USGS Professional Paper 144 Butler & Burbank pp 203-211

      North of Portage Lake the Arcadian lode, from its position relative to No. 8 conglomerate which lies below it, has been regarded as equivalent to the Isle Royale lode. The character of the lode material as seen on the Arcadian dumps shows a close similarity to the Isle Royale lode; the Arcadian, like the Isle Royale, is highly and coarsely brecciated and carries numerous inclusions in the lower part; its oxidation is comparable to that of the Isle Royale; and both sericitization and pumpellyitization are factors in its alteration. It was long supposed that the beds on opposite sides of Portage Lake were offset by a fault concealed by the lake. Marvine¹⁰ assumed a horizontal throw of 720 feet, the north side being displaced  westward. Hubbard,¹¹ however, has shown by the position and strike of No. 8 conglomerate at two points 15,000 feet apart on opposite sides of Portage Lake, that the horizontal displacement on this supposed fault could not exceed 275 feet. Furthermore, it is now known that the fault already mentioned as occurring north of the Isle Royale No. l shaft intervenes between the two positions used by Hubbard and accounts for 175 feet of the total, thus reducing to a maximum of 100 feet the displacement of a fault under Portage Lake. So small a displacement over so great a distance may be due simply to a slight bend in the strike of the formations.

     Farther north the "St. Louis" lode is at about the same horizon, but neither this nor any other known lode north of the Arcadian bears close resemblance to the Isle Royale.

     To the south a lode in the stratigraphic position of the Isle Royale has been opened in the Elm River (Contact) property near Twin Lakes, and also in the Winona mine, where it is known as the Winona lode and has a thick breccia top with pumpellyitic alteration like that of the Isle Royale. Identification still farther south is less certain; the Evergreen and succeeding lodes and the Forest ("Victoria") lode are at about the same distance above the supposed No. 8 conglomerate as the Isle Royale lode. These lodes are considered separately.

     Thickness of Isle Royale flow

 Arcadian:                                                                Feet 
     Amygdaloid portion ------------------------------ 27
     Trap portion ------------------------------------ 117
                                                                                144

Isle Royale:
     Crosscut, sixth level north of No. 4 shaft----------- 66
     Diamond-drill hole No. 4, near No. 6 shaft -------- 93
     Diamond-drill hole No. 5, south of No. 7 shaft----- 81

     CHARACTER OF LODE

     The Isle Royale flow as represented at the Isle Royale mine is of the fragmental-top type so well illustrated elsewhere in the Baltic, Kearsarge, and Osceola lodes. The lode or top portion of the flow may in turn be separated into several phases or types that are easily recognized but that show gradation from one to another. These may be designated, from the top down, fragmental zone, banded amygdaloid, "vein trap," and foot-inclusion zone. The last forms the transition to the main trap portion of the flow. In addition, irregular small tongues or stringers from the overlying flow extend down in many places from a few inches to a few feet into the upper portion of the lode and are distinguishable only with some care from the lode proper.

     FRAGMENTAL ZONE

     The fragmental portion of the lode consists of irregular fragments of amygdaloid and fine-grained trap ranging from small grains to tabular blocks several feet in greatest dimension. In form the fragments range from sharply angular through subangular to fairly well rounded. In general, the larger fragments are more angular than a large proportion of the smaller fragments. The top portion of the fragmental zone is composed of fragments of a little smaller average size than the bottom portion. In any given section of the lode, however, there is a mingling of fragments of varying size. In texture the fragments may be very finely amygdular, or more coarsely amygdular with distinct "chilled" or finer-grained margins, or trappy and virtually uniform in character. The more traplike fragments are in general commoner near the base than near the top of the fragmental part of the lode.

     As a rule the vesicles are filled with minerals, though they may be only partly filled, especially if epidote is the filling material. The spaces between the fragments likewise contain minerals. In general, the color of this mineral filling, whether in the interstices or as amygdules, contrasts plainly with the usual red color of the rock fragments themselves; but in many places near the overlying trap, especially where the lode is thin, the rock, although clearly brecciated, contains little mineral filling of contrasting color, and the lode has in consequence a dead, "burned" appearance, mottled with alternating small patches of brick-red and darker brown, dark gray, or dark green.

     BANDED AMYGDALOID

     Relatively short stretches of the lode contain banded amygdaloid similar in general to what has been called "intermediate lode" in the Kearsarge but less abundant and persistent. This material may be overlain by the hanging-wall trap and grade below into the foot inclusion zone, or elsewhere banded amygdaloid a few feet thick may persist over considerable areas immediately below the fragmental zone. In other places large slabs of banded or cellular amygdaloid lie in the midst of the fragmental material. Most commonly the fragmental zone gives place to the foot inclusion zone without a distinct intervening layer of cellular rock.

     FOOT INCLUSION ZONE

     Directly below the fragmental zone, or the banded amygdaloid, is a brownish-gray rock of rather fine trappy texture, known in the mine as "foot trap." It commonly contains somewhat indefinite patches or inclusions of amygdaloidal rock; the amygdules in these fragments consist usually of chlorite, though less commonly of calcite or some other light-colored mineral, with or without chloritic amygdules near the outer margins of the inclusion. Some inclusions are sharply angular, essentially like the amygdular fragments of the brecciated portion of the flow, but far more are rounded, in consequence of partial melting by the inclosing material, and in extreme cases of melting or resorption the position of the original inclusion is marked by an area containing amygduler cavities but possessing no recognizable boundaries. The inclusions range, in general, from an inch or less to 6 or 8 inches in diameter; a few reach 2 or 3 feet. They are most abundant and most conspicuous in the upper part of this zone and decrease in number and become more completely resorbed and smaller farther down. They cease to be abundant 10 to 12 feet below the upper limit of the zone. These amygdular inclusions are regarded as pieces from the fragmental zone sunk or dragged into the lower, still molten, and moving portion of the flow and partly remelted by it.

     In contrast with the amygdules of these inclusions, the trappy rock of this zone contains amygdules of its own, but these are commonly sparsely distributed and noticeably large, as is usual in the lower portions of the lodes throughout the district.

     VEIN TRAP

     Slabs of trappy rock included in the fragmental zone are known as "vein trap." In general this material shows no notable differences from that of the foot inclusion zone, though in places it exhibits a slightly chilled border. Some slabs have an under margin marked by notably elongated or "pipe" amygdules, and, as a rule, the large amygdules that occur sparsely in the foot inclusion zone are lacking. These slabs range from 1 foot to 4 or 5 feet in thickness and from a few feet to several scores of feet in extent. The smaller slabs may lie parallel to the walls of the lode or be tilted at any angle to them; the larger masses of necessity are essentially parallel to the walls. Fragmental material lies both above and below the slabs. These slabs of vein trap probably represent parts of the flow that solidified without brecciation just under the fragmental layer but were then broken by the continued movement of the flow and had fragmental material dragged or formed underneath them.

     MAIN TRAP

     The foot inclusion zone passes gradually into a monotonous trap rock, practically devoid of amygdules and of amygdular inclusions, which constitutes the main lower part of the Isle Royale flow; this trap is greener and less brown than the trap of the foot inclusion zone.

     HANGING-WALL TRAP 

     The trap immediately overlying the Isle Royale flow is a dark greenish-gray and distinctly crystalline rock that commonly may be distinguished readily from the more brownish and finer-grained trap of the flow itself. Pipe amygdules in the trap just above the Isle Royale lode are present in places, but they are by no means so common or so characteristic as in the corresponding position above the Kearsarge lode. The hanging-wall flow above the Isle Royale lode ranges in thickness from 195 to 225 feet. At Winona the flow above the Winona lode is 310 feet thick.

     The trap has invaded the top portion of the fragmental zone as small tongues and irregular stringers that range from less than an inch to a few inches or even a foot or more in width. These tongues and stringers are distinctly reddish, and in this respect as well as in their finer grain they differ notably from the main mass from which they branch; they are thus more like the Isle Royale trap, from which, however, they may be distinguished by their denser texture, their amygdules, and their red rather than brownish color. These invading tongues are not everywhere readily distinguishable from the fragmental material that they penetrate, for the two are not unlike in texture and are closely similar in color. Ordinarily, however, the tongues are slightly redder near their margins, and at their very edges they may be marked by a faint lighter-colored line apparently due to incipient bleaching or else to the deposition of a colorless mineral. These invasions are rarely of such regularity as to be mistaken for dikes; they penetrate irregularly into and ramify through the fragmental material, parts of which they may surround. Although irregular and variable in direction, their average attitude is likely to approximate right angles to the hanging-wall contact, and they gradually fork and pinch, so that most of them terminate less than 4 or 5 feet below the contact. They probably represent places where the relatively fluid melt of the overlying flow has locally broken through the chilled lower contact of that flow and thus found access into the loose and jumbled material of the fragmental zone.

     DISTRIBUTION OF TYPES

     The distribution of the several phases of the lode seems most unsystematic. Throughout the mine areas of thick fragmental rock are interspersed with areas of thin, denser fragmental rock, of banded amygdaloid, or of vein trap. The foot inclusion zone is everywhere present, though its thickness and the number of its inclusions may vary.

     The areas of thick fragmental rock are very irregularly distributed; at many places where the fragmental material was piled above the general level of the flow surface it now bulges into the overlying trap. Where these bulges are pronounced there may be an apparent local change in the dip of the lode, the dip seeming to flatten on the up-dip side of these bulges and to steepen on the down-dip side. A similar local modification of the footwall dip is effected by abrupt reentrants of the thick fragmental rock into the basal trap, the footwall dip being steeper on the up-dip side and flatter on the down-dip side of these reentrants. On the other hand, where the fragmental material is notably thin, the hanging wall appears to bend in toward the footwall and the footwall to rise from its normal position. In these thin places the fragmental rock commonly loses something of its brecciated appearance, becoming more dense and massive, but it retains its red color, which chiefly serves to distinguish it from the rock of the footwall.

     There is thus a tendency toward alternate divergence and convergence of hanging wall and footwall-that is, toward alternate thickening and thinning of the lode But not every swing in one wall is accompanied by an opposite swing in the other wall. Moreover, the bulges into the hanging wall, which may measure as much as 10 or even 20 feet, are more extreme than the bulges into the footwall, which ordinarily do not exceed 5 to 8 feet. The bulges of fragmental material into the lower part of the lode may be more abrupt, so that locally the boundary between the fragmental zone and the foot inclusion zone may be almost at right angles to the plane of the lode. The bulges into the hanging wall are marked by slopes which appear to have been, when the flow was in its original horizontal attitude, of less than 40° and which thus may have been limited to the angle of repose of loose fragments. Vein trap is naturally commonest where the fragmental layer is thick but is not present in all thick places.

     The various kinds of lode rock are not only irregularly intermingled but occur in decidedly different proportions in different parts of the mine. In the workings of No. 4 and No. 5 shafts the fragmental zone is relatively thick and the areas of thin lode and banded lode are comparatively small. In the No. 7 shaft region, down to the 7th or bottom level, the lode is relatively thin, and areas of thick fragmental rock are small. The ground south of No. 6 shaft is intermediate in character between that of No. 5 and that of No. 7.

     In the old Huron workings, especially about No. 6 and No. 8 Huron shafts, the lode averages thinner than in the neighborhood of Isle Royale shafts Nos. 4 and 5, and the alternations between thick and thin lode are numerous and abrupt, especially in the upper levels. At certain of the thicker places, especially south of No. 8 old Huron shaft, the lode is double through the presence of a relatively large mass of "vein trap" between two layers of fragmental rock. Here both the upper and the lower fragmental layers were ore-bearing, but as a rule only the upper layer was thoroughly explored by the Huron management.

     South of No. 2 shaft, in the upper levels, the lode is of good thickness and grade, but in depth, down to the Grand Portage fault and especially northward to the limits of the workings beyond No. 1 shaft, thin lode predominates, though with numerous local exceptions. The Grand Portage or faulted portion of the lode has essentially the same character as the main section south of the fault, but the average thickness of the fragmental zone seems to be a little greater than around No. 1 and No. 2 shafts, on the main section of the lode.

     The distribution throughout the mine of areas of thick and thin fragmental material interspersed with cellular rock and of areas of trappy material overlain and underlain by fragmental rock, though suggesting complicated and varied modes of origin, is believed to result from comparatively simple conditions of movement and solidification while the flow was in progress, as is discussed in connection with character of tops on page 31.

     STRUCTURE

     ISLE ROYALE SYNCLINE

     Aside from the tilting of the entire series, the largest structural feature revealed in the mine is the Isle Royale syncline. This is a gentle fold which accounts for the curvature of the lode. Its axis is at about No. 4 shaft. The fact that the best ground centers about the axis suggests that the presence of the fold had something to do with the localization of the copper. The dip of the lode at the north and south, as near No. 2 and No. 6 shafts, respectively, is steeper, about 56°, than near No. 4 and No. 5 shafts, where the dip is about 51°. If a similar difference in dip were maintained downward the fold would become less and less marked, and at a depth of about 10,000 feet down the lode from the outcrop it would disappear. Indeed, at the present deepest levels the curvature of the lode is notably less than at the surface, that is, the levels are straighter than those above. On the other hand, the beds higher in the series, exposed to the west as far as the Ashbed or Atlantic lode, retain the synclinal structure. As these higher beds where they reach the surface overlie (stratigraphically, or perpendicular to the bedding) deeper parts of the Isle Royale lode and therefore are likely to reflect the structure existing at depth on that lode, it would seem probable that the Isle Royale syncline persists downward for a long distance.

     FAULTS

     The only large fault recognized in the mine is the Grand Portage fault, encountered in the north end. It strikes N. 35°-60° E. and dips 60°-80° NW. The section north of the fault is the "West lode," known as the "Grand Portage lode" before its identity with the Isle Royale was established. Where each section of the lode meets the fault, it is bent or dragged around so as to point somewhat in the direction of the other section. This attitude confirms the conclusion as to the direction of displacement caused by the fault and strengthens the correlation of the two sections as parts of the same lode. The fault offsets the northern section of the lode about 175 feet toward the west, as measured on the level. This is the distance at right angles to the lode, but the actual amount of displacement must have been greater, ranging from some 300 feet to very much more, depending on the direction of the movement. On the assumption that the fault is an overthrust the Grand Portage section moved relatively upward and slightly to the southwest. If that is assumed as the true direction of displacement, then any point in the lode south of the fault would find its equivalent in the Grand Portage block some four or five levels higher up. But in any case, except that special one in which the displacement had been entirely in a horizontal direction, equivalent portions of the lode on opposite sides of the fault would not be at the same level. This relation may explain differences in the lode on the two sides of the fault on any given level.

     The character of the Grand Portage fault varies from place to place. In the upper levels north of No. 1 shaft it is a clean-cut break, carrying from a few inches up to 2 feet of gouge and brecciated rock, with pronounced dragging of the lode but only minor fracturing parallel to the fault. Toward the bottom of No. 1 shaft, however, and in the workings north of No. 2 the fault is less definite and simple; it tends to fray into a series of fractures of somewhat variable dip and strike but conforming on the whole to the general direction of the fault; these fractures are more likely to carry veins of either light or dark colored calcite than to contain a notable amount of gouge.

     Another fault of approximately the same strike and dip and with the same direction of displacement was observed from the eighth level of No. 4 shaft close to the old Huron workings to the sixteenth level south of No. 4, or beyond. Its horizontal displacement is not over 30 feet as measured on the levels. In character it resembles the Grand Portage fault, being well defined and only a few inches to a foot in width in the upper levels but breaking into branches that form a fractured zone in the deeper workings. It carries quartz, calcite, pumpellyite, copper, and chalcocite.

     In contrast with these two faults, which appear to belong to the same series, there are other faults and fractures that strike nearly parallel to the lode but generally a little more to the north. More than half of these dip against the lode; the others dip with it, all at angles steeper than the inclination of the lode. The small divergence in strike between these fractures and the lode might permit considerable displacement along them to escape recognition, but so far as can be seen there is only slight displacement along any of these breaks and none whatever along many. They are rendered conspicuous chiefly by the vein minerals deposited in them, which consist mainly of laumontite, calcite, and quartz, in places with a little chalcocite or copper.

     All these faults for which the direction of displacement is evident appear to be reverse faults - that is, the hanging-wall side of the fault has been raised relative to the footwall side. This is the same type of faulting as that shown by the Keweenaw fault. Diamond drilling has disclosed, according to Lane, several other faults between the Isle Royale lode and the Keweenaw fault and parallel to the Keweenaw; these he has interpreted as probably reverse faults also. It thus seems reasonable to assume that the hanging-wall block of the Keweenaw fault, up to and beyond the Isle Royale lode, suffered distortion at the time of the main faulting and had relatively subordinate sympathetic faults and fractures developed in it.

     The faults carry minerals of the general period of copper deposition and thus are believed to have been formed before the mineralization.

     FISSURES

     Fissures are abundant in the mine and vary greatly in their attitude. Some of the larger ones are discussed above in connection with faulting. Most of the fissures can be grouped into a few systems. Those of one system strike N. 50°-65° E., and most of them dip southeast at rather steep angles, although some of them dip northwest. This system is best developed in the south end of the mine and is more sparsely represented in the north end. Another system strikes N. 25°-45° W. and has steep dips. It is represented in both the north and the south ends of the mine. A third series strikes nearly parallel to the lode and dips in the same direction as the lode but perhaps on the average a little more steeply. This system apparently changes strike with the main curvature of the lode and so keeps essentially parallel with the lode throughout the mine. A slight amount of displacement of reverse-fault direction can be seen along some of these fissures.

     These fissures are present in the main trap under the lode, and some of them pass through into the hanging-wall trap where this is exposed by the workings. As a rule only the stronger fractures persist from the trap of the footwall into the fragmental zone, and many of these are deflected upon entering the lode to a course approximately parallel or more nearly parallel with it. In general, then, fissures are most common in a zone a few feet thick that occupies the basal portion of the lode and constitutes the main copper zone. In places this concentration of fissures near the base of the lode gives to it a distinct appearance of sheeting approximately parallel to the walls, as is well shown in many of the old stopes, where in the blasting the rock has broken clean along these joints. On this account, and because of the position of the best copper along the base of the lode, in contrast with most of the other amygdaloids in which the best ground is near the hanging wall, it is desirable to ascertain the nature and cause of this fracturing and its effect on copper localization.

     Copper mineralization is often found to have terminated downward at a fracture plane essentially parallel to the lode, known by the miners as the "foot slip" and regarded as the boundary between the lode and the underlying "foot trap". The readiness with which the blasting breaks to such a fracture plane would seem to indicate that sheeting nearly parallel to the lode is indeed a characteristic of the zone just at the bottom of the fragmental layer.

     The assumption of this condition of sheeting along the foot of the lode must not be carried too far, however. For example; where the fissuring essentially parallel to the lode is best shown, two or three other systems of joints are likely to be well developed also, and the question arises whether, if there were occasion to carry the mining in a different direction, one of these other joint systems might not seem the prominent one to the miners and be as conspicuous after blasting as the "foot slip" is now. In the crosscuts from shafts to lode it can not be seen that fractures or joints parallel to the lode are notably more common or stronger than those in other planes, and when the lode is reached it is not commonly marked by a conspicuous concentration of fracture planes parallel to it. Where the lode pinches and the footwall rises perceptibly from its normal position the "foot slip" tends to swing with the footwall, but it is less conspicuous where the footwall is high and the lode is thin than it is where it lies farther down from the hanging wall under a thick fragmental zone. The change in direction of the slip is also less extreme than that of the bottom of the fragmental lode. The slip is therefore deeper in the foot inclusion zone where passing over the upward bulges of the footwall than where the lode is thick, for there it is either close to or actually within the fragmental part of the lode.

     The relative importance of the fractures that are approximately parallel to the lode as contrasted with fractures of other directions is not very definitely known. Their effect, however, in guiding and in limiting the flow of mineralizing solutions to the horizon close to the footwall can not be overlooked. It is not unreasonable to suppose that fissuring along the base of the lode and especially in a plane about parallel to the lode may be noticeable in the Isle Royale mine because of proximity to the Keweenaw fault to which this fissuring may be related. Another possible cause of the fracturing parallel to the lode is slipping of the beds when the Isle Royale syncline was formed.

     Altogether, the origin and the importance of the foot slip at the Isle Royale mine are not at all clearly understood.

     Many of the fissures are mineralized; the details are given under the next heading.

     ROCK ALTERATION AND ORE DEPOSITION

     The effects of two distinct periods of alteration can be recognized in the lode - one earlier than copper deposition and independent of it, the other believed to be closely associated with the formation of the ore.

     FIRST PERIOD

     Oxidation was the earliest alteration to affect the lode. This change probably took place before the lode was buried by the overlying flow. All of the fragmental part of the lode was decidedly oxidized and reddened, apparently about the same degree throughout its thickness. This porous material was apparently rather strongly affected by oxidation, even where overlain by a layer of "vein" trap. The cellular amygdaloid and the denser material that characterizes places where the lode is thin were also oxidized and reddened though to a less intense degree, and the oxidation and reddening extended with further decrease of intensity into the trap of the foot inclusion zone. In general, the fragmental portion of the lode has been more highly oxidized than the nonfragmental portion, but the fragments in the foot inclusion zone rarely show more oxidation than the surrounding trap.

     The tongues of overlying trap that extend into the lode, as well as the bottom portion of the main hanging wall flow for a few inches above the lode are red and oxidized, in distinct contrast to the main portion of the mass. This oxidation may have been accomplished at the time of or shortly after the outpouring of the hanging-wall trap by means of the oxygen included in the porous lode on which it was spread.

     Although the oxidation of the main lode undoubtedly produced a notable change in color of the rock, through the development of hematite, the texture of the rock was unaffected, and feldspar, the chief constituent, was not altered.

     SECOND PERIOD

     The second period of alteration came long after the surrounding rocks had been formed, and probably after the lode had been tilted into approximately its present position. In contrast with the early alteration, a main feature of which was the conversion of ferrous iron to ferric, the alteration of the second period was complex and more intense, causing in many places a profound change in both appearance and character of the rock. Embraced in this composite alteration were the successive but somewhat overlapping developments of epidote, pumpellyite, quartz, sericite, calcite, metallic copper, and copper sulphides and arsenides, as well as several other minerals of less common or less abundant occurrence.

     This second period itself seems divisible into at least two and possibly three stages - an early stage intimately connected with the deposition of most of the metallic copper and producing epidote, pumpellyite, quartz, and calcite; a later stage subsequent to the deposition of most of the copper and characterized by the development of sericite with quartz, calcite, anhydrite, gypsum, and a little barite; and a final stage or else the concluding phase of the second, in which the minerals of the sericite stage occur along with copper sulphides and arsenides or arsenical copper, chiefly in veinlets through the lode.

     The stage of copper deposition is marked by the production of a greenish or grayish rock, a result of the quartz-pumpellyite type of bleaching, in which pumpellyite is the most prominent mineral produced, with less abundant quartz, calcite, and epidote. A little prehnite, alkali feldspar, or laumontite may be present here and there. This greenish rock, dense and usually harder than the unbleached red breccia, is the characteristic associate of most of the metallic copper of the lode. It may occur in patches and streaks or may persist in long stretches of the lode, but it is generally confined to a layer, embracing the lower part of the fragmental zone and the uppermost part of the foot inclusion zone; and it is from this layer that most of the copper is obtained in the mine. Little copper occurs in this layer that is not inclosed in the green, pumpellyitized rock, but there may be considerable masses of the green rock that contain little or no copper. As a rule, however, copper and the bleached rock are close companions.

     This green copper-bearing layer coincides closely with the zone of fracturing along the footwall described above under "Fissures." Many of the fissures either carry copper or are marked by especially pronounced pumpellyitic alteration along them. It may be that the concentration of copper and the accompanying bleaching was localized near the base of the lode by reason of this belt of fracturing, which added to the normal permeability of the fragmental part of the lode.

     In places in this pumpellyitic zone the minerals and the texture of, the original rock are entirely destroyed. Such alteration with attendant copper deposition is scanty in the thin parts of the lode, where there is little breccia, and is strongest where well-marked little altered fragmental rock lies above the pumpellyite zone.

     On the under side, where the mineralized zone is in contact with the trap of the foot inclusion zone, the bleached rock is likely to terminate abruptly and in places is bounded by a break or fracture that separates it sharply from the darker rock underneath. It is probable that this green rock is formed mainly by alteration of the breccia and to a less extent by alteration of nonfragmental amygdaloid or of trap.

     At many places residual traces of the fragmental structure can be seen in the green rock, and it fades out gradually upward to typical red fragmental amygdaloid; in the main the alteration has obliterated the breccia structure, so that the part derived from breccia and that derived from the underlying trap are indistinguishable.

     In places this pumpellyitic zone and the accompanying copper extend above their usual upper limit into the main fragmental portion of the lode, even to the hanging wall and rarely for a few inches into the overlying trap. In such places the alteration is less, the bleaching or removal of the red color and replacement by green not so marked, the destruction of the brecciated structure less thorough, and the amount of copper smaller than in the principal zone. These upward extensions of the copper-pumpellyite zone are generally in places where the lode is widest and the fragmental layer thickest - in short, probably places of greater than average permeability. In some of the old Huron workings, where a tendency toward double lode is to be seen, this bleaching is present in both hanging-wall and footwall portions, though it is not very intense in either.

     Sericite is the most characteristic or distinctive mineral of the later stage of rock alteration. It is very abundant in some of the red fragmental parts of the lode, but its distribution is irregular, contrasting in this respect with the pervasive distribution of the pulverulent hematite that gives the red color to the breccia. In parts of the Isle Royale mine the breccia is well sericitized, as in the upper and intermediate levels of No. 6 shaft and in parts of No. 7 shaft and of the Grand Portage lode. In other parts of the mine, as in No. 2, No. 4, and No. 5 shafts, the sericitization was much less intense. The mineral occurs chiefly as a cement to the fragments, formed in part by filling of interfragment spaces and in part by replacement of the smaller fragments that lay between the larger ones. In many places where sericite was deposited, rock solution went on more rapidly than mineral deposition, the result being to produce a vuggy texture in the brecciated material. This is shown particularly well in the lower levels of No. 2 shaft, where pumpellyite and later sericite have replaced the breccia, leaving vugs that are now lined with crystals of calcite and quartz. Where present abundantly the sericite may be identified by its softness and rather greasy feel, also by its light color, ranging from white to pinkish or pale green, which contrasts with the strong red color of the breccia fragments cemented by it. In smaller quantity also it has partly replaced the larger fragments, but generally this replacement was not so extensive as to destroy either the texture or the red color of the fragments.

     By deposition chiefly in the spaces between the fragments, the sericite appears to have luted or plugged up the breccia and thus to have decreased the permeability of those portions of the lode in which it is plentiful. Because of its soft and somewhat plastic character, the sericite would probably be more effectual in this respect than quartz, calcite, or the other brittle minerals that were deposited in similar relation to the breccia fragments.

     It was at first thought that the deposition of the sericite with its plugging effect had taken place in the highly brecciated portion of the lode because that was the most permeable, before the pumpellyite-quartz-copper mineralization. Subsequent study, especially with the microscope, shows that this conclusion was ill founded. Nearly or quite all the sericite is found to be younger than the pumpellyite and to have replaced it in part. It is thus necessary to conclude that the solutions from which pumpellyite and metallic copper (and accompanying minerals) were deposited entered the lode before the fragmental portion was sealed up by sericite, that they therefore chose the zone near the base not by necessity but by preference, and consequently that this zone, because a site of fracturing, was actually more permeable at the outset than the highly fragmental but less fissured material overlying it. At places where the fragmental stuff was especially permeable, however - and apparently this is where it was thickest - some copper and zoisite did find their way up into it toward the hanging wall. Later, when the sericite-forming solutions came along, they had to content themselves with the main mass of the breccia, which was then the most permeable material available; they were excluded from the fractured zone near the base, which had already been occupied by copper and pumpellyite; they attacked the margins of the pumpellyite zone but were unable to penetrate far into it; and they gained little access to those thick places in the fragmental lode where copper and pumpellyite had been formed.

     In much of the best ground in the mine, especially near the No. 4 and No. 5 shafts, calcite, accompanied by more or less quartz and by smaller amounts of prehnite, strongly predominates over sericite as the cementing material of the fragmental zone. In certain areas, moreover, notably on the fourteenth level south and nineteenth level north of No. 5 shaft, on the twenty-sixth and twenty-seventh levels of No. 2 shaft, and on the fifteenth level at the south end of the Grand Portage section of the lode, anhydrite, with or without gypsum, occurs in a similar cementing relation to the breccia, locally forming patches several inches across. Scattered here and there through the lode also is a very little barite. The possible significance of these sulphate minerals is considered on page 136 in connection with the general hypothesis that the metallic copper was deposited as a result of the oxidation of copper sulphide solutions by hematite.

     Copper sulphides and arsenical copper, accompanied by calcite, sericite, quartz, chlorite, and specular hematite, occur in numerous veinlets that cut the lode. Possibly contemporaneous with these are the copper arsenides that have been found on the twelfth level at the very north end of the Grand Portage section. These occur in fragmental material near but not actually in veinlets. The arsenide patches are bordered by rock that shows bleaching of the iron-removal type and that under the microscope is found to contain sericite in addition to the usual quartz, epidote, calcite, and laumontite.

     In general the sulphide veinlets are narrow, being rarely more than 3 inches in width and commonly much less. In some, as one on the fourteenth level north of No. 5 shaft, the chalcocite occurs in short lenticular masses from which specimens weighing a pound or two may be collected. The gangue constitutes the chief filling or replacing material of the vein, so that it is frequently necessary to search carefully before finding the sulphide which the gangue minerals suggest is present. The wall rock of these veinlets is chloritized, sericitized, and calcitized. For a width of an inch or two the red wall rock may be bleached by the bodily removal of the iron and replacement by sericite and calcite; or the immediate walls may be dark green, owing to the removal of the ferric iron and the development of chlorite. In the vein proper calcite and quartz are the chief minerals, although ankerite, a carbonate of lime, magnesium, and ferrous iron, is also common. Where ankerite forms the gangue, both bornite and chalcopyrite may be very sparingly developed and specular hematite, magnetite, and metallic copper are also to be found. The central part of the vein may carry massive sulphide, which has replaced the carbonate.

     As a rule the chalcocite is confined to the veins, but in the Grand Portage section as well as on the second and third levels north of No. 7 shaft it has been found sparingly developed within the sericitized lode.

     The distribution of arsenides and arsenical copper has not been studied in detail. At one place arsenical copper, domeykite, and whitneyite are associated with chalcocite; the age relations are not clear, but there is a suggestion that the arsenides preceded the sulphide. Arsenical copper has been found in highly sericitized fragmental rock; there may be in this mine a relation between arsenic and sericite, such as is shown by the sericitic alteration along the Mohawkite fissure in the Ahmeek mine, but the Isle Royale sericite is by no means invariably associated with arsenide or arsenical copper.

     CHEMISTRY OF ORE DEPOSITION

     The chemical changes that accompanied the deposition of copper in the Isle Royale lode appear to have tended in the same general directions as in the other important amygdaloids. In the main, the red lode rock seems to have been favored for replacement by copper. The rock alteration attending the precipitation of copper has been relatively profound. There has been a notable decrease in total iron and a reduction of ferric to ferrous iron. As compared with the rock alteration accompanying the deposition of copper in the Kearsarge lode, there has been less removal of the iron, and more of what remains has been converted from the ferric to the ferrous condition; there has also been more deposition of quartz. The physical result is seen in the difference between the Isle Royale type and the Kearsarge type of bleaching.

     Potash was a notable constituent of the solutions, as indicated by the abundant sericite and the smaller amount of orthoclase feldspar, the former more and the latter less plentiful than in the Kearsarge and Osceola lodes.

     Sulphides are more abundant in this lode than in the Kearsarge and Osceola lodes and indicate that in the later stages of the ore-depositing period, when subordinate open fractures were followed by the solutions, sulphur compounds were the stable form for copper deposition. It is not clear why hematite was destroyed on a large scale at the time when most of the copper was deposited and was later precipitated in minor amounts along with chalcocite. Arsenic as arsenical copper and arsenides may, like the sulphides, be products of the later stages of the period of mineralization, but the evidence on this point is meager and inconclusive.

     Throughout the second period of mineralization as described above, various rock elements were taken into solution as the rock was attached. The deposition of copper that bodily replaced the rock would also naturally force into solution such rock constituents as calcium, sodium, aluminum, and silica. It seems altogether probable that the gangue minerals formed by the mineralization, such as quartz, epidote, pumpellyite, sericite, calcite, and laumontite, represent in the main recombinations of these rock constituents in forms that were stable at the successive stages of the mineralizing period.

NOTES