsin the presenceof sodiumchloride attack cole- manite to form ulexite and calcium chloride. Ulexite is the com- mon occurringborate in the playa lakeswhere it is found with borax and salt. Ulexite hasbeen synthesized by Van't Hoff 7 by the action of a borax and sodium chloride solution on Ca.oB60•, 9H20, oneof the membersof the colemaniteseries. At 83ø ulexite splitsinto. its componentborates according to the equation 2NaCaB509.8H20: Na2B407q- Ca•B60•. 7H•O q- 9H•O. This splittingtakes places at lower temperaturesin the presence of sodiumchloride. It •vill be noticedthat one of the products of the above reaction.is the mineral meyerhofferite,one of the colemanite series. Kraut s obtained another member of the cole- manite series,Ca2B60•.6H20, by simpleleaching of ulexite. Van't Hoff found further that the pentahydrate,colemanite, was less soluble in sodium chloride solutions than the hepta- hydrate,meyerhofferite. By treatingthe heptahydratewith a sodium chloride solution he obtained the pentahydrate, cole- * Sitzb. Acad. Wiss. Berlin, p. 566, •9o6. 8 Arch. Pharm., 2, •2. ORIGIN OF COLEM.4NITE DEPOSITS. 209 manite. By leachingulexite with sodium chloride solutionsin- stead of pure water he obtainedcolemanire in place of meyer- hofferite. The reaction 2NaCaB•O9.8H20 = Na2B•O7.IoH½O -+- Ca•BoOn. 5H20 -+-H•O. is completelyreversible and by varying the amountsof borax the reaction can go in either direction---colemaniteand borax may result from ulexite or ulexite from colemanite and borax. If an excesso.f borax is present any colemaniteis converted into ulexite while if the borax is removed as fast as it is formed the ulexite will break down completely,leaving the mineral meyer- hofferiteif no sodiumchloride is presentbut giving colemanite if the solutionsare salt-bearingones.

ORIGIN OF THE BORON. Three hypotheseshave been put forward to explainthe source of the boron in the variousborate deposits:(I) origin from sea water, (2) from the decompositionof borosilicateminerals, and (3) volcanicexhalations. •. In such depositsas those at Stassfurt the sourceof the boron is without doubt from sea waters. The amount of boric acid in the waters of the ocean is sufficient to account for the rela- tivelysmall amounts of boratesin thesesalt beds. 'For the Cali- fornia deposits,however, all evidencepoints to an origin other than the evaporationof sea water, so that this possiblesource may be dismissed. 2. The averageboron trioxide contentof igneousrocks is very smallbut may reachan appreciableamount in suchrocks as tour- malinegranites. Theserocks, upon decomposition, may give up their boron.which may then be leachedout and concentratedin suitable basins. But borosilicates such as tourmaline are resistant minerals and no accumulationsof borates are known that can be definitelytraced to this source. 3. The volcanicorigin of theboron has much evidence to favor it. Boron hasbeen detected in the vaporsof many volcanoes. It 210 WILLIAM F. FOSHAG. is one of the important mineralizersof igneousrocks. But most importantof all is the actualaccumulation of boronin the craters of certain volcanoesand in the hot springsand solfatarasof volcanicregions. Thus sassolite,B(OH)3, is sometimesfound at Vesuvius' while at Volcano, borates were at one time present in suchquantities that it becameprofitable to work them in the very crater. At SteamboatSprings, Nevada, a hot springconnected with rhyoliticrocks, the boroncontent of the dissolvedsolids is as highas 9 per cent.B2Oa. At SulphurBanks, in Lake County, California, there is a solfatara of large dimensionsdischarging highlyboriferous waters and close by areborax deposits of actual commercialimportance. The waters of the hot springscarry over25 per cent.of B•Oa (of the dissolvedsalts) while waters collected in the old Parrot Shaft contained solids with over 40 per cent.boric oxide. Thusin the caseof SulphurBanks a large quantity,of borax is dischargedyearly into the watersof Clear Lake. These springsare connectedwith fairly recentbasalt flowsand have a directgenetic connection with the sulphurand cinnabardeposits from •vhichthey isssue. The Tuscanysoffioni have long been cited as an example of the volcanicorigin of the boron,although the evidencehere is not conclusive,since the waterspass through strata of sandstonean_d othersedimentary rocks, and a possiblesource of theboron in theserocks is not entirely precluded. It is significantthat all depositsof boraxand'other borafes are situatedin regionsof past or presentvolcanic activity. The Borax Lake in California is adjacentto the solfatarasof the dis- trict, the depositsof the desertregion are connectedwith numer- ous volcanicflows. The borate depositsof South America are, accordingto Chamberlainø largely confined to thoselakes which lie closeto the volcanoesof the western Cordillera. Away from these volcanoesthe borates rapidly disappear. The original sourceof the boron in our Western depositsis most probablyto be foundin the hot springsand solfatarasconnected with the 9 ]our. Geol., 20, 763, 1912. ORIGIN OF COLEMANI•E DEPOSITS. 2II

'tremendousvolcanic activity that characterized the Tertiary Period when these depositsfirst accumulated.

PREVIOUS THEORIES.

To account for these accumulations of calcium bo.rate two theorieshave been proposed. The older, and the one accepted for many years,is that of the chemicalprecipitation of the cole- manitefrom thewaters of inclosedbasins d'uring periods of great desiccation. This theory was first put forward by W. H. Storms•ø and later supportedby Campbell,u Keyes,•2 Baker,•a and others. It fails, however,to explain the reactionby which suchmasses of colemanitemight form. In order to producebeds of any thicknessthe amountof water would have to have beenvery great, becauseof the slight solubilityof the colemanire. Since the borate-bearingbeds have all the appearanceof shallow-water deposits,it is difficult to conceive any means whereby such massescould be depositedby simpleevaporation. In the playa lakesof todayno precipitationof colemaniteis evident. In fact, under ordinary playa conditions,where there are large amounts of borax present,the formation of colemanireseems impossible; for under these conditionsany simple calcium borates are pre- vented from forming, any calcium in solution separatingas the double sodiumcalcium borate, ulexite. Colemaniteforms only from solutionsfree of sodiumborate, a conditionwhich is appar- entlynever'reached in the playa lakes. ' These relations have been discussedunder the separateminerals. The secondtheory is oneput forth by H. S. Gale.TM From his studiesof the Ventura depositshe reached'the conclusionthat the colemanitewas formed by metasomaticreplacement of lime- stone beds intercalated with beds of shale between basalt flows. xoCalif. State Min. Bureau, Ixth Ann. Rep., p. 346, I893. axU.S. Geol. Survey Bull. 20o, p. 8, I9o2. • Am. Inst. Min. Eng. Bull. 34, I9o9. aaUniv. Calif. Pub. Dept. Geol. Bull., vol. 6, No. •5, •9xI. x• Prof. Paper 85, U.S. Geol. Survey, p. 3, •914. 212 WILLIAM F. FOSHAG.

In supportof thishypothesis he callsatt.ention to the irregular characterof the depositsand the vein characterof the associated gypsum. Furthermorehe statesthat "certain specimenscollect4d on the ore dumpsand in the minesshow that at leasta part of the colemanireis a replacementdeposit. Irregular portionsof the limestoneare surroundedby white crystallinecolemanite and minute fractureswhich traversethe limestonethroughout are also filled with this mineral. These veinlets are observed to have been enlarged irregularly within the limestone. Small rounded massesof limestoneare also included.within the solid portions of the colemanite,indicating that in placesthe sameenlargement of the intersectingveinlets has been carried to a furtherstage and the separated portions of the limestone are residual within t•e depositedcolemaniteJ' These reactions are possible, for, although boricacid is a muchweaker acid than carbonic,the greatervola- tility of the latter may allow replacementto take placein concen- trated and hot solutions of boric acid. The reverse reaction takes placein nature,however, to a great extent and colemanitecrys- tals altered to calcium carbonate are common in the borate de- posits. It is doubtful if the boric acid emanationsfrom the neighboringbasalts were of sufficientvolume and of long enough duration to effectthe replacementof suchlarge amountsof lime- stones. Being surfaceflows, their periodof activitymust have been short. What replacementof limestonehas taken placeis more probablydue to boratesolutions of vadoseorigin. Eakle•5 arrived at a similar hypothesisto the abov.e. He con- sideredthe Lang depositsas probablyderived from marlsby the action of boric acid solutions.

0R•GIN OF THE DEPOSITS. Van't Hoff's work clearly demonstratesthat, under the con- ditions existing in playa lakes, ulexite forms instead of cole- manite,and in orderto accomplishthe splittingof the ulexiteinto Univ. Calif. Pub. Dept. Geol.,vol. 6, No. 9, I9II. ORIGIN OF COLEMANITE DEPOSITS.

its component calcium borate and sodium borate the sodium borate must be removed as fast as it is formed. In the closed basinsof the playa lakeswith their clay floors this removalof the sodiumborate cannottake place to any large extent either by surfaceor undergrounddrainage, so that the borateaccumu- lationsconsist wholly of ulexiteand borax. When thesedeposits are later coveredover and uplifted sufficientlyto allow free drainage from the beds, the percolation of chloride solutions graduallyconverts the ulexite to colemaniteand other members of the colemanite series. The evidenceto supportthis hypothesisis: I. The total absenceof colemaniteor any memberof the cole- manite seriesin any of the playa lake deposits. 2. The occurrenceof ulexite in large quantitiesin the borate bedsat Lang, where it lies for the most part near the foot wall. 3. The beddedcharacte• of the deposits. 4. The structuralfeatures of the deposits,especially the nodu- lar and geodalcharacter of a large part of the coelmanite. The first three points have already been treated. The fourth item will now be discussedbriefly. It has beenpointed out that nodular and geodalmasses are commonin the deposits,in fact at the Calico bed•sand presumablyat those at Death Valley a large proportionof the ore consistsof nodulargeodes embedded in the clays. This suggestsstrongly LeConte's description of the occurrenceof the "cotton-balls" in the claysof RhodesMarsh. If the theory of the derivation of colemanitefrom ulexite be accepted,there is a ready explanationof this type of ore. The ulexite embeddedin the claysis actedupon by salt solutions. The light, fluffy "cotton-balls"are convertedinto the morecompact ½olemanite,giving rise to the more or lessspherical geodes and allowingthe free crystallizationof the colemanitein the center. It is inconceivablethat the cavitieswere original in the shales and that they were later filled•in with the colemanite. Where ulexite was aggregatedin more compactmasses the colemanite took a more massive form but the contraction of volume still allowed for a large number of drusy cavities such as are so I/VILLI.4M F. FOSH.4G. abundantin the deposits. In somedeposits, as at Lang, the pres- sure exerted was sufficientto closethe cavities and compactthe mass. The selenitegypsum and perhapssome of the limestone also, resultedfrom the action of sulphatedand carbonatedwaters uponthe colemanite. SMITHSONIAN INSTITUTION, U.S. NATIONAL MUSEUM, WASHINGTON, D.C.