SUPPLEMFMTARY MAT La .HL IN BACK OF BOOK ABSTRACT GEOCHEMISTRY AND PETROLOGY OF THE NlPISSlNG DIABASE IN ONTARIO by John A. Colwell The Nipissing diabase occurs in the area between the eastern end of Lake Superior and Lake Temiskaming as sill-like bodies (sheets) intrusive into the flat-lying or gently folded Huronian strata and as dikes in the steeply dipping pre-Huronian rocks and in the massive pre-Huronian granites. Sheets range in thickness from 250 feet to at least llOO feet. Moderate differentiation of the diahase, most noticeable in the sub-horizontal sheets, is shown by the range in rock types from olivine-hypersthene diabase to aplite. Detailed investigation of the differentiation of the diabase in the Cobalt area has been accomplished by petro- graphic and chemical study of samples from a vertical mine shaft and a diamond drill core. Analyses of both major and minor elements were made by the interrupted arc method of emission spectrography. This method has given analyses of sufficient precision for the study, and its speed and low cost, relative to chemical methods, has permitted a more complete investigation of the chemistry of the rocks than is usual in studies of this type. Field and petrographic data indicate that the John A. Colwell intrusions were emplaced essentially in their present form, and that the highly fluid magma had undergone little, if any, crystallization before intrusion. Samples of the chilled phase of the Nipissing in the Cobalt area indicate that the magma had a uniform compos- ition - differentiation took place after intrusion. The chill composition is similar to that of other tholeiitic intrusions except for the low titanium content, observed in only a few other cases. The Nipissing diabase in the Cobalt area is a partially-eroded undulating sheet forming a number of basins separated by arches. The trend of differentiation in a typical basin-arch section has been shown to be as follows: I. A chilled phase of quartz diabase was formed at the margins of the sheet. 2. Fractional crystallization and vertical grav- itational settling resulted in the accumulation of minor amounts of olivine in the lower parts of the sill which were not completely resorbed by the quartz diabase crystallizing around them. 3. Lateral, upward movement of acidic constituents toward the arches shifted the_magma composition in the lower part of the basins so that olivine and hypersthene became stable phases and together with the plagioclases became more basic for a time, toward the top of the sheet. Later crystallization followed the common trend of iron and sodium enrich- John A. Colwell ment. As a result the olivine hypersthene zone is succeeded by a quartz-hypersthene zone, a quartz diabase zone (hypersthene replaced by pigeonite), and, a varied texture zone, where irregular patches of coarse to pegmatitic diabase indicate enrichment of volatiles. Later phases, postulated to occur mainly in the now-missing upper parts of the arches are granophyric diabase and aplite. This is support- ed by the occurrence of these phases elsewhere. The behaviour of minor elements during different- iation shows definite trends in their incorporation into various phases of the diabase, similar to those found in other studies. Nickel and cobalt are enriched in the more basic phases, copper in the quartz diabase, vanadium titanium and manganese in the varied texture and granophyric diabase. Petrographic and spectrographic study of samples collected throughout the region of Nipissing diabase intrusions shows that intrusions of the area along the north shore of Lake Huron are similar to those in the Cobalt area. Intrusions in the Sudbury area are dike-like and show accumulation textures and more equilibrium conditions than in the other areas, but chemical similarity to the Nipissing as well as field evidence indicates that they too are part of the Nipissing. Regional variations in metalliferous deposits, while not studied in detail, are thought to be due to circumstances arising during differentiation, not to original differences in the composition of the magma. GEOCHEMISTRY AND PETROLOGY OF THE NlPISSlNG DIABASE IN ONTARIO by \ John A? Colwell A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geology l967 ACKNOWLEDGMENTS The writer acknowledges the assistance of several staff members of the Geology Department at Michigan State University in the accomplishment of this study. Special thanks are due to Dr. Harold B. Stonehouse, who directed the research. The late Dr. J. Zinn, and Drs. W. Hinze, S. Romberger, R. Sandefur and J. Trow were other members of the guidance committee, and gave freely of their assistance. Dr. R. Thomson of the Ontario Department of Mines kindly provided samples from a diamond drill core and loaned some thin sections. Acknowledgment is also made to the Geological Survey of Canada for the preparation of thin sections used in the study. TABLE OF CONTENTS ACKNOWLEGEMENTS . . . . . . LIST OF TABLES . . . . . . LIST OF FIGURES . . . . . . INTRODUCTION Chapter I. GEOLOGY OF THE NIPISSING DIABASE. . . Regional Setting. . Pre-Huronian . . . Huronian . . . . Nipissing diabase . Post-Nipissing . . Age determinations . II. METHODS OF INVESTIGATION Regional sampling . Detailed sampling . Petrographic methods Chemical analyses . Spectrographic analyses Sample preparation External standards Analytical procedure Precision . . . Ill. DIFFERENTIATION . . . Petrography . . . Phases of the Nipissing diabase. Distribution of phases. . . . Variation in mineral Spectrographic data. composition Composition of the chilled phase Analyses of the sections . . . Calculation of weighted moving averages in the sections. Mafic and felsic indices . . . Trends of elements Page m(D\J \4 -poa©\qOL> A Chapter Differentiation of the diabase. . Introduction. . . . . Composition and nature of the magma . . . . Trend of differentiation. . Composition of the "acid” section. Behaviour of minor elements during differentiation. . . . IV. REGIONAL VARIATION . . . . . . . Introduction. . . . . . . North Shore of Lake Huron . . . Sudbury area. . . . . . . . Cobalt area . . . . . . . . Comparison of areas . . . . Relation of metalliferous deposits to the Nipissing diabase . V. SUMMARY AND CONCLUSIONS . . . . . Summary . . . . . . . . Conclusions . . . Suggestions for further work . . LIST OF REFERENCES . . . . . . . . . . I03 I04 Table I3. Table of format Duplicate chemi LIST OF TABLES ions . . . . . . cal analysis of sample 2l2. . Analytical conditions . . . . . Major mineral constituents of the phases of the phases of the Nipissing diabase . . Analyses of the Nipissing Spectrographic analyses - Rix section chilled phase of the diabase . . . . . Spectrographic analyses - Nasco section . . Composition of intrusions Calculation of extended Nasco section chilled phase of some Composition of acid section - from other seCtionS . C I I I O I D O C . Analyses of granophyre and aplite . . . . Comparison of acid sections . . . . . . Unit cell size and melting point of some olivines . . . . . . . . . . . Spectrographic analyses - Nipissing diabase - north sh ore of Lake Huron. . Average analysis of north shore of Lake Huron and average analysis of quartz and varied tex Spectrographic - Sudbury Spectrographic - between Spectrographic - Cobalt a Average analyse ture diabase of Nasco analyses of Nipissing area D O O O C 0 analyses of Nipissing Cobalt and Sudbury . analyses of Nipissing rea . I C B I O s of various areas . V section. diabase diabase diabase Page 23 27 40 $2 45 48 58 64 65 70 7I 74 79 82 84‘ 87 93 Figure LIST OF FIGURES Geology of the Nipissing diabase. . Location of samples - Rix section . Location of samples - Nasco section. Form and composition of the Nipissing diabase at Cobalt . . . . . Chemical variation - Rix section. . Chemical variation - Nasco section . Page .(in pocket) Chemical variation between Rix, Nasco, and "acid” sections. . . . . . Hypothetical section showing distribution of phases in the Nipissing diabase at Cobalt. . . . . . . . vi l9 20 33 47 SO 66 68 I“ (n A... I ' ‘1‘ I INTRODUCTION Basic intrusives usually referred to as the Nipissing diabase, (locally as Sudbury gabbro) occur as sheets and dikes in the area between the eastern end of Lake Superior and Lake Temiskaming. They are generally closely assoc- iated with areas of Huronian metasediments and appear to be related to the period of gentle folding and faulting of these rocks shortly after their deposition. Radioactive age determinations have shown that the intrusions in question are contemporaneous throughout the area. Previous studies of the intrusives, particularly in the Cobalt area, have shown a variation in lithology, with rock types ranging from olivine—hypersthene diabase through quartz—hypersthene diabase, quartz diabase, to granophyre and aplite. Differentiation due to fractional crystallization has been generally used to explain the existence of these phases, but several authors have consid- ered assimilation of enclosing rocks to be responsible for at least some of the granophyric phase. There is a regional variation in mineral deposits associated with the Nipissing. The Cobalt area is famous for its siIver-cobalt-nickel-arsenic ores, which are consid- ered to be genetically related to the diabase. Minor copper-nickel sulphide deposits occur in or near the diabase in the Sudbury area, while copper deposits were mined for some time in the Bruce Mines area. This investigation was originally planned to study 2 the regional petrologic and chemical variation in the Nipissing diabase to determine if trends could be established that might explain the variation in mineral deposits. In order to provide a "standard" for comparison, it was decided to do more detailed work in the Cobalt area, where mine shafts and diamond drill cores provide an opportunity to examine the vertical and lateral variations through the intrusive from top to bottom. Studies, mainly petrographic, by Satterly (l933) and Hriskevich (l952), had shown a regular pattern in the distribution of the phases of the diabase which was ascribed to differentiation of the diabase. As the present study has progressed it became obvious that the additional data on major and trace element variations might generally support their conclusions and further elucidate the mechanism of differentiation and behaviour of elements during crystallization in the Nipissing in particular, and by comparison, in basic intrusives in general. As a result the emphasis of the investigation was shifted somewhat and a large portion of this paper is devoted to the detailed work in the Cobalt area. The study of differentiation is presented separately from the regional study, although the conclusions drawn from the former are considered as they apply to the latter. Lack of data on sulphur, arsenic and other probable key elements have prevented a proper evaluation of the regional variation in the diabase as it may be related 3 to the variation in ore deposits but some provisional conclusions are drawn from the present results. CHAPTER I GEOLOGY OF THE NIPISSING DIABASE Regional Setting The region of Nipissing diabase intrusions lies along the southern margin of the Superior Province of the Canadian Shield between the eastern end of Lake Superior and Lake Temiskaming. Except for a small area of Lower Paleozoic strata north of Cobalt, all of the consolidated rocks of the area are of Precambrian age. While the geology of parts of the region, particularly in the mining areas of Elliot Lake, Sudbury, Cobalt, and Gowganda, has been studied in considerable detail, much of the region has been mapped only on a reconnaissance basis. Publications of significance to this study may be fOund in the list of references. The discussion that follows is based largely on the work of Moore (I955), Robertson (l963), J. E. Thomson (l962a, l962b), and R. Thomson (I962). For the purposes of describing their relations to the Nipissing diabase the rocks of the region may be simply divided into three groups, corresponding to the stratigraphic divisions pre—Huronian, Huronian, and post-Nipissing. Figure I shows the extent of these groups as well as the distribution of Nipissing diabase intrusions. The ubiquitous olivine diabase dykes of post-Nipissing age are not shown. The stratigraphic columns in Table | show the :Fuezmmx :wumzmwx mewmemewe newcogzz-mgn camcog< A.s.n m.NV cc. A.s.a m.NV muwcmgm ceeomF< muwcmLm :mEom—< .cmgocwx vacmwcogzz cm_:oL:: cmwcogsz cmwnwza< A.s.. mao.mv A.s.n mm_.mv omenmev mc_mmwawz OLnnmm sznuzm mmmnmwu ucwmmwavz A.s.a me._v mercegm Lmeau Acawxocmav A.>.n mm._v cmvcomnzx owoNoquoga m>PpQ=LLw zgznusm A.s.a mNN.FV mmmamwu mce>w_o mmmnmwu wcw>wpo omenm_n wcw>wpo sewxv—m: cmwuw>ougo curbs—em uFoNomea o_o~ogmcm:a pFenou zxsnuzm aces: memo Co meazs ubzezmeomo <1“ zoo meowmeLom mo anme ._ mommxmflsm smpcav .paacoo pm amended mgemmflmez one mo goflpflmOQEOU com Show .: mmDlo pmmw Doom 0 q H d d I J .mpompgoo pm amended Nehmsw oopoflpgmsmmmficgs “macaw poz moflgmoao> cwmemmm \‘Ill'llllll \ “"-'-||-"l'|'|l"l"|i‘V \ o“ ‘ \ QSOHm pa no "l.‘ O O o \ O O 00 on” mhdpxoa cmwmo> \\ 1/ w \ lull“ " _ 34 known contacts) -- medium to dark gray. Intergranular to subophitic texture. The "chilled marginal layer of quartz diabase" of Hriskevich. Olivine-hypersthene diabase -- medium grained, dark gray. Generally subophitic. Characterized by presence of olivine. The lower part of the "hypersthene zone” of Hriskevich. Quartz-hypersthene diabase -- medium grained, dark gray. Generally subophitic. Large poikilitic plates of orthopyroxene; olivine absent; quartz usually present. The upper part of the "hypersthene zone" of Hriskevich. Quartz diabase -- medium grained, gray to greenish- gray, but with more obvious color difference between mafics and felsics than in the above types. Subophitic. Hypers- thene and olivine absent. Quartz as individual grains and in micropegmatitic intergrowths. May be part of Hriskevich’s "varied texture" phase. Varied texture diabase -- medium to coarse grained, mafics greenish-black, felsics white to pink (granophyric diabase). Hypidiomorphic granular. Minerals similar to quartz diabase. Opaques more abundant than in other phases - mainly magnetite-ilmenite intergrowths. While no aplites were analysed in the present study, a thin section was obtained from aplitic meterial north of Gowganda. The aplite bordered a carbonate vein containing cobalt mineralization and appeared to grade quickly into normal diabase. The description of aplite 35 that follows includes data from this sample as well as that of Hriskevich (I952). Aplite -- medium to fine grained. Pink to dull red. Hypidiomorphic to alIotriomorphic-granular. Mafics minor or absent. Mainly plagioclase, quartz and micro- pegmatite. A small amount of microcline is present. Carbonate present in some aplites, partly in miarolitic structures. Distribution of phases The distribution of phases according to the writer's classification is shown in Figure 8. This is an hypothetical cross section based on the petrographic and spectrographic data from the Rix and Nasco sections and projection of this data to a so called "acid" section, the derivation of which is discussed later. The following description of petrographic variation is based on the Colonial Mine section of Hriskevich and the Rix and Nasco sections of the present study. The chilled diabase at the base of the sheet is composed of plagioclase and pyroxene crystals arranged in sheaf-like or radiating form, along with up to ID per cent of phenocrysts of plagioclase, pyroxene and olivine. The plagioclase of the groundmass occurs in needle—like crystals about 0.l mm long while the altered lath-shaped phenocrysts are about 0.5 mm long. The plagioclase compos- ition, according to norms calculated by Hriskevich, is about A055. The pyroxene of the chilled phase appears to 36 be all augite - the grains in the groundmass are too small to be definitely identified. The presence of olivine is inferred from fine grained aggregates of antigorite and chlorite with anhedral to euhedral outline. Iron oxides occur as tiny rounded grains throughout the groundmass. Grain size coarsens to about I mm a few feet from the contact, and the texture becomes subophitic or diabasic as one passes from the chill phase into the fine grained quartz diabase, which in both the Colonial and Rix sections extends for about l00 feet from the lower contact. Plagioclase occurs as lath-shaped crystals about I mm long and generally appears to have crystallized earlier than the augite, although the reverse is true in some cases. Pyroxenes include augite, pigeonite, and pigeonite inverted to orthopyroxene. the latter being identified 'by the Iamellae of clinopyroxene exsolved along the DUI direction of the original pigeonite and the common severe gray alteration called "turbid alteration" by Hriskevich. Quartz and micropegmatite occur interstitially. Iron ore, in some cases showing magnetite - ilmenite exsolution occurs as small skeletal crystals. Clusters of rounded aggregates of antigorite and chlorite occur throughout this phase in the Rix section. They are usually rimmed by inverted pigeonite and are interpreted by the writer to be altered olivine. The olivine-hypersthene phase begins with the rather abrupt appearance of large untwinned plates of orthopyroxene 37 in sample 222 of the Rix section, and extends for over 600 feet. With the exception of these plates which average about 5 mm in length (ranging, according to Hriskevich, to l5 mm) the grain size is only a little coarser than in the fine grained diabase, plagioclase crystals having an average length of about l.5 mm. Pigeonite and inverted pigeonite disappear rapidly, their place taken by the hpyersthene. The hypersthene shows little or no pleochroism. The grain rims usually have a slightly higher birefringence than the core. The plates typically contain numerous well-formed plagioclase crystals and occasional grains of augite. Olivine occurs as clear anhedral to euhedral grains and clusters of grains. The lack of significant magnetite in altered olivines attests to its high magnesium content. Iron oxides occur in the lower part of this phase but are almost totally absent in the upper 450 feet. Olivine disappears at about 750 feet above the base of the Rix section and quartz reappears in increasing quantities approaching the quartz-hypersthene phase. Other minerals generally remain the same, alth0ugh their character is a little different in some cases. The ortho- pyroxene shows stronger zoning and exsolution of clino- pyroxene appears around the margins of the plates, eventually extending throughout some grains. Pigeonite and inverted pigeonite reappear in the upper part of this phase, as iron oxides. The amount of quartz and micropegmatite increases. The disappearance of hypersthene at about 38 950 feet above the base in the Rix section marks the beginning of the quartz diabase. Its mineralogy is quite similar to that of the fine-grained diabase above the lower chill zone except that olivine is absent. Quartz, micropegmatite and iron oxides are generally more abundant than in the fine-grained phase. In both the Colonial and Rix section patches of coarse-grained material appear almost immediately in this phase (varied texture diabase). In the Nasco section, on the other hand, it appears from the hand specimens and a few slides that about 290 feet of medium grained (about I.5 mm) quartz diabase occurs between the hypersthene zone and the varied texture phase. Grain size in the varied texture phase of the Rix section reaches a maximum of about 5 mm while Hriskevich describes pegmatitic patches with pyroxenes up to 5 cm long. Micropegmatite and iron oxides seem to be most abundant in the coarser phases. Variation in mineral composition Olivine: The range in olivine composition noted by Hriskevich in the Colonial Mine was from F072“5 to F085' but in contrast to other sills of similar composition the most magnesian olivine was not nearest the base. Instead the composition varied from F072“5 at l00 feet above the base to F085 at 400 feet and back to Fo72.5 at the point where olivine disappeared, 700 feet above the base. The maximum amount of olivine - 26.9 per cent - was noted at 600 feet. Rough determinations indicate the same trend 39 to be present in the Rix section. Plagioclase: A similar situation to that of olivine occurs in the case of plagioclase norms calculated by Hriskevich from his chemical analyses show An66 in the chill zone, An74.5 at l80 feet, An80.5 at 600 feet and An68 in varied texture diabase 920 feet above the base at the Colonial Mine. Determinations on minerals show this same trend. The An content drops sharply in the granophyric diabase and aplite - one aplite norm gives AnO. Plagioclase varies in amount in the sections from about 30 per cent to a little over 50 per cent, being least abundant in some olivine - hypersthene diabases, and most abundant in the quartz and varied texture phases. Pyroxenes: Hriskevich found remarkably little variation in the composition of hypersthene or augite in the Colonial Mine, although the former was a little more iron rich near the base and again toward the top (especially the rims of the grains) than in the middle of the hypersthene zone. Compositions given were - hypersthene about M973:Fe27, augite about C539M95IFel0' The data for the major constituents are summarized in Table 4. Minor constituents - primary minerals: Biotite occurs as small reddish-brown flakes in all phases of the diabase; amounts are rarely over I per cent. Apatite occurs in all phases except the olivine - hypersthene phases and is most abundant in the coarser phases of the 40 x x x x amas._a e.___;a Lagos x x x . x x mmmamwo umcwmcm me?“ x x x . x mmmnmwu mcmcum Icmaxciwcw>wpo x x x x mmmaawc mcmgemgmgxzi~ugaso x x x x _mmmnmwu Nucmso x x x x mmmnmwc mcszmp umwgm> x x x mmmnm_u a.__u;a .maas . omega mcs>rpo mcmcumcquz opwcommea mp_m:< wmmFUmeaFa Npcmzo mpacm:_z .mmmamwc mcwmmwawz one we mommca on» mo mbcmapwpmcoo pagmcwe game: .e momwro wmmnmwu.wcspxmw wwwga> .am upsmzo mpwpam,u:w szcnocagw coc< :wmmm a “‘5? - several 69 substantiated by the presence of micrographic qaartz-feldspar intergrowths throughout the quartz-diabase phase. With these poor guide lines, the writer has arbit— rarily assumed that the Nipissing contains abOut 5 per cent total aplite and granophyre; l per cent aplite and 4 per cent granophyre, distributed as shown in Figure 8. This means that the acid section would have about 30 per cent aplite and granophyre. As no complete analyses of either phase were made in the present study, analyses quoted by Hriskevich were used to determine the average composition of these phases, as shown in Table II. The remaining 70 per cent of the acid section was first assumed to have the same composition as the quartz diabaSe of the Nasco section. Combination of this with the aplite, and granophyre yields a slightly more basic rock than that previously calculated for the "acid” section. Since it is likely that the quartz diabase would be more acid near the upper part of the section another calculation was made assuming the 70 per cent to have the composition of the varied texture diabase of the Nasco section. It will be seen from Table I2 that the result is remarkably close to the "acid” section calculated from the chemical data alone, in other words, the latter corresponds to a possible combination of observed phases. In view of the lack of an actual " cid” section it is not possible to assess whether the assumed section corresponds with nature. Several suggestions however may be considered. 70 TABLE II. Analyses of granophyre and aplite GranophyreI Aplite2 MgO % 3.9 .9 Feo3 % 10.5 2.2 CaO % 3.7 .9 Na20 % 5.97 6.90 Ti ppm 5400 3000 (Hriskevic I952) I. Average of analyses 7 , I h, l (Hriskevich, I952) 2. Average of analyses 3. Total iron as FeO 7| TABLE l2. Comparison of acid sections Acid section Acid section calculated from 30% granophyre and aplite other sections 70% Nasco varied texture MgO % 6.I 6.4 FeO % 8.6 8.9 CaO % 7.9 7.9 NaZO % 2.64 3.29 Ti ppm 2900 4.00 72 l. The "acid" section before any differentiation would most likely have had the average magma comp- osition. 2. Some vertical differentiation seems to have occurred in the Nasco and Rix sections before the lateral differentiation became effective. These two points suggest that the acid section might contain some material more basic than the varied texture diabase of the Nasco section, and it would seem that is an error has been made in the nature of the phases in the ”acid” section it has been in assuming the lower 70 per cent to be more acid than It probably would be. This in turn indicates that the assumed amount of granophyre and aplite is probably not excessive. Behaviour of minor elements during differentiation If it is assumed that the trend of differentiation shown by the petrography and major elements has been correctly interpreted, some inferences may be drawn about the behaviour from their distribution of minor elements during different- iation. Nickel shows a greater range of concentration in the sections than any of the other elements studied. Its concentration in the original magma as judged from the chilled diabase, was 8| ppm. The trend of nickel is much like that of magnesium but more pronounced as is obvious from Figure 5, indicating that nickel is preferentially incorporated in the more basic rocks, particularly the 73 olivine-hypersthene diabase. Studies on individual minerals in other intrusions by Nockolds and Mitchell (I949), DeVore (I955), and Wager and Mitchell (l95l) indicate that most of the nickel occurs in the olivines, lesser amounts in the orthopyroxenes and clinopyroxenes and very little in the feldspars. Values given by DeVore suggest the relative distribution to be about olivine l00; orthopyroxene 30; clinopyroxene 20; with higher concentrations in all cases in the more magnesian members. The enrichment of nickel relative to magnesium in early olivines has been noted by Wager and Mitchell (l95l), and suggests that Ni-olivine is more stable than Mg-olivine. This is supported by data on the unit cell size of Ni-olivine given in Table I3, but both of these observations are surprising in view of the lower melting point of Ni23i04- l62000 compared to |89OOC for MgZSiO4. Ringwood (I956) has suggested that in natural olivines the presence of iron may lower the thermal stability enough that the nickel enters the structure in preference to magnesium and replaces some of the iron. Clearly, the problem is not concluSively resolved, but the observed pattern cannot be denied, for it has appeared in every study of basic intrusions. The entry of nickel into the basic rocks in excess of its concentration in the liquid results in its rapid decrease in later phases, as is clearly shown in the trend. Hriskevich (I952) found very little nickel in the aplites, but some is present in the carbonate veins, which he interpreted 74 TVHBLE l3. Unit cell size and melting point of some olivines . Cell parameters m.p. ax bR cg OC NizSiO4 4.7l I0.II 5.9I I620 lfig28i04 4.76 l0.20 5.98 I890 COZSiO4 I345 Fe28i04 4.82 l0.48 6.II I205 MnZSiO4 4.86 l0.60 6.22 Sources: Deer, Howie and Zussman (I963) Ringwood (I956) Handbood of Chemistry and Physics, 47th ed. (I986) 75 635 the final differentiate. It would appear from this that nickel may be excluded from late minerals, to the extent that its concentration in the liquid rised again in the final stage. Wager and Mitchell (l95l) noted this same late rise in the Skaergaard. Cobalt exhibits somewhat the same trend as nickel and magnesium but its enrichment in the basic rocks is much less than either of these. Studies by others on minerals show the same tendency. DeVore's data suggest that the relative distribution of Co in olivine, orthopyroxene and clinopyroxene is in the proportions l00:50:40. Wager and Mitchell found no cobalt in plagioclases while noting its presence in coexisting pyroxene, ilmenite, magnetite and olivine, the latter having the higher concentration, especially in the earlier phases. They noted a similarity in the trends of cobalt and iron during the course of differentiation. The present data suggest that Couolivines and pyroxenes should be a little more stable than the iron forms but less stable than those containing Mg. This is supported by melting point data on olivines given in Table I3. On the basis of the unit cell sizes one would expect manganese to enter olivine and pyroxenes less readily than iron. This is supported by the data of Wager and Mitchell - they note enrichment of manganese in the late olivines and pyroxenes, with both containing about the same amount. The presence of minor amounts in plagioclase and 76 apatite led them to the conclusion that manganese may substitute for calcium as well as for iron. The results of the present study confirm that manganese enters the early phases in lesser amounts than its concentration in the magma, with the result that it builds up in the magma and appears in greater concentrations in late phases. The trends of copper in the Rix and Nasco sections agree well with the findings of Wager and Mitchell (l95l) and show that copper in early minerals is less than the magma concentration but soon rises to a level about the same as the magma. The few values for acidic rocks indicate a sharp drop in copper in late phases. Wager and Mitchell concluded, from their mineral analyses, that copper enters pyroxene, plagioclase, olivine, magnetite, ilmenite and apatite, probably proxying for ferrous iron and sodium, as well as forming sulphides, in late stages of crystallization. While no chalcopyrite was noted in any of the rocks in the sections, it does occur in places in the Nipissing. The trends of V and Ti are remarkably similar - both seem to have been excluded from the early rocks and, especially in the case of Ti, enriched in the varied texture diabase. Wager and Mitchell found that in the Skaergaard some Ti entered early clinopyroxenes but the amount was less than that in the magma, with the result that Ti accumulated in the magma until ilmenite was precipitated. The same situation seems to apply here. The early rocks - olivine and hypersthene phases contain less Ti than the chilled 77 diabase. It may occur in the augites, but some probably ialso occurs in the rare iron oxides. The latter increase considerably in the varied texture diabase, where exsolution phenomena in the iron oxides indicate the presence of ilmenite. The chemical analyses of granophyric diabase, quoted by Hriskevich, indicate that the above average Ti continues into this stage, while the lower values in the aplite suggest that, as Wager and Mitchell indicate, the magma became depleted in Ti in its final stages of crystallization. The same explanation probably applies to vanadium as this is indicated by Wager and Mitchell’s data. They found, however, that most of the V in the oxides occurred in the magnetite, presumably replacing ferric iron - in one instance 2000 ppm, was found in magnetite compared to 6000 ppm - 200 ppm andl<5 ppm in coexisting ilmenite, clinopyroxene and olivine. in view of the much greater abundance of augite than oxides in the rocks of the present study, it is likely that much of the V and Ti in the rock is in the pyroxene even though the oxides are richer. CHAPTER IV REGIONAL VARIATION Introduction In this section the petrographic and spectrographic data for the regional samples are examined and compared with the results of the differentiation study in order to determine whether there are any regional differences in the Nipissing. The data has been divided into the three areas mentioned earlier - north shore of Lake Huron, Sudbury area and Cobalt area (Figure I). North Shore of Lake Huron The diabase intrusions in this area would seem to be well correlated with the Nipissing on the basis of the age determinations. Folding is more definite here than in the Cobalt area, although still quite gently, and the rough alignment of the larger intrusions with fold axes is obvious, especially in the case of the mass north of Thessalon which appears to lie on the south limb of an anticline. Examination of the thin sections shows that quartz in present in all samples. Samples 29 and 30 contain hypersthene, 25 may have been hypersthene bearing but is so altered that it is difficult to be sure, I7 is a sample of granophyric diabase, 27 and l5 are fairly fine grained (near chill zone ?), while all of the others belong to the quartz-diabase and varied texture diabase phases. The spectrographic data is presented in Table I4. The results in general, and the mafic indices in particular, 78 TABLE Sample number 25 24 23 26 3l 27 28 30 Z 5‘8 -"l\) OLOOQ-bLOOODOLOQCOGJOVCOOOLO—N 0 O U 0 O 0 C O O O a O O O O O 0 O O O I O O O O O O O O 0 O 0 O O O B I C 0 C 0 U mflo—V—ONLOLOLO—NVVODCDVLOGV 0 B O D ll 0 0 0 O O 0 B I O I 0 0 O O 0 0 @QVb——VQQNMOWOGOO#A®# “WQAQOMWMQWLOCDVOV450KDUIO iron as FeO NaZO 3.00 I.70 2.40 2.20 I.70 I.82 2.70 I.60 I.60 l.82 2.70 2.65 2.I5 3.90 2.00 2.40 2.l0 Ni ppm I30 l25 72 83 I49 93 3l5 I46 ll8 56 42 66 90 38 93 27 82 88 72 96 I22 aphic analyses - Nipissing diabase - north Lake Huron Co PP. . SI 44 59 44 56 SI 43 42 53 56 49 47 59 l7 6! 66 an 44 62 Si 54 Mn ppm lI80 950 i700 8l0 I400 I440 l8l0 [1240 l 220 n. \J U'l O r 950 ~ \I ~03 cu: fl--.-N- ONV‘NLOM—Ol mmNNOCflOO OOOOOOOO :3 pm 2430 I640 4600 4400 l7l0 2380 2630 I220 l560 6900 9900 4900 4000 I5000 5900 6200 2l40 6400 6400 2880 3600 80 V ppm I96 l5l 256 247 ISO I75 I94 I39 l55 252 630 I76 I94 400 280 267 I37 235 29l I95 203 Mafic Index 39 4I 57 46 45 48 SI Felsic Index 20 I3 2! Sample number 25 24 23 26 3I 27 28 L. ‘ avg? "“33"" “‘Yiwm". ‘3" ¢ 8| agree with the conclusion from the petrographic evidence that the diabase at the surface is a little more acidic than that in the sections at Cobalt. Comparison of the average of the analyses for this area with the weighted average of the quartz diabase and varied texture phases of the Nasco section (Table l5) shows good agreement. In view of the unusually low average titanium content <>f the Nipissing at Cobalt as compared to other basic i ntrusions (Table 8) the similarity in these samples is noteworthy. The high titanium and vanadium content of sample I'7 (granophyric diabase) was expected from the amount of i ron oxide in the thin section, and conforms well with ”Elne trend seen in the differentiation study, which showed <:omcentration of these elements in later differentiates. ‘TThe low cepper content of this sample is in agreement with 1:he view that copper is concentrated in the intermediate Cquartz-bearing phases and depleted in later fractions. Sudbury area The diabase in this area is usually referred to as SSudbury gabbro and has generally been thought to correlate hiith the Nipissing. This would seem to be confirmed in 'tlfie eastern part of the area, where intrusions in the SL-Idbury group extend into the overlying Huronian rocks in ‘tlfie area mapped by J. E. Thomson (I960). Examination of the thin sections shows that while ‘F<>ur of the samples (32, 9, l0, and 79) have textures 82 TABLE l5. Average analysis of north shore of Lake Huron and analysis of quartz and varied texture diabase of Nasco section North shore of Quartz and varied Lake Huron texture diabase-Nasco MgO ‘7. 8.4 8.4 FeO i, 8.7 8.9 (ZaCD % 9.7 l0.2 H.320 7. 2.2 l.95 Ni ppm l00 62 CO tbpm SI 53 I“? F>pm I460 I390 Cu ppm HS 84 T" me 4600 3070 V C>F3m 236 I94 83 typical of the Nipissing elsewhere, the others seem to be unusual. Most of these are highly altered but sample 5 is quite fresh and shows texture that can be interpreted in the other samples. Sample 5 has stubby euhedral grains of augite and pigeonite (some hypersthene ?) enclosed in large plates of plagioclase. A modal analysis of the rock shows it to contain only I8 per cent plagioclase as crompared to the usual 30 to 55 per cent in the rest of 1:he Nipissing. Some of the other samples in this area éippear to have even less. The amount of pigeonite is l5 FDeer cent; the remaining 67 per cent is all augite. Zoning 5 8 absent in the plagioclase as well as in the pyroxene. The nature of this sample would seem to leave little <fl<3ubt that it is an accumulation of pyroxene crystals. ‘TThe lack of zoning in the pyroxenes or the large unzoned FD lagioclases enclosing the pyroxenes and the simple miner- 63 hogy suggest that near-perfect equilibrium attended the C:rystallization, something not seen in the sub-horizontal SSheets or sills. The most likely place for such conditions tn; occur would seem to be in more deeply buried intrusions, Phaving a shape giving a long vertical extent to the differ- ‘Erwtiating magma. The dike-like nature of the intrusions if? this area is thought, therefore, to be responsible for tifie unusual nature of the rocks. The spectrographic data, presented in Table l6, Sifiow that these rocks are among the most basic of any of the Nipissing samples - as basic as any of the olivine- TABLE I6. Sample number 32 OVVGDOLO 84 Spectrographic analyses of Nipissing diabase - Sudbury area MgO FeOa CaO NagO Ni Co ".7. 7. 7. I. 9 pm ppm 7.5 7.3 8.5 l.54 74 59 9.5 l0.2 l3.0 2.09 78 79 7.6 8.5 l0.4 2.45 82 53 2l.4 I0.0 9.3 .40 830 l04 [8.3 9.4 9.3 .80 620 l03 l4.3 7.6 l3.2 I.60 242 53 l6.6 9.2 l6.0 .83 470 79 l4.3 7.6 8.0 l.20 400 70 9.7 9.9 ll.9 l.52 93 6| l3.I 7.3 Il.4 I.27 240 65 l3.7 7.7 l0.8 I.37 340 7i ll.8 6.5 l2.4 l.95 I76 47 I4.0 8.I l3.9 l.09 264 79 II.5 7.7 l3.6 l.56 l65 57 9.5 9.3 l0.8 I.30 l04 42 l2.6 7.7 ll.6 I.I8 l89 55 l3.I 9.4 l2.8 l.43 l29 59 II.5 9.5 I3.4 l.37 I04 59 a - Total iron as FeO 85 Mn Cu Ti V Mafic Felsic Sample ppm ppm ppm ppm Index Index number II50 I50 3400 I37 - 49 I5 32 l380 57 5900 280 52 I4 9 l240 I3 2880 I84 53 I9 I0 I340 5 550 203 32 4 8 IISO 54 I55 234 34 8 7 II50 75 I420 22I 35 II 5 l240 I9 I270 227 35 5 77 I520 SI 2520 2I0 35 I3 5 I500 I38 5500 220 5I II 79 IISO 82 550 I49 35 I0 3 II20 I4 I030 I84 35 II 4 I250 43 750 II7 35 I4 78 I040 I23 930 I84 37 7 I 930 47 I490 l7l 40 I0 2 '260 77 II50 II2 49 II 72 I220 I08 750 l33 38 9 73 I460 I23 I580 I95 42 IO 7I I550 l33 20I0 2I4 45 9 74 86 hypersthene phase in the Rix section. If these rocks originated from the Nipissing magma a considerable portion of the upper part of the intrusions is missing. This would support Hawley’s (I962) suggestion of post Nipissing domal up)lift of the Sudbury area and the removal, by later erwosion, of Huronian rocks that may have covered the area. Thee arcuate arrangement of Huronian outcrops around the nor~thern part of the area supports the idea of domal up! 5ft. Sample 79 is surprising in the midst of these bas;i c rocks, although it is a little further south than an»! of them. It is similar to the typical quartz diabase botlfw in petrography and element concentrations. The writer IS 5 nclined to suggest that it may be part of a down TGLJ l ted block. Sample 32 is interesting in that it is the only NIP)? ssing sample to show bent and broken grains, especially fel cjspars, and thus provides evidence of post Nipissing deT=<>rmation in this area. Cobalt area For convenience of presentation the samples in this areas have been separated into those between Cobalt and the eeastern end of the Sudbury area and those at Cobalt and the the north and west. Analyses for the area are glVen in Tables l7 and I8. Samples 75 and 76 (Table I7) are of interest in that: they seem to continue the trend of more basic rocks Co Ni ppn ppm Na 0 2 %. ’9 iv...- 87 CaO ‘1 FeOa ‘7. Spectrographic analyses of Nipissing diabase - ‘4 between Cobalt and Sudbury TABLE I7. M90 Sample number iiIF/h’iail II.I'III 47477.3890l34809862432558l9024l03l529 I4444444.6555475455454643345545440443 36940075I4830l4302639485754800480492 l23988663052I6523098069767434237mwmwl2 2" 'I' II "II' ' III I 020550094.6622383545684206508728802l5 I0I888826608722977002076l78l96424l7l o a o o o o o a D o o O o a o O o o o o o a o O O a o a o o o I o o a o Inn/.2 lllllll 2ll22lll2222 lllll 2ll22ll22 cocoo-coco-nooooooaooooooooooooooooo 52498i9443i02i9i39ii309770268288i765 38203369764l269440763l67702883l88l50 O O D 9 O O 0 O O O a G O o O O O O O D O 0 O O D O O O O O I O l D O O I778887897l9909998707375587868962687 78026332i230709726743928806535962829 0 O o O O O O O l O a O 0 O O O D O O 0 a O O O O O O O o D O 0 O 0 O O 5009797229I098790088979668O76i777i64 iron as FeO a - Total 88 Mafic Index Felsic Index N—®NLD<flCAU1-bmmQWQWQNMOVN#m-OQOA\IQW43V —l\)f\)—'—l’\)——-—————————N—-———— NO!)— LOCO-9 Sample number TABLE Sample number 83 84 BI I8. 89 Spectrographic analyses of Nipissing diabase - Cobalt area MgO Reg6 050 NaZO Ni Co % %. %. % ppm pun 8.0 9.5 I0.I 2.25 78 44 9.7 7.4 7.8 2.68 78 40 9.3 7.0 9.4 l.45 ll2 40 7.6 6.2 8.6 I.64 62 34 7.3 8.8 9.8 I.89 59 53 5.4 0.6 7.2 2.55 26 68 9.2 8.4 l0.5 I.83 l25 57 5.4 8.6 8.0 2.28 27 45 l0.6 7.9 l3.0 l.55 I78 60 l7.7 8.4 I2.2 l.26 460 73 l0.0 6.5 l3.7 l.53 I26 53 8.7 5.9 I2.0 I.68 I46 45 ll.6 8.6 l0.5 I.60 l84 65 8.0 7.2 8.I 2.55 93 35 8.8 6.3 9.6 l.90 ll2 44 8.9 8.9 8.6 2.60 88 49 9.9 7.9 II.5 I.80 ll8 50 a - Total iron as FeO Mn ppm I380 2500 860 Il20 l650 I800 l340 I600 I260 I240 960 800 I340 I900 ll00 I440 ll50 Ti ppm 5400 |870 I700 I420 3200 9300 2I40 3800 I640 940 I6I0 l240 l930 3000 880 2430 2270 90 V ppm 300 I49 I80 I25 l5l 374 I95 l68 205 I49 I43 I37 I60 I67 II3 l96 l80 Mafic Index 54 43 43 45 Felsic Index I8 26 I3 I6 I6 26 IS 22 II 9 l0 l2 I3 24 I7 23 I4 (fl hi ma 53 50 dCC nor 86 87 chi are anal 9| seen in the Sudbury area, although texturally they are like the normal Nipissing. The other samples in the area between Sudbury and Cobalt include representatives of nearly all of the phases of the Nipissing - 57 and 33, while highly altered, appear to be olivine-hypersthene diabase, 55 is a quartz-hypersthene diabase, the others appear to be quartz and varied texture diabases. With the few exceptions noted below the spectrographic data are similar to those for the same phases in the Rix and Nasco sections. The high titanium content of sample 6l is clearly due to the concentration of iron oxide in the sample. Since vanadium appears to occur in the oxides as well, its high concentration is to be expected. The same explanation may apply to the cobalt enrichment. Examination of this sample and sample 64 under reflected light showed that a sulphide mineral, probably chalcopyrite, was present, accounting for the abnormal copper values in these samples. The samples from the area around Cobalt and to the north and west also show considerable variation (Table l8). 86 and 88 are olivine-hypersthene diabase. 59, 60, 85, 87, and 89 are quartz-hypersthene diabase. Except for the chill phase samples (8|, 82, 83, 84, 94) all of the others are from the quartz and varied texture phases. Again the analyses correspond well with similar phases in the sections. Comparison of areas The foregoing indicates a basic similarity in the 92 Nipissing intrusions over the whole regions, in that the phases present and the concentrations of the various elements in them are similar to those observed in the detailed work in the Cobalt area. The unusual rock in the Sudbury area is compositionally similar to the olivine- hypersthene phase of the sections; their different mineralogy and texture is due to differences in the extent to which equilibrium between the various minerals was attained, not to any basic difference in the process involved. That there are regional differences in the samples is shown by consideration of the averages of the analyses for the different areas (Table l9). The average for the north shore of Lake Huron is slightly more acidic than the average chilled diabase, the area around Cobalt and to the north and west is a little more basic, than the chilled diabase, the area between Cobalt and Sudbury is more basic yet, and the Sudbury area is composed largely of very basic rocks. There are two possible explanations for this. I. There are real differences in the overall composition of the Nipissing. 2. The composition of the Nipissing is uniform, but erosion has exposed different average levels of the differentiated intrusions in the different areas. That the latter is a possible explanation is indicated in the Cobalt area from the results of the present work, TABLE l9. Average analyses of various areas No. of samples M90 ‘7. FeO % CaO % N520 % Ni ppm Co ppm Mn ppm Cu ppm Ti ppm V ppm MAFIC INDEX Chilled Phase I0 8.5 8.2 9.5 2.0 8| 44 I450 95 2370 I85 49 93 Cobalt Cobalt- l7 9.I 7.9 l0.0 2.0 I22 55 l380 83 2630 l82 46 Sudbury 36 9.3 9.I I0.5 II0 55 I500 I33 2850 235 49 Sudbury North shore of l8 l2.8 8.5 ll.7 l.4 256 66 I260 73 I970 I88 40 Lake Huron 2| 8.4 8.7 9.7 2.2 l00 5| I460 II3 4600 236 Si 94 as well as from the studies by Hriskevich (l952) and R. Thomson (l962). Erosion in this area has removed the most acidic phases of the diabase, exposing a considerable proportion of the more basic rocks. Random surface samples, therefore, have a more basic composition than the magma. Extending the hypothesis to the other areas, it would then appear that between Cobalt and Sudbury erosion has been a little deeper than at Cobalt, especially near Sudbury. The suggested doming of the Sudbury area, has resulted in the almost exclusive appearance of very basic phases on the present erosion surface. The rather abrupt transition from the Sudbury area to the north shore of Lake Huron is reasonable, occurring as it does across the major fault zone. The latter area has had the least amount of erosion, with the result that the phases seen are mainly acidic. (Granophyre is often mentioned in reports on the area, eg. Robertson, I963). Relation of metalliferous deposits to the Nipissing diabase As noted in the introduction, one of the purposes of this investigation was to determine whether the variation in types of deposits associated with the Nipissing diabase could be related to any regional differences in the Nipissing. The deposits in the Nipissing can be grouped into three main types. Quartz-carbonate veins carrying chal- copyrite, pyrite, and minor specularite are typical of the north shore of Lake Huron area. Small deposits of massive and disseminated pyrite, chalcopyrite, and nickel-bearing 95 pyrrhotite seem to be typical of the Sudbury area (eg. Ginn, l965). The Cobalt area is well known for its unusual silver-cobalt-nickel-arsenic mineralization which occurs mainly in carbonate veins, although a few of the "north shore type” copper showings have been noted, expecially in the Timagami area (Simony, l964). As detailed studies of the various deposits have not been made, it is not possible to relate them positively to specific phases or structures in the diabase, but some inferences about their variation would seem obvious in view of other results of this investigation. Thus it is not surprising to find copper deposits along the north shore of Lake Huron as the "average" diabase here is quartz diabase, the phase found to be most enriched in copper in the differentiation study. The presence of quartz and carbonate suggests that while the sulphides may have formed as immiscible droplets in the quartz diabase, they were segregated into the late liquids trapped in this phase. Fracturing of the nearly solid diabase would have allowed the escape of this hydrothermal fluid, and the formation of veins extending through the diabase and into the country rock. The presence of nickel in the sulphide deposits in the Sudbury area would appear to be related to the very basic composition of the diabase, as the differentiation study showed considerable enrichment of nickel in the most basic phase. The massive and disseminated nature of 96 most of these deposits may indicate that they represent varying degrees of accumulation of sulphide droplets formed early in the crystallization process. The slight enrichment of copper at the base of the Rix section suggests that this might have occurred elsewhere as well. The silver deposits of the Cobalt area have been explained as having formed at the last stage of crystallization of the diabase (Hriskevich, I952). The occurrence of nickel and cobalt is thought to be due to their exclusion from the late phases of the diabase (granophyre and aplite), and the resulting concentration in the final fluids. It is apparent that the different types of deposits seen in the north shore of Lake Huron and Sudbury areas may be due to the level of erosion of the diabase (i.e. nickel-copper deposits may occur in buried basic phases along the north shore, and hydrothermal copper deposits may have occurred in the eroded upper parts of the intrusions of the Sudbury area). All phases of the diabase are seen, however, in the Cobalt area, and no deposits of either type are known to occur. In view of the similar copper and nickel contents of similar phases of the diabase in all three areas the writer suggests that the absence of copper or nickel sulphide deposits is due to a deficiency of sulphur. This is also supported by the low sulphur content of the silver deposits themselves, where such elements as arsenic, antimony, and bismuth, which are usually much subordinate to sulphur in amount, become 97 important constituents. There are two possible reasons for the dificiency of sulphur in the Cobalt area - I. The original magma was deficient in sulphur in this area. 2. Sulphur was lost from the magma before the formation of the metalliferous deposits in the Cobalt area. To accept the first explanation would be to assume that there are regional differences in the Nipissing. This has not been demonstrated for any of the elements determined in the present study, but since the sulphur content of the rocks was not determined, it remains a possibility. Because of the constancy of the other elements, however, the writer prefers the assumption that the original magma in the Cobalt area had the same sulphur content as in the other areas, and would offer the following suggestions in explaining its preferential loss. I. The Huronian rocks are thinner in the Cobalt area than in the north shore of Lake Huron area. The Nipissing magma may have risen closer to the surface at Cobalt before it was able to spread out into sheets. The lower confining pressure of over- lying rocks could have been low enough to allow much sulphur to escape. 2. Lineaments possibly related to pre and post Nipissing faults seem to be more common in the 98 Cobalt area than in the north shore area, and may also have allowed greater sulphur loss in the Cobalt area. 3. The intrusions in the Sudbury area may have been more deeply buried than those at Cobalt, or the much smaller horizontal surface area of the dikes may have allowed less sulphur loss than from the sub-horizontal sheets. The resolution of the questions raised here obviously requires further studies, beginning with sulphur analyses. The significant point is that regional variations in types of deposits associated with a given type of igneous rock may be due to minor changes in physical conditions during differentiation and not to a change in "metallogenetic province". CHAPTER V SUMMARY AND CONCLUSIONS Summary The purposes of this study, as noted in the intro- duction have been to investigate the local variations in the Nipissing diabase of the Cobalt area, on a chemical, petrographic and structural bases, to determine whether similar intrusions in adjoining areas are related to the Nipissing and, if so, whether there are any regional differences in the character of the Nipissing that might explain the regional variation in types of metalliferous deposits associated with this rock. The emphasis in the study has been on chemical variations in both major and minor elements. Analyses were made by a somewhat novel method (interrupted arc) of emission spectrographic analysis which appears to have given satisfactory results for both the major and minor elements involved. The speed and low cost, relative to chemical methods, of obtaining analyses in this was has permitted a much more complete investigation of the chemistry of the rocks than is usual in studies of this type. The field and petrographic studies show that the Nipissing was intruded as a fluid magma, generally forming sill—like bodies (sheets) in the flat-lying or gently folded Huronian rocks and dikes in the steeply dipping or vertical pre-Huronian and in the massive granites. The level of the sheets was probably determined in part by the 99 ‘VM- ‘1': ”F.3— l00 then-existing erosion surface and in part by the Huronian- pre-Huronian unconformity. The result in the Cobalt area was an undulating single sheet having a number of basins separated by arches. The thickness of the sheets ranges from less than 200 feet to at least ll00 feet, the latter being typical of the Cobalt area. Erosion has apparently removed part or all of the diahase in many places so that it is not possible to determine the original areal extent of the sheets. The largest mass exposed is west of Lake Timagami and covers an area of about l00 square miles. Samples of the chilled phase of the Nipissing in the Cobalt area indicate that the intruded magma had a fairly uniform tholeiitic composition - differentiation of the magma took place after intrusion. The chill compos- ition is similar to that of most other tholeiitic intrusions except in its low titanium content, which has been observed in only a few other cases. Petrographic and spectrographic (major and minor elements) studies of samples from a mine shaft and an inclined drill hole in the Cobalt area have been combined with the earlier surface and underground data, and show that the trend of differentiation in a typical basin-arch section is as follows: I. A chilled phase of quartz diabase was formed at the margins of the sheet. 2. Fractional crystallization and vertical gravit— I0l ational settling resulted in the accumulation of minor amounts of olivine in the lower parts of the sill which were not completely resorbed by the quartz diabase crystallizing around them. 3. Lateral, upward movement of acidic constituents toward the arches shifted the magma composition in the lower part of the basins so that olivine and hypersthene became stable phases and together with the plagioclases became more basic for a time, toward the top of the sheet. A limit to this was reached and later crystallization followed the common trend of iron and sodium enrichment. As a result the olivine hypersthene zone is succeeded by a quartz-hypersthene zone, a quartz diabase zone (hypersthene replaced by pigionite), and, a varied texture zone, where irregular patches of coarse to pegmatitic diabase indicate enrichment of volatiles. Later phases, postulated to occur mainly in the now-missing upper parts of the arches are granophyric diabase and aplite. This is supported by the occurrence of these phases elsewhere in the region. The behaviour of minor elements during different- iation shows definite trends in their incorporation into various phases of the diabase, similar to those found in other studies. Nickel and cobalt are enriched in the more basic phases, copper in the quartz diabase, vanadium, III. I I! I. . . ‘I- ___:l‘ I I..1';_""é-. . * l02 titanium and manganese in the varied texture and grano- phyric diabase. Data from other studies on the unit cell size of nickel, cobalt, and manganese olivines indicate the stability of these minerals relative to magnesium and iron olivines, and the observed trends fit this data very well, even though these elements are probably not exclusively related to the olivines. The regional study shows that intrusions of the area along the north shore of Lake Huron are similar to those in the Cobalt area, and their contemporaneity is confirmed by age determinations. Intrusions in the Sudbury area are dyke-Iike and show accumulation textures and more equilibrium conditions than in the other areas, but chemical similarity to the Nipissing as well as field evidence indicates that they too are part of the Nipissing. Regional variations in metalliferous deposits, while not studied in detail, are thought to be due to circumstances arising during differentiation, not to original differences in the composition of the magma. Conclusions l. The Nipissing diabase intrusions are contempor- aneous and uniform in original composition in the region between the eastern end of Lake Superior and Lake Temiskaming. 2. Differentiation of the magma occurred after intrusion, and produced rock types more basic and, later, more acidic than the magma. l03 3. Differentiation was lateral as well as vertical, due to the inclined attitude of most of the intrusions. 4. The behaviour of minor elements during differ- entiation follows the same rules as those that govern the behaviour of major elements. 5. The type and occurrence of associated metall- iferous deposits may also be governed by variable conditions during differentiation. 6. Surface sampling of differentiated igneous intrusions may not give an accurate picture of their overall composition, nor of the extent of their variation. Suggestions for further work The Nipissing diabase is well suited for studies of various aspects of the problems of petrogenesis of basic intrusions. The varying degrees of erosion of the sheets, and the mine shafts and drill cores, permit unusually thorough examination of most levels of the intrusions from the lowest parts of basins to the top of arches. Further study might well result in the discovery of some of the feeders of the sheets. The extent of the area intrusions makes significant studies of regional variation. It is hoped that the regional variation of sulphur and other elements related to the problem of variation in metalliferous deposits can be studied in the near future. Relation of these deposits to specific rock types in the diabase would also be of value. LIST OF REFERENCES Ahrens, L. H., I950, Spectrochemical analysis, New York, Addison Wesley. Collins, W. H., l9l7, Onaping map area, Geol. Surv. Canada, Mem. 95, |57 p. , I925, North shore of Lake Huron, Geol. Surv. Canada, Mem. I43, I60 p. Deer, W. A., Howie, R. A., and Zussman, J., I963, Rock- forming minerals, vol. I, Ortho and ring silicates, New York, John Wiley and Sons, 333 p. DeVore, G. W., I955, Crystal growth and the distribution of elements, Jour. Geol., vol. 63, pp. 47I-494. Fairbairn, H. W., Hurley, P. M., and Pinson, W. H., I960 Mineral and rock ages at Sudbury-Blind River, Ontario, Proc. Geol. Assoc. Canada, vol. l2, pp. 4l-66. , I965, Re-examination of Rb-Sr whole—rock ages at 'Sudbury, Ontario, Proc. Geol. Assoc. Canada, vol. l6, pp. 95-l0l. Faure, G., Fairbairn, H. W., Hurley, P. M., and Pinson, W. H., Jr., I964, Whole-rock Rb-Sr age of norite and micro- pegmatite at Sudbury, Ontario, Jour. Geo|., vol. 72, pp. 848-854. Fleischer, M., and Stevens, R. E., I962, Summary of new data on rock samples G-I and W-l, Geochim. et Cosmochim. Acta, vol. 26, pp. 525-544. Ginn, R. M., I965, Nairn and Lorne townships, Ont. Dept. Mines, Geol. Rept. 35. Goodwin, A. M., I965, Mineralized volcanic compleses in the Porcupine-Kirkland Lake-Noranda region, Canada, Econ. Geol., vol. 60, pp. 955-97l. Grant, J. A., Pearson, W. J., Phemister, T. C., Thomson, J. E., I962, Broder, Dill, Neelon, and Dryden townships, Ont. Dept. Mines, Geol. Rept. 9. Grout, F. F., l948, Origin of Granite, Geol. Soc. Am.,Mem. 23' pp. 48-54- Harvey, C. E., I957, Spectrochemical prodedures, Glendale, Applied Research Laboratories. l04 l05 Hawley, J. E., I962, The Sudbury ores, Their mineralogy and orIgIn, Can. Mineralogist, vol. 7, pt. I, 207 p. Hess, H. H., I960, Stillwater igneous comples, Montana, a quantItative mineralogical study, Geol. Soc. Ah. Mem. 80, 230 p. Holmes, A., I965 Principles of physical geology, 2nd ed., New York, Ronald Press l288 p. Hriskevich, M. E., I952, Petrology of the Nipissing diabase sheet of the Cobalt area of Ontario, Unpublished PhD thesis, Princeton University. McDougall, I., I962, Differentiation of the Tasmanian dolerites, Red Hill dolerite - granophyre association; Bull. Geol. Soc. Am., vol. 73, pp. 279-3l6. Miller, W. G., I9l3, The cobalt-nickel arsenides and silver deposits of Timiskaming; Ont. Bur. Mines, Ann. Rept. l9, pt. 2, 279 p. Mills, F. C., I955, Statistical methods, 3rd ed., New York, Holt, Rinehart and Winston, 842 p. Moore, E. 5., I955, Geology of the Miller Lake portion of the GOwganda silver area, Ont. Dept. Mines, Ann. Rept. 64, pt. 5, pp. l-4l. Nockolds, S. R., and Mitchell, R. L., I948, The geochemistry of some Caledonian plutonic rocks; Trans. Roy. Soc. Edino’ VOI. 6', pp. 533-575- Olmsted, J. F., I966, Petrology of a differentiated anorthositic intrusion in northwestern Wisconsin; unpublished PhD thesis, Michigan State University. Phemister, T. C., I928, A comparison of the Keweenawan sill rocks of Sudbury and Cobalt, Ontario; Proc. and Trans. Roy. Soc. of Canada, 3rd ser., vol. 22, sec. 4, pp. I2l-l97. Ringwood, A. E., I956, Melting relationships of Ni-Mg olivines and some geochemical implications; Geochim. et Cosmochim. Acta, vol. IO, 99- 297-303. Robertson, J. A., I963, Geology of the Iron Bridge area, Ont. Dept. Mines, Geol. Rept. l7. Satterly, J., I933, The Nipissing diabase of Cobalt, South Lorrain, and Gowganda, Ontario; Unpublished M.A. thesis, Univ. of Toronto. " ' I. .7713!!! l06 Simony, P. 5., I964, Northwestern Timagami area; Ont. Dept. Mines, Geol. Rept. 28. Stockwell, C. H., I964, Fourth report on structural provinces, orogenies, and time-classification of rocks of the Canadian Precambrian Shield. Geol. Surv. Canada, Paper 64-I7, pt. 2, pp. l-2l. Stonehouse, H. 8., I954, An association of trace elements and mineralization at Sudbury; Am. Mineralogist, vol. 39, pp! 452-474- Thomson, J. E., I960, Uranium and Thorium deposits at the base of the Huronian system in the District of Sudbury; Ont. Dept. Mines, Geol. Rept. no. I. , l962a, The Proterozoic of the original Huronian; Roy. Soc Canada, Spec. Pub. no. 2, pp. 63-65. , l962b, Extent of the Huronian systems between lake Timagami and Blind River, Ontario; Roy. Soc. Canada, Spec. Pub. no. 4., pp. 76-89. Thomson, R., l962, The Proterozoic of the Cobalt area; Roy. Soc. Canada, Spec. Pub. no. 2, pp. 40-45. Turner, F. J., and Verhoogen, J., l960, Igneous and metamorphic petrology, 2nd ed.; New York, McGraw Hill, 694 p. VanSchumus, R., I965, The geochronology of the Blind River- Bruce Mines area, Ontario, Canada; Jour. Geol., vol. 73, pp. 755-780. Wager, L. R. and Deer, W. A., I939, The petrology of the Skaergaard intrusion, Kangerdlugssuaq, East Greenland; Med. om Gronland. Bd. l05, no. 4, 352 p. Wager, L. R., and Mitchell, R. L., I95I, The distribution of trace elements during strong fractionation of basic magma - a further study of the Skaergaard intrusion, East Greenland; Geochim. et Cosmochim. Acta, vol. I, pp. l29-208. Walker, F., l940, Differentiation of the Palisade diabase; Bull. Geol. Soc. Am., vol. 5|, pp. I059-Il05. Walker f., and Poldervaart, A., I949, Karroodolerites of the Union of South Africa; Bull, Geol. Soc. Am., vol. 60, pp. 59l-706. Yule, G. U., and Kendall, M. G., I950, An introduction to the theory of statistics, l4th ed.; New York, Hafner, 70l p. lug I Ejj F0 LI”. 9““ S_OI_O>—z mdhdm cz_amw 22222222 J) /+ I . r... / I/ 9. 0 .. (Iv ./