Relations of zoned pegmatites to other pegmatites, granite, and

American Mineralogist, Volume 75, pages 631455, 1990

Relations of zoned pegmatites to other pegmatites, granite, and metamorphicrocks in the southern Black Hills, South Dakota

Jnvrns J. Nonrox*Branch of Central Mineral Resources, U.S. Geological Survey, Federal Center, Denver, Colorado 80225, U.S.A.

J.lcr A. RnonrNBranch of Central Mineral Resources, U.S. Geological Survey, Federal Center, Denver, Colorado 80225, U.S.A.,

and South Dakota School of Mines and Technology, Department of Geology and Geological Engineering,Rapid City, South Dakota 57701, U.S.A.

Ansrnlcr

The pegmatite field and the Harney Peak Granite of the southern Black Hills, SouthDakota, form an igneous system that progrcsses from slightly biotitic muscovite granitethrough layered pegmatitic granite, with alternating sodic and potassic rocks, to simpleplagioclase-quartz-perthite pegmatites, and on to zoned pegmatites. The zoned pegmatitesrange from rather simple ones, which have been mined for sheet muscovite or potassiumfeldspar, to complex ones with minerals of Li, Be, Nb, Ta, Sn, and Cs.

Most of the country rocks are Lower Proterozoic mica schists that were isoclinally foldedand metamorphosed to biotite and almandine grades before granitic activity began. At1700 Ga, intrusion of the Harney Peak Granite created a large dome in these rocks, athermal aureole with a staurolite, a first sillimanite isograd, and a small area of metamor-phism above the second sillimanite isograd. The estimated pressure is 3.7 kbar.

The granite has an area of 104 km'z, mostly in a single pluton consisting of thousandsof sills and dikes. The pegmatite field is exposed over an area of 711 km'?. An isogrammap of the abundance of pegmatites has many local irregularities. Calculations from theisograms show a total of 24000 pegmatites. About 2o/o of these are zoned pegmatites.

The zoned pegmatites have a strong tendency to occur in clusters, and the types ofpegmatites are different in diferent clusters. A less obvious tendency is a regional zonationin which rare-mineral pegmatites become more abundant and muscovite pegmatites lessabundant toward the outskirts of the region. This trend is parallel to a drop in crystalli-zation temperatures from more than 660 "C in parts of the granite, some of which haveprimary sillimanite instead of muscovite, to less than 600 'C in inner zones of a few rare-mineral pegmatites, many of which are outside the first sillimanite isograd.

The composition of the granite indicates that its magma originated by partial meltingof metasedimentary mica schists similar to those at the present surface. The pegmatiticnature of most of the granite probably reflects exsolution of an aqueous phase. The re-sulting processes continued into the simple pegmatites and zoned pegmatites, which are4o/o and 0.50/o of the exposed granitic rocks. For rare-element pegmatites, residual concen-tration in granitic magma can account for their modest contents of Be, Sn, Ta, Nb, andCs, but Li presents complications because a large share of the Li in the exposed parts ofthe system is in Li-rich zoned pegmatites. The clustering of zoned pegmatites shows themto be products oflocalized subsystems that behaved differently.

fNrnooucrroN

The Harney Peak Granite pluton and its several satel-litic plutons in the southern Black Hills, South Dakota,are surrounded by a region containing about 24000 peg-matites. Approximately 300 of these are zoned wellenough to have been mined or intensively explored forcommercially valuable minerals. The zoned pegmatitesare also the ones that have attracted the most petrologic-

t Present address: 3612 Wonderland Drive, Rapid City, SouthDakota 57702, U.S.A.

0003404x/90/0506463 l $02.00

interest. A major purpose of this article is to examine thedistribution of zoned pegmatites across the region andthus to add to the body of information that bears on theissue of how they emerge from the evolution of the granit-ic system.

The pegmatites and granite are exposed over an areaof 824 km'z(3 18 sq. mi.). Granite in the main pluton hasan area of 84 km2 (32 sq. mi.) and outlying bodies add20 km2 (8 sq. mi.). The pegmatite region is narrow aroundthe north side ofthe granite but very broad to the south-west and south (Fig. l). The pegmatite field goes under

63r

NORTON AND REDDEN: ZONED PEGMATITES IN THE BLACK HILLS

R.sE. 103030'

T . 1 S .

tU.. tl

;i *rllX9*tc,*f.,Y..fi "ii ii itiii

NATIONAL PARK I

Paleozoic sedimentary rocks on its southwest, south, andeast sides.

The granite is a sodic leucogranite with textures rangingfrom fine grained to pegmatitic. The chemical composi-tions of the granite and the pegmatites are very similar.Pegmatites of the variety generally called layered peg-matites closely resemble the more pegmatitic parts of thegranite. About 800/o of the pegmatites, however, are sim-ple coarse-grained bodies consisting almost entirely of al-bite, quartz, perthite, and muscovite. The simplest zonedpegmatites have wall zones rich in muscovite or innerzones rich in perthite. Others are more complex; a fewhave numerous zones and contain concentrations of sev-eral economically important minerals. The locations ofzoned pegmatites are in Figure 1, and the names of theprincipal ones are listed in the caption.

Long ago Van Hise (1896, especially p. 687-688) saidthe Black Hills have a regional change from a granitecenter to pegmatites nearby and on to banded veins (pre-sumably zoned pegmatites) and ultimately to highlyquartzose veins. Regional zonation has since been de-scribed in many articles about pegmatite districtsthroughout the world. Though Van Hise was right in ageneral way, the actual distribution in the Black Hills hasmany irregularities and complexities that cannot be rec-onciled with the simplicity of the concept of regional zo-nation.

More recently, the tendency in discussing regional zo-nation has been to treat the distribution only of zonedpegmatites, especially those containing certain rare ele-ments. This, for example, has been done in the BlackHills on the basis of Li, Rb, and Cs by eernj' and True-man (l 978, p. 372-3'1 3).

Our own approach depends heavily on the distributionand abundance of unzoned pegmatites as indicators ofthe intrusive pattern because these are far more commonthan the zoned pegmatites. The literature about other re-gions is vague about whether unzoned pegmatites aregenerally as abundant as in the Black Hills or whethersome regions contain proportionately many more zonedpegmatites.

O J J

The Black Hills district has an unusually great varietyof zoned pegmatites. This diversity allows the zoned peg-matites to be divided into several categories. Differencesin the distribution of the pegmatites of these categoriessuggest that genetic processes were somewhat diferentfrom one place to another.

Rncrox.ql sETTTNG

The Black Hills are an elongate Laramide dome havinga north-northwest-trending Precambrian core flanked byPaleozoic and Mesozoic sedimentary rocks. The regiontreated in this report (Figs. I and 2) occupies the southernhalf of the Precambrian core. Pegmatites are absent tothe north except in two small localities so far away as tohave no interpretable relation to the Harney Peak system.

The most recent summary of the Precambrian geologyof the Black Hills was published by Redden and Frenchin 1989. The region has folds of several ages, numerousfaults, and significant facies changes. The most obviousstructural trend is in a north-northwest direction.

The latest known Precambrian event was the crystal-lization of the Harney Peak Granite and its pegmatites at1.7 Ga (Riley, 1970). The rest of the pegmatite regionconsists almost wholly of Early Proterozoic metamorphicrocks, predominantly of sedimentary origin. By far themost abundant rocks are metamorphosed graywackes andseveral kinds of metamorphosed shales. Other rocks in-clude quartzite, amphibolite from basalt flows and fromgabbro, and metamorphosed conglomerate, arkose, ironformation, and dolomite. Figure 2 shows the distributionofthe rock types.

The last major Precambrian tectonic event was the de-velopment of domes during the intrusion of granite in thesouthern Black Hills. The largest of these is the HarneyPeak dome, which resulted from the growth of the mainpluton by the emplacement of a succession of sills andother intrusions. The surrounding area has one small domewith granite and several others that gravity data show areprobably underlain by granite (Duke et al., 1990).

The Bear Mountain area, which is in T.2S., R.3E., l3km (8 mi.) west of the Harney Peak Granite, is also a


Fig. 1. Map showing distribution and abundance of pegma-tites and granite in the southern Black Hills, South Dakota. Is-ograms show the number per sq. mi. of pegmatites and smallbodies of layered granite. Symbols show the zoned pegmatitesby category: l, dot; 2, circle; 3, x; 4, solid square; 5, open square;6, triangle; 7, plus sign. The letter f signifies a zoned. fracturefilling in a small body of layered granite. Short dashed linesoutline the large clusters A to F of Table 7. Small clusters of thesame table are labeled a to j. The names of numbered pegmatitesare as follows (asterisks designate large mines of Tables 6 and8): l. Cowboy 2. Mohawk 3. Mateen* 4. Hardesty 5. Hunter-Louise 6. Oreville Spar 7. High Climb 8. Tin Queen 9. Dewey10. Gap 11. Westinghouse No. 4 12. November 13. Bob Inger-soll No. 2* 14. Bob Ingersoll No. 1* 15. Sitting Bull 16. DanPatch* 17. Peerless* 18. Hugo* 19. Etta* 20. White Cap* 21.

Edison* 22. Hesnard 23. Dyke 24. Eureka 25. Big Chie{x 26.Mountain Lion (Soda Spar) 27. Barker-Ferguson 28. Wood Tin29. Star (Nichols) 30. Wilhelm 31. Old Mike* 32. Tin Moun-tain* 33. Warren Draw 34. Highland* 35. Rachel D. 36. Crown*37. Lost Bonanza* 38. White Spar* 39. Victory* 40. Florence41. Harbach 42. Climax* 43. Shamrock and L5 No. 3* 44. Ag-new* 45. Custer Mountain 46. Glenwood 47. Bull Moosex 48.Burt 49. Earl 50. Elkhorn 51. Hot Shot* 52. Triangle A 53. St.Louis* 54. Michaud Beryl 55. Helen Beryl 56. New York* 57.Tiptop* 58. Buster* 59. Pine Top 60. Silver Dollar 61. WhiteBear 62. MacArthur 63. Jack Rabbit 64. Ruby Reef 65. Midas66. White Elephant* 67. Beecher No. 2* 68. Beecher Lode* 69.Beecher No. 3 -Black Diamond* 70. Number 9 71. Rose Quartz72. Red Deer* 73. Roadrunner 74. Ann 75. Charles Pringle 76.Townsite* 77.Dakota Feldspar 78. Greene 79. Smith*

634 NORTON AND REDDEN: ZONED PEGMATITES IN THE BLACK HILLS

1 03030' R.6E.

T.3S.

T.4S.

EXPLANATIONPz Paleozoic rocksXh Hamey Peak GraniteXmg Metagabbro(amphibolite)Xbs Biotite-gamet schistXms Mica schistXgw Quartz-mica schist and metagraywackeXif Meta'carbonate-silicateiron formation

Xcq Metaconglomerate,quartzite,gamet schist, and metabasalt

Amphibole schist, calcsilicate gneiss,and biotite schist

Quartzite, marble, amphibole schist,and biotite schist

Quartzite and mica schistBiotite schist and pegmatitic graniteFaultContact

Xac

xqm

g . *'t5t*t

-uMILES

Xq

=

Fig. 2. Generalized geologic map of the southern Black Hills, South Dakota. The Harney Peak Granite and the Precambrian-Paleozoic contact are shown in greater detail on Figure 1. Metamorphic units shown by rock type; stratigraphic order not impliedby list of units.

large dome and probably has a 1.7-Ga granite beneath it(Duke et al., 1990). At the surface, mica schist is intrudedby very late Archean granite and pegmatite. Unconform-ably above these Archean rocks is metamorphosed con-glomerate containing clasts of granite and pegmatite.Overlying rocks include arkose, quartzite, amphiboleschist, dolomitic marble, and garnetiferous schist (Red-den, 1968; Ratt6, 1986). Similar rocks are exposed inscreens and inclusions in granite in the center of the Har-ney Peak dome. Whether Archean rocks are also exposedis unknown. If they are, the lower part of what is mappedas Harney Peak Granite may include Archean granite,equivalent to the granite exposed at Bear Mountain, whichis similar to the Harney Peak Granite.

All of the metamorphic rocks shown on Figure 2 areabove the almandine isograd. Metamorphic intensity in-creases to the south. A staurolite isograd and then a firstsillimanite isograd roughly follow the northern boundaryof the pegmatite region. Andalusite first appears near thestaurolite isograd but slightly farther from the granite.Cordierite also occurs in some of the schists of suitablecomposition. Staurolite disappears a few kilometers in-side the first sillimanite isograd. The highest grade ofmetamorphism is marked by the second sillimanite iso-grad on the southwest side of the Harney Peak plutonand has been noted in isolated places to the southeast inthe area of quartzite and schist. Sillimanite also occurssparsely in the granite in southerly parts of the plutonand in some nearby pegmatites. Kyanite is absent in themetamorphic rocks near the granite but occurs in the BearMountain dome and in a few places to the south.

The metamorphism consists of a thermal event and anearlier regional event. The effects of the regional meta-morphism, which is almost entirely below staurolite grade,extend throughout the Precambrian core of the Black Hills.The thermal metamorphism is related to the emplace-ment of the granite. Not only is this apparent from theregional relation to the main pluton, but also small areasaround outlying granite bodies show local increases inmetamorphic grade.

H.l.nNBv PEAK GRANTTE

The Harney Peak Granite pluton is a multiple intru-sion that consists ofperhaps a few dozen large sills, whichhave gentle dips in accord with the domal pattern, andthousands of smaller sills, dikes, and irregularly shapedintrusions. The largest sills are probably no more thanabout 100 m thick and extend laterally for only a fewkilometers. Detailed mapping has not been practicablebecause of the multiplicity of the intrusive bodies andtheir similarities in composition. The extreme complex-ity of parts of the granite is shown by the work of Dukeet al. (1988) on one of the satellite localities. The prin-cipal minerals of the main pluton and the smaller outly-ing plutons are, in order ofabundance, albite-oligoclase,qvartz, perthite, and muscovite. The most common ac-cessory minerals are tourmaline (schorl), biotite, garnet,and apatite. The texture ranges from aplitic to granitic to

63s

pegmatitic, but nearly all outcrops have at least somepegmatitic aspects. Pegmatitic segregations and fracturefillings or dikes are common. Most of the perthite occursas megacrysts in a phaneritic plagioclase-quartz matrixand in lenses and layers parallel to the nearest contact.The granite has recently been described in some detail byRedden et al. (1985), and further information is in Shear-er er a l . (1987).

Several indicators point to a trend from less evolvedgranite in the central and southeastern parts of the mainpluton to more evolved and more pegmatitic granite tothe northeast, north, and west, and in the outlying plu-tons. Shearer et al. (1987) showed that the K-Rb ratiosof samples from interior intrusions are generally higherthan in samples from several marginal intrusions. Un-published K-Rb data from bulk samples of Norton andMclaughlin confirm these results. A parallel change is anincrease in the amount of coarse perthitic lenses and lay-ers and in the abundance of pegmatitic fracture fillings.The muscovite content increases from less than 50/o inmuch of the central and southeastern parts of the mainpluton to l0-l5o/o to the north and northeast. The biotitecontent is low everywhere but does exceed lolo where themuscovite content is low, yet becomes almost nil to thenorth, where tourmaline is the principal mafic mineral.The median Li content in the granite is 30 ppm (Norton,198 1), and the richest analyzed sample, with l7l ppm(Shearer et al., 1987, Table l8), is at the north border ofthe main pluton. The anorthite content of plagioclase,which ranges from Ano to at least Anrt (Shearer et al.,1987, p.475), also tends to be highest in the less peg-matitic parts of the granite. All of these remarks are gen-eralizations. Detailed mapping of the granite accom-panied by thorough sampling to find differences betweenintrusions and within single intrusions would show farmore complex patterns.

The most conspicuous source of heterogeneity is thelayering caused by the alternation ofplagioclase-rich rockwith pegmatitic perthite-rich lenses and layers. Locally,the plagioclase-rich rock is aplitic and layered on a muchsmaller scale, forming line rock in which layers 0.5-10cm thick have sharp differences in the abundance oftheirminerals, especially tourmaline, feldspar, garnet, mus-covite, and quartz. Most of the line rock is near the edgeof the main pluton or in satellitic plutons or pegmatites.

The heterogeneity of the granite causes obvious diffi-culties in determining bulk composition. In about 130samples used by Redden et al. (1985) and Shearer et al.(1987), the total of SiO', AlrOr, NarO, and KrO is gen-erally 97-98o/o. The remainder is mostly HrO, CaO, andiron oxides. The median muscovite content is 80/o (Red-den et al., 1985, Fig. 2) and the total of feldspar andquartz is about 900/0. This allows the rock to be repre-sented on Ab-Or-Q ternary diagrams, which has beendone by Redden et al. (1985, Fig. 5) and Shearer et al.(1987, Fig. 4). The diagrams show a wide scatter. Onereason for the scatter is that many of the samples do not,and were not intended to, represent the entire rock.


636

Ab orFig. 3. Ternary diagram showing Ab-Or-Q contents of small

layered intrusions calculated from visually estimated modes.Sources ofmodes are: Redden (1963b, Table l0), Lang and Red-den (1953, Table l), and unpublished notes. In the calculations,all plagioclase is assigned to Ab, perthite is assigned 25o/o to Aband 75o/o to Or, 700/o of the muscovite is assigned to Or, andquartz is assigned to Q. Where graphic granite is reported in themode, 900/o of it is assigned to perthite and 100/o to Q.

Another reason is differences in different parts of singleintrusions, probably caused mostly by an increase in Kand a decrease in Na from the bottoms to the tops. Sev-eral samples of Redden et al. (1985, Fig. 5) that are fromnear the edge of the main pluton have a high content ofalbite relative to orthoclase, much like the albitic wallzones of zoned pegmatites. Most determinations are be-tween 30 and 45o/o qvartz,25 and 500/o albite, and l5 and350/o orthoclase, but probably the bulk compositions ofthe larger sills and dikes have a smaller range than this.The average normative composition is approximatelyplagioclase (Ab + An), 40o/o; quartz,350/o; and orthoclase,25o/o.

Slrnr,r. LAYER-ED TNTRUSIoNS

The pegmatite region contains many small intrusionsthat previously have been called layered pegmatites (e.g.,Redden, 1963b) but that actually differ from pegmatiticgranite mainly in their size. We will refrain from callingthem pegmatites in the text of this article, but it is im-possible to avoid treating them as pegmatites in Figure Iand in some of the tables.

In total quantity ofrock the layered bodies far exceedthe true pegmatites. Some of them are hundreds ofmeterslong and tens of meters thick. At the other extreme aresmall layered bodies only about a meter thick and a fewmeters long. Most of the small ones are dikes, cross-cutting foliation and bedding of the enclosing rocks. Somenarrow dikes, especially near Keystone, are more than a


kilometer long. The large intrusions, however, are highlyirregular and both concordant and discordant.

Microcline is most abundant in coarse-grained layers,plagioclase is most abundant in the rest of the rock andis of low anorthite content, tourmaline is common, andbiotite is rare. The layering results principally from theparallelism of lenses and layers of coarse-grained perthiticrock surrounded by finer-grained rock consisting ofpla-gioclase and quartz. Contacts between layers are ordinar-ily gradational, but some are sharp. The intrusions canhave hundreds of lenses and layers from the footwall tothe hanging wall. In the outer parts of the intrusions thelayering gradually fades out through the disappearance ofcoarse layers. The proportions ofthe two kinds oflayershave a wide range; the average is probably about one-fourth coarse grained to three-fourths finer grained.

Line rock is more abundant than in the large plutonsor in true pegmatites. Though it is a rather uncommonrock and of small areal extent in individual localities, itoccurs in hundreds of places. Most of it is near contacts,especially footwall contacts.

Some of the small layered bodies, like the large plutons,are multiple intrusions. They have a structure that is sim-ilar but not identical to layering if the separately intrudedmembers have perthitic cores and albitic border zones,or if they are perthite-rich at the top and albite-rich atthe bottom (cf. Redden, 1963b, p. 235). Spectaculardrawings of such multiple intrusions have been publishedby Orville (1960, Figs. 4-6) ar.d Duke et al. (1988, Fig.5).

Bulk compositions of layered bodies plotted on Figure3 show a wide scatter, as in the granite. They are closerto the albite apex, but this is partly because plagioclaseof the granite has more anorthite.

Srprpr,n PEGMATITES

The simple pegmatites, which in previous publicationshave been called homogeneous pegmatites, differ fromlayered granite and zoned pegmatites by lacking easilydiscernible internal structure, though they do ordinarilyhave a fine-grained border zone a few cm thick. Theyconsist of the same minerals as the layered bodies but aresomewhat coarser grained. Perthite again forms the larg-est crystals. Some of the simple pegmatites have vaguelayering and others have rudimentary zoning, for they aretransitional between pegmatitic granite and zoned peg-matites.

Most of the simple pegmatites are small, and thoughthe total number of such bodies is large, the quantity ofrock contained in them is modest. They generally aretabular to lenticular sills concordant with the foliation orbedding of the host rock. Most of them are between l0and 100 m long, and only a few are more than 300 mlong.

Figure 4 shows that the estimated compositions of sim-ple pegmatites have a much wider scatter than the layeredintrusions of Figure 3. The pegmatites are homogeneousin the two horizontal dimensions, which is why they have

NORTON AND REDDEN: ZONED PEGMATITES IN THE BLACK HILLS 637

previously been called homogeneous pegmatites, but theyalmost certainly have vertical changes in composition.Hence the erosion level influences the observations andis probably a major cause of the scatter. The open circleson Figure 4, which are from study of an apparently simplepegmatite, show how compositions from one pegmatitecan extend from the orthoclase-poor to the orthoclase-rich parts ofthe field containing the dots on this graph.

Zoupn PEGMATITES

The zoned pegmatites have attracted the most geologicattention because of their commercial importance andtheir petrologic peculiarities. They have also yielded themost information because extensive mining and explo-ration have facilitated detailed study, Thus they are thebest-known kind of pegmatite, even though by far theleast abundant. The principal descriptions ofthe individ-ual zoned pegmatites are in Page and others (1953), Sher-idan (1955), Sheridan etal. (1957), Redden (1959,1963a,1963b), Norton et al. (1962), Staatz et al. (1963), andNorton and others (1964).

The zoned pegmatites are, with few exceptions, rathersmall. The mineable concentrations within them, whichare zones or parts of zones, are smaller still, generallyvery much smaller. Most of the pegmatites are betweenl0 and 100 m long and less than 15 m thick. The smallestshown on Figure I are sheet muscovite pegmatites thatare as little as 20 m long and I m thick. Countless zonedpegmatites of still smaller size are scattered through theschists ofthe region but are ignored for obvious reasons.Thousands of small zoned bodies in granite or in simplepegmatites are also ignored because they are merely unitsof the larger intrusion that is their host, but the map doesshow several zoned pegmatites in the granite and sevenzoned fracture fillings in small layered intrusions becausethey are large enough to have been ofeconomic interest.The four largest Black Hills zoned pegmatites that arereasonably well known in three dimensions are the Hugoand Peerless at Keystone and the Helen Beryl and Bee-cher No. 3-Black Diamond near Custer. Each of thesecontains at least 500000 t ofrock, and the largest hasabout 3000000 t. These arelarge by any standard, butthe world does have a few zoned pegmatites containingtens of millions of tons of rock.

The compositions and arrangements of the zones re-flect differentiation from the outer to the inner parts ofthe pegmatites, but this is accompanied in many placesby a vertical component ofdifferentiation that causes anupward increase in the abundance of K relative to Na(Norton, 1983). Fracture-fllling units are numerous butsmall. Maps and descriptions of individual pegmatitesindicate the existence of fewer replacement bodies thanare said to exist in many complex pegmatites, but theBlack Hills have had an ample share of advocates of mas-sive replacement (cf. discussion in Norton et al., 1962, p.100-103).

In zoned pegmatites the many differences in the com-positions and sizes of zones are major obstacles to esti-

Ab Or

Fig. 4. Ternary diagram showing Ab-Or-Q contents of sim-ple pegmatites. The dots are calculated from modes in the sameway as for Figure 3 and from the same sources. The open circlesand the lines connecting them are from Norton (1970, Fig. a).They show changes in composition in two drill holes from thefootwall to the hanging wall of a pegmatite that at the surfaceappears to be simple.

mating bulk composition. Nevertheless, the overall com-position even of a very complex pegmatite can beestimated if surface exposures, together with under-ground workings and perhaps drill holes, allow reason-ably complete three-dimensional interpretation. At theHugo pegmatite the size of each of seven zones and twosmall replacement bodies was calculated from a series ofvertical and horizontal cross sections, and the results werecombined with estimated modes of these units to com-pute the bulk composition of the pegmatite. Norton et al.(1962, p. lll-118) described the procedures and dis-cussed the improbability of having significant errors. Asimilar calculation of the bulk composition of the Peer-less pegmatite, which is also highly complex, was basedon vertical cross sections made with the assistance of datafrom numerous drill holes (Sheridan et al., 1957, p. 17-18). The points on Figure 7 for the bulk compositions ofthese two pegmatites show them to be more silicic thannearly all the compositions of simple pegmatites. TheHugo, which has been alarge producer ofpotassic feld-spar, is much nearer the orthoclase-qvafiz sideline thanthe Peerless, which is less potassic than most pegmatitesand has been commercially notable mostly for beryl andscrap muscovite mined from an outer zone.

The elaborate geometric work done on the Hugo andPeerless pegmatites is not possible nor practicable withmost other zoned pegmatites because of difficulties in de-termining the geology above the surface or below thedeepest mine workings. The principal difference fromsimple pegmatites and granite is the presence in numer-


Fig. 5. Ternary diagram showing Ab-Or-Q contents of wallzones and cores of pegmatites that have only two zones. X :wall zones. Dots : cores. Calculated as for Figure 3. Sources ofmodes are: Redden (1963b, Table 7), Lang and Redden (1953,Table 2), Page and others (1953, p. 82), Staatz et al. (1963, p.1 64), and unpublished notes.

ous zoned pegmatites of large quartz-rich inner zones-either quartz cores or quartz-feldspar zones or quartz-spodumene zones. Sn pegmatites near Hill City containso much quartz that they are often called quartz veins.On the other hand, many zoned pegmatites, expeciallyamong those mined for muscovite or potassic feldspar,lack highly quartzose zones and appear to be near thesimple pegmatites and granite in bulk composition.

Figure 5, which shows the compositions of the zonesin pegmatites that have only two zones, offers a way ofestimating the range of bulk compositions. In these peg-matites, most wall zones are near the albite-quartz side-line, and most cores are rich in orthoclase or quartz orboth. Ifthe exposed parts ofthe two zones are represen-tative of the entire pegmatite, the bulk composition issomewhere along the line between the points for the twozones. If a second pegmatite has the same bulk compo-sition, the intersection of the lines for the two pegmatitesshows that composition. About 800/o of the intersectionson Figure 5 are between 35 and 500/o Q, which indicatesthat many of the pegmatites are more silicic than simplepegmatites and granite, and most of the others are in thesilicic parts of the fields for simple pegmatites and gran-ite. The wall zones and the inner zones ofFigure 5 eachhave a median quartz content of 4lo/o, which is notablyhigher than the medians of simple pegmatites and granite.

Figure 6 shows that the compositions of the zones of26 pegmatites with more than two zones are scatteredacross nearly all of the Q-Ab-Or diagram. Again, com-positions of wall zones are near the albite-quartz sideline.

Fig. 6. Ternary diagram showing Ab-Or-Q contents of zonedpegmatites that have three or four zones (A) or five or morezones (B). X : wall zones. Squares : spodumene-rich zones.Triangles : lepidolite or lithian muscovite zones. Dots : otherzones. Calculated as for Figure 3. Several line patterns are usedso that the pegmatites can be distinguished from each other.Sources ofmodes are Redden (1963b, Table 7), Lang and Red-den (1953, Table 2), Page and others (1953), Norton and others(1964),Staatz et al. (1963), Sheridan (1955, Table l), Sheridanet al. (1957, Table 3), Redden (1959, p. 543), Redden (1963a,p. 2-8), Norton et al. (1962, Tables 19 and 20), Jollitret al.(1986, Fig. 20), and unpublished notes.

These generally are followed by orthoclase-rich interme-diate zones and then by silicic zones, but the graph showsmany twists and turns along the way caused by innerzones rich in albite andquurtz, some of which have spod-umene. Three pegmatites-the Bob Ingersoll No. l, theHugo, and the Peerless-have lepidolite or lithian mus-covite cores that follow silicic inner zones.


The median quartz content of the zones of Figure 6 is41010, just as it was for the bizonal pegmatites of Figure5. The quartz contents ofthe Hugo and the Peerless areslightly greater. Probably pegmatites with many zoneshave about the same range of bulk composition as peg-matites with two zones, but one cannot determine thisfrom Figure 6 because the zones have a wide range inS1ZE.

Figure 7 shows an appraisal of how zoned pegmatitescompare in bulk composition with the other pegmatitesand with granite. The most notable feature is the simi-larity in the composition of the four kinds of rocks; everyfield on Figure 7 overlaps every other field. The graniteminimum is within all fields. The granite field juts outtoward the quartz-orthoclase sideline, but probably wouldnot do so if its An content were added to its albite so asto include all the plagioclase and thus make the calcula-tion more like that of the pegmatites, in which all plagio-clase is treated as albite. The diameter of the field forsimple pegmatites indicates a Ereater spread of bulk com-positions than is likely to be correct, for reasons alreadydiscussed. The zoned pegmatite field overlaps the moresilicic parts of the other fields and extends from theretoward the quartz apex. It would reach much more siliciclevels if the highly quartzose Sn-bearing pegmatites or"veins" were not excluded from the diagram.

Some of the zoned pegmatites have been mined for Li,Be, Ta, Nb, Sn, and Cs. A few pegmatites have enoughLi to have been mined for spodumene almost from wallto wall. Beryl has been a major product from several peg-matites. Nonetheless, the amount of rare elements is smalleven in ore bodies and is so much smaller elsewhere inpegmatites that few analyses are available. As for non-commercial trace constituents, anions as well as cations,the body of data is still slight. Knowledge in this field isbeing increased by Papike and others in numerous pub-lications from the Institute for the Study of Mineral De-posits of the South Dakota School of Mines and Tech-nology (e.g., Jolliffet al., 1986, 1987; Walker et al., 1986a,1986b; Shearer et al., 1985, 1987).

Clrnconrns oF zoNED pEGMATITES

The region has a remarkable variety of kinds of zonedpegmatites, perhaps more so than in any other pegmatitefield in the world. It has muscovite-rich pegmatites likethose mined in sheet mica districts the world over; it hasLi deposits that dominated the industry for many de-cades; it has major sources of beryl; it has many feldsparmines; it has pegmatites that have been intensively ex-plored for cassiterite; and it has produced white and roseqtartz, tantalum-niobium minerals, and pollucite. TableI shows the characteristics of the pegmatites that havebeen used to separate them into seven categories. Eachof the zoned pegmatites shown on Figure t has been as-signed to one of the categories. Figures 8 and 9 showhorizontal and vertical sections of pegmatites that areexamples of the various categories.

Category 1 includes most of the sheet muscovite peg-

Ab Or

Fig. 7. Ternary Ab-Or-Q diagram showing the fields thatcontain most of the granite in the large bodies and approximately900/o of the small bodies of layered granite and the simple andzoned pegmatites. The field for large bodies ofgranite is basedon Redden et al. (1985, Fig. 5) and Shearer et al. (1987, Fig. 4).The other fields are based on Figures 4-6. H and P are the bulkcompositions of the Hugo and Peerless pegmatites (Norton etal., 1962; Sheridan et al., 1957). The numbers l, 3, 5, and 10are granite minima and eutectics from Luth et al. (1964) in kbar.

matites and nearly all the important mines. The musco-vite is in wall zones with quartz and sodic plagioclase.Most of the mineable concentrations are on the hanging-wall side of the pegmatite, and the greatest concentrationsare in rolls in the contact, where the muscovite contentcan be as great as 500/o (e.g., the Crown mine, Page andothers, 1953, p. 95-96). Muscovite crystals are generallymost abundant and largest near the contact and decreasein abundance and size toward the inner part of the peg-matite. Some of the pegmatites have quartz cores, andsome have as many as three feldspar-quartz inner zones.Most of these pegmatites have a lenticular shape.

Category 2 differs from Category I only by havingperthite as well as plagioclase in the wall zone. Many ofthese are similar to simple pegmatites except for the en-richment in muscovite near the contact. Others are well-zoned pegmatites, and in some of these the muscovitebooks are largest and most abundant in the inner part ofthe wall zone, which may be mapped as the first inter-mediate zone. A feature of several pegmatites in Category2 is an unusual abundance of schist remnants in the wallzotle.

Category 3 consists of pegmatites in which the mostprominent unit is an intermediate zone or core rich inlarge crystals of perthite. The mining of potassic feldsparhas centered on these deposits. The most productive peg-matites are oflarge size and oflenticular or irregular shape,but much thicker than the pegmatites of Categories I and2. The wall zone is ordinarilv thin. and its muscovite


Fig. 8. Horizontal sections of zoned pegmatites. Essential minerals of the zones are shown in the order of their abundance. Q:quartz.Ab:albite.Cl:cleavelanditevarietyofalbite.P:peflhite.Mi:microcl ine.Mu:muscovite.LiM:l i thianmuscovite.

Sp : spodumene. Am : amblygonite. Bi : biotite. Stippling indicates replacement, mostly by albite, lithian muscovite, andmicrocline.

sparse and of low quality. The perthite commonly is con-centrated in a hood-shaped unit in the upper part ofthepegmatite; such units carry quafiz and albite as well asperthite, but they grade downward into units with lesseramounts of perthite or none at all. The geologic sectionof the Dan Patch pegmatite (Fig. 9) has an unusual sharpro11 in the hanging wall that divides the pegmatite intotwo nearly independent segments, but each of the seg-ments has the uncomplicated nature of most of these de-posits. The Dan Patch is much larger than most pegma-tites of Categories I ar.d 2, but many of the otherpegmatites of Category 3 are far larger than it. All of theones labeled as large mines in the list on the page facingFigure I are in large pegmatites. In contrast, many of themajor muscovite mines are in very small pegmatites be-cause the high price of sheet mica at times in the past hascaused small deposits to be important mines. The Climaxpegmatite (Fig. 9), which is only 45 m long, mostly lessthan 3 m thick, and was mined to a depth of 50 m, wasa major source of sheet mica, but a potassic feldspar peg-matite of this size would never have been importantenough to be listed by name, and might have so com-

pletely escaped notice as not to be shown at all on Fig-ure l.

Not everywhere is it certain whether a pegmatite be-longs to Category I or Category 3. Vertical sections ofthe White Spar muscovite pegmatite (Page and others,1953, Plate 45) show that the upper part ofit is occupiedmainly by a perthite-rich unit and that the muscovite-rich wall zone is well developed only in the lower part ofthe pegmatites. The Buster has a similar arrangement: itwas first developed as a feldspar mine, and the muscovitedeposit was not discovered until later. When it was stilla feldspar mine, the available exposures would havecaused it to be assigned to Category 3, as a small feldsparmine, rather than to Category l, as the large muscovitemine it later became.

Potassic feldspar pegmatites that have inner zones withLi minerals are assigned to Category 4. The prime ex-ample is the Hugo (Figs. 8 and 9), which, in addition tobeing a very large source of feldspar, has also been minedfor amblygonite-montebrasite and spodumene, and haslithian muscovite in its core. At the Bob Ingersoll No. Ipegmatite, the core has lepidolite, which has been the

CIETTA P-Adit level

COWBOY TIN1O0-foot levelCategory 7

and cassiterite)

SILVER DOLLARWHITE SPARCategory 1

CLIMAXCategory

Ab-O-Mu

P-O

DAN PATCHCategory 3

Q-Ab-Mu

MATEENCategory 5


Fig.9. Vertical sections of zoned pegmatites. Symbols are the same as in Figure 8.

641

most important commercial product. It is the only suchpegmatite known in the Black Hills and is grouped withthe pegmatites of Category 4 because it is geologicallymost similar to them.

Pegmatites with abundant spodumene are in Category5. The Etta (Fig. 8) is the most widely known and alsorather large. Some pegmatites of Category 5 have sizeable

perthite-rich zones and difer from pegmatites of Cate-gory 4 only in having much larger spodumene units.

Category 6, which emerged unexpectedly during com-pilation ofthe data used to establish the categories, con-sists of beryl-bearing pegmatites in which the beryl is con-centrated in the inner part ofthe wall zone or in the firstintermediate zone and is accompanied by abundant mus-

642

covite saleable only as scrap mica. The Peerless pegmatiteis the most notable example, for it has been the largestsource of beryl and scrap mica among Black Hills peg-matites (Sheridan et al., 1957). A smaller and thus moretypical example is the High Climb (Fig. 9). Because mostof the pegmatites of Category 6 have units rich in platyalbite (cleavelandite variety), perthite, or amblygonite-montebrasite, and some have spodumene, they resemblepegmatites of Categories 4 and 5, but perthite is muchIess prominent than in Category 4 and Li minerals aremuch less abundant than in Category 5.

Category 7 consists of tabular quartz-rich pegmatitesthat have enough cassiterite to have attracted interest assources of Sn. The Cowboy is one of the larger ones,though it is smaller than the other examples of pegmatiteson Figures 8 and 9. The quartz-cassiterite bodies are sim-ilar to quartz veins and much different from the otherpegmatites, but their content of plagioclase, beryl, andmuscovite, as well as their proximity to other pegmatites,suggests that they are part of the pegmatite system. Sev-eral small quartz veins northwest and northeast of theHarney Peak Granite have been prospected for W, butthey are ignored in this article.

The pegmatites of Categories 4-i are the ones that canmost appropriately be called rare-mineral pegmatites.They have yielded the spodumene, amblygonite-monte-brasite, tantalum-niobium minerals, cassiterite, and pol-lucite produced in the region, as well as most of the beryl,and have many other rare minerals that lack commercialvalue.

Every kind of zoned pegmatite can have monomin-eralic quartz units, and all except the Sn category canhave quartz-potassium-feldspar zones. The supposition,perhaps still widespread, that all zoned pegmatites havequartz cores is incorrect. Only about 20o/o of the zonedpegmatites in the Black Hills are known to have quartzcores.

DrsrnrnurroN AND ABUNDANcE oF pEGMATTTES

Figure I shows the distribution of the pegmatites bymeans of isograms that indicate the abundance of peg-matites in various parts of the area. The area containsabout 24000 pegmatites, according to a calculation to bedescribed later. The most numerous of these are simplepegmatites, and most of the rest are layered granite. Fig-ure I shows 278 zoned pegmatites.

The isograms show the number of pegmatites per sq.mi., which is the size of the sections in the land grid ofthe western United States shown on all topographic maps.The interval between isograms is 50 pegmatites per sq.mi., which is equivalent to about 20 pegmatites per km2.

The isograms are most accurate near Custer, Hill City,and Keystone because these areas have been mapped atscales of l:24000 and larger. Elsewhere the isograms arebased mostly on reconnaissance and on the study ofaerialphotographs. The pegmatites form resistant outcrops andare generally easy to detect. Information from Custer StatePark is scant.


The most precise isogram is the one for the outer bor-der of the pegmatite region. The isogram goes from theKeystone region west to Hill City and then southwest tothe Paleozoic contact. One small outlying group of peg-matites, l0 km northwest of Custer, is surrounded by azero isogram. Another outlier, 5 km west of Hill City,has no isogram because the site contains only a singlepegmatite, a zoned one shown on the map. Three inlierssurrounded by the zero isogram are northwest, west, andsouthwest of Custer. Though these areas contain no map-pable pegmatites, they may contain smaller pegmatiles.On the other hand, each of these closed zero isograms isin a place where the zero and 50 isograms are widelyseparated, and the top of the pegmatite field may well dipbelow the surface.

The pegmatite region is narrow along the north side ofthe Harney Peak Granite. Elsewhere the pegmatites ex-tend over a much broader area; the greatest abundanceis southwest of the granite. If one visualizes the isogramsas topographic contours, they can be said generally todescribe a rolling surface with modest hills in severalplaces, a sharp ridge south of Custer with a steep slopeon its east side, and local highs around some of the smallgranite plutons. How far the pegmatite region extendsbeneath the Paleozoic rocks is, ofcourse, not known, butit could be small.

Though the isograms show how many pegmatite intru-sions are in a locality, they are not a measure of the quan-tity of pegmatitic rock because a single large intrusionmay contain as much rock as dozens of smaller ones.Data on quantity are far more difficult to obtain, thoughthey might be more meaningful. In the Fourmile andBerne quadrangles, which are southwest and northwestof Custer, enough information is available to comparequantity of pegmatitic rock and number of intrusions persq. mi. (Fig. 10). In several places the two kinds of iso-grams are drastically different; the greatest differences arecaused by the presence oflarge bodies oflayered granite.The isograms for 0.50/o and lolo pegmatitic rock have ageneral similarity to the isograms for 50 and 100 peg-matites per sq. mi. Above the 10/o isogram, discordanciesbecome conspicuous. About 600/o of the pegmatite fieldof the southern Black Hills is below the isogram for 100pegmatites per sq. mi., and probably about the sameamount is below the l0lo isogram for quantity of pegma-tite. The higher isograms for quantity of pegmatite mayindicate centers ofintrusive activity that perhaps are re-lated to irregularities in the distribution of zoned peg-matites that will be discussed later.

Calculation of the total number of pegmatites dependson measuring the area in sq. mi. between each pair ofisograms and then multiplying by the number of peg-matites per sq. mi. The numbers to use as the multi-pliers-that is, the number of pegmatites per sq. mi. be-tween each pair of isograms-were obtained from thedetailed examination of the distribution of pegmatites inthe Fourmile and Berne quadrangles shown in Table 2.These data lead to the calculations in Table 3 that show

a total of 23700 pegmatites in the 275 sq. mi. (71I km,)between the outermost isogram and the granite bodies.The area between the 0 and 50 isograms, which includesabout 400/o of the pegmatite region, has only about 1800pegmatites. Between the 50 and 200 isograms, which isabout 54o/o of the region, there are about 17400 pegma-tites. The small area above the 200 isogram has 4500pegmatrtes.

Table 2 also shows that 800/o of the pegmatites in theFourmile and Berne quadrangles are simple and most ofthe rest are layered. Table 4, however, shows that thequantity of rock in simple pegmatites is far smaller thanin layered intrusions. Zoned pegmatites are at the samelow level in both methods of measurement.

Further calculations yield an estimate of the abundanceof rock in simple and zoned pegmatites relative to allgranitic rocks at the present surface. The area ofthe dif-ferent kinds of pegmatites was obtained by multiplyingthe totals in the Fourmile and Berne quadrangles (Table4)by 6.2, which is the ratio of the entire pegmatite areato the area in these quadrangles. The result added to thearea of granitic bodies on Figure I is a total of l2l km'z(47 sq. mi.) of granitic and pegmatitic rock. The largegranitic bodies constitute 860/o of this total. The percent-age is 9.6 for the small layered intrusions, 3.9 for simplepegmatites, and 0.5 for zoned pegmatites. The 0.5 per-centage for zoned pegmatites is divided into about 0.10/omuscovite pegmatites, 0.20lo potassic feldspar pegmatites,and 0.2o/o rare-mineral pegmatites.

These percentages are not necessarily representative ofthe entire igneous system above and below the presentsurface. Information is also lacking for the part of thepegmatite field covered by Paleozoic rocks, but such in-formation probably would not cause significant changesin the calculations, for the exposed sample is a large one,comparable in size with many other major pegmatitefields.

As for how far the system extended above the presentsurface, the appearances are in accord with the orthodoxview that pegmatites are above granitic plutons. The Har-ney Peak pluton could be the highest of these plutons,and others may be in the subsurface elsewhere in theregion. Presumably pegmatites once existed above themain pluton, and pegmatites near its border are above itsdown-dip extensions. There is a slight possibility that Sngranites and Sn deposits once lay far above the pegmatitesystem, because numerous Sn districts in the world haveabundant Li and some have concentrations of Be. TheCornwall Sn district even has a remarkable Li-rich dikecalled the Meldon Aplite dyke, which according to Beeret al. (1978, p.286) is 3.5 km long and 12 m thick andhas 1951-7400 ppm Li in seven samples. It consistsmostly of medium-grained albite, qtrartz, potassium feld-spar, and lepidolite but has accessory petalite, spodu-mene, and amblygonite-montebrasite. At Macasuni, Peru,samples of rhyolitic glass contain about 3000 ppm Li(Barnes etal.,1970,p.1545; Pichavant et al., 1987, Tablel) and otherwise resemble Li pegmatites in composition.

1 03045',

643

103037',30"43052',30"

43037'30"

Fig. 10. Map of the Fourmile and Berne quadrangles show-ing relations among (l) the quantity of pegmatite rock, (2) thenumber of intrusions per sq. mi., and (3) the locations of zonedpegmatites. Quantity is expressed by solid-line isograms at 0,0.5, l,2,5, 10, and 20010, based on measurements of areas of y4

sq. mi. Dashed lines show the number of intrusions per sq. mi.,and symbols show the zoned pegmatites in the same way as inFigure 1.



TaBLE 1. Categories of zoned pegmatites in the southern Black Hills

Numberof peg-matitesshown

onCate- Figuregory I

Dominant economic units Other large units

Structural Major industrialposition minerals

Other abundantminerals

Abundantminerals Abundantmineralsof outer units of inner units Distinctive characteristics

1 t l

49 Wall zone Sheet mica

Wall zone orfirst inter-mediatezone

Sheet mica

QuartzAlbite (rarely oli-

goclase)Tourmaline

QuartzAlbite (rarely oli-

goclase)PerthiteTourmaline

Quartz+ Albitet Muscovite

Quartz+ AlbiteI Muscovite

Quartz+ Albite (general-

ly cleavelandite)+ Perthite or mi-

croclinet Amblygonite

OuartzAlbite (commonly

cleavelandite)+ Perthite

QuartzMuscovite

Quartz+ Albite+ Perthitet Muscovite

QuartzPerthiteAlbite+ Muscovite

Quartz+ Albitet Muscovite

QuartzAlbite (commonly

cleavelandite)+ Perthite or mL

croclinet Muscovite+ Amblygonite+ spodumene+ LepidoliteNo large inner

units other thanthe spodumeneunits

Quartz+ Albite+ Perthite or mi-

crocline+ Spodumene+ Amblygonitet MuscoviteQuartz+ Albitea Muscovite+ Spodumene

Quartz and plagioclase arethe dominant minerals.Perthite is ordinarily alsopresent, but only in innerzones, in which it is ac-companied by quartz orby quartz + albite.

The main difference fromCategory 1 is that perth-ite occurs in the micazone. Quartz and oerth-ite are ordinarily muchmore abundant than al-bite in inner zones. Zon-ing is commonly less dis-tinct than in Category 1.

Contain minable depositsof potassium feldspar,mostly in hood-shapedunits. Li minerals are ab-sent or rare. Generallycontain few zones andhave relatively simplestructure.

Resemble pegmatites ofCategory 3 but carry Liminerals and generallynave many zones.

Contain abundant Li min-erals. Many have numer-ous zones.

Beryl concentrated at inneredge of wall zone. Innerzones ordinarily containLi minerals.

Ouartz-rich. Contain Sn.

Intermediatezone or

Intermediatezone orcore

Intermediatezone orcore

29 First interme-diate zoneor inner partof wall zone

19 Wall zone orentire peg-matite

Potassium teld-spar

Potassium feld-spar (lepidoliteat Bob Inger-soll No. 1mine)

Spodumene

Beryl and scrapmtca

Cassiterite

OuartzAlbite (rarely oli-

goclase)PerthiteMuscovite

QuartzAlbite+ PerthiteMuscoviteTourmaline

QuartzAlbite+ PerthiteMuscoviteTourmaline

QuartzAlbite (comonly

cleavelandite)+ Perthite or mi-

croclineMuscovite+ AmblygoniteTourmalineQuartzAlbite (commonly

cleavelandite)+ PerthiteMuscoviteTourmaline

These samples show that magmas with the compositionof Li pegmatites can reach the surface. The Black Hills,however, have no known physical evidence for the passageofsignificant quantities ofgranitic magmas to higher levelsor for thermal conditions that would create the biotiticgranites characteristic of Sn districts and drive the mag-mas to shallow depths.

The amounts of granite and pegmatite in the subsurfacemay be the largest cause of error in the percentage cal-culations. The world literature on pegmatite districts con-

tains examples ranging from localities with few or noproved granitic associates to localities with few zonedpegmatites and huge quantities of contemporaneous gran-

itic rocks. The Black Hills are between these two extremes.Pegmatites probably are overrepresentated at the present

surface because granite bodies may exist as far down asthe magmatic sources. If so, the zoned and simple peg-

matites are an even smaller percentage of the entire systemthan the calculations show.


TABLE 2. Abundance of zoned, simple, and layered pegmatites in the Fourmile and Berne quadrangles

645

Number of pegmatites Percentage of each kind of pegmatite Number ofpegmatitesper sq. mi.'lsograms Simple Layered Total Simple Layered sq. mr.

0-5050-1 00

1 00-1 501 50-200200-250250*300

>300Entire area

1 97384

13326825z5

627

44ee

oe

o24

98

283562346387665166483

2892

3466684JO

523939193512

3617

12.84.91 . 40.6u,o1 . 00.82.7

81 .784.279.474.070.886.094.380.0

5.510 .919 .325.428.513 .04.9

17 .3

55.425.19.17.8

10.91 . 83.9

114.0

21.4v . I

3.53.04.20.71 . 5

44.O

't670

125175225275340

Note. Determined from maps by Redden (1 963b, 1 968) and isograms of Figure 1 .'Rounded to nearest 5 except between 0 and 50 isograms.

DrsrnrsurroN oF zoNED PEGMATTTES

The map (Fig. l) shows that the distribution of zonedpegmatites is much more complex than simple regionalzonation around the exposed granite or around any hy-pothetical unexposed bodies of granite. Though zonedpegmatites are scattered widely across the region, theyhave a strong tendency to congregate in clusters, and theclusters difer in the kinds of pegmatites they contain. Astill broader aspect of the distribution has at times causedthe region to be regarded as three separate mining dis-tricts: a muscovite district at Custer, Sn at Hill City, andLi and Be at Keystone. The two categories of muscovitepegmatites are in different parts of the Custer region,though the only difference between them is the presenceor absence of perthite in the wall zone. Nearly all peg-matites of Category 2 are in arr area extending 2 to l0km west of Custer State Park, and nearly all pegmatitesof Category I are farther west and northwest. The potas-sic feldspar pegmatites (Categories 3 and 4) are abundantin a small area near Keystone and in a broad area southand southwest of the granite. Some spodumene pegma-tites ofCategory 5 are near Keystone; others are in a beltnear the zero isogram from Hill City southwest to the TinMountain mine; and still others are along the east side ofthe steep ridge in the isograms south of Custer. The beryl-scrap muscovite pegmatites of Category 6 are chiefly alongthe west side of the pegmatite region, but the two thathave produced the most beryl are the Peerless mine atKeystone and the Beecher No. 3 mine south of Custer.All the Sn pegmatites are northwest and north of thegranite. Some of these characteristics of the distributionreflect a regional zonation in which rare-element peg-matites tend to be at low isogram levels and muscovitepegmatites at high isogram levels. The zonation is some-what vague but no more so than it appears to be in mostof the districts reviewed by Heinrich (1953).

Tables 5 and 6 allow assessment of regional zonationand Tables 7 and 8 show the clustering behavior of thezoned pegmatites in Figure 1. The map shows no zonedpegmatites in Custer and Wind Cave parks; some do existand a few were once worked, but almost nothing is knownabout them because prospecting and mining have been

prohibited for many decades. The 278 zoned pegmatiteson Figure I include only those large enough for miningor prospecting. The region undoubtedly contains morezoned pegmatites than are on the map. Yet it is clear thatonly a few hundred, or about 2o/o, of the pegmatites inthe southern Black Hills are of the zoned variety.

Table 5 shows that the ratio of zoned pegmatites toother kinds of pegmatites increases sharply at the lowerisogram levels but that the abundance of zoned pegma-tites per sq. mi. increases at a much more modest rate.That is, the distribution does reflect regional zonation, atleast in the sense that anyone searching for an undiscov-ered zoned pegmatite would have to examine far fewerpegmatites and a smaller area if he looked only in placesbelow the 100 isogram.

In the pegmatite literature, the kinds of regional zo-nation most commonly discussed depend on changes inmineral content and in the abundance of rare elements,which here would cause changes in the abundance ofthecategories ofzoned pegmatites at different isogram levels.Table 6 has the pertinent data. The abundance of mus-covite pegmatites (Categories I and 2) telative to otherkinds of zoned pegmatites increases from the zero iso-gram to above the 150 isogram. A large share of the po-

Traue 3. Calculated abundance of pegmatites in the southernBlack Hil ls

Area Number of Calculatedpegmatites number ofper sq. mi.- pegmatites--sq. mr.

lso-grams

0-5050-1 00

1 00-1 501 50-200200-250250-300

>300Total

284.6150.5123.01 1 0 . 117 .613.212.2

711 .2

109 .958.1+ I - 3

42.56.85.1i a

274.6

1 80041 00590074001 50014001 600

23700t

1 670

125175225275340

- From Table 2.*. Rounded to nearest 100.t Does not include pegmatites in the 2.6 sq. mi. of metamorphic inliers

in the Harney Peak pluton. Only a few pegmatites are exposed in theseinliers, perhaps partly for geomorphic reasons. The total number of peg-matites in the inliers is probably only about 100.


TABLE 4. Area of pegmatites of different kinds and sizes in the Fourmile and Berne quadrangles

Layered intrusions. Simple pegmatites Zoned pegmatites All pegmatites

sq . f t .Area

(sq. mi.)Area

(sq. mi.)Area

Number (sq. mi.)Area

(sq. mi.)

1 0-500s00-5000

5000-1 50001 5000-40000

>40000Total

1 00-50005000-50000

50000-1 50 0001 50 000-400 000

>400 000

0.010.010.02

8495

98

456 0.02 274580 0.07 13752 0 .18 830 0.24 19 0.21 1

627 0.72 2892

0.150.090.020.010.02o.29 0.04

3285 0.18226 0.176s 0.2231 0.2510 0.23

3617 1 .05. Excludes bodies shown as granite on Figure 1.

tassic feldspar pegmatites of Category 3 is below the 100isogram. Most of the rare-mineral pegmatites are belowthe 100 isogram. The distribution of large mines, alsoshown on Table 6, has variations from the general pat-tern, but none ofthem appears to be significant.

The map has many clusters of zoned pegmatites, andit also has large areas in which zoned pegmatites are ab-sent. The clusters and the barren areas can extend acrosslow or high isograms and can be near granite or far awayfrom it, and their positions have no obvious relation toother known aspects of the regional geology.

The clustering effects are expressed quantitatively inTable 7 for the six large clusters that are outlined onFigure 1 and labeled with uppercase letters, and for thethe small clusters labeled with lowercase letters enclosedin circles. The large clusters include all areas with l0 ormore zoned pegmatites in which each is within I km ofits nearest neighbor. The small clusters have five or morepegmatites at a spacing of 0.6 km or less. These are notoutlined on the map because of their small size. The se-quence for each set of clusters in the table is from the onewith the smallest to the one with the largest percentageof rare-mineral pegmatites. All except three of the smallclusters are a part of one of the large clusters. Only onelarge cluster does not contain a small cluster.

The six large clusters contain a total of 135 zoned peg-matites, which is about half of all the zoned pegmatites,but the total area of these clusters is only 39 kmr, whichis a trifling share of the pegmatite region. The ten small

TABLE 5. Regional abundance of zoned pegmatites

Location sq. mi.

Per- NumberNumber centage of zoned

Number ol zoned zoned pegma-of peg- pegma- pegmatites permatites tites tites so. mi.

0-50 isograms50-100 isograms100-150 isogramsAbove 150 isograms and

outside graniteIn granite or layered

pegmatites

Nofe; Calculated as in Table 3. Excludes areas in Custer State Park andWind Cave Park.

clusters occupy a total area of only about 18 km'z, yetcontain 97 zoned pegmatites, which is about a third ofall the zoned pegmatites. Table 8 shows that the six largeclusters contain about half of the pegmatites in Categoriesl, 2, 4, and 6. The ten small clusters contain about athird of all these categories except number 6, which ismore abundant in the small clusters because Cluster i(Table 7) contains five pegmatites of this kind. The sim-ple feldspar pegmatites (Category 3) are somewhat fewerthan half the pegmatites in the large clusters and fewerthan a third of those in the small clusters. The Li and Snpegmatites (Categories 5 and 7) are the only kinds of peg-matites that have more tendency to be in the clusters thanelsewhere, for most of them are in the large clusters andmore than half of them are in the small clusters. The 33large mines, as Table 8 also shows, behave in much thesame way as smaller pegmatites, except that most of themajor muscovite mines are outside the clusters.

The clusters have startling differences. Cluster A, withl6 pegmatites south-southwest of Custer, has 570lo micapegmatites, and the only rare-mineral pegmatites are twosmall ones of the beryl-scrap muscovite category. ClusterB, containing 25 pegmatites east-southeast of Custer, alsohas many muscovite pegmatites, but they are accom-panied by several rare-mineral pegmatites. This largecluster contains two small clusters (a and b) that consistmostly of muscovite pegmatites. Cluster C, east-south-east of Keystone, is the largest in Table 7, for it has 34pegmatites and six of them are large mines. lt has 620/opotassic feldspar pegmalites,29o/o in the rare-mineral cat-egories, and only a few small muscovite pegmatites. TheKeystone district is best known for its rare-mineral peg-matites, but it also has numerous feldspar mines. In twosmall clusters in this district, the one near Keystone (f)has nearly as many rare-mineral pegmatites as feldsparpegmatites, and the one farther away (c) consists predom-inantly of feldspar pegmatites. Clusters D and d, l0 kmwest of Custer, cover virtually the same area. Muscovitepegmatites are most common, but this is misleading be-cause the muscovite pegmatites are small and the area isbest known for its rare minerals. Clusters E and g, on thewest side of the granite, also cover much the same area,and each consists mostly of small rare-mineral pegma-tites. Cluster F, at Hill City, has nine Sn pegmatites and

86.641.434.5

43.9

1 40029004300

9300

1 1 377

8.1 1 .32 .7 1 .90.6 0.8

0 .5 1 .0


TABLE 6. Distribution of the varieties of zoned pegmatites in relation to the isograms

647

Location

Category

Percentage of muscovite, potassicfeldspar, and rare-mineral

pegmatites

Potassic Rare-Muscovite feldsoar mineral

Total pegmatites pegmatites pegmatites

0-50 isograms50-100 isograms100-150 isogramsAbove 150 isograms and

outside graniteIn granite or layered

pegmatitesTotal

0-50 isograms50-100 isograms100-150 isogramsAbove 150 isograms

Total

113 12 38 5077 23 36 4027 37 44 19

4 2 5 2 4 3 5

1 72

1 2 2 4 310 8 289 1 ' , t 2

1 3 9 1 8

1 9 4 2 5 3 5278 26 40 34

15 13 33 5310 20 30 506 5 0 1 7 3 32 100

33 27 27 45

24

2

6

1 01 1 1

31

9

323

1

I

c

49

213

8

All pegmatites1 0 1 41 0 84 -

123

33 large mines

2

1 51 1'I

2

29

11

2

1 q

0

Note. Muscovite pegmatites are Categories 1 and 2. Potassic feldspar pegmatites are Category 3. Rare-mineral pegmatites are Categories 4-7.

one rather important spodumene mine, and Cluster j isthe same except that it lacks the spodumene mine andtwo of the Sn pegmatites.

The three small clusters that are not a part of any largecluster are e, h, and i. All have very few pegmatites. Nonecontains a muscovite pegmatite. One of the clusters (e)has one more of the feldspar pegmatites than of the rare-mineral pegmatites, but the others consist mostly (Clusterh) or entirely (Cluster i) of rare-mineral pegmatites. Clus-ter h has three of the most important rare-mineral minesin the Black Hills.

To sum up, both the large and the small clusters have

TABLE 7. Zoned pegmatites in the clusters

wide ranges in the kinds of pegmatites they contain. Theclusters have as much as 830/o muscovite pegmatites, asmuch as 750lo feldspar pegmatites, and as much as 1000/orare-element pegmatites. Most of the spodumene and Snpegmatites are in the clusters. One small cluster consistsentirely of Sn pegmatites, and another consists entirely ofberyl-scrap muscovite pegmatites. The clustering effect isa major aspect of the distribution. The clusters may benear the exposed granite or in the farthest reaches ofthepegmatite region, where granite is either absent or is be-neath the surface. They may be at high or low isogramlevels. The distribution suggests that ultimately nearby

Location."

Numberof peg-matites 3 4

Number of pegmatites in each category

Percentage distribution

Potassic Rare-Muscovite feldsDar mineralpegmatites, pegmatites pegmatites

0 2 0 5 7 3 1 1 21 0 0 4 4 3 6 2 05 2 0 9 6 2 2 91 4 0 4 8 1 9 3 8

1 1 6 6 7 1 0 8 31 0 I 0 0 100

19 14 15 26 31 43

0 0 0 8 3 1 7 00 0 0 5 7 2 9 1 41 1 0 6 75 191 4 0 5 0 1 1 3 90 1 0 0 5 7 4 33 0 0 0 5 5 4 55 3 4 1 3 7 8 02 1 0 0 2 0 8 00 5 0 0 0 1000 0 7 0 0 100

12 15 11 22 30 48

1 6 62 5 53 4 32 1 1 029 11 0 0

135 25

6 07 4

1 6 11 8 I7 0

1 1 01 5 1C U5 07 0

97 15

Large cluste?s5 09 4

2't 34 23 10 0

42 10Small clusters

AE

DET

I oral

aD

defshI

Total

5 1o 2o 1 2o 20 40 61 10 10 00 06 2 9

e60010

1 0

1-4 mi. SSW of Custer1-4 mi ESE of Custer0-4 mi. ESE of Keystone6 mi. west of Custer4-8 mi. SSW of Hill CityAdjacent to Hill City

4 mi. ESE of Custer2 mi. east of Custer3 mi. SE of Keystone6 mi. west of Custer6 mi. WSW of Custer1 mi. south of Keystone6 mi. SSW of Hill City4 mi. SSE of Custer5 mi. WNW of Custer1 mi. west of Hill City

01'I

22201U

o

. The sequence is according to the percentage of rare-mineral pegmatites in each cluster." Locations are shown on Figure 1.


Trele 8. Comparison of the abundance of the various categories of zoned pegmatites in the entire region and in the clusters

Number of pegmatites in each category Percentage in each category

Total

1 8 8 4 0 I I 1 0 71 9 7 3 1 7 1 4 1 0 1 11 5 6 3 0 9 1 2 1 5 1 1

2 4 3 2 7 1 8 2 1 6 01 3 0 3 3 2 0 2 7 7 08 0 3 1 1 5 3 8 8 0

1 91 51 1

000

All pegmatites23 291 9 1 41 2 1 533 large mines7 24 15 1

278 49 23 111 24135 25 10 42 109 7 1 5 6 2 9 9

3 3 8 1 9 61 5 2 0 5 31 3 1 0 4 2

Entire regionLarge clustersSmall clusters

Entire regionLarge clustersSmall clusters

unzoned intrusions will also be found to have localizeddifferences.

Rnr,,c,rroN oF DISTRIBUTToN oF pEGMATTTES roSTRUCTURE, LITHOLOGY, AND

METAMORPHISM OF HOST ROCKS

Comparison of the pegmatite distribution map (Fig. l)and the geologic map (Fig. 2) shows little evidence for arelationship between the distribution of pegmatites andthe structure of the metamorphic rocks. Some of the iso-grams follow the domal structure around the granite, butelsewhere the isograms cross structural trends nearlyeverywhere. An exception is the east side of the area ofabundant pegmatites south of Custer, which is almostexactly along the faults that run north from Pringle.Structural control by the faults seems unlikely, but themetagraywackes to the west may have been more suitedto intrusion by pegmatites than the quartzites and micaschists to the east. Generally, however, any tendency tofavor one kind of rock type over another is slight. Oneexample is the small low in the isograms northwest ofCuster, which coincides with a narrow strip of biotite-rich schist. Another is the indentations in the isograms 3km southwest of Custer, where they cross a thin unit con-taining various mica schists, amphibole rocks, andquartzites, all ofwhich seem to have been unreceptive tothe emplacement of pegmatites. Elsewhere any prefer-ences for different kinds of host rocks are not obvious,but we do believe that pegmatites tend to be more abun-

Tasle 9. Distribution of zoned pegmatites in relation to the firstsillimanite isograd

Pegmatites above Pegmatites belowsillimaniteisograd sillimaniteisograd

dant in rocks of moderate competency (the quartz-mica-feldspar schists of the metagraywackes) than in the verycompetent rocks (the quartzites and amphibolites) or thevery incompetent rocks (the mica-rich schists). Such dif-ferences, however, have no effect on the broad aspects ofthe isogram patterns.

The distribution of zoned pegmatites also has little orno relation to the kinds of host rocks. Most of them arein metagraywacke, but they occur in other metamorphicrocks also. The many zoned pegmatites below the 50 iso-gram can be in any host rock, regardless ofrock type orstructure. No known evidence indicates that the host rockhas an important part in determining whether an intru-sion is zoned, simple, or layered.

Metamorphism has at least some correlation with theabundance of pegmatites. The outer isograms tend to beparallel to isograds. North of the granite the stauroliteisograd is very nearly parallel to the zero isogram andgenerally to the north of it. The first sillimanite isogradis mostly between the zero and 50 isograms. The areaabove the second sillimanite isograd has only two smallzoned pegmatites and a low in the isograms. Metamor-phism in the southeastern part ofthe region is not knownwell enough to judge its relationship with the distributionof pegmatites.

The abundance ofthe various kinds ofzoned pegma-tites above and below the first sillimanite isograd is shownin Table 9. About 700/o of the zoned pegmatites are abovethe isograd and about 300/o below it. Most of the mus-covite pegmatites and an only slightly smaller share ofthe feldspar pegmatites are above the isograd. Half of thespodumene and a majority of the beryl-scrap mica peg-matites are below the isograd, and most of the rest areeither near it or are on the east side of the ridge in theisograms south of Custer. Three-fourths of the Sn peg-matites are below the isograd, and the rest are only slight-ly above it.

Pnrssunr AND TEMPERATURE

In Figure I 1, which shows various field boundaries thatlimit the P-T reglme, the haplogranite minimum shouldbe about equivalent to the solidi for the Harney PeakGranite compositions because they are almost entirelyalbite, orthoclase, and qvafiz. The most abundant of theother constituents is excess Al"O". A solidus fewer than

Categoryof zoned Total in

pegmatites category

Percentof total

tnNumber category

Percentof total

tnNumber category

12

4

b

7Entire area

49 38 78 11 222 3 2 3 1 0 0 0 0

111 91 82 20 182 4 1 7 7 1 7 2 923 12 52 11 4829 11 38 18 621 9 5 2 6 1 4 7 4

278 197 71 81 29


20 "C lower was obtained by Huang and Wyllie (1973)for a muscovite-rich bulk sample of Harney Peak Granitein which the content of normative corundum is 4.80/o andthe composition of the normative plagioclase is Anr.Burnham and Nekvasil (1986) found a solidus that isnearly identical to the haplogranite minimum for a bulksample of a feldspar-quartz-muscovite pegmatite at SprucePine, North Carolina, which has 1.25o/onormative corun-dum and normative plagioclase of An,,. Neither of thesesolidi is shown on Figure 1l because of their similarityto the haplogranite minimum.

The boundary between muscovite + quartz and silli-manite * potassium feldspar + HrO (the second silli-manite isograd) crosses the haplogranite minimum at 3.7kbar and 660 "C. This is likely to approximate the naturalconditions in the Black Hills because granite crystallizedon both the muscovite and sillimanite sides of the bound-ary. Confirmation comes from a temperature of 670 'C

determined from calcite-dolomite intergrowths in an in-lier near the southwest edge of the main pluton (Reddenet al., 1985, p. 238).1 The surrounding granite containsprimary sillimanite and primary muscovite, and a shortdistance to the west the second sillimanite isograd abutsthe contact of the main pluton.

The AlrSiO, field boundaries and triple point on FigureI I are from Robie and Hemingway (1984, p. 305), whoput the triple point at 4 kbar and 517 oC. In the graphsofa review article, Essene (1982) placed the point at theslightly lower pressure and temperature of Holdaway(1971). Salje (1986, p. 1369) obtained generally similarresults with coarse-grained sillimanite, but with a sampleof fibrolite the boundary with andalusite was at muchhigher temperatures, yielding a triple point at 5.9 kbarand 663 oC. In contrast, Kerrick (1987, especially Fig. 4)found that fibrolite derived mainly from biotite in a ther-mal aureole appeared at a temperature about 100'C low-er than the coarse sillimanite isograd and about 0.3 kmaway from it. In the Black Hills, coarse sillimanite isabundant near the southwest border of the granite andhas been seen adjacent to the first sillimanite isograd atthe Custer-Pennington County line west of the granite,but most of the sillimanite is fibrolite that formed frommuscovite and other micaceous minerals (Redden, 1963b,p. 260). According to Holdaway (197l, p. 125-128) thereaction with muscovite would be at approximately theequilibrium temperature, unlike the reaction in which fi-brolite forms from andalusite, which would be higher by200 "C or more. Because the first sillimanite isograd iswell inside the staurolite isograd in the Black Hills, thelow temperature of the triple point with respect to thestaurolite curve in Figure I I may mean that the isograd

tThe data were inadvertently omitted from the original pub-lication. Hand-picked separates ofthe calcite contain 48.83 and50.760/o CaO and 4.11 and 4.000/o MgO; the calcite-dolomitesolvus used was thg one in Essene, I 982, Figure 6. C. K. Shearerobtained similar results from microprobe analyses (oral com-munication, 1987). The samples were from a marble quarry lo-cated at SW Yr SE 7+ sec. 33. T.2S. R.5E.

Temoerature in oC

Fig. ll. Field boundaries relevant to the environment ofpegmatites in the southern Black Hills. The reaction of petaliteto spodumene + quartz is from London (1984). The reactionyielding staurolite is from Hoschek (1969); it represents thestaurolite isograd according to Winkler (1974, p. 199). The kya-nite-andalusite-sillimanite boundaries are from Robie and Hem-ingway (1984). The reaction ofmuscovite + quartz to form po-tassium feldspar + sillimanite (the second sillimanite isograd)or andalusite is from Chatterjee and Johannes (1974). The so-lidus for the Harding pegmatite is from Burnham and Nekvasil(1986). The granite minimum is from Luth et al. (1964).

for fibrolite in the natural conditions is at a somewhathigher temperature than on the graph and that stauroliteformed at a lower temperature than the reaction used onFigure 11. On the other hand, Milton (1986) has shownfrom the study of a North Carolina locality that the stau-rolite curve is indeed at a higher temperature than theAlrSiO, triple point. In short, the AlrSiO. field boundariesof Figure I I are likely to be close to the natural condi-tions in the Black Hills. Because andalusite and silliman-ite are the AlrSiOs minerals in the Harney Peak aureole,the maximum pressure is less than that of the triple point.

The minimum possible pressure is above that of thepetalite-spodumene reaction (Fig. I l) because all of thespodumene of the southern Black Hills is primary andpetalite has never been found. Burnham and Nekvasil(1986) showed that the HrO-saturated solidus of a com-posite sample from drill holes in the Harding pegmatite,Dixon, New Mexico, which is rich in lepidolite and spod-umene, is at 610 oC at pressures above 2 kbar (Fig. I l).Stewart (1978) found a eutectic at 640 "C and 2 kbar inthe system albite-quartz-eucryptite-HrO. The tempera-ture would be depressed to somewhat below 600 "C bythe presence of potassium feldspar, excess AlrOr, and oth-er constituents (Stewart, 1978, p. 976-977). London(1986) traced the evolution of the fluid to lower temper-atures by a study of fluid inclusions in spodumene in theTanco pegmatite, Manitoba, where nearly all the spodu-mene is from the breakdown of petalite, but some is pri-mary. The inclusions contain a dense hydrous fluid thatis exceedingly rich in B, also rich in Li, Na, and Cs, and

6

C

o-@

L

650

has 140/o AlrO, and 380/o SiOr. Experiments in the systemLiAlSiOo-NaAlSi3O8-SiOr-LirB4O7-HrO at 2 kbar indi-cated that the composition proceeded from Stewart's eu-tectic at 640'C to 500 "C at a point near the NaAlSirOr-Lr2B4Oj sideline of the tetrahedron (London, 1986, Fig.4). London showed that this fluid probably formed thealbitites that are the source of the Ta mined at Tanco andare also enriched in Sn, and that it did so between 470oC and 420 "C when the B went into tourmaline in thewall rocks and the wall zone of the pegmatite. Londonsaid the B-rich fluid was the final stage of primary mag-matic crystallization, and he regarded a subsequent lowdensity fluid as the beginning of hydrothermal subsolidusactrvlty.

The circumstances are reminiscent of the replacementunits in the Hugo pegmatite (Figs. 8 and 9), which hasthe largest such units in the southern Black Hills. Theyconsist largely of introduced albite and lithian muscoviteand contain cassiterite and tantalite-columbite, and thewall rock and the outer part of the wall zone near thesereplacement bodies have a much higher content of tour-maline than elsewhere along the border of the pegmatite.The replacement units extend from a small albite (cleave-landite varietylmicrocline-lithian-muscovite core acrossall intermediate zones and into the wall zone, but mostof the replacement is in three quartzose intermediate zones(Norton et al.,1962, especially p. 86-96 and 102-103).The replaced minerals include large spodumene crystals.The Hugo magma was emplaced at a temperature highenough to cause fibrolitic sillimanite to form in adjacentquartz-mica schist, though the pegmatite is I km outsidethe first sillimanite zone. About 7 5o/o of the magma crys-tallized as feldspar-quartz outer zones (Norton etal.,1962,Table l9) before spodumene appeared in inner zones, atpresumably about 600 'C. The replacement bodies, ifgenerated in the manner described by London, formed attemperatures extending down to less than 500 "C. Thereplacement units constitute only aboul 2o/o of the rockin the Hugo pegmatite, and thus their formation and tem-perature regime were only a minor part of the history ofcrystallization of the pegmatite.

The Etta pegmatite, which is only a few hundred me-ters from the Hugo and is in the same unit of quartz-micaschist, consists mostly of spodumene-rich zones (Fig. 8).Sillimanite is absent from the wall rocks. Thus the max-imum magmatic temperatures for the Etta and Hugo peg-matites were on each side of the sillimanite-andalusitefield boundary, and the pressure for both pegmatites wasabove that of the petalite-spodumene boundary. The Ettapegmatite lacks replacement bodies like those in the Hugopegmatite, and it has a small quartz core rather than acore of aluminosilicates, but it has heavily feldspathizedrocks along its contacts, and at least some ofthem containenough Sn and Ta to have caused expensive efforts aimedat mining them. Thus the geology is compatible withLondon's (1986) scenario, though the normal suppositionwould be that the feldspathization preceded rather thanfollowed crystallization of the pegmatite.


The sillimanite-andalusite and the petalite-spodumeneboundaries on Figure I I intersect at 550 'C and 3.25kbar. If actual crystallization temperatures of the Etta andHugo spodumene were above 550 oC, the fibrolite stabil-ity field would be at higher temperatures than shown onthe figure. The same applies for the other spodumenepegmatites of Category 5 that are outside the sillimanitezone (Table 9). Minimum pressures indicated by the pet-alite-spodumene curve are between 3.25 kbar at 550 'C

and 3.6 kbar at 600 'C.

The evidence points to crystallization temperaturesranging from somewhat more than 660 "C in parts of thegranite to well below 600 oC in inner units of a few of thezoned pegmatites at a vapor pressure of about 3.7 kbar.If the load pressure were 3.7 kbar and if the overlyingrocks were similar to those at the present surface, thedepth would be about 13 km. The vapor pressure is un-likely to be much less than the load pressure, for other-wise, according to Barton (1986, especially Fig. 9),chrysoberyl instead of beryl would be the stable Be min-eral, and beryl is common and chrysoberyl very rare,though it does occur (Roberts and Rapp, 1965,p.64).

At the inferred depth and maximum regional temper-ature, the geothermal gradient would be 50 oC per km,which is exceedingly high but not as high as that suggest-ed for a similar terrane in Colorado (Nesse, 1984, p. 1158).Because most of the Black Hills outside the Harney Peakaureole is at almandine and biotite grades of metamor-phism, the temperature prior to heating by granite mag-ma may be as much as 200 'C lower (Winkler, 1974, p'199 and 210). This yields a gradient of 35 "C per km,which is plausible (Barker, 1983, p. 7). A gradient of 35oC per km would intersect the muscovite solidus and liq-uidus (Burnham, 1979, Fig. 3.a) at a depth of l9 km, anda gradient of 30 'C per km would do the same at 22 krn.These depths are 6 to 9 km below the present surface,which is possible for the deeper parts of metasedimentaryschists in Proterozoic synclines of the southern Black Hills.

Evor,urroNl oF THE rcNEous SYSTEM

Grn!'and Meintzer (1988) have recently reviewed cur-rent information and interpretations about the geologicprocesses that yield pegmatitic magmas. They trace thepetrogenesis from anatexis in highly deformed metasedi-mentary troughs to granite magmas and on to pegmatiticgranites and thence to zoned pegmatites. Their reviewshows that the literature has a shortage of the kinds offield data that have been presented in this article. Thediscussion here will center on interpretations of specificaspects of the field data, with a minimum of commentabout how they are related to interpretations in otherpegmatite regions.

The high initial SrtT/Sr86 content (Riley, 1970, Fig. 2),the low calcium content, and the peraluminous compo-sition indicate that the granite is an anatectic product ofmetasedimentary micaceous schists. The low content ofFe and Mg in the granite shows that the anatectic tem-


peratures were mostly below the biotite solidus or thatthe source rocks themselves had a low content of biotite.One consequence of melting below the biotite soliduswould be the observed contrast between about 30 ppmLi in the granite and about 50 ppm Li in the mica schists(Norton, l98l). If the melting left biotite behind, it islikely to have left much of the Li behind, for biotite prob-ably attracts more Li than muscovite in the schists (e.g.,Shearer et al., 1986).

Walker et al. (1986a) concluded from Sm-Nd isotopicdata that the source rocks cannot be among the metasedi-ments exposed at the present surface. The mica schistunit that is exposed over a large area west of Hill Cityand extends south almost to Custer (Fig. 2) seems to usto be an attractive candidate as one ofthe source rocks,but even if it is not, it is useful in describing how theanatectic process might have functioned. Modes of I Isamples of this schist show an average of 400/o muscovite,260/o quarrz,160/o biotite, 8o/o plagioclase (oligoclase), and100/o other minerals (Redden, 1968, Table 5). The amountof combined HrO would be about 1.80/o from the mus-covite and 0.60/o from the biotite, for a total of 2.4o/o. TheHrO in the muscovite would allow melting of about 200/oof the rock, and the melt would contain about 8.40lo HrO(Burnham, 1979, p.98-99; and Burnham and Ohmoto,1980, p. 2-3). Unless the temperature went to the higherlevels required to melt biotite, the magma would be lowin Fe and Mg.

Some small plutons are so far away from the mainpluton that their locations imply the existence of severalmagma sources. The best examples are near the south-west corner ofCuster State Park and northwest ofPringle,but the granite southeast of Keystone may be added tothe list ifit extends a long way beneath Paleozoic cover.All of these are accompanied by highs in the isogramlevels. Other highs in the isograms and small domes invarious places may be surface expressions ofgranite cen-ters. The isolated cluster of Sn pegmatites at Hill Cityand its proximity to gently dipping rocks at the edge ofthe Harney Peak dome may also reflect the existence ofsubsurface granite. Gravity modeling by Duke et al. (1990)from the data ofKleinkopfand Redden (1975) shows thatgranite probably underlies the Hill City area and severalother outlying areas in the pegmatite region.

The outer and structurally higher parts of the mainpluton and at least one small pluton are generally moreevolved and younger than the inner parts (Shearer et al.,1987; Duke et al., 1990). Shearer et al. (1985) showed,using trace-elements, especially K-Rb ratios, that zonedpegmatites are still more evolved and thus probablyyounger than the granite. These results are in accord withthe long-standing belief in an age sequence from essen-tially ganitic rocks to simple pegmatites and then to zonedpegmatites with concentrations of muscovite and potassicfeldspar and on to the rare-element pegmatites. Graniteprobably overlaps in age with simple pegmatites and withmuscovite and feldspar zoned pegmatites, for it containsmany fracture fillings of simple pegmatites and zoned

pegmatites, some perthite-rich and some with muscovite-rich wall zones, and also has monomineralic quartz frac-

ture fillings. Overlap in age with rare-mineral pegmatites

is less evident. Beryl and cassiterite occur in the granite.

Amblygonite-montebrasite has not been recorded and

spodumene is probably absent in fracture fllings in the

main pluton. One satellite located just south ofthe countyline (T.2S. R4E.) has a spodumene pegmatite (Fie. 1). A

much smaller body, called the Mountain Lion or Soda

Spar, which consists largely of layered pegmatitic rock

but has been suspected of also having zonal structure,contains spodumene-bearing fracture fillings (Norton and

others, 1964, p.329-331). Dozens ofthe zoned pegma-

tites have spodumene; most are in Categories 4-6, but at

least one (the New York) is a muscovite pegmatite.

Crosscutting relations have been unhelpful in establish-ing relative ages because, in most exposures where onepegmatite cuts across another, the two pegmatites are

similar to each other. No large zoned pegmatite of any

category has been conclusively shown to cut a pegmatite

of another kind. Confirmation that simple pegmatites

overlap in age with granite comes from a few places where

they are cut by small layered bodies.Evidence bearing on temperatures of crystallization may

be useful in inferring age relations. Both the metamor-phism, as shown by the isograds, and the distribution ofpegmatites, as shown by the isograms, are indicators of

the temperature regimes. Though the age of one is not

necessarily the same as that of the other, they do produce

similar patterns in the outskirts of the region. Granite

crystallized both above and below the temperature of the

second sillimanite isograd. Several pegmatites above the

second sillimanite isograd have muscovite books that arepartly to completely altered to sillimanite, but more com-monly the sillimanite is primary and in some places has

been partially converted to fine-grained muscovite by ret-

rograde reactions. The pegmatites with muscovite altered

to sillimanite obviously formed before the peak of meta-

morphism, and those with primary sillimanite formed

when crystallization temperatures were above the stabil-ity field of muscovite.

The spodumene pegmatites are in the outer and thus

cooler parts of the region, which is in accord with data

showing them to have low crystallization temperatures(Stewart, 1 978; London, 1984, 1 986). Other rare-mineralpegmatites also tend to be in the apparently cooler areas.

The position of the Li and other rare-element pegmatites

implies that the magma bodies moved into an environ-ment cool enough so that the diferences in temperature

between them and the host rocks were too small to allowfurther movement, and they stopped and crystallized. The

Beecher rare-element pegmatites are an especially good

example of temperature control because they are east of

a very steep gradient in the isograms, which implies a

temperature gradient also. Sheet mica pegmatites, in con-

trast, tend to be in the apparently hotter areas. Their crys-

tallization temperatures may have been similar to those

of unzoned pegmatites, though somewhat lower because

652

their more peraluminous composition would depress thetemperature at which crystallization began.

The apparent control by temperature can be turnedinto a means of surmising the age sequence in places thathave a variety of kinds of pegmatites in a small area.South of Keystone the Peerless beryl-scrap muscovitepegmatite, the Etta spodumene pegmatite, the Hugo feld-spar pegmatite with Li minerals, several ordinary feld-spar pegmatites, and a few simple pegmatites (includingthe one from Norton, 1970, used in Fig. a) are within afew hundred meters of each other. It seems plausible thatthese pegmatites were emplaced at different times, underthe control of declining regional temperatures. The Ettapegmatite is likely to have had the lowest crystallizationtemperature and to be the youngest. The Peerless alsohad a low temperature and late age, for it caused theretrogression of staurolite and andalusite to micaceousaggregates (Sheridan et al., 1957, p. 4-5 and Plate l). TheHugo, which has much less Li than the Etta, had a highertemperature and presumably a greater age. The feldsparpegmatites and the simple pegmatites probably are some-what older.

It is now widely agreed, from Jahns and Burnham(1969), that exsolution of an aqueous phase marks thebeginning of pegmatitic crystallization processes (Cern!and Meintzer, 1988, p. 200). In the exposed rocks of theBlack Hills, pegmatitic processes began with the forma-tion of coarse-grained layers in the Harney peak Granite,and an exsolved aqueous phase almost certainly was in-volved (Duke et al., 1990). Pegmatitic processes contin-ued among the thousands of unzoned pegmatites. Lay-ering, large crystals, and an upward increase of K relativeto Na have all been attributed, probably correctly, to ac-tions ofthe aqueous phase. The vertical differentiation ofK and Na continued into two of the major assemblagesof zoned pegmatites (Norton, 1983, Table 3). The sepa-ration of potassic from sodic pegmatite was the largestand most widespread step in the evolution of the peg-matitic part of the system.

Differentiation from the contact inward is the domi-nant feature of zoned pegmatites. In the muscovite andfeldspar pegmatites (Categories l-3), this differentiationis from muscovitic wall zones to feldspathic zones and,in some pegmatites, quartz cores. Muscovite and plagio-clase of wall zones are normal early precipitates of peg-matitic magma, according to experimental data of Burn-ham and Nekvasil (1986), but the amount of muscovitemay become greater as the ratio of Al to alkalis is in-creased by the loss of alkalis to the wall rock via theaqueous fluid (Shmakin, 1973; Morgan and London,1987). In the rare-element pegmatites, the higher con-tents of Li, Be, and Sn lead to zones in which these ele-ments form independent minerals. This is a major effectonly in pegmatites of Category 5, where spodumene isknown from Stewart (1978) to be best interpreted as re-flecting an increase in the Li concentration in the restmagma to a level that requires it to be precipitated as aLi mineral.


Rare-element pegmatites are not as different from theother kinds of zoned pegmatites as might be supposed.Except for the Sn category, they too are of essentiallygranitic bulk composition. They acquire their status asrare-element pegmatites by having enough Li, Be, Sn, Cs,Ta, Nb, or U to form minerals in which these elementsare an essential constituent. The most Li in any of thepegmatites is about 7000 ppm; the most Sn certainly isless than 5000 ppm and probably is about 1500 ppm(Page and others, 1953, p. 59,93,94); the most Be isabout 300 ppm; Cs appears to be in the low hundreds ofppm in the Tin Mountain pegmatite (Walker et al., 1986b),which is the only one known to have pollucite; and Ta,Nb, and U are probably much below 50 ppm in all thepegmatites. The total of rare elements is well under 10loin every pegmatite. The low atomic weights of Li and Because their amounts to be deceptively low when mea-sured by weight instead of by atomic abundance (and thecontents ofheavy rare elements are deceptively high), buteven so, the totals are small. Li is by far the most abun-dant of these elements, and because it forms spodumene,amblygonite-montebrasite, and lithian micas, it addsgreatly to the number and variety of zones by giving riseto the existence of assemblages 5, 6, and 9 of Norton(1983, Table 3).

The quantity of Li and Be, and presumably the otherrare elements in the rare-element pegmatites, is only asmall share of the quantity in the entire system. The mainpluton contains an average of about 30 ppm Li and 8ppm Be, which are close to the abundances in averagegranite, and the highest analyses are 171 ppm Li and 17ppm Be (Norton, l98l; Norton et al., 1958, p. 25; Sheareret al., 1987, p. 477). If the Li-rich pegmatites have anaverage of 5000 ppm Li and beryl-rich pegmatites havean average of 200 ppm Be, they contain, respectively,about 1 00/o of the Li and 2o/o of the Be in the exposed partof the granitic system. The percentages probably are sub-stantially smaller for the three dimensions of the entiresystem, but it does seem clear that much more of the Lithan of the Be in the system went into zoned pegmatites.

Li and Be behaved quite differently during the evolu-tion from granitic rocks to zoned pegmatites. Spodumeneformed in only a few dozen pegmatites, and beryl occursin probably hundreds of pegmatites and in the granite aswell. One of the remarkable properties of beryl is that itcan appear in granite ofvery low Be content (Norton eta l . , 1958, p. 25) .

The content of beryl probably increases gradually andcontinuously from low levels in granite to higher levelsin various kinds of pegmatites until it reaches its highestlevel in a few zoned pegmatites. In contrast, Li must beconcentrated to several thousand ppm before it begins toform spodumene or petalite, and the behavior can beshown on simple phase diagrams (Stewart, 1978). Am-blygonite-montebrasite, which is the principal Li phos-phate, forms at much lower concentrations of Li, for it isan accessory mineral in numerous pegmatites that lackLi silicates (Page and others, 1953, p. 54; Redden, 1963b,

Table 7). These pegmatites probably contain less than500 ppm Li. This is a likely limit because 500 ppm isapproximately the content of the Peerless pegmatite, whichhas nearly all of its Li in phosphates but also has lithianmuscovite and a few crystals of spodumene (Sheridan etal., 1957).

The processes that gave rise to Li-rich and Be-rich peg-matitic magmas seem to have operated separately morethan they operated together. The clusters ofTable 7 showa wide range in the ratio of spodumene pegmatites (Cat-egory 5) to beryl-scrap muscovite pegmatites (Category6). Many rare-mineral pegmatites have both Li mineralsand beryl, but the most Li-rich pegmatites tend to havemodest beryl contents and the most Be-rich pegmatitestend to have low amounts of Li minerals (with the ex-ception of the Bob Ingersoll Nos. I and 2). The BeecherNo. 2 pegmatite, which is one of the richest in spodu-mene, has no known beryl, yet the adjacent Beecher Lodeand Beecher No. 3-Black Diamond pegmatites are majorsources ofberyl. At the Edison spodumene mine, the onlyberyl ever noticed was two crystals a few millimeters indiameter, but the nearby White Cap mine has been asignificant producer of beryl, and its only known Li min-erals are accessory triphyliteJithiophilite. It may be thatin these places the concentrating processes for Li and Beoperated at different times.

Among the possible processes by which pegmatiticmagmas become enriched in rare elements, the most ob-vious is that the low ability of the principal minerals ofthe granite to accept these elements causes them to beconcentrated in residual magmas that ultimately crystal-lize as rare-element pegmatites. This may indeed be thechief process for Be, which in pegmatites reaches no morethan 40 times its level in the granite, and also for suchelements as Cs, Sn, Ta, Nb, and U. But for Li a concen-tration factor of 40 would yield only I 200 ppm. As Stew-art (1978) has emphasized, most pegmatites contain eithervery little Li or several thousand ppm. The traditionalview that Li-rich magmas formed by concentration inparental granitic magmas is still the dominant opinion(Cernj' and Meintzer, 1988). Suggestions that the concen-trating process was influenced by behavior in the anatec-tic zone (Norton 1973, p.368-369; Stewart, 1978, p.979)or by aqueous fluids that transported Li in schists andmay also have interacted with pegmatitic 4agma (Nor-ton, 1981) have been vigorously opposed (Cernj', 1982,p.441442 Cern! and Meintzer, 1988, p. 170). Unlikemuch other work on pegmatites-in which researcherscan, for example, examine zones within pegmatites andtrace the changes and test the possible reasons for them,or can do the same for the various kinds of layering-theissue about Li-rich magmas suffers from a lack of obser-vational data from sites where the process took place be-cause the sites have not been found, or at least not rec-ognized.

Possibly, however, the sites can be found, for Li-richpegmatites are disproportionately abundant in the clus-ters, and the clustering itselfindicates that localized pro-

653

cesses are more important than region-wide processes rncausing the development of zoned pegmatites. Graniticmagma bodies of surprisingly small size may be adequateto supply the constituents ofeven the largest and richestlithium and beryllium pegmatites of the Black Hills. Ifsuch a magma is able to give up 100 ppm of its Li, whichis well below the contents of parts of supposed parentgranites elsewhere in the world (e.g., Siroonian et al., 1959,especially Fig. 3; Pye, 1965, p. 52-54), a sill 100 m thickand 300 m in the other two dimensions (0.01 km3) couldbe the source of a 400000 t pegmatite containing 6000ppm Li, which probably is larger than any spodumene-rich pegmatite of the Black Hills. A sill of the same sizewould have to give up only 5 ppm Be to form a 400 000t pegmatite containing 300 ppm Be, which would beequivalent to the Peerless pegmatite. The 6000 ppm Liapproaches the maximum found in pegmatites of theworld (Stewart, 1978), and the 300 ppm Be is about themost contained in pegmatites (Griffitts,7973, p. 89). Inthe places in the world that have much larger rare-ele-ment pegmatites, as at Tanco, Manitoba, and KingsMountain, North Carolina, the volume of the source rockwould have to be larger by a factor of as much as 100,which would require a sill of only about 3 km'z and 300m thickness. Though this argument is only arithmeticshowing that small bodies of granite might yield rare-element pegmatites, it does bring out the possibility thatkey processes in small subsystems are only slightly under-stood or have escaped notice entirely (cf. Walker et al.,I 989) .

The Harney Peak system, from the beginning of re-gional heating and doming to the end of the emplacementof magmas and their crystallization, yielded tens of thou-sands of various kinds of granitic and pegmatitic bodiesover a long span of time. Obviously conditions must havebeen different in different places at any single time andmust also have changed through time in any single place.When the circumstances are viewed in this way, thevagueness of the regional zonation pattern is unsurpris-ing, and the regional zonation concept itself, though sim-plistic, retains some validity. Yet the clustering of zonedpegmatites and the differences in the kinds of pegmatitesin the clusters are more significant. The clustering behav-ior points to the existence of subsystems, and indicatesthat the subsystems behaved differently at different places.Because large areas have no known zoned pegmatites andother broad areas have very few, it is obvious that someparts of the system were unable to create magmas thatformed zoned pegmatites. Hence the pegmatitic magmasgenerated in some layered sills crystallized only as frac-ture fillings within granite or as simple pegmatites in thecountry rock. Others went to the stage of ejecting magmasthat formed the simpler kinds of zoned pegmatites. Andit is likely that only a few reached the extremes of expel-ling fluids that crystallized as rare-element pegmatites. Ina single subsystem the age sequence almost certainly isthe conventional one from granitic to pegmatitic rocks,but subsystems in different places may be of different



ages, and some localities may have two or more subsys-tems of different ages.

Geology has not yet reached an end to its decades ofdebate and its multitude of theories about how zonedpegmatites crystallized, and it clearly is only at the begin-ning of its understanding of the genesis of the magmas ofzoned pegmatites. The localized diferences brought outin this article probably result largely from differences inthe granitic phase of crystallization, but some of themmay have been influenced by differences in the compo-sitions and processes in the anatectic furnaces.

AcxNowr,BocMENTS

We are thankful for reviews of the manuscript by E. E. Foord, P. J.Modreski, D. B. Stewart, J. J. Papike, David London, and J. L. Munoz.We also are indebted to colleagues who have worked on Black Hills peg-matites and to prospectors, miners, and mine operators; the total of theseis in the hundreds We especially thank J. J. Papike and his associates atthe South Dakota School ofMines and Techlology, whp lavs significantl]advanced knowledge of the geochemistry of Black Hills pegmatites.

RrrnnnNcps crrnoBarker, D.S. (1983) Igneous rocks, 4 I 7 p. Prentice-Hall, Englewood Cliffs,

New JerseyBarnes, V.E., Edwards, George, Mclaughlin, W.A., Friedman, lrving, and

Joensu, Diva (1970) Macusanite occurrence, age, and composition,Macusani, Peru. Geological Society of America Bulletin, 81, 1539-r546

Barton, M D. (1986) Phase equilibria and thermodynamic properties ofminerals in the BeO-AlrOj-SiOr-HrO (BASH) system, with petrologicapplications. American Mineralogist, 7 l, 277-300.

Beer, K.E., Edmund, W.M., and Hawkes, J R (1978) A preliminary lookat lithium in the United Kingdom. Energy, 3, 281-292.

Burnham, C.W. (1979) Magmas and hydrothermal fluids. In H.L. Barnes,Ed., Geochemistry ofhydrothermal ore deposits (2nd edition), p. 7l-136. Wiley, New York.

Burnham, C.W., and Nekvasil, Hanna (1986) Equilibrium properties ofgranite pegmatite magmas. American Mineralogist, 7 1, 239-263.

Burnham, C.W., and Ohmoto, Hiroshi (1980) Late-srage processes of fel-sic magmatism In S. Ishihara and S. Takenouchi, Eds., Granitic mag-matism and related mineralization. Society of Mining Geologists of

- Japan, Mining Geology, Special Issue 8, l-l ICernj', Petr (1982) Petrogenesis ofganitic pegmatites. In Petr Cernj', Ed.,

Granitic pegmatites in science and industry. Mineralogical AssociationofCanada Short Course Handbook, 8,405-461.

Cernj', Petr, and Meintzer, R.E. (1988) Fertile granites in the Archeanand Proterozoic flelds of rare-element pegmatites: Crustal enwiron-ment, geochemistry, and petrogenetic relationships In R.P. Taylor andD F. Strong, Eds., Recent advances in the geology of granite-relatedmineral deposits. Canadian Institute of Mining and Metallurgy SpecialPublication 39, 170-207.

Cerni, Petr, and Trueman, D.L. (1978) Distribution and petrogenesis oflithium pegmatites in the westem Superior province of the CanadianShield. Energy, 3, 365-377.

Chatterjee, N.D., and Johannes, Wilhelm (1974) Thermal stability andstandard therrnodynamic properties of synthetic 2M,-muscovite, KAl,{AlSi3O,o(OH),}. Contributions to Minerology and Petrology, 48, 89-114

Duke, E.F., Redden, J.A., and Papike, J.J (1988) Calamity Peak layeredgranite-pegmatite complex, Black Hills, South Dakota: Part I. Structureand emplacement. Geological Society of America Bulletin, 100,825-840.

Duke, E.F., Shearer, C.K., Redden, J.A., and Papike, J.J. (1990) Protero-zoic granite-pegmatite magmatism, Black Hills, South Dakota: Struc-ture and geochemical zonation In J.F. Lewry and M.R. Staufer, Eds.,

Geological Association of Canada Special Paper on the Trans-HudsonOrogen, in press.

Essene, E.J (1982) Geologic thermometry and barometry. In J.M. Ferry,Ed., Characterization of metamorphism through mineral equilibria.Mineralogical Society of Amedca Reviews in Mineralogy, 10, 153-206.

criffirts, w.R. (1973) Beryllium. In D.A. Brobst and w P Pratt, Eds.,United States Mineral Resources U.S Geological Survey ProfessionalPaper 820, 85-93.

Heinrich, E.W. (1953) Zoning in pegmatite districts. American Mineral-ogist, 38, 68-87.

Holdaway, M.J. (1971) Stability of andalusite and the aluminum silicatephase diagram American Joumal of Science, 27 l, 97 -131.

Hoschek, G. (1969) The stability of staurolite and chloritoid and theirsignificance in metamorphism of pelitic rocks Contributions to Min-eralogy and Petrology, 22, 208-232.

Huang, W L., and Wyllie, P.J. (1973) Melting relations of muscovite-granite to 35 kbar as a model for fusion ofmetamorphosed subductedoceanic sediments. Contributions to Mineralogy and Petrology,42, l-t 4

Jahns, R.H., and Bumham, CW. (1969) Experimental studies of peg-matite genesis: I. A model for the derivation and crystallization ofgranitic pegmatites. Economic Geology, 64, 843-864.

Jollifi B.L., Papike, J.J., and Shearer, C.K. (1986) Tourmahne as a re-corder of pegmatite evolution: Bob Ingersoll pegmatite, Black Hills,South Dakota. American Mineralogist, 71, 472-500.

-(1987) Fractionation trends in mica and tourmaline as indicatorsof pegmatite internal evolution: Bob Ingersoll pegmatite, Black Hills,South Dakota. Geochimica et Cosmochimica Acta. 51. 519-534.

Kerrick, D.M. (1987) Fibrolite in contact aureoles of Donegal, Ireland.American Mineralogist, 7 2, 240-254.

Kleinkopf, M.D., and Redden, J A (1975) Bouguer gravity, aeromagnetic,and generalized geologic maps of part of the Black Hills of South Da-kota and Wyoming U S. Geological Survey Geophysical InvestigationsMap GP-903.

Lang, A.J, Jr., and Redden, J.A (1953) Geology and pegmatites ofpartofthe Fourmile area, Custer County, South Dakota. U.S. GeologicalSurvey Circular 245,20 p.

London, David (1984) Experimental phase equilibria in the systemLiAlSiO.-SiOr-H'O: A petrogenetic grid for lithium-rich pegmatites.American Mineralogist, 69, 995-1 004.

-(1986) Magmatic-hydrothermal transition in the Tanco rare-ele-ment pegmatite: Evidence from fluid inclusion and phase-equilibriumexperiments. American Mineralogist, 7 1, 37 6-395.

Luth, W.C., Jahns, R.H., and Tuttle, O.F. (1964) The granite system atpressures of 4 to l0 kilobars. Journal of Geophysical Research, 69,7 59-773.

Milton, D.J (1986) Chloritoid-sillimanite assemblage from North Caro-lina American Mineralogist, 71, 891-894.

Morgan, G.B., VI, and London, David (1987) Alteration of amphiboliticwallrocks around the Tanco rare-element pegmatite, Bernic Lake, Man-itoba. American Mineralogist, 72, 1097-ll2l.

Nesse, W.D. (1984) Metamorphic petrology of the northeast Front Range,Colorado: The Pingree Park area. Geological Society ofAmerica Bul-let in. 95. I 158- l 167.

Norton, J.J. (1970) Composition of a pegmatite. American Mineralogist,55, 98 1-1002.

- (1973) Lithium, cesium, and rubidium-The rare alkali metals. InD.A. Brobst and W.P. Pratt. Eds.. United States Mineral Resources.U.S. Geological Survey Professional Paper 820, 365-378.

-(1981) Origin of lithium-rich pegmatitic magmas, southern BlackHills, South Dakota. Geological Society of America Abstracts with Pro-grams, Rocky Mountain Section meeting, 1981,221

-(1983) Sequence of mineral assemblages in differentiated graniticpegmatites. Economic Geology, 7 8, 854-87 4.

Norton, J.J., Griffitts, W.R., and Wilmarth, Y.R. (1958) Geology andresources of beryllium in the United States. In Survey of Raw MaterialResources, United Nations International Conference on the PeacefulUses ofAtomic Energy,2nd Proceedings, Geneva, 1958, 2,2l-34.

Norton, J.J, Page, L.R., and Brobst, D.A. (1962) Geology of the Hugopegmatite, Keystone, south Dakota. U.S. Geological Survey Profes-sional Paoer 297 -8. 49-126

Norton, J.J., and others (1964) Geology and mineral deposils of somepegmatites in the southern Black Hills, South Dakota. U.S. GeologicalSuwey Professional Paper 297 -E, 293-341.

Orville, P.M. (1960) Petrology of several pegmatites in the Keystone dis-trict, Black Hills, South Dakota. Geological Society of America Bulle-tin, 7 l, 1467-1490.

Page, L.R , and others (1953) Pegmatite investigationsl942-1945,BlackHills, South Dakota. U S. Geological Survey Professional Paper 247,228 p.

Pichavant, Michel, Valencia Herrera, Jacinto, Boulmier, Suzanne, Bri-queu, Louis, Joron, Jean-Louis, Juteau, Martine, Marin, Luc, Michard,Annie, Sheppard, S M F, Treuil, Michel, and Vernet, Michel (1987)The Macusani glasses, SE Peru: Evidence of chemical fractionation inperaluminous magmas. In B.O. Mysen, Ed., Magmatic processes: Phys-iochemical principles. The Geochemical Society Special PublicationNo. l , 359-373.

Pye, E G. (1965) Geology and lithium deposits of Georyia Lake area.Ontario Department of Mines Geological Report l3l, I I 3 p

Ratt6, J C (1986) Geologic map of the Medicine Mountain quadrangle,Pennington County, South Dakota U.S. Geological Survey Miscella-neous Investigations Map I-1654.

Redden, J.A. (1959) Beryl deposits ofthe Beecher No. 3-Black Diamondpegmatite, Custer County, South Dakota. U.S. Geological Survey Bul-letin 1072-I, 537-559.

-(1963a) Diamond drilling exploration of the Beecher No. 3-BlackDiamond pegmatite, Custer County, South Dakota. U.S GeologicalSurvey Bulletin I 162-E, 1l p.

-(1963b) Geology and pegmatites of the Fourmile quadrangle, BlackHills, South Dakota. U.S. Geological Survey ProfessionalPaper 297-D, r99-29r

-(1968) Geology of the Berne quadrangle, Black Hills, South Da-kota U S. Geological Survey Professional Paper 297 -F, 343408.

Redden, J.A., and French, G M (1989) Geologic setting and potentialexploration guides for gold deposits, Black Hills, South Dakota. U.S.Geological Survey Bulletin 1857-8, 845-874.

Redden, J.A., Norton, J.J., and Mclaughlin, R.J. (1985) Geology of theHarney Peak Granite, Black Hills, South Dakota. In F.J Rich, Ed.,Geology of the Black Hills, South Dakota and Wyoming (2nd edition),p.225-240. American Geological Institute. (Also issued in 1982 as U.S.Geological Survey Open-File Report 82-481.)

Riley, G.H. (1970) Isotopic discrepancies in zoned pegmatites, Black Hills,South Dakota. Geochimica et Cosmochimica Acta. 34. 713-725.

Roberts, W L., and Rapp, George, Jr. (1965) Mineralogy of the BlackHills. South Dakota School of Mines and Technology Bulletin 18,268 p.

Robie, R.A., and Hemingway, B.S. (1984) Entropies of kyanite, andalu-site, and sillimanite: Additional constraints on the pressure and tem-perature of the AlrSiO, triple point. American Mineralogist, 69,298-306.

Salje, Ekhard (1986) Heat capacities and entropies ofandalusite and sil-limanite: The influence of fibrolitization on the phase diagram of theAlrSiO, polymorphs. American Mineralogist, 7 l, 1366-137 1.

655

Shearer, C.K, Papike, J J., and Laul, J.C. (1985) Chemistry ofpotassiumfeldspars from three zoned pegmatites, Black Hills, South Dakota: Im-plications concerning pegmatite evolution. Geochimica et Cosmochim-ica Acta, 49,663-673

Shearer, C.K., Papike, J.J., and Laul, J.C. (1987) Mineralogical and chem-ical evolution ofa rare-element granite-pegmatite system: Harney PeakGranite. Black Hills. South Dakota. Geochimica et Cosmochimica Acta,

51,473486.Shearer, C.K., Papike, J.J., Simon, S B , and Laul, J.C. (1986) Pegmatite-

wallrock interactions, Black Hills, South Dakota: Interaction betweenpegmatite-derived fluids and quartz-mica schist wallrock. AmericanMineralogist, 7 l, 5 1 8-539.

Shendan, D.M (1955) Geology of the High Climb pegmatite, CusterCounty, South Dakota. U.S. Geological Survey Bulletin 1015-C, 59-98 .

Sheridan, D.M., Stephens, H G., Staatz, M.H, and Norton, J.J. (1957)

Geology and beryl deposits of the Peerless pegmatite, PenningtonCounty, South Dakota. U.S. Geological Survey Professional Paper 297 -

A., 147.Shmakin, B M. (1973) The genetic types and the geochemical specializa-

tion of Precambrian pegmatites of India. In Recent researches in ge-

ology, l, 165-173. Hindustan Publishing Corp. and Department ofGeology, University of Delhi, Delhi, India.

Siroonian, H A., Shaw, D.M., and Jones, R.E. (1959) Lithium geochem-

istry and the source of the spodumene pegmatites of the Preissac-La-motte-Lacome region of westem Quebec. Canadian Mineralogist, 6,

320-338.Staatz, M.H., Page, L.R., Norton, J.J., and Wilmarth, V.R. (1963) Explo-

ration for beryllium at the Helen Beryl, Elkhom, and Tin Mountainpegmatites, Custer County, South Dakota. U.S. Geologrcal SurveyProfessional Paper 297 -C, 129-197 .

Stewart, D B. (1978) Petrogenesis of lithium-rich pegmatites. AmericanMineralogist, 63, 970-980

Van Hise, C.R (1896) Principles of North American PreCambrian geol-

ogy U S. Geological Survey 16th Annual Report, Part 1, 581-843'Walker, R.J., Hanson, G.N., and Papike, J.J. (1989) Trace element con-

straints on pegmatite genesis: Tin Mounuin pegmatite, Black Hills,

South Dakota. Contributions to Mineralogy and Petrology, l0l, 290-

300.Walker, R.J., Hanson, GN., Papike, J.J., and O'Neil, J.R. (1986a) Nd,

O and Sr isotopic constraints on the origin ofPrecambrian rocks, south-em Black Hills. South Dakota. Geochimica et Cosmochimica Ac1a, 50,2833-2846.

Walker, R.J., Hanson, G.N., Papike, J.J, O'Neill, J.R., and Laul, J.C.( 1986b) Internal evolution of the Tin Mountain pegmatite, Black Hills,

South Dakota. American Mineralogist, 7 1, 44M59.Winkler, H.G.F (1974) Petrogenesis of metamorphic rocks (3rd edition).

Spnnger-Verelag, New York.

Maruscnrpr R.EcErl'ED Apnn 10, 1990Maxuscnrvr AccEprED FesnuARY 5, 1990


Relations of zoned pegmatites to other pegmatites, granite, and

Documents

Transcript of Relations of zoned pegmatites to other pegmatites, granite, and