The iron oxide in the finer part of the conglomerate and in the sandy bands is of two kinds -- included plates of hematite in the grains of felsite, as just described, and irregular clastic grains of iron oxide that once were titaniferous magnetite. In the sandstone lenses the grains of oxide are in large part concentrated in layers of "black sand," giving the rock a banded appearance; some of these grains, as seen under the microscope, are made up of limonite crossed by bars of ilmenite. In the thin conglomerate beds that mark the lower part of the series these grains consist mainly of hematite with similar ilmenite bars; very few of the grains are attracted by an ordinary magnet. Some of the iron oxide grains in the Great conglomerate, however, are distinctly magnetic and under the microscope are found to consist mainly of magnetite with bars of ilmenite; the magnetite is partly oxidized, but the ilmenite is unattacked. These facts suggest that the elastic grains of iron oxide were derived from the erosion of areas of basalt and that although the amygdaloid and trap were largely destroyed and dissipated (though perhaps represented in part by the red basic mud rock underneath the felsite conglomerates) the compact iron oxide grains, because heavy and chemically stable, were preserved and were accumulated with the felsitic debris.

     Hematite, the prevailing oxide in the conglomerates, is not a by-product of copper mineralization; it is present in all the beds whether mineralized or not, and mineralization instead of producing it has destroyed it. (See p. 103.) The conglomerates, indeed, though not as rich in hematite as the amygdaloids, contained before mineralization a notable amount of ferric iron but, except for a few amygdaloid pebbles, very little ferrous iron,

     The average amount of iron in the conglomerates has not been closely determined. A felsite pebble from the Allouez conglomerate was found to contain 4.88 per cent of Fe₂O₃ and 0.6 per cent of FeO. The iron content of the conglomerate is probably very close to 5 per cent of Fe₂O₃ and 0.6 per cent of FeO. Certain porphyry pebbles and boulders much higher in ferric oxide are of interest because of their especial susceptibility to copper replacement; those that have been peripherally or entirely replaced by copper form, respectively, the copper "skulls" and "boulders" so often described. One of these iron-rich pebbles gave 12.68 per cent of Fe₂O₃ and 1.30 per cent of FeO.

     MINERALIZATION

     MINERALOGY

     The minerals that have been introduced into the conglomerate since its consolidation are few and of simple character as compared with the corresponding minerals of the amygdaloids. They all belong to the same general period of mineralization. as the copper. The more abundant minerals, named in the order of their deposition, are red alkali feldspar, early; epidote and pumpellyite, mainly earlier than copper; calcite and quartz, throughout; copper; chlorite, associated with the alteration accompanying copper, especially in the iron-rich pebbles.

     The zeolites and the allied minerals like prehnite are strikingly absent from the conglomerate; laumontite, though present in fissures in the adjacent trap, almost never persists where these fissures continue into the conglomerate. It is evident, therefore, that the zeolites were not a characteristic and necessary product of the mineralizing solutions but that their formation was primarily dependent on the nature of the rock through which the solutions passed.

     ROCK ALTERATION

     The copper occurs chiefly in the finer cementing material of the medium to coarse grained conglomerate. Undoubtedly in part it filled pore spaces in the sandy matrix of the pebbles, but mainly it replaced the cement.

     The pebbles, especially those of dense felsite, are generally unreplaced, though here and. there bleaching may affect their borders or extend through them along planes of permeability. A few pebbles of quartz porphyry were attacked by copper, but even where advanced replacement has occurred the phenocrysts of quartz and of feldspar remain. In some pebbles the feldspar crystals contain minute copper flakes, and the feldspars may have been attacked thus before the groundmass was replaced. Where iron-rich pebbles were partly replaced by copper, the ferric oxide was in part removed and in part reduced and recombined into chlorite, which is a conspicuous alteration product of such pebbles, as described by Pumpelly and Lane. In these pebbles, chloritized and partly converted to copper, crystals of barite may be present.

     The most conspicuous alteration associated with the copper was a pronounced bleaching of the conglomerate from a rich brownish red to a light pink or salmon color. This change is recognized by all those working on the lode as an accompaniment of good ore and is so intimately and faithfully associated with copper from the largest masses down to microscopic particles as to leave no doubt that an intimate causal relation exists between the two. It resulted from the removal of a part, commonly a large part of the hematite of the conglomerate with no very pronounced change in the other minerals. In the mineralized portions of the rock the fine material and the small pebbles are generally bleached entirely, but the larger pebbles may be bleached only at their margins, the centers remaining dark. This bleaching, as exemplified in the Calumet & Hecla lode, is somewhat more conspicuous in the lower part of the mine than in the upper part but is to be seen practically everywhere.

     Alteration of another type, whose effects are most conspicuous in certain lenses of sandstone, converted a large part of the rock into pale yellowish-green epidote. This change resulted from a recombination of the materials originally present, together with a considerable addition of ferric iron. Evidence of alteration of this kind is to be seen especially in the sandy margins of the Calumet & Hecla conglomerate where it is thinning down.

     DISTRIBUTION OF COPPER IN THE LODE GENERAL CONTROLS

     The following discussion applies directly to the Calumet & Hecla conglomerate, but in its general features it relates to the other conglomerate lodes in so far as they are known.

     The copper occurs chiefly in the sandy matrix of the medium and coarser conglomerate. It may be present as small isolated grains, but more commonly it forms a spongy mass through the matrix; and in especially rich places the matrix has been largely replaced, so that the ore consists of abundant felsite pebbles in a cement of metallic copper. Deposition of the copper in the matrix, more than in the pebbles, probably was due to physical causes rather than to differences in composition. The contacts between the sand grains of the matrix and the smooth surfaces of the pebbles are much more permeable than the pebbles themselves and notably more so than the masses of closely packed grains in the midst of the sandy matrix or in the larger sandy lenses. Higher permeability means greater facility for the passage of solutions, and all the evidence indicates that the intensity of mineralization varied in proportion to the quantity of solution that passed through a given volume of rock. The rate at which copper was deposited by replacement of rock was proportionate to the rock surface exposed to replacement, and the ratio of surface to mass is higher in the sandy matrix than in the pebbles. These two controls then, were favorable to deposition of copper in the sand mixed with abundant pebbles but unfavorable to deposition where sand occurred alone.

     Not only in detail but in a larger way the copper is distributed irregularly through the lode. The conglomerate is distinctly a banded rock, made up of layers or thin lenses that differ in texture. Certain of these bands are well charged with copper; others near by and apparently similar may contain little. This difference in intensity of mineralization from place to place is again not due to variation in composition but is an evident consequence of difference in permeability; the more permeable layers were channelways for the solutions, and in these channelways more copper was precipitated.

    
FIGURE 15.-Variation in copper content per foot of depth, Calumet & Hecla mine

     In a given cross section of the lode, however, the most permeable layer may not always be the one most largely replaced. This may be the result of local causes. For example, solution that has entered a fairly permeable layer lying near the middle of the lode and bounded on each side by less permeable bands may continue along this channelway for some distance to a place where other layers are more permeable but are dammed off by less permeable bands. Farther along these more permeable beds may be reached by the solution and may eventually become the principal channelways in a given stretch of the lode. The consequence is that the copper is found in overlapping streaks or lenses through the conglomerate bed.

     The well-mineralized streak may occupy any position in the lode from footwall to hanging wall, and at places two or more good streaks may be present with lean conglomerate between. These well-mineralized layers may persist for long distances or they may be of relatively short extent.

     RICHNESS IN RELATION TO DEPTH

     In the deeper parts of the Calumet & Hecla mine, where the Calumet & Hecla conglomerate lens is broad and thick, the mineralized streaks and lenses constitute a smaller proportion of the conglomerate and the poor lenses a larger proportion. Higher up, where the conglomerate body is smaller, and thinner, the richer streaks predominate over the poor ones. In consequence, the conglomerate as a whole is distinctly richer in the upper than in the lower part of the mine.

     The difference in grade, however, is due mainly to the fact that in the lower part a smaller portion of the lode has been. mineralized rather than to any material difference in intensity of mineralization or abundance of copper in the part that has been mineralized. That is, if the mineralized lenses of the lower levels could, be mined without inclusion of material from the unmineralized parts of the lode, the ore so obtained would compare favorably in copper content with that from higher levels, where the lode as a whole is much richer.

     The richest rock in the Calumet & Hecla conglomerate, averaging about 80 pounds to the ton for yearly yield, was essentially at the present surface (see pl. 38), and there has been a fairly steady decrease in richness of the rock mined till at the present depth of operations the yield is 35 to 40 pounds. The axis of maximum mineralization occupies a position about parallel with and a little above the axis of greatest thickness of the conglomerate lens. (See pl. 38.)

     There is no considerable change in total quantity of copper at any given horizontal section front the top to the bottom of the mine -- that is, horizontal sheets of unit thickness at the twenty-fifth, fiftieth, and seventy-fifth levels would each contain about the same quantity of copper. (See fig. 15.) But as depth is attained and the conglomerate lens increases in size, essentially the same quantity of copper is distributed through a greater volume of rock, with resulting lower grade for the lode as a whole at the deeper levels.

     In the upper levels, where the lode was relatively thin and the ore rich, the entire width of the lode was stoped, and all the copper that it contained was recovered, even though part of the thickness of the lode was too lean to pay provided it could conveniently have been left unmined. As the lode thickened and decreased in average grade lower down, streaks that were barren or too poor in copper to pay for handling were left unmined wherever they were thick enough or so related to the good ore as to make this possible. Because of the copper thus left the recovery from the lode at depth has been less than the recovery at higher levels, where all the lode was mined.

     The actual decline in richness from top to bottom, viewed geologically, results apparently from the fact that as the solution rose through the conglomerate mass that contracted upward it encountered a constantly decreasing volume of conglomerate and consequently traveled through and mineralized a larger and larger proportion of it, until, near the present surface, the cross section was so reduced that nearly all the conglomerate was traversed and converted into ore. The inference is that had the Calumet & Hecla conglomerate maintained to the surface the same cross section as it has at depth, the ore near the surface would be of essentially the same grade as that at the deep levels. The inverted-funnel structure afforded by the conglomerate lens is a special case of the barrier condition, which is considered in more detail on page 115.

     The Allouez is the only other conglomerate that has been explored to any great extent. It has been opened and mined at three localities - at the Franklin Jr., Allouez, and Delaware mines. The mineralization is very similar to that of the Calumet & Hecla conglomerate, but in all three places the conglomerate is rather thick, and, only portions of it are mineralized. It has yielded on the whole a low-grade ore, though it contains lenses of good grade. A notable feature of the Allouez conglomerate is the unusual amount of chalcocite in it wherever it has been opened. This sulphide occurs characteristically intergrown with calcite in veinlets along joints; the calcite is dark because of its chalcocite content, and the metallic luster and streak of the sulphide are obscured by the calcite. Chalcocite occurs similarly though in lesser amount in the Kearsarge conglomerate and, indeed, in all others examined, including the Calumet & Hecla. In the latter it, is apparently most abundant near the margins of the ore shoot, for it is relatively abundant on the dumps of the Centennial shafts and, to a less extent, on those of the Osceola shafts; at the north and south ends of the shoot, respectively.

     SANDSTONE AND SHALE LODES GENERAL RELATIONS

     Sandstone and shale become increasingly abundant in the upper part of the Keweenawan sediments. They are plentiful in the Copper Harbor group and become dominant in the overlying Nonesuch, Freda, and Jacobsville formations.

     The only formation that needs special consideration as a copper bearer is the Nonesuch. It consists prevailingly of red sandstone and shale, but in the Porcupine Mountain region it includes a considerable thickness of black, fine-textured shale that covers a considerable area. Immediately beneath this black shale the sandstone, in several localities at least, is gray instead of red. This gray sandstone and the black shale are the principal copper-bearing rocks of the Porcupine Mountain region. The old Carp Lake mine is high up on the north side of the uplift and at a lower horizon - namely, in red sandstone just below the Lake Shore trap.

     Most of the exploration at the Nonesuch horizon has been done at the White Pine, Nonesuch, and White Pine Extension mines. These mines are several miles apart, south and southeast of the Porcupine Mountain uplift. The three deposits are near pronounced faults, and the ore bodies at White Pine, at least, are associated with minor cross faults and fissures.

     The Porcupine Mountain uplift is probably due to a laccolithic intrusion either of the felsite that forms most of the range, which, however, is commonly regarded as a lava flow, or of coarser porphyry, of which small. exposures are known. The fracturing and faulting in the surrounding rocks was, like the doming, a result of this intrusive action. Along the openings so produced the ore solutions probably ascended, and they may be regarded as having come from the same source as the igneous rock.

     MINERALIZATION AND ROCK ALTERATION

     In the sandstone, which is arkosic, copper occurs in the more permeable layers, especially those immediately under impermeable covers of shale. It has filled the pore, spaces and partly replaced the less resistant grains and is present also along small joints and fractures. The proportion of silver to copper is higher than in most other deposits of the district, and, in exceptional patches a few inches across, silver may quite displace copper and make up a considerable fraction of the rock.

     Part of the gray sandstone of the Nonesuch formation has clearly been derived from the red by bleaching, as it repeatedly grades into the red. It is not so clear that the layer immediately under the shale was once red and has been bleached. This alteration can not be as definitely ascribed to copper mineralization as the bleaching in the conglomerates and amygdaloids lower in the series. It is believed, however, that the bleaching of the sandstone has been accomplished by sulphur-bearing solutions that deposited the copper. The sandstone that lies immediately below the shale is persistently gray wherever exposed, whether mineralized or not, though sandstones at other horizons are characteristically red except near mineralized rock, where they are bleached gray. The cause for this change of color can not be definitely assigned, but three possible causes have been recognized --- namely, the action of bituminous matter from the overlying shales, of hydrocarbon entering the sandstone through the fault fissures, and of the solutions that deposited copper and chalcocite.

     In the White Pine mine, where the gray rocks were most thoroughly examined, they contain a considerable amount of black asphaltic hydrocarbon, as well as of copper and of chalcocite. The hydrocarbon or the agencies that deposited it may have exerted a bleaching effect on the rocks. The distribution of both hydrocarbon and copper seems to be controlled by fissuring and by the permeability of the beds, so that in this small area wherever copper is present hydrocarbon is present also, and the influence of each on the iron oxide has not been clearly distinguished. The hydrocarbon appears to be earlier than the copper, for some of it has been partly replaced by chlorite, and some of the chlorite has been in turn replaced by copper. There is also the possibility that the black shale itself, which at the White Pine mine is bituminous, reduced the iron of the sandstone immediately underneath it and thus accounted for the bleaching in that layer. Hydrocarbon in the shale so far as known is confined to the White Pine area, where the fractures extend into the shale. It is possible that the hydrocarbon may have been distilled from the shale by the igneous intrusion, but it may have been introduced into both shale and sandstone from some outside source.

     Another erroneous impression that may be gained from the literature is that the copper in the Nonesuch formation occurs mainly as chalcocite; this belief probably resulted from mistaking hydrocarbon for chalcocite. Samples of all grades of concentrates from the White Pine mill in December, 1918, as reported by Mr. George L. Heath, indicate that about 98 per cent of the copper is native and about 2 per cent occurs as sulphide. This proportion of sulphide, however, is much more than is shown in the conglomerate and amygdaloid lodes.

     At the Carp Lake mine the red sandstone immediately below the Lake Shore trap has been mineralized locally from a few inches to a few feet below the base of the trap; and in these places it has been bleached to a gray color. The copper in the bleached sandstone is metallic copper, at least in large part. The bleaching of the red sandstone seems clearly related to the copper. On the dump of the tunnel is fissure material consisting of ankerite or some mixed carbonate containing plentiful chalcocite. In the Carp Lake occurrence no hydrocarbon was noted.

 NOTES

     4 The term "lode" as thus used in the region denotes an ore zone parallel to the rock bedding and made up largely of the rock itself - a restriction of the general structural significance of the term as applied to an ore deposit of tabular form.

    5 Irving, R. D., Copper-bearing rocks of Lake Superior: U. S. Geol. Survey Mon. 5, p. 132, 1884.

     6 Nisbio, Keijiro, Native copper and silver in the Nonesuch formation, Michigan: Econ. Geology, vol. 14, pp. 124-134, 1919.