A:SHHSHMJGRAPHK:ANALfiflSKfl?THEIXKES(fi7THE (KNEHflCIUUfiBE.NfiCHKHUN Thuhfhrdnlkunncfilfl.s. NHCHKHU'STATEIHHVEEETY DumuficL.Ddena 1959 Eff-'- LIBRARY briichigan State University (J ' SPEC’IRCBRAPHIC ANALYSIS OF THE BIKES OF THE GOGEBIC RANGE, MICHIGAN A Thesis Presented to the Faculty of the Department of Geology Michigan State University In Partial Fulfillment of the Requirements for the Degree Master of Science by Domenic L. DeMarte June 1959 gcpominnsmwms The writer is deeply indebted to Dr. H. B. Stonehouse of the De- partment of Geology, Michigan State University, whose knowledse, guid- ance, and encouragement have made the completion of this study possible. Acknowledgement is also extended to Drs. James W. Trow and Justin Zinn of the Department of Geolozy for their helpful criticism durirq the writing of this thesis and to Dr. C. Prouty for his personal in- terest in this work. The author also extends his thanks to the personnel of the Peter- son mine and the Pickands and Mather District Office for their infor- mation and assistance in obtaining samples used in this study. Deep appreciation is also extended to the author's sister, Miss Mary Ann DeMarte, for the typing of this manuscript. -11- I. II. III. IV. VI. VII. VIII. IX. XI. XII. XIII. XIV. XV. XVI. XVII. TAB LE OF C ONTENI‘S ACKNOWLEDCEMENTS......................................... ABSTRACT................................................. GENERAL GEOLOGY - THE GOGEBIC RANGE...................... STRUCTURAL HISTORY....................................... INTRUSIVES AT THE PETEISON MINE.......................... LOCATION AND FIET PROCEDURE............................. DE ERMINATIONS OF STRUCTURAL ATTITUDES OF DIXES.......... AGE DIFFERENTIATION...................................... ETROGRAPHY.............................................. QUALITATIVE AND QUANTITATIVE SPECTROSCOPY................ PREPARATION OF SPECIMEN PRIOR TO ARCING.................. SPECTROGRAPHIC (FEC(JEU\JIQIF£SOOOOOOOQIOOOOOOOOOO...0.0.0.0... BURNING OF THE SAMPLE.................................... GEOCHEMICAL NATURE OF TRACE ELEMENTS..................... INTERPRETATION OF SPECTROGRAPHIC RESULTS SUMMARY.................................................. RECOMMENDATIONS.......................................... .APPEITDIX.0.0000000000000000000.00.000.000000000000.0.0... BIBIDIC'GRAPIEY...0....0.0.0..0OOOOOOOOOOOOOOIOOOOOOOOOOOOOO -iii- - DIFFERENTIATION Page ii v l 7 16 £6 us 66 77 83 93 I 0 u - 9 Q n o a o y o o I u o C O 9 0 g o 9 . o u a l I " o u v - O a ‘ ~ I n I U - u a I I O 0 I I n I - c O i Q C '0. OOOQIDQ' L ST OF FIGURES Figure 1. Map of Central Gogebic Range Sho in; Former and Present I'fine LocationSOOOOOOOOO0.0......OOOOOOOOOOOOOOOOOOOOOOOO Stratigraphic Column - Gogehic Range...................... Generalized Cross Section of the Gogebic Iron Formation... Sample Location I‘hp " DaVis Dike.............o............ Stereographic Projection of Intersection of Davis Dike and Iron Formation at Sample Locations #1, #3 and fill....... Photographs of Spectra.................................... -iv- Page DJ \n 147 ADCTPACT The primary ore controls on the Cogehic iron ranfie are a series of eastward pitching dikes which intersect the footwall quartzite and im- permeable layers of the iror formation to form plunging troushs, the locus of the majoritv of the ore bodies. The dikes are mineralocically similar, bein’r classified essentially as diabaso intrusives in their unaltered phases. At depths, the dikes form coMpleX patterns due to their intersectinc nature, and are fvrther complicated by the effects of the TeoloTic processes of faultinc, alteration hr leachinf solutions, and possibly metamorphism. Fence in sub—surface diamond drillint and mininv developmert, the identification and correlation of specific dikes, which is of consider- able economic importance, has become a difficult prohlem. It was the author's contention that whereas the dikes were essen- tially mineraloqically similar, a soectrochemjcal analysis misfit reveal considerable differences in trace element content. either "vantitative- ly and/or queljtotively Hnnn which identification and correlation could be made. In an effort to identify and correlate the dikes of the Gomebic iron range, a Spectrochemicsl analysis was made on 90 dike samples, 15 from the Davis dike, which is associated with major ore production on the ronre, and 5 from two associated dikes, the Tenevi dihe uni the Ironton dike, which are also related to ore production. -v- The Davis dike was set up as a control to test the homorenity of a single dike as to its trace element content and to serve for correlative purposes in the identification of the other dikes of the range. The 15 samples taken from the Davis dike were chosen so as to include a lateral coverage of anproximately 3,300 feet along strike and l,ooO feet alone dip. A qualitative analysis showed virtually no variations in tre ele— ments present with the exception of minor traces of scandium and lith- ium distributed erratically among the samples analyzed. The same ele- ments found in the Davis dike were also found in the Ironton and Geneva dikes. A semi—quantitative analysis revealed varying derrees of inten- sity of the spectral lines of the elements identified, thus indicatinm varyinr concentrations of these elements in the samples studied. Fur- ther investigation along these lines, utilizinq ratios of chemically related elements (Ms/Fe, Sr/Ca etc.) revealed marked differences in elemental ratios between varying locational samples of the Davis dike and also between the Davis dike and the Tronton and Geneva dikes. These variations appear to be related to the defiree of alteration, environmental control and structural trends 0f the dikes. It anpears that though a sincle dike may have been homoceneous at the time of its intrusion, the seologic factors mentioned above, operating either sinmu- larily or in combination with each other, have succeeded in destroyinc this innate homosenity. Hence, any identification or correlation of the dikes on these grounds must await further study. -Vi" GE‘TRAE e101 92!. TEE CQQPRIF_RANQE. The Goqehic district lies south of the west half of lake Superior in the states of Michigan and Wisconsin, extending from Lake Numakafon in Wisconsin approximately N3OOE to Lake Goeebic in Michigan, a total distance of about 80 miles. (Figure l) The major topographic features of the district are dependent upon the relative resistance of the formations, the strike of the harder formations being the primary directional controls of the ridges on the range. (Van Hise, lQll) Two main ridges trending somewhat south of west and separated by a valley from one to two miles in width exist. The crest of the south- ern-most ridge is formed by the iron formation or the formations imme- diately underlyinzit, quartzite and Archean granite and preenstone, while the northern ridge is formed by the Keweenawan trap rocks (lava flows) (the copper bearing series of Michinan). The valley in between is underlain by the Tyler eraywacke and slate formation. (w. o. Hotch- kiss, 1933) The present area streams cut an unorthodox path almost directly across the ranges and the longitudinal valley between them and pursue a staggered course of a few miles to Lake Superior, which is about one thousand feet lower than the iron ranee. a 52858... .3: 2m... 3 .0 IN... o. czoioox E 2 .m. oflI co> SE: .00 umfloco . \ . 6:2 33:96 20:55.3 I oooom \ _ I WumomboEcom Ufizocm flash \ \ - :3on .H m . coioEcom wE_on_ o..— oES- LO Q41). 55;: mozqz 0538 .2528 cozoEco... nooSco: l e H \ $35 3:: E eve V._ocourem,mo>o: r motwm 533$sz 33: . D l W \ \ I 16380 1 ~ , \ c x / k z r x f f \ \ Imwjmfisias . .. . i an - .... flflpfifihux 42"! F Ilfibemaflflr .Wuu! brim-lush... ‘ g i [leverages/1452212? _ . 'r... awurmmgal ‘ i. ' a I The rocks of the Gogebic range dip about 6OON and form the south limb of the Lake Superior syncline, with the conformable overlying Keweenawan lava flows and sandstones extending continuously from the northeast end of the Keweenawan Peninsula into Minnesota. (W. O. Hotch- kiss, 1933) A generalized stratigraphic column after W. O. Hotchkiss (1933) is shown in Figure 2. The oldest rocks of the district, occurring to the south, are pri- marily granites, greenstones, and green schists, igneous in origin and immediately underlying the Huronian rocks. The Bad River cherty dolo- mite lies unconformably on this base and is generally found as isolated remnants, most of it having been eroded away before the succeeding for- mation was deposited. In the eastern part of the district, the dolo— mite has a rather pure quartzite base, known as the Sunday Lake quartz- ite. Upon the dolomite, or lower formations where it is absent, lies the Palms formation. It is generally #00 to 500 feet in thickness, reaching a maximum thickness of 800 feet at Sunday Lake. It is marked by a slight unconformity with only thin lenses of conglomerate being dis- cernible at its base. The lower part of this formation is a quartz slate displaying thin to irregular ripple marked beds. The upper por- tion of this formation is a hard, glassy quartzite, 30 to 100 feet thick, forming the footwall of most of the ore-bodies found on the range. The Ironwood formation, which is conformable on the Palms formation, is the economically important iron bearing stratum of the Huronian series. It is dominantly a chert formation; interbedded with the chert are -3... STRATIGRAPHIC COLUmg Go sebé_9...F~.aIJ-::s -H. —- -,,,- “~'~WWA* .- -—..—o— --—-m Archean Post Pleistocene Keweenawan Cambrian s. s. (N. E. part of district) __ .Qfisafimflhspsigsmity 5 Basic Intrusive g Sandstone, Shale and Condlomerate g Acidic Lava Flows 8 Basic Lava Flows 5 S. S. and Conmlomerate i s . 3} Un conformity 551' ,i-“_a__.,_tw_ ._.-i_ _ 2‘ $4 Fraywack. Slates 8’ ‘Tyler Slates — iIron Carbonate Slates :g' iPabst (cherty and fragmental) LEerruginous slate beds Unconformityfi?) C Upper F_Khvil-wavy bedded m ferrudinous chart 'H m Ponce-even bedded a E: Basic Intrusives (some mamnetite) O '3 Ironwood Iron Formation -— Norrie-wavy bedded “ E3 Palms Quartzite ferruqinous chert 5 Lower Yale-even bedded “4 thin ferruainous chert and slate Plymouth-irregular wavy bedded chert .ii-_i_illn“iiiEzzefiisazlnfipcesfermity a Bad River Dolomite (cherty, limestone) ‘ 6 Sunday Lake quartzite .~‘*_ T:H _*____ __h‘ Great Unconformity_ l Laurentian Granite and Granitoid Gneiss Granite (Intrusive into Keewatin) Keewatin Greenstones and Green Schists Fipue 2. -u- thinner beds of iron minerals. The iron minerals in the unaltered part of the formation are siderite, magnetite and hard blue hematite, while in the productive part, the siderite is completely replaced by or altered to hematite. This formation is divided into five members based on differing lithologic characteristics of the material, as demonstrated by both iron content and specific gravity. In order of succession from the footwall quartzite, they are the Plymouth member, characterized for the most part by irregular wavy-bedded chert with iron minerals occur- ring both in the bands between the chert layers and also in small par- ticles throughout the chert; the Yale member which is chiefly a thin, even bedded ferruginous chert and slate member; the Norrie member, a dominantly wavy—bedded ferruginous chert; the Pence member, preceded by a slight erosion interval, which is unusally even bedded throughout most of the range and contains a moderate amount of magnetite; and the Anvil member, which is also a wavy-bedded ferruginous chert composed Largely of rounded chert granules and differing from the other wavy- bedded members in this respect. The Tyler formation lies upon the Ironwood formation throughout most of the district except in those places where it has been removed throurh erosion. It is primarily a slaty formation reaching a maxi- mum thickness of 10,000 feet and divisible by material content into a ferruginous slate member, cherty and fragmental member (Pabst), iron carbonate slate member and a graywacke slate member. The deposition of the Tyler was followed by the formation of the great Sunday Lake fault which moved the beds east of it up and to the northwest. This was -5- I.-- "*"‘ ”‘rx . ‘ . .rsrwq r. r5 'i- .1 ~u-v J27 u; r -u- «t ‘.‘ “v P)“ M‘“ .r- -‘-o followed by an erosion interval of sufficient length to remove the Tyler and part of the Ironwood east of Sunday lake. The Keweenawan sediments were laid down on this unconformable surface, the lower por— tion varying from sandstones in some localities to quartzites in others, followed by successive volcanic flows, mostly dark green diabases and amygdaloids with porphyries and syenites. smuemm. 3182031 A discussion of the geologic processes of folding, faulting, and intrusive activity which have occurred in the Gopebic range and their relationship to ore development and its localization now follows: Eoldingi Extensive folding in this area has occurred only in the non-productive east and west ends of the range. Folding in the produc- tive portion of the range is of a minor character, being confined to small drag folds only a few feet across. These folds as far as known, have not played any part in ore localization. Faulting: Faulting throughout the extent of the iron formation and the rocks associated with it is very common. (w. 0. Hotchkiss, l937) The influence of faults in ore development has been cited previously and hence, th’ ‘ir“us inn Vh‘ch folluvs ‘" in reference to the types of faults and their aye reletionsrins. 0n the basis of the evidence on hand, four types of faults of rela- tive age occrrr‘ng throughout the range have been described. An adapted description of these faults after Wotchkiss, 1937 follows: 1. Transverse faults - Faults striking nearly perpendicular to the formntion and nearly vertical in din. f0 ?urei a type faults - Faults striking nearly parallel to th (I) strike of the formation or parallel to eastward-pitchinf dikes, and nearly perpendicular to the beds. 3. The great Sunday lake fault, the onlf one of it: tyre. -7— h. Pedding faults - Faults oarcllel to tle Veda of the irrn for- mation. The transverse faults are the only ones krown to affect the Ke- weenawan volcanic rocks and are probably the latest faults to be formed. Horizontal disnlacement by this tyne of fault rabies upward to lSOO', thounh in deneral the disnlacement bv these faults throughout the Iron- - - E . . wood formation is of the order of 20 ~ 100'. Vertical movement is un- determinable. mbe Eureka faults are ramed after a type of fault found in the Eureka mine, which dips toward the footwall and strikes nearly parallel to it. This tyne of fault which offsets the iron formation has also been observed in the Asteroid and Mikado mines to the east, and in the Plymouth and Wakefield mines. The TPureka fault is offset by the transverse faults and so is 4L1 older. It in turn offsets the bedding fault and the sill so that these must be still older. (W. O. Fotchkiss, 1937) The Sunday'lake fault, which represen s the greatest fault movement on the range, has a northwest strike and is nearly vertical in dip. The base of the iron formation on tte east side of the fault has been dis- placed horizontally one and one-half miles to the northwest relative to the base of the iron formation on the we: side. (M. C. lake, 19l7) The movement along this fault plane, which took place while the beds of the iron formation were still horizontal, was such that the beds on the east side of the fault were over-thrust unward in a westerly direction. (w. o. Notchkiss, 1937) —8- 1". l|:,”‘ Following the formation of this fault, there was a period of ero- sion which almost completely removed the Tyler slates on its eastern side allowing the outpouring of the Keweenawan lava flows to rest directly on the iron formation in many nlaces. As far as known, this fault does not cut the Keweenawan volcanics and is not affected by the bedding fault. Rather, it seems probable that the Sunday Take fault must have disnlaced the bedding fault since smaller bedding faults believed contemnoraneous with the main beddinq fault are known in the iron formation east of the Sunday Lake fault. Bedding faults within the iron formation are found near the base of tte Plymouth member, in the Yale member, the Pence member and also in the Pabst member of the Tyler formation. The movement on these fault planes is relatively small excent for the great bedding fault found in or near the black ferruqinous slate of the Yale member. (W. 0. Hotcb— kiss, 1937) The main movement alone this fault plane was such that the beds above were moved to the east, with minor movements up or down the dip of the beds. The main movement resulted in displacements, where measur- able, un to 950 feet. The bedding fault is displaced by the transverse faults and by faults of the Eureka type which indicates this fault was older than the two latter tynes. In turn, the main beddinq fault is known to displace at least one intrusive in the iron formation and possibly affects others. '1‘-" A discussion on the intrusives and extrusives originating in the Huronian and the Keweenawan geologic periods in this area now follows. Particular emphasis is given to this phase of the geologic setting due to its intrinsic relationship to the problem at hand, namely, that of dike identification and correlation. A real problem arises because of the similar mineralogic and structural relationship of these rocks. Although the author is primarily concerned with those dikes on the range which have been instrumental in the formation of the ore—bodies, the innate geologic behavior of the sum total of all intrusives is so integrated that a solution to this immediate problem necessarily in- volves postulation as to the origin and age relationships of all in- trusives on the range; because of their interrelated characteristics, extrusives are also included in this discussion. Geologists have yet to agree on any absolute means of differen— tiating Keweenawan intrusives from those of the Huronian, although par- tial evidence related to metamorphic effects, rank and mineral assem- blages, chemical analyses, and to some extent structural behavior and the concept of polarity have been introduced in this respect. W. O. Hotchkiss (1933) describes the intrusives in the Huronian as numerous basic dikes whose structural behavior is that of near per- pendicularity to the beds of the iron formation which they intrude. This structural relationship has also been observed by the author. Two of these dikes are known to intrude both the Ironwood forma- tion and the Keweenawan lava flows, but most of them cannot be traced into the Keweenawan. -10- At Bessemer, a thin sill,.which thickens toward the east until it reaches a maximum thickness of #50 feet at the old Mikado mine, is found near the great bedding fault. It is intruded by a dike, but is not known to cut any dikes. The other significant intrusive is found at the east end of the area near Lake Gogebic, where Huronian forma- tions are intruded by Presque Isle granite. (w. 0. Hotchkiss, 1933) Hotchkiss reserves any opinions as to definite age relationships of the sum total of intrusives found in this region though suggesting two different periods of intrusive activity and hinting that the Presque Isle granite may be pre-Keweenawan in age. Van Hise (1892) proposes that the Keweenawan eruptives above the Penokee (Huronian) series are the volcanic equivalents of the Penokee dikes and sheets, their contacts being sometimes intricately related. 'These facts strongly suggest that the Penokee dikes are the pipes through which the volcanics passed. The identical lithological char- acter of the dikes and sheets suggests that the former fed the latter. We are thus led to the hypothesis that all of these eruptives belong to a single period, the Keweenawan." (C. R. Van Hise, 1892) The author at this time wishes to reserve any personal opinions in regard to this controversial issue and instead present a more detailed description of these intrusive rocks, particularly of those found in the Ironwood formation. These dikes are essentially typical diabase rocks in their un— altered form, but within the range they are commonly altered to what is referred to by mining personnel as soapstones. -11- The lithologic significance of this term (soapstone) is that of a rock which is soft, friable and greasy to the touch in which little, if any, of the original constituent minerals of the primary basic rock remains. These dikes, however, even in their most altered form, frequently retain a relict diabasic texture and can be traced in many instances into comparatively unaltered phases which are true diabases. Alteration of these rocks has extended furthest in those portions of the iron formation which contain the ore-bodies. The feldspars in the dike having been altered to kaolin or replaced by chlorite and the pyroxene being converted to hornblende, biotite, or chlorite. The con- tinuation of these dikes into their unaltered portions, usually finds them cutting relatively impermeable beds of the upper slates. Van Hise (1892) suggests that the contrast between the altered and unaltered portions of the dikes is a result of the influence of environ- ment upon decomposition, the diabases enclosed in the impervious upper slate being kept in a well preserved condition since their intrusion, whereas the portions of these same dikes in a formation which contains evidence of having been subject to the leaching action of percolating waters have been almost totally decomposed. It is assumed that the alteration by leaching solutions is post- metamorphic in origin. Because of their intimate association with the iron ores, the Structural trends of the dikes are given with reference to the iron formation in which they occur. The iron formation has a general east- -12- west strike and dips to the north at 600. The dikes, however, though varying considerably in their dip and strike at various mines, always dip to the south with a southernly component of 200 - #00 and generally pitch to the east with a pitch as high as 350, but this amount may vary to horizontal or even a slight western pitch. (Van Hise, 1911) From these observations, it can be concluded that if the iron-bearing strata were returned to a position of horizontality, the dikes would be near- ly vertical. This suggests that the dike intrusion took place before the sediments were tilted to their present attitude. The intrusives vary greatly in thickness ranging from a few inches to as great as 90 feet; at depths some of the main dikes are known to split into tongues or stringers and ore is occasionally found between these components. The function of these dikes as ore controls has initiated diversi- fied opinions among geologists. Van Hise (1912) preposes the leaching action of downward percola- ting waters as the main contributary cause to ore formation and hence assigns the role of the dikes to be due to their orientation and rela— tionship to the land surface and the forces of gravity acting upon meteoric waters. S. A. Tyler (19h9) suggests that the escaping water from the dike magma plays the more important role in ore formation, whereas J. w. Gruner,(l930) prOposes that the dikes served as conduits for ore-concen- trating hydrothermal solutions. In either of these latter cases, the structural position of the -13- . \ ' O I . , I x q . . - . . . . 0 l \ . ‘ 4 I g ‘ \ . ‘ , ' O . . O \ i . dikes in secondary concentration is not of primary importance. Structural and mineralogical data and ore—dike relationships tend to support Van Hise's theory of downward percolating waters as the primary enriching agent contributing to the formation of ore-bodies. For this reason the discussion of ore development in relation to struc— tural controls which follows is intricately related to this theory. The dikes in intersecting the more impermeable layers of the iron formation and the footwall quartzite form eastward pitching structural troughs which acted as the main circulatory controls for downward per— colating surface waters, the agents which enhanced the formation of the iron-ores. Structural controls are also formed by the intersection of two or more dikes. The channelways which carried these solutions were generally fault and fracture zones or the more permeable members ofthe iron-bearing series. In almost every case the ore-bodies are found immediately overlying the dikes in the apices of these troughs. The alteration of these dikes (through decomposition) appears to be taking place from their structural tops downward, thus supporting the hypo— thesis of enrichment of the iron formation by downward percolating sur- face waters. A generalized cross section across the Gogebic iron for- mation is shown in Figure 3. -1n- _ . o M ’ - ' \\\\\ (I _ “WV \ '0‘.. Figure 3. GENERALIZED CROSS SECTION OF THE GOGEBIC IRON FORMATION SCALE I: ma (PICKANDS a MATHER COMPANY) am—A smegma"- PW Baa are The dikes of the Peterson mine are typical of most of the dikes of the range as previously described. They have a diabasic texture which is persistant even in their altered phases, and correlate closely with the general mineralogic description of the sum total of dikes on the range as described by Leith and Van Hise having approximately 51% feld- Spar, LO% ferromagnesium minerals, 1% quartz, 5.3% magnetite, 1.5% il- menite and small amounts of calcium-magnesium carbonates and apatite in their less altered portions. These dikes are approximately parallel to each other, generally forming low angles in those instances where they do intersect. Their horizontal traces on the footwall are inclined about lho east, these intersections forming the axes pf eastward pitching troughs as pre- viously described. The main ore carrying dikes at the Peterson Mine are the Davis, Geneva, Ironton and Puritan dikes. Ore production along the Davis and Puritan have now ceased. The Davis, Geneva and Ironton dikes were selected for this study and sampling was confined within the limits of their boundaries. Their structural positions are described as follows: The Davis dike has an average strike of N 650E, a variable dip ranging from 250 to 50° southeast and pitches (rake on iron formation) from 00 to 200 northeast; the Ironton dike has a strike varying from NHOOE to N620E, has a dip of 23° southeast and pitches 13° to 2&0 northeast; the Geneva -16- 1 o . I . \J ‘ . . . x o . . . \ I u \ , p \ V ' ‘ .a o -' A . v .1 I I c 4 I u I I ' \ «>1- 7' l ‘ . ' ‘ 9 ~ I - l ‘ - n . ~ - . -. . I __. . u .- .A " v ' . 1 k4 ‘ v -. - l\ _ _ _, .1 5.4 , _ ~‘J d _ ‘_ >_ .. .A . ‘ ' V I I ‘- L '7 h ‘7 - _ ‘A . . AA , _ . . . > 9 A - . w a ‘ v , ‘ _ .. . ‘J I ‘ i ‘ 1 a I > I . - , I a -s ' 9 . \ I A l ' ‘ - - I . ‘ . . . . ' v -.A > ' J -‘ 1 ' I. v , _1 i \ , - s» ,- -- \_ .. ' - . I w J . k A ., ~' ‘- ‘ 1 > ‘ I I o ‘ ‘ . h ‘7 ‘ - J ‘J o—\‘ _~ '3 ~ I a > . .. ' . ‘ ) ’ r . .. . ‘ ‘ . .’, k, . ,g ..at.-— . t . '. . l , A ' y . - ‘ V “ ~ . I. ‘ ~ ‘ ~ / I - _ t . \, .r , , kl .‘ J.) . ‘ ‘ . . ‘ ‘ l . . . . . ‘ ' ' “ ' ' I -' \ - Z. . , ,. - :1 . _ " 1 . I ' 3' ' . |. I ‘ ’~ '1 ‘ v ‘. ‘ ' _ . ‘ J U .7 ‘ - ‘ _ 1 » , , _ L A . . dike has a strike varying from N56OE to N780E, dips hlo southeast and has a pitch which varies from near horizontality to 160 northeast. -l7- 1.00M; QM. _ mantra £139ch IRS. The samples utilized in this study were obtained from diamond drill core and underground workings at the Peterson Mine located in Bessemer, Michigan, in the heart of the Gogebic iron range, seven miles from the eastern border of northern Wisconsin, and 16 miles south of the south- ern shore of Lake Supeior. This mine, one of the six still active on the range is operated by Pickands Mather and Company affiliated with Bethlehem and Youngstown Steel Corporation. Due to the nature of the ore-bodies as to shape, size and manner of formation, the majority of the mines on the range are confined to the underground type utilizinrr sub-level caving mining methods for ex- traction of the ore. Since the dikes are the main ore controls and hence the locus of the majority of ore bodies, most of the mining de- velopment is close to the dikes making them accessible to geologic in- vestigation. Because of the otherwise inaccessible nature of the dikes, samples had to be confined to those portions of a dike exposed by un- derground workings and diamond drill core which had intersected the dikes during preliminary exploration. The Davis dike has been one of the largerW e producing" dikes of I ‘ ’ f‘ '\ C T‘ the range, its lateral and vertical extent and structural treads 081 a 4" ' "49. confirmed through numerous exposures as a consequence on prelimi ry exploration and mining development. Mining along this dike has been . , p. k. Carried on extensively at the Peterson Mine. hence the choice of t-is particular dike for preliminary sampling. -18- /‘ vd' The spacinq of the samples was not as selective as the author would have preferred, but they were taken at all accessible exposures to include a lateral coverage of 3,300 feet along the strike and approxi- mately 1,000 feet along the dip of this particular rock unit. Two addi— tional samples of the Ironton and three of the Ceneva dike were taken for correlation. The locations and field descriptions of the various samples are given below: Davis ‘11 he. v..-’ «n- -. =fiu 27th Level "A" shaft x-cut gfi?5§, lfgfif, lZ:3 Elevation strike TSOE, .-~.-A .- - Dip 59:33, pitch 9:. Hard, fire rraired, dark :ray—black, fresh appearing, unaltered diabase rock. Some red—lematite stain alcn: joint planes and fractures. Overlain by iron formation above and quartz slate be- low. Sample taken h feet from iron formation contact. ¥2. 27th Level "A" slaft x-cut, E. Urift, 1P5 feet from x-crt. “3&0 E., 17ho V.. Elevation 17cc, strike UYoOt. Dip ?5” at g... —~... —-———. - pitching 60pm. Yard fine-freired, dark fray-black, unaltered diabase rock. Sample taken l2 feet from lower apart? slate contact. r3. 97th level shaft x—cut, F. drift, 75 feet from 2600 E loading sub. 2595 F., lBhS N, Tlevption TYCl, strike V6GOV, bin ‘1‘.” h90 SE, pitching 90-10' W?, Semi-bard, fine grained, partially altered, light green-firey dia- base rock with red-hematite stain alone fractwres and evidence of -19- I .1:- fl)- ;[r/ 6 . #11. kaolinization and chlcrite alteration takinc place. 27th level. PYhO X—Q"t North 97?0 R. 1040 N, Elevation l700, strike ~— m-‘ NRnO a, Bio is” es, ?itchinc a“ we. Soft, friable, red—hematite stained altered diabase rock (span— stone). Green arfillaciovs material along joints and fractures. Sample taken 3 feet from iron formation contact. hth Sub above 97th Level, 1800 a. Raise 1800 r iflho M, Elevation .I .....-’ £239: strike g§93_nip goo Cs, Pitchflgff Semi-hard, fine grained, red hematite-stained partially altered diabase rock. 90ft white clay-like mineral insinuated along fractures and joints planes. U9rd blue hematite ore abcve dike. Sample taken h feet from ore contact. 7th sub above 27th level, 1550 E. Raise, lFSO F, lRGO N, Eleva— tion 1h70, strike nRoo F, Dip too be, Pitch of. Soft, friable, extremely fine trained, red hematite-stained al- tered diabase rock. Soft white clay-like mineral insinuated along fractures and joints planes. Word blre hematite ore above dike. Sample taken h feet from ore contact. 28th revel fogs drift, 3ooo r, eofifi N, Flevation TQOR. strike N502, Dip noose, Pitchin: no“. —__ Hard, fine to medium grained, partially altered, dark gray to green-black diabase rock. Upper contact with quartz slate. Sample taken 7 feet from this contact. strike —\ m “2‘ 28th Level, "A” shaft x-cpt 9739?, VQYCW, Elevation 100?, NSOOE, Dip homes, Pitching 200- -90- #12. Hard, fine to medium grained, unaltered, fresh apnearinfi, green- black diabase rock. Soft, orange-brown translucent mineral alonr a joint nlane (talc). Contact with granite at its base and qvartz slate at its ton. 9ample taken 15 feet from cranite contact. #1h. 29th tevei "A“ shaft x—cut, 200 feet 5. of car-shifter cutout e180 a, lfiRflF, Elevation nine. Strike neat, Dip 950cm, Pitch is undeterminable (dike intruded into granite) Hard, fine to medium crained, eenerally freon anpearing dark nray to green-black diabase rock. Slight kaolinization along fractures. Sample taken 2 feet from unner contact with Prnnite. Davin “ike (UiOmond Drill Uole Camnlerl 315. Diamond Frill Ft1e 512, 3075? eaefiw, tievetion 1179, fitrike NGSW, Din 222:, Pitching loom”, dike from 9l3-T7l feet. Semi-bard, fine to medivm grained, partially altered, dark gray diabaee rock, with thin (l/6h inch) white, clay-like laminae l/h inch apart throughout. Fli“ht red—hematite stain discernible. Ore immediately above dike, cherty iron formation below. Camnle taken at PhO feet. #16. Diamond Drill Hole 508, 13533, iaian, Flevation 1203, Strike neon, Din ggfg, Pitching 13OVE, dike from nah-526 feet. Hard fine to medium grained, unaltered, preen~black diabane rock, intrusive within footwall quartz slates. Famnle taken at 510 feet. #17. Diamond Drill Hole 507, 22002, 17925, Elevation 1102, Strike N630n, -91- l- 1. #17. #18. Dip 32:3, Pitching 110nm, dike from see-sue feet. Hard, medium grained, unaltered, fresh anyeerinr, dark gray- black diebase rock. Some quartz discernible but only in minor amounts. Dike intrusive within the onartzite phase of the Palms formation. Fault zone immediately ebove. Sample taken at 590 feet. Diamond Drill Vole 500, eeooe, 23SSN, n1,vetion 1185, Strike N62E, Dip §ES§2 Pitching_11f-3ont, dike from o-3u feet. Fard, fine to medium grained, sligbtly altered, dark gray to green—black diabase rock. Wavy-bedded chert? iron formation at bottom and top. Cample taken at 15 feet. Diamond Drill Hoie 510, 391st, 17RGN, nievntinn-iiflu, Strike N65e, Dip 390s, Pitching 100nm, dike from 205~ee6 feet. Semi-bard, medium grained, partially altered, gray-green, diabase rock. Some kaolinization, but primarily chlorite alteration throughout. (A green mineral wricr may be epidotc, present.) Contact with quartzite slate above and qrartsite below. Samnle taken at 2lO feet. Diamond Drill Hole 511, 301og, eéoev, Flevetion 1177. Strike nése, O . O . . . . Dip QO'R, P1tcb1n; l0 NR, dike from l?8-JRY feet. Hard, fine to medium Crained, sliettly altered, gray—black dia— base rock. Some chlorite alteration and red-hematite stain alonj fractures. wavy-bedded iron formation above and below. Sample taken at 155 feet. Esaexafliia # 7. 27th Level, w. Drift of l3OOE. x-cut, §§g_§, giégg, @lexatignnlégg, Strike §1§E§, Dip ggfsa, Pitching gigs, Semi-hard, fine grained, altered red hematite-stained, diabase rock. Extensive kaolinization throughout. Sample taken h feet from contact with iron formation. #13. 29th revel, x-cui N. , 229E. 2.21-5.5“... Bieyeiise 2923) Strike 91.9.52 Dip ggfsg, Pitching gfgp. Semi-hard to soft, medium grained, altered, green-dark gray dia~ base rock. Chlorite alteration common throughout. Schistose like structure along a joint plane. (Caused by alteration of a fiberous mineral?) Sample taken 10 feet from upper contact with Yale slates. #20. Diamond Drill Hole 505. Lia—ii). amen. iieyei.iee_i.i2<2. Strike ggéfg, Dip ggfs, Pitching iéfyp, dike from 61h-7o3 feet. Semi-hard, medium to coarse grained, partially altered light to dark gray diabase rock. Kaolinization and chlorite alteration common throughout. Some red-hematite granules scattered through— out, believed to be alteration of magnetite crystals. Hard blue- hematite ore immediately above dike, cherty iron formation below. Sample taken at 660 feet. Boater}. Bee. “- # 8. 27th Level, main drift, uoow, 1280N, Elevation 1675, Strike whoop, Dip usosE Pitching 2u9ug. -23- # 8. Semi—hard to soft, medium grained, altered light to dark green— gray diabase rock. Extensive kaolinization and chlorite altera- tion throughout (talc present along joint plane). Sample taken 2 feet above floor i feet from quartzite contact. the 28th Level feet drift 9a emit Eiaiieiieriiéii?’ Strike mega Dip EQS§§, Pitching lgggg. Hard, medium grained, generally fresh appearing gray—black to green-black diabase rock. Some chlorite alteration along with transversing veinlets of white argillaceous-like material. Some red-hematite stain along fractures and joint planes. Sample taken 3 feet from upper contact with quartz slates. The samples taken from underground workings were carefully removed from exposures with extra precaution being taken not to contaminate the sample with foreign matter. This was done in the following manner. First, a portion of the immediate wall exposure was pried away with a geologic pick exposing a section of the dike free from contamination likely to occur during mine operations. A sample was then removed from this exposure, the portion of it touched by the pick being broken off, using another piece of the rock as an improvised chisel and striking it with the pick. Similar procedures were used on all other samples collected. A representative sample approximately 2 inches by 2 inches was obtained from all locations. These were placed in doubled Paper bags, marked and stored. (See Figure h for sample locations) -oh- 26 25 24 23 2! .ga_ N O I i T ' \ g 51 Kf/ l [ I 1 . I I ’3 / { «ref/K ' V D f l # W A» 3’ ‘ i I 1 , ,1 1 1 1 1111.. 1 1 1 1 1 111- 11111. ' ’ : } /// g 2 “239“. l 11 l,l 1 ,1 _ i111 ‘31". "1i 11,—, 1 1 1 1 i 1 l i 7‘“, l ‘ l i t A T I 1 1 g’ i l i ‘ l L 4411111 ,1l11 e» * ~ : t ;:\,/‘(\ a 1 i e T T i i fitter l V . 1 45 i l 11.// g L 3 “5/ II 4 P”' / ’ J v s A!” I 4, whqy a, E ‘ , l 45' 53' eff R 40° : {/L‘./ :90 1 16° I l l 1 1 1 l 1‘ . 11l 111 1 1- i plan 11 _ 1,1 l, 1 l rev. 11 1. 1‘ ‘K o #‘b V ‘0 1‘ ‘S 39¢ 4"‘_¥_~ iA—‘i V— ‘>‘ V fib—r—v 11111 1- < l1111.¢l1‘1/.,1 77’“ A V M —‘ / N 6° 2 l4 is 16 l7 l8 I9 20 2| 22 23 24 25 26 27 28 29 30 3| 32 33 Figure 14-. SAMPLE LOCATION MAP DAVIS DI K E ‘ Scots l"=200' DETERMINATIONS OF STRUCTURAL ATTITUDES 0F DIKE§_ Determinations were made of the structural trends and thicknesses of the dikes at their various sample locations. No direct thicknesses were measured in the field, but rather computed from mine map data. All rakes (pitches) and plunges of the dikes were computed by the sterographic projection method as described by Walter H. Bucher (19th). In every instance, the rake (pitch) computed, was that of the dike on the intruded formation. The results disclosed that these dikes all pitch to the northeast, contrary to the southeasterly pitch as described in previous literature. See Figure 5 for an illustrative example of the method used to derive these results. -26- Figure 5. ___ ~‘ 8 Stereographic Projection of intersection of Davis dike and iron formation at sample loca- tions #1, #3 and #11. A-B, C—D, E—F are plunges of dike, O—A, o-c, O—E are pitches of dike as measured by its trace on iron formation. -27_ ACT DTTTTDEletTTPW As was indicated previously, the determination of afie relationships of these dikes still remains a controversial issue. Some of the evi- derce uncovered in this respect based on particular mineral assemblajee, more common to either the Huronian or Keweenawan intrusives is related below. I .' Huronian Dikes: Tb Huronian dikes are generally metamorphosed, the original plagioclases being recrystallized with a greater Ab/An ratio since calcium is removed during metamorphism. Hence, a metamorphic min- eral assemblage, dependent on the metamorphic rank attained (in this case, most likely that of the chlorite sub facies rank) is to be expected in dikes of Hurorian age. These dikes presumably never reached the pre— existing surface, and hence, most alteration is concluded to be the re- sults of the effects of downward percolating surface waters. The two contributory causes to the present physio-chemical makeup of these dikes are thus believed to be the effects of metamorphism with resulting new mineral assemblages and the alteration of these minerals through the action of leaching solutions. Keweenawan Dikesz- The Keweenawen dikes are unmetamorphosed and gen- erally fresh appearing, with the original mineral content being preserved. The altered phases are a result of weathering and not due to metamorphic effects. A number of these dikes reached the surface and no doubt served as conduits for the Keweenewan lava flows. The original mineral content typical of diabases is present with an abundance of calcium rich plagio- _28- clase, monoclinic pyroxene, some olivine, marnetite and/or ilmenite and biotite. EEECQREBEIX The extent of the petrographic study of these dikes was confined to that of a supporting role in an effort to predict the gee-chemical behavior of the trace elements found therein and deduce their origin and also possibly supply evidence in regard to age relationships of these rocks. For a more complete discussion of the petrography of these dikes, the author makes reference to literature by Irving and Van Hise (l892) and the work of George A. Hoffman (1950). An adapted description of the diabases of the Gogebic range after Irving and Van Hise (1892) is as follows: "The diabases usually have a well developed ophitic structure, the augites being of large size and including many somewhat idiomorphic lath- shaped plagioclases. In the diabases in which this structure reaches the extreme, the feldspars have a tendency toward two generations, there being aside from the smaller lath-shaped plagioclases, larger, somewhat porphyritic appearing ones. The rocks vary from ophitic diabases to true gabbro, all grades of variation being observed. The gabbro occurs in only a few localities and is of little importance as compared with the diabases. The original minerals are apatite, magnetite, olivine, plagiocLase and monoclinic (augite) and orthorhombic pyroxene. The latter occurs only in one exposure and in the most wideSpread phase of rock, the only important original minerals are magnetite, plagioclase and augite. The -30- . . a . - o 1 . v s l .1 order given is that of crystallization. In some of the rocks, this suc- cession can be made out with a good deal of sharpness, each mineral present having nearly completed its crystallization before the succeed— ing one began to separate. This is particularly true of the ophitic diabases and becomes less and less true in passing toward the gabbros.” The environments were peculiar to each sample utilized in this study, selected in such a manner as to permit a petrographic investiga- tion from the most altered to unaltered phases of these rocks, thus allowing the observation of the mineralogic changes taken place through processes (and possibly metamorphic effects). Four thin sections were prepared from samples of the Davis dike. The samples utilized came from three environments; namely, granite, footwall quartz-slates and the iron formation. An additional thin section of the Geneva dike was prepared from a sample taken close to reLatively unaltered iron formation. The degree of alteration showed a marked increase in going from the granitic environment through the footwall quartz slates and into the iron formation. In the slides examined, the minerals present consisted of labrador- ite, olivine (questionable), augite, magnetite, some ilmenite and biotite, along with leucoxene, hematite, biotite, chlorite, sericite and kaolinite. The iron oxides were for the most part (except in altered phases) magnetite and some ilmenite, occurring as subhedral crystals or linear masses scattered throughout the thin section. They are primary consti- tuents of the rock and fairly abundant. The ilmenite is generally -31- represented by its alteration nrodvct, lencoxcre, and is intimately asso- ciated with the magnetite. Some oxidation to red lematite is discern- ible witlir individual maénetite trains. In the more altered phases, the iron oxides consist wholly of red hematite, the magnetite having been completely oxidized. T’~eceu:se of the deep penetration by red- hematite stain, the other constituent minerals of the thin sections representing the altered phases of this dile'could not be determined, with only the Opbitic strncture being preserved. The feldspars are restricted to the calci-rich end of the nlacio~ clase series being essentially labradorite with the possibility of some andesine. They are generally found as small idiomorphic, striated, lath-shaped individual crystals displaying polysynthetic twinning and on rare occasions also Carlsbad twinning. In the fresher portions of this rock they are well preserved and extremely abundant, whereas in approaching the altered phases they are frequently altered to sercite and in some cases kaolinite (the determination of kaolinite was based on x-ray study). Determination of specific plagioclase was based on indicative refractive indices and eytinction angles. The nyroxenes are essentially those of the audite-ferroaugite series with some indications of associated pigeonite, although this variety is minor in the rock slides examined. They are roughly equi— demensional, stubby to elongated crystals diSplaying prismatic cleavaze on the 110 plane and 100 parting. In the unaltered portions of this rock the individual grains are fairly well preserved, their crystal boundaries being sharply defined in some instances, but for the most -39- part controlled by the plagioclase lathe. In the altered phases, the pyroxenes have been converted almost wholly to a pale green chlorite. In some instances augite can be found growing around nlacioclase lithe indicating that the mineral is later than the feldspars. Riotite was present in these rock specimens in which alteration had begun to pass from that of intermediate to an advanced stage. In these slides no pyroxene is apparent, magnetite is abundant and chlorite and kaolinite alteration is extreme. The biotite which is quite pleochroic in shades of brown frequently surrounds magnetite "rains. Some of this mineral appears to be secondary after the pyroxenes and for the most part is an alteration product or a reaction mineral. Biotite is con- spicuously absent from the sections of the unaltered rock that were examined. The presence of olivine is suspected in only one instance where an area of what appears to be serpentine alteration maintains a roundish boundary suggesting the former presence of an olivine granule. The exact determination of the secondary mineral serpentine, however, is uncertain due to the obliterating effects of a redehematite stain. The order of crystallization of these minerals as indicated by petrographic evidence, appears to be, ilmenite, magnetite, olivine (?), labradorite, pyroxene. Some biotite may have been formed from the resi- dual liquid phases of the magma or as a reaction mineral during the final stages of solidification. The results of this petrocraphic study provided no conclusive evi- dence that might be used in age differentiation of these intrusives. The -33.. apparent lack of olivine in the sections studied suggest dikes of Huronian age, since Keweenawan intrusives and extrusives (if the dikes are related to the flows) are generally reported to be quite rich in olivine. A Huronian age for these dikes is also supported by structural evidence. Yet, the abundance of calci-rich plagioclase and monoclinic pyroxene and the lack of amphibole suggests dikes of Keweenawan age. Biotite offers no clue to the solution of this problem, as it is con- spicuously absent in one section and particularly abundant in another. No flow structures were noted in any of the sections that might postu- late intrusives of Keweenawan age, such as might be expected had the dikes furnished the pipes through which the Keweenawan lavas poured out. This suggests that age determinations based on petrographic evi- dence alone is inconclusive and hence a more fundamental approach in- volving research on the physical chemistry of magmatic differentiation and crystallization in these rocks is necessary in the final analysis. -3h- QUALITATIVE AND QUANTITATIVE SPECTROSCOPY The basic theory of spectroscopy is outlined briefly in the fol- lowing paragraph, succeeded by a discussion on qualitative and quanti- tative methods of analysis. For a more detailed discussion on the theory of Spectroscopy, the author makes reference to the works of Ahrens (1950), and Willard, L. L. Merritt, J. A. Dean (19h8). The quantum theory introduced by Planck to explain the emission of radiant energy from the excitation of an atom or ion, predicts that each atom or ion has definite energy states in which the various elec- trons can exist. In the normal or ground state, the electrons are in the lowest energy state, but upon addition of sufficient energy by thermal, electrical, or other means, one or more electrons may be re- moved to a higher energy state farther from the nucleus. This condi- tion is an unstable one and the electrons upon returning to their ori- ginal orbital position emit radiant energy of Specific wave lengths which can be recorded as Spectral lines on a photographic plate. Since radiant energy when diSpersed produces a Spectrum unique for each particular atom or ion, identification of a mixture of unknown elements is made possible by interpretation of their characteristic spectra emitted during excitation. Final identification is achieved by visual comparison with known standard charts and tables. Quantitative analysis of these elements is achieved through the measure of the intensity or Optical density of their characteristic spectral lines which is reflected on the photographic plate as varying -35- degrees of blackening of a particular line. This immediately follows since the radiation emitted by any element is directly proportional to its concentration in the sample and the intensity or blackening of a Spectral line on the plate is in turn proportional to the radiation emitted. Various techniques utilizing this concept have been employed in the semi-quantitative determination of trace elements from Spectro— chemical analysis. Harvey (19h?) describes a method by which weighed amounts of samples are completely volatilized and selected line to background ratios are determined on a microphotometer. The background continuum is used in principle as an internal standard and the conditions of arcing must be closely controlled. This method is applicable to the analysis of a wide number of elements. Van Tongeren (1938) has made use of a method involving cathode layer excitation on carefully weighed samples diluted with N2CO3 in which successive exposures were made on each of the samples at varying amperages and exposure times. Semi-quantitative determinations were made visually by comparing line responses of the unknown with those of standards which had been prepared in a base of quartz and NaECOB. Another general method which employs cathode layer excitation is that described by Mitchell (19h8). This method which has been widely applied to the analysis of soils, minerals and rocks, requires that a small accurately weighed quantity of specimen be loaded into a cath- ode cavity and arced at a predetermined amperage. Three successive -36- exposures of about one minute are made. To facilitate visual compari- son of the line response, a step sector is employed and the plates are examined in a comparator. The semi-quantitative methods just described are designed to handle a relatively large number of elements and are applicable to a wide variety of materials. The value of these methods is that of en- abling the analyst to obtain much information about many_elements in a short time. The method employed by the author is similar to that as described by Mitchell with some modifications, and is described Later on in this work. Strictly quantitative methods utilizing internal standards and possessing a greater degree of accuracy have been described by Ahrens (1950). In these methods, intensity ratios are measured, that is, the ratio of the analysis line to that of a line of an element, the "in- ternal standard". The advantage of using an internal standard is that it affords compensation for various factors affecting the intensity of line emission such as are wandering, change in length of arc column, failure to time the exposure exactly, and lack of uniformity in the technique of photographic development. The main disadvantage of inter— nal standardization is that each element or a limited number of elements IIByrequire a separate internal standard, and a complete and accurate analysis becomes very tedious. Since this study was designed to cover a wide range of elements, it did not seem feasible to employ internal standardization, but rather -37.. utilize semi—Quantitative tectrii”es with close crntrol on ttcse factors \.. inflrencin" the intensity o” line emjscior. EBEPARATION OF SPECIMEN PRIOR TO ARCING The laboratory techniques and procedures utilized in this study followed a preliminary investigation of crushing and grinding methods as described by Ahrens (1950). Because it was necessary to obtain a small and accurately repre- sentative powder sample from a rock Specimen whose physical-chemical properties made it quite resistant to ordinary crushing and grinding techniques, certain modifications in these techniques had to be employed in the procedure followed by the author. First, the original rock specimen was cracked into two smaller pieces approximately 1” x l" in size with the aid of a porcelain pestle and mortar. One of these pieces was then selected for further prepara- tion of the powder sample, and washed with dilute HNO3 in order to re- move the oxide minerals (particularly red hematite dust which had col- lected on dike exposures from mine operations) as soluable nitrates. The specimen was then rinsed in de-ionized water, placed in an electric furnace and heated to 6000(L for 25 minutes. The specimen was removed, chilled in de—ionized water and placed in an electric oven for drying (30 minutes). This procedure not only facilitated secondary crushing and grinding but also served to drive off Volatile components (primarily H20) which are often the cause of the erratic behavior (spattering effect) of sili- cate minerals during initial arcing. After being removed from the dryer, the specimen was crushed and -39.. ground in a mullite mortar. Final grinding in an agate mortar was continued until the bulk of the sample was homogenous and reduced to about - 150 mesh. A one-half gram weight of this sample was accurate- ly weighed out, mixed with an equal portion by weight of purified powdered carbon and stored in a clean glass vial. All samples were prepared in this manner with particular care being taken to cleanse all laboratory equipment used in their preparation between samples. Such procedures limited possible contamination. Control samples were then run in an effort to determine moisture loss during preliminary heating in the electric furnace. The results disclosed an average moisture loss of 3.5% by weight. The addition of powdered carbon to the samples in a 1:1 ratio was made to facilitate arcing. Ahrens (1950) states that "The presence of powdered carbon speeds up volatilization rates considerably, without affecting total line intensity, provided the specimen is arced to com- pletion and the change in composition does not alter the arc temperature significantly". Arcing to completion, is referred to as the total energy method since it allows the total intensity of emission to be recorded. For this reason, the total energy method can be effectively employed in semi-quantitative analysis and was utilized by the author. The main requirements imposed upon this method are accurately weighed samples, a carefully controlled excitation source, a smooth burning arc, and arcing to completion. sgpcmnoanprc mapprcqu The fpectrographic techniques employed for the respective sample analyses were designed to maintain uniform rate of volatilization and a constant temperature range which insured equal levels of excitation and hence the consistency of treatment of each particular sample. Prelim- inary samples were first run to determine the optimum Operating condi- tions of the spectrograph relative to the unknowns to be analyzed. The instrument used for these analyses was a Ranch and Lomb large Littrow Spectrograph possessing a ranfie of 2100 to 80002 and capable of recording this entire range in three photographs on 25.h cm. photo- graphic plates. The dispersing media is a quartz Littrow prism, with o o o the diSpersion being about o.u mm/A at 2500A and 0.10 mm/A at hOOOA. An interchangeable glass prism can be inserted for increased dispersion in the visible region. For the purposes of this study the quartz Littrow prism was used exclusively and the computed dispersion for the particular spectral range used was 0.7 mm/AO . A direct current arc energized at 70 volts and 3 amperes was used as the excitation source with an average exposure time of two minutes. The reason for D.C. excitation is as stated by Willard, Merritt, Jr., Dean (l9h8), "The D.C. arc is a very sensitive source and is used for the determination and identification of substances present in very small con- centrations. A comparatively large amount of the substance being analyzed passes through the arc, and on this account, an average or more representa- -h1_ tive value of the concentration is shown provided the complete sample is burned”. The recording device used were Kodax Spectrum Analysis Number One plates of size 10.2 by 25.h cm, which were developed by use of Kodak Developer D-IQ for h minutes at 700F, followed by a 30 second stop bath, a 15-20 minute acid fix, a 30 minute washing period and a final rinsing with distilled water. -hg- l3? WEI’IQPE LEE :3.“ WI? The procedure involved in the actual burning of the sample is outlined as follows: Eight milligrams of a sample were accurately weighed out and placed into the drilled out recess of the lower electrode. Transferal of the sample to the recess was done by means of a beveled carbon rod of elec- trode material, thus eliminating the introduction of any foreign im- purities. The upper electrode was beveled to a point to prevent arc wandering and maintained at a distance of l/h inch from the lower elec- trode during all arcinqs, a condition which is described by many spectro- analysts as optimum for a smooth, continuous burning arc. A power supply unit furnished the amperage and induced voltage, con— trolled by a rheostat to the particular reouirements deemed necessary for optimum results in these analyses; arcing was induced by means of a striker electrode. The spark image was then centered on a plastic screen and the shutter opened with exposure time necessary for complete consump- tion of the sample. The shutter was closed and the plate racked up 3 millimeters, allowing an interval spacing of 1 millimeter between suc- cessive spectra. A slit width of 2 millimeters was used in all exposures. The orifiinal procedure was then repeated until the spectra of a sufficient number of unknowns were recorded on one plate. A copper spectrum and FeCl spectrum were then recorded immediately above and below the unknowns to act as a guide and external standard for qualitative purposes. Final identification of the recorded spectrums was achieved by -h3- comparative methods with known standard charts and tables. The definite presence of any particular element was confirmed throuch the identification of at least two of its most persistent lines. A semi-quantitative evaluation of these elements was achieved by setting up an intensity scale (l-lO) designed to measure the relative concentration of an element in a particular sample. A persistent line of an element was chosen from the spectrum in which it appeared to be most intense and given an intensity ratinn of ten. The same line was then examined in the other spectra ard an intensity ratins from l-lO assigned to it accordingly. In this manner, the relative concentra- tion of each element, as it appeared in a particular sample, was de- termined. The emulsion factor between a number of spectrum analysis plates has been known to vary, causing an apparent deviation in the concen- trations of elements recorded on these plates. Since the author used two plates in recording the total samples analyzed, and since the in- tensity of the recorded iron Spectra between the two plates varied, the relative concentrations of the elements in one plate had to be multi— plied by a variance factor (l.h), computed as the proportional differ- ence in the intensities of a persistent iron line as it appeared in the iron spectra of the two plates. The trace elements listed in the following table along with their identifying spectral lines were studied semi-quantitatively in this work. Photographs of the spectra used in this analysis are depicted in Firure 6. -hh- SPECTRAL LTWEQ USWD TN ETFM?VT IDHVTTFTCATTON Magnesium (ms) Aluminum (Al) Vanadium (V) Silicon (Si) Cobalt (Co) Chromium (Cr) Nickel (Ni) Copper (Cu) Titanium (Ti) Potassium (K) Sodium (Na) Calcium (Ca) Lead (Pb) Platinum (Pt) Beryllium (Be) Antimony (Sb) Cadmium (Cd) , . . o fipectral Lines - Wavelepcthnip A 3090-90, 3096.02, 3082.16, 2801.66, 3703-57 2881.59, 3158.76, 3578.69, 32u3.06, 3610.89 32h7.ss. 3088.03, 36u2.68, 3217.1 3302.3h, 3158.87, 3730-95 3628.11, 3130-79, 3232-52, 3261.05, 3100.31, 3336-89; 3002.72, 20c8.81, 3905.52 325u.20, 3503.u8, 3380.56, 3273.06, 3199.02, 3685.19, 3302.0h 3179-33, 3672.00 3131.06, 3267.u7 3h66.20, -h5- —.-*- 3100.67 3832.31 3suu.03, 3061.5u 2023.63, 3118.38, 3ld3.09, 3377.06, 3385.23, 3L53.61 h25h.3u, u289.73 3h33.57, 3uu6.26, 3h61.66, 3307.96, u062.76 3222.8h, 33h1.87, 3372.80, 3986.3u, 3058.21 36hu.39, h302.53 3321.08, 3321.35 3h67.66 3233633 Molybdenum (Mo) Strontium (Sr) Barium (8a) Scandium (Sc) 7irconium (7r) Lithium (Li) Spectral_§ines : wavelength in A? 3112.12, 3132.60, 3327.31 h077.3h h30h.00 3589.66, 3900.82, 36L2.81 3106.57, 3165.97, 3h38.23, 3L06.09 3232.67 Py relative concentrations, silicon, .cnesium, aluminum, calcium and iron are by far the most abundant of the elements studied. Some potassium and possibly some barium micht be attributed to the presence of potash feldspar and barium feldspar which are known to occur within \ the limits of the intermediate placioclase feldspar series. Some manqanese may be present as a specific mineral or on rare occasion re- placing the ferrous iron in the alteration product of chlorite. Petro- Vraphic study also indicates that a greater pronortion of titanium is Up result of the presence of ilmenite in these rocks. The remaining elements do not form common rock making minerals and their presence is hence concluded to be due to substitution in the silicates, oxides and possibly sulfides (pyrite) of the rocks. The occurrence and behavior of some of these elements in inpeons rocks follows. -ug- I I III I I I I IIIIII IIIIIIIIIIIIIIIIIIIII IIIIII II I IIIII IIIIIIII II IIIIIIIIIIIIII III ‘ ’ HIEI 'l|||l I Geocmmnrcnt unaware or WW swarms ,J LJ Because of the important role assigned to trace elements in this work, a discussion on their occurrence, abundance and physio—chemical relations with rocks of manmatic orisin is deemed necessary before presenting the results and their interpretation as derived from this spectrochemical study. The investisations of Clark and Washington (lQQQ, lth) based on 5,159 analyses of igneous rocks from all parts of the world, confirmed that the eight elements constitutinq the bulk of the upper lithosphere were also the primary constituents of all irneous rocks. These main elements, oxygen, silicon, aluminum, iron, calcium, sodium, potassium and macnesium comprise 98.28% hv volume of the elemental composition of all isneous rocks; all other elements as a srorp, accessory or minor or trace elements forminr only l.79% of the total mass of igneous rocks. Of these latter elements, titanium, phosphorous, hydrogen and manganese comprise l.l3% and hence only 0.50% is left for by far the greatest part of the elements. Rankama and Sahama (l9hQ) define trace elements as those elements not common in the upper lithosphere, that is, all elements other than those which make up the bulk of the upper lithosphere according to the findings of Clark and‘Washinston, and proposes a division of the trace elements into two groups according to their manner of occ“rrence in idneous rocks. The first of these arovps is comprised of the trace elements which ordinarily form independent accessory constituents in -ts_ igneous rocks: and the second group heinfi composed of those elements which rarely, if ever, form independent minerals, their presence in igneous rocks heind the result of incorporation in solution in other minerals. The author is primarily concerned with the second group of trace elements as it was mostly elements comprising this group that were used in the final nuantitative analyses Regarding the incorporation of trace elements in mineral struc- tures, Go‘dschmidt (lghh) states that, "Three structurally different types of diadochy resulate the dispersed manner of occurrence of the trace elements". These are further described as: (l) camouflace, which occurs when a trace element replaces a common element of similar valence (2) cap- turing, which is the replacement of a common element by a trace ele— ment of higher valence, and finally (3) admission, which occurs when a trace element replaces a common element with a valence hisher than itself. 73 .. - =actors (ovarri ..... erals are: (l) the effective sine of tie atom or ion in the structure, which depends upon the bond stren7th to neisthcrinr atoms or ions, on the electronic ccnfinnraticn of the atom or ion ard on its co—ordira- tion (2) the affinity of an element for a particular phase, t'nt is: e <“Vlicate, ' ride rn"r‘lnli8«2~uv*sn and *flwe ahnndnrwwa of a“f'<39 t“03€! phases in the particular melt (3) the relative ah ndance of n parti- crlnr trace element in the oritinal melt from whence the miter“? _li(1_ crystallised. (Goldschmidt, loht) The sires of the ions or atoms involved in rpnlncnment, m'st he such as not to disrubt the risidifi? Of tLe orininal crystal strnnt"rc. the tolerance in sire difference hetween any two atoms or ions sut- stitntina for each other heins shout 134 of the Jarrer of the two, as measnrcd bf their atomic or ionic radius in tnrns of envstroms as (fnrkama and Caters, lflhfll. :chntinns to the unit of measurement. this are protnhlv due to tte infl‘ence of temperature and pressure at lmqgl. the time of crystallisation (Ttonehrnoc, he decidinf factor coverrink initial s“hstitution is t’e relative bond strenoth of two atoms or ions in identical position in a strrc- tune, t“e atom or ion providing hreater bond strenoth being substitrted in preference to those contrihrtinf washer hoods. Tn Teneral, for those atoms or ions zith similar valency cta Te, the lar7er tends to the one with a _..l hood, while for atoms or ions of similar s ., a hifher volenn" ohnrne will tend to strenrthen the hord (moldschmidt, 1037). CafitHre of trace elements occurs d“rip" the initial stase of CT?- stalli7ntion of a manna and hence tlese are fovnd in the early crystallized minerals, whereas admitted trace elements are incorpor- ated into the structure of later crystallizins miner"ls. An illustrative example of this phenomena is that of trivalent O . 7pxlry) an]? 9 . scandium (Tc +0.93RK) which is captnrcd by m?”nositm (Us' \ is concentrated in hasic roots, whereas (Ti 1 0.7an) is admitted h" mannncium and concentrated in aCid.rOCkS° -50- 7‘ 7'3 "'1 r4- Ge new . b6 0 Wrmmni tn. WHfiite 0 9.? Granite 7.? 17“.h (Ueferennes - ”tene*nnre. 19R?; CJTd‘etmiflt, 7037: i7rcnk, 70?‘\ E ‘ 3m ~+ m +mom «+mom fiIItIIslllltlualimlx4|I\\1LlelllaWMI ‘ .mB 9..on +m:H .+mwo K$5. ixIIIitIlIlltlilttIIIIIHMIIIH1II| M + om 3% ”$5 3&3 4% ml a a .HN ta £8 {mg +% {4.3 E i I ‘ ~ a . mo 13 . m +9 is page £2 5% LE 4 i ~ ~ . +3 €me +mom Jew: +3 +mO +OE ‘ i ‘ a . +wz «+Ndo +mw A+mmo +mm< +m> i+mm ‘ q t . R w . Hm +Nwm «+3Hm +mS +x +:He +: a +3 .\ a a a ‘ x a a h a ~ $6 a 2 +mm +m2 1mm fism +x +m> +mmm +m o +maz +m +m am Duo Rwo +mpm ~+mpm .+mmm ~+x a a n , a omnudo mduwuouomn mwpdfio maamwmponH .nfiz uofisfipmpsm oEoodfln _ mwpdso downdsoao . £2qu 5 833 $386 03% no 33398 assoc $533 83d gm 23 g 533 033 33348 was -52- Q 2 J ' ‘\ wanvni vi“* (0 78%y\ onfi Pote‘t n (m.wdkv) . . i I ; ——-———-——— Nickel Dnfl ”noo‘t ere oreseot in normal ivono e rooP" of t“o vein stnfe of mcfflntic crfotellirntion, tteir occrre“ce teinr due in tort f‘ v to tte nrerence of cm°ll qreotitins of common Rwlfifie mineraTo snob e, nentlendite, oyrrtotite nni pyrite and of mnflor inoorteoce being owe to ’T‘V- 1 '1 their incoro(retion into t“e FtrnetVres of silicate Fi“er~1q. letter cq"e iofiicctes ttpt nichel Poi cokflit ere ovrntile in toe rroer littonohere. The nickel ion, with tte some raoiwfi nofl orcrje ad w“"reoium. i1 ordinarilv cnmowfieccd in MONecsi"m minerci: nlttov*5 the ti"5 Vi23” retio ion 87?]? formed crfictolr and t}o rteofl" ceilirc in ioter formed rocks ard minoraTe iniinetes its cnotnrw in come maéneciwm minorais, evicent‘y the nictel ion witt a velooco of creeter than 4? is centurnd. (Vfison, 10:9). ..... ”Fe bivolert nobelt ion on tte other band is Dracticallf 9i7e we the ferrous ion (0.83“) and it thwo conerfiliy cemonflnceJ in ferrown comoonndo. Tendell and ”oldich (10L?\ fhwpfl that corth varies linearly with mayhenivm 0V8? a wioe renfe Of OO“CQWfi?QtioD1 and hence su""e°+ ttot this reletiooorip in more eorircot tren tte one betvoeo J; «I v cobalt and ferrous iron. not“ cobalt and nicPel are relotiveiv abundant in hnoic rocks and deficient in silicic rocks, the abnrdnoce o? cohdlt in tbe latter rocks increqeihj ht the exoonee of trn nickel. (Vankmmn anc Tetzme. IOLL) moo enrichment of nickel in tte earl? cryotallized me*reqiwm "nfl ferromacnesinm minerals is one to t“o onitcbility of it? incornorntion into toe cry tel strnctnre of olivine nnfi 1*7“enrichene. Vickcl is lots ebuniant in anjite, smokinole and biotite (chkama and Cannon, lnhl) The content of nickel and cobalt nrcocnt in plegiocleco, byners- theno, ancite and biotite from a fiiorite is indicsted in the following table after chkolds and Mitctell (iOhE). n W 113:: Placioclsee - 15 anersthcno 100 100 Annite 000 7o Biotite lGO AG The following tables after Water and Mitchell (10:1) show the relative contents of nickel and cohelt in pyroxene, ilmenite end mafnetite prom on olivine free snooro. Vi nrm Co rnm Plecioclsse - - Prroyere 30 6o Marnetite so 80 Ilmenite lOO lFO T“ese minerals are ell present in the dinnsoc dikes sampled. Roth of those elements also occur in hornblende. O . gengcneso Mn“+ (0.9lkx) Though manrnnosc forms rnocific minercls in igneous rocks, ttey are of minor imnortcnce in comnorison to the mnnrcneco occnrinc in silicates and oxides. Sulficos ere rc‘ctjvpiy free of manepngge (stonehonfie; 1/5?)° .1 G; + ‘ o , ‘ Divalent menfienese reelsces Ns' (0.76kx) T-‘e' (0.33%x) end Ce (l.Ookx). It is admitted in plece of moreesinm and ferrous iron and is captured in nlace of Celcinm. (Pinon, lQ‘Q) Nockolds 1nd Nitcncll (19h?) show a relative increase of Fn:Fc in later differentistes in- dicating that removal of mnngnncse from mosmss is due to a lerre ex- tent to their admittance to ferromesncsinm minerels. Only traces op msnfanese ere found in nlaoiocleses Moore tter replace calcium. This ’34. is attributed to the feet ttat tte fin“ ion is too small to give (Mason, lQSP) stability to toe feldsper structnre. Figures by Nockolds and Nitcnell (lOHR) end Wcser eni Mitchell (19:1) Stow trot pyroxene is a more innortent rest mineral for tte acceptance of manganese t“en is riotite. Chromium Crai (0.6hkx) Chromium occurs in igneous rocks 3 chromite or other corome sninels end as traces in the structnres of silicate and oxide minere_s. Its most imrortent occurrence is t‘at apneerins in siliccte minerals where it is cemouflesed Hy nlnminrm (Al' 0.57kx) and iron (Fe 0-67VX) r7 + A o‘*' and captured by mejnesinm (M“( 0.7okr) End iron (We‘ 0.9?kx). (”trnn- house, l95?) Chromium is hichlv concentrated in the early differentiate” sili- cates occnrrins in magnesium-rich olivine ard(flinonyroxcne tto'fr stov- in: a preference for the letter. Tt elso occurs in common encite end hornrienie. (3”ide“srdh, thG) chroniwm contents in the nrimnry mineral: of a aioritp 73 stovn -55.. by Nocko‘dszrvvil‘itc‘ell (lCuQ) ire lintrd Molten Cr rrm _,.a_,_ Plagioclnse - anerstTene POO Alflffite 1‘30”) Fiotite 900 "4‘ ‘ xtrn 9 Pt")? Peletive en tents of chromifim in en olivine Free ”flier enfl Viteeell (lOGll Fellovc: Cr “rm —.-. 4- ~..-u—u Pl34ioclere — Pvrcxen? - l'z‘. metite mm W . Tlmenitc 100 02‘- C' (A ("x1 Conner Cu \u.OjKX) The Hifh attinity of Conner for erl?“r i9 t”: nrimerf factor ”e- termining the marner of oeenrrenee o? Conner in ifT°“HS roeve. Tt id almoet entirely nreeert es ehelcfinvrite in uneltererq ieneous reeks, but small amounts .ev also ren!ace ferrous iron in minerel structured in the ebsenee of sufficient srlfur. Remflohr (lghfl) believes this reelieemert wenlfl te“e nleee in anfitee. Cerohti and Pieruceini (10h?) su¢~eet tnet conner can re- 9 + r + place both Fe and Ni , for exemnle in torrmeline. The content of Conner in the mirerele of e fiiorite (Noctolfie end Mitchell, iQuR) is listed relow: -56- Cu pnm Plagioclase — Hypersthene lSO Aunite loo Biotite 3o0 Its nresence in minerals of an olivine free cabbro (wager and Mitchell, 1951) follows: CB. mi Placioclase lS Pyroxene SO Mannetite 50 Iljnenite SO n+- lgtgniym Ti (0.6uxx) Titanium tends to become Strongly enriched in the early products of crystallization and hence occurs in igneous rocks mainly as oxides, the most important of which is ilmenite, although it can replace Al and appears, therefore, in pyroxene, hornblende and biotite in which case it is captured by such minerals due to its higher charge. In highly siliceous mammas it is removed as titanite, which is probably the reason for its lack of occurrence in muscovite (Mason, 1952). Though titanium is not generally considered as a trace element, its inclusion here is due to its important geochemical relationship with vanadium. . h+ Xanadium_v (0.6lkx) Vanadium generally does not form independent minerals in ieneous -57... rocks and hence is usually found in the structures of other minerals, essentially as the quadrivalent ion. It is similar to titanium in that it tends to become concentrated in basic rocks. The highest vanadium content in igneous rocks is found in those formed during the initial steps of the main stage of crystallization. (Rankama and Sahama, lth) Ramdohr (lQHO) believes the major content of vanadium found in igneous rocks is due to its presence in magnetite, whereas Lebedev and Levedev (193k) concluded that it was primarily concealed in ilmenite. Vanadium occurs in additional titanium minerals in igneous rocks such as sphene and rutile where the phenomena is evidently a replace- .u'i- ll+ ' ment of T1 by V . . it . . Quinquevalent vanadium replaces phosphorous (P‘ ) in apatite. (Rankama and Sahama, l9hl) FeldSpars are nearly entirely deficient in vanadium, whereas pyroxenes, amnhiboles, micas almost always carry some vanadium as h+ 5+ . . . 3+ 3* V and V ions which replace Fe' and Al . (Rankama and Bahama, lghl) Vanadium, hence occurs in nearly all of the primary constituents of the diabase dikes. The distribution of vanadium in a diorite and an olivine free gabbro is as follows: Vllilfl Ellfifll Plagioclase 2O Plaaioclase l0 Hypersthene 100 Pyroxene 100 -58- V ppm V-ppm Auqite 200 Machetite 800 Piotite hOO llmepite 300 (Nockolds and Mitchell, louS) (Wager and Mitchell, lgsl) §tr93tium Sr2' (1.27kx) and BariumP-ag+ (l.h3kx) Strontium and Farium ordinarily do not form independent minerals in igneous rocks, but are rather concealed in the rock making minerals. C‘trontium generally accompanies calcium (1.06kx) in these minerals, whereas barium is usually substituted for potassium (1.33kxl. The behavior of barium in isneous rocks is similar to that of strontium and the presence of one usually indicates that of the other, althouph, in general, the content of barium in calc-alkali rocks de- creases rapidly towards the granites, whereas strontium, though in low concentrations, is more evenly distributed in all ieneous rocks. The most important occurrence of strontium and barium in irneous rocks is in the feldspar structure. Here strontium is admitted by calcium or captured by potassium with an increase in the Sr:Ca ratio in the later differentiates. Barium being too larae to replace calcium or sodium is eenerally only found in potash feldspars replacinq potassium. It also appears in biotite and muscovite. Strontium also replaces calcium in apatite and the calcium bearinq pyroxenes and amphiholes. 2- aerzlllum Pe (0.3ukx) The studies of Goldschmidt and Peters (lQBE), Randell and Goldich -59.. (19h3) and Sahama (l9h5) show that beryllium tends to hecome enriched toward the late stares of mavmatic differentiation, and hence displays a maximum content in both cranites and nenheline syenites. In alkalis rocks low in silica, the degree of enrichment is considerably hieher than in silica-rich granites. Sahama and Vahatalo (1939) have also established the concentration of beryllium in the residual solutions of dishase macmas. The most important occurrence of beryllium in igneous rocks is that contained in other mineral structures since too little of this element is contained in the rocks to allow the formation of its inde— pendent minerals. 4 This occurrence is based on the suhstitution of Fe for Si (0.39kx) in the Sioh tetrahedra. The alkali feldsnars, micas, alkali amnhiholes and alkali nyroxenes provide suitahle structures for the renlacement of silicon by beryllium. Because of its small size and low charce which wives a weaker bond, the main content of beryllium becomes concentrated in the resi- dual solutions and hence enriched in peamatites. u .. 2.139933% Zr ‘ (0. 87m ) Thouqh Zirconium resembles titanium chemically in manv resnects, their manner of occurrence differs considerably. Zirconium, because of its hinh charce and relatively hich radius, does not enter into any of the common rock-forming minerals, hut rather appears exclusively as zircon. Unlike titanium, the content of zirconium is low in the early -60- ‘n crystalates but tends to become enriched in the last rocks to crystal- lize. The following table after Ranhama and Bahama (lQLl) indicates this tendency. UthO¥ Peridotites, eclocites, dunites 6o Gabhros lMO Diorites 280 Granites hflfl SCfinfiiUW Sc3*'(o.8?kx) Goldschmidt and Peters (l931) show that the bulk of scandium ore- sent in the upper lithosphere is that concealed in ferromacnesium mineralscf the early crystallates and of such nltrabasic and basic rocks as pyroxcnites and gabbros. T . . , 2% C, 2+-( 0 _ Fere scandium is captured by Mr (0.7»ax) and Fe 0.02kx) in ferromagnesium mineral structures. This indicates a concentration of scandium in the nyrorenes, amnhiboles and biotite in the basic fractions of a crystallizinc magma. The anparent lack of scandirm in the earlier formed olivines is due 3-+ 2+ to the fact that a renlacement of Sc for Mg here would not balance the excess positive charge thus introduced by other suitable replace- ments. (Mason, 195?) 3A+ Scardium.may also replace Al (0.5 kx) in aluminum minerals, but onlv if the Al3+ is in octahedral coordination and in general does not follow aluminum durinj differentiation. -61- Only traces of scandium were discernible in the diabases studied. Militia-“Vt?! V0 Molybdenum is nrefcrentially concentrated in the lest differen— tiates during macmatic concentrations and hence anneirs in notable smourts in sranites, and less so in hahhros and ncrites, the order of mannitude beinn indicated by the following table: 'j_ton Cnbbros and Norites (”evesy and Vohhie, lQ3ll 3 Subsilicic rocks (Randell and Coldich, lvh3) 2 ' Granite (Hevesy and Robbie, lfi33) l2 9!. L- The incorporaticn of No in mineral structures i In rare and it is usually found in the form of the sulfide molybdenite l ( 0 To 30. The hlgl rx ‘ 1 ’ e "-'L o n o ~ f 1‘ 1‘ s ( P :Ir. 1" .‘f‘ .C‘lj.-‘ (‘uw 1A \ ~+M“.(1t“l ill ". pr‘3'Y'61r {31(371 (VP fj‘“ ‘f‘fi :6 ‘1 ( f. l . . n... < _., a 4 - . . . _. . o , , l .V.-. .L _ J " ‘ \ v -. 1 ‘ ‘1 4 q. l — ,, ‘ crqur mdlfijfll"39¢8 einiflén e «leur is Iu"" t. . ) 0!- Cadmium cd‘ (1.03le “admium closely follows zinc in its manner of occrrrence in isneous rocks, althouch it is much less abundant than the letter. Tt remains lnrécly in the residual melts and solutions throu~b- out the main sto~e of crystallizatirr. flccordinfi to Candell and Coldich (l0h3) cadmium seems to be con— centreted in ferromagnesian minerals of acidic i~neour rocks ferti- culerly in biotite althOV"h traces 2 OF P”3mium I‘eve al"o been Tfiflhrfnd in nnatite. Coldschmidt rWV” Irormann (1027) found notnble concnntrstinnq of cadmiwm in certain plagioclase rocks, but have been unable to deter- mine the manner of occurrence of this element in these rocks, and report that actually cadmium rarely becomes enriched in igneous rocks. 2+ .2333. Zn (0.83m) Zinc is concentrated in a greater proportion than cadmium in the silicate rocks formed during the main stage of differentiation. The manner of occurrence of zinc is determined by its property of dia- dochically replacing ferrous iron (0.83kx) and magnesium (0.78kx) in mineral structures. (Rankama and Bahama, lyhl) This similarity of size between zinc and ferrous iron causes the presence of zinc in magnetite and ilmenite. Amphiboles, pyroxenes and some plagioclases are also carriers of zinc. Rankama and Bahama (lth) state that the main carrier of zinc in normal igneous rocks is biotite. Only traces of zinc were found in the diabases studied. 4. Antimony_s The content of antimony in ieneous rocks is generally low. Pnti- mony was absent in a composite sample of gibbros analyzed by Preuss (19hl) and in a composite analysis of granites only 0.3 ppm was found. Silicate minerals of antimony are very rare and it is generally found in the oxides or sulfide phase. It forms antimcnides with various heavy metals, preferably with copper, iron, nickel and cobalt. The strong affinity of antimony for sulfur srnnests that it is predominantly present in this phase in igneous rocks. Elatinutht Platinum is generally found in nature in the native state due to -63- its reluctance to combine with other elements. It does, however, form such independent minerals as cooperite (Ptfi), Eraenite (Pt, T3d, Ni 8) and Sperrylite (Pt A32). Due to its reluctance to combine with other elements and its hioh melting point, platinum is preferentially enriched in the early-sepa— rated fractions durirq maematic crystallization and hence is often found in considerable quantities in dunites, pyroxenites and serpentinites. In the ultrabasics, platinum may freouently be concentrated in chromite. (Rankama and Bahama, lghl) P3339291“. Li+ (0.78.4) Since the lithium ion is smaller than any of the other alkali ions (Nal', 0.98A, K", 1.33A) it does not follow these elements during magmatic crystallization, bvt rather follows magnesium since their ionic sizes are identical. Because of its lower charge, lithium is admitted into magnesium minerals with an increasing Li/Mg ratio in later formed rocks and minerals. Lithium has been found in pyroxenes, amphiboles and particularly in the micas. SM'WITAHYE- Mégfls. (Davis, Ironton, Geneva Dikes) Elements. ,. A l -__ T _ “___ Emlsflgi ELkél—l-lflill_l§ii_mi§_mfle_gs_§a_QT._lfi._lli___B_e_Qil_~EQ_le Zr-M_FNJN 17" n .32 ,0 l C. L CCATION S AMPLF. TRATIONS s 14 IS 5 g: v E 5 g 3 4 H /4 15 16 17 l8 )9 £0 21 22 23 24 as 26 27 28 as 30 3! 32 SAMPLE LOCATIONS -87- Plate 2 .Mn .En Q: A w , ___-...___._________ 0 9 8 7 Cu 5 4 3 9— wzo-h‘KFZM—ozoo w>-LI(JUN- H 32 3O 23 28 Z 26 25 24 ’IPLE 8AA Plate 3 -88- lllllillll 1. N. .I Iv 5 4n 3 2 mzo-x—I‘mhuzmnvzoo N>—LI(J.WK 17 :6 12 ES 2 3 20 / 15 JG I7 I4 7. I. 4 5 6 .I O I I w W Z ZO-F<>NJN I7 ‘8 2.9 26 E1 23 I7 LCCATILNS S AMPLE Plate 1+ -89— pun-.n.‘ . NuLnllv: a; ~‘\\ —/' <. flO—T_. _z_; \-\./ ~ ~ 5‘ / / / / / / / / / / / / / / / / / / S _________ h. Sc 0- ——————— —o--——-——————————--—-—-—————— ———————————— — ['4 1 1'2 I7 ,9 H 1.2 :3 14 m 16 s7 )8 19 20 21 aa 23 a4 25 as 27 28 29 I ,4 “.2 r . [.7 /' . . . [,8 ll be )3 14 15 16 17 I8 IS .20 a! 22 23 24 25 25 a7 28 2.9 SAMPLE LOCATIONS Plate 5 CONCENTRATIONS 0- -$~ v- m \- RELATIVE ru I4 12 I7 I 18 H 12 13 I4 IS ‘6 I7 18 \8 20 Z! 22 ‘ £3 24 25 26 27 28 23 SAMPLE LCCANCNS -91- Plate 5 Oat E / I ISOPACH MAP OF E; ,/ a- Dyer quwfi mv l. strens. l. V. 269nm (1050) fncctrocbemical analysis. Addison-Wesly Press, 2. Pucher, Walter V., (19th), ”The fitereorraphic ?roiectiop, fl “and? TOO-l- for (JV? Prictieal pe(j]mfiig;t"’ UT, rm. 3, mhe Iovrnel of Geolo"", Vol. 7‘“ 101 " P1,). 3. Carobbi, Cnide, and Pieruccini, Dense, (lOEYF, "Pnectrocrarhic iralysis of Toprmalincs from Island of leq with Forreletion of Color and Fomnnsitjon" Am. Mineral 3?, p. lPl. h. El.an, Stanley Vireld, (1059) "A Cnectrochemical Aralysis of the Insoluble P9313“9“ Of the Dundee Limestone of Presoue Isle OOH-“33’: NiCthn". N. " mites-1'8: x.) 0 Nichiran State University. 5. @Oldschmidt, V. N. (l99h). Geochemistry, Clarendon Press, 73flnn. 6. Goldschmidt, V. M. and Peters, C., (1039a), "” ee Inr ”euphemie dos Feryllivms", Nath.— Phrsik. Vlas , Ill. 93 IV. ?5, p. BHO. 7. Crvner, John W., (lOQQ), "The Origin of Sedimenfiry Tron Formations: The Biwabik Formation of the hessbi Paste”, Economic Ceoloey 17, #07 pp. 8. Fevesy, G. and Hobbie, R., (l933) "Die Ermittlnn~ des Molybdsn and Wolframrehalteo Von Gesteinen. 7.", Anorc. Allrem. Chem. 2lQ, p. 13h. 9. Hoffman, Ceorse flbert,(lQSO),"The Correlation bv PetrOjraphic Nethois of Dikes designated Five and Ci? in the Vewpcrt Vine, Ironword, Vichiéan”. ”. ?. mhesis, Vichifin ¢tatc University. 1“. Vctchki as, Y. 0., (lClQ), "Geology of the roebic Panda ard its Relationship to Recent Vining Develonments", Tusineerinf and P’fini T1 ,7 JOUI'D'J. 10 ll. Fotchkiss, ’. 0., (l03?), "Take Superior Region", XVI International Geolofiical Confress, Guidebook 27, pp #9 - 65. 1?. Irving and Van Hise, (TRQR), "Penokee Iron Pearirc Ceries of ”ichi— fan and Wisconsin”, ”. 9. moolocieal Srrvey, MonoTraph XIV, J. V. Powell, Director, pp 1 — th. l3. Lebedev, P. and Lebedev, A., (193M), ”bu the Geochemistry of Tita- nium and Vanadium in Western Siberia", Compt. Rend..Acad. Sci. U.S.S.R. 3, p. 298. 14. Lundegardh, P. H., (l9h6), 'Rock Composition and Development in Central Roslagen, Swedan. Arkiv Kemi. Mineral Geol. 23 A, No. 9. 15. Mason, Brian, (1952), Principals of Geochemistry, John Wiley and Sons Inc., New York, pp. 1 - 25M. 16. Nockolds, S. R. and Mitchell, R. L., (19h8), 'The Geochemistry of Some Caledonian Plutonic Rocks: A Study of the Relationship Between the Major and Trace Elements of Igneous Rocks and Their ,Minerals'g Trans. Roy. Soc. Edinburgh 61, pp. 533 - 575. 17. Ramdohr, Paul, (19h0), 'Die Erzmineralien in Gewohnlichen Magma- tischen Gesteinen'g Abhandl. Preuss. Akad. Wiss., Mhth.-Naturw. Klasse, No. 2. 18. Rankama, Kalavero and Bahama, Th. G., (l9h9), Geochemistry, Uni- versity of Chicago Press, 912 pp. 19. Bahama, Th. G., (19h5b), 'Epurenelemente der Gesteine im Sulichen Finnisch — Lapplandffl Bull. Comm. Geol. Finlande 135. 20. Sahama, Th. G. and vahatalo, veikko, (1939a), 'The Rare Earth Content of Wiikite", Compt. Rend. Soc. Geol. Finlande 13; Bull. com. Geo—lo FMde 125, p. 970 21. Sandell, Ernest B., and Goldich, Samuel 8., (19h3), "The Rare Metallic Constituents of Some American Igneous Rocks", I.J. Geol. 51, p. 99; II. Ibid., p. 167. 22. Stonehouse, H. 3., (1952) “Trace Elements and their Association with Mineralization ". Ph. D. Thesis, University of Toronto. 23. Streak, Lester w., (1936), "Eur Geochemie des Lithiums", Nachr. Ges. Wiss. Gottingen, IV, N. F., 1, No. 15, p. 171. 24. Van Hise, (1911), 'The Geology of the Lake Superior Region", U. 8. Geological Survey, Monograph 52, pp. 225 - 250. 25. Van Tongeren, W. B. C. (1938), 'Un.the Occurrence of Rarer Elements in the Netherlands East Indies'g D. B. Centens, Amsterdam. 26. wager, L..R., and Mitchell, R. L., (1951), “The Distribution of Trace Elements During Strong Fractionation of Basic Magma: A .Further'Study of the Skaergaard Intrusion, Bast Greenland", Geochim. Cosmochim..Acta 1, pp. 129 - 208. -9h- 27. Willard, L. L., Merritt, J} A., and Dean, (19h8), Instrumentation in Spectroscopy, John Wiley and Sons Inc., New York, pp. 18 - 183. -95.. .buhr..nr...r ..