A vein may be defined as an aggregation of mineral matter, more or less tabular or lenticular in form, which was deposited from solution and is of later origin than the inclosing rock. This definition differs from the one found in many textbooks in that no assumption is made as to the origin of the space occupied by the vein or the manner in which the mineral matter was deposited. In the present paper the writer purposes to discuss the mechanics of vein formation, and does not wish to begin by begging the question.

The origin of ore deposits has been diligently investigated in recent years, but, while much has been written concerning the chemistry of the process, the physical side of the question has received less attention. The source of the ore minerals; the chemical composition of ore-bearing solutions; the relative importance of magmatic and meteoric waters as agents of concentration; and the causes of mineral precipitation have all been discussed at length. On the other hand, little has been written concerning the mechanics of vein formation, and the theories given in textbooks today are essentially the same as those developed by the early Cornish and Saxon miners.

Early investigators often regarded both veins and dikes as igneous in origin, but today the theory that vein minerals have been deposited from solutions, either liquid or gaseous, is firmly established. The latter view is supported by the fact that many vein-forming minerals are not found in unaltered igneous rocks; by evidence obtained in the laboratory regarding the solubility of minerals; by the presence of many ore minerals in hot-spring waters; by the occurrence of some veins in regions consisting exclusively of unaltered sedimentary rocks; by evidence as to the temperature at which some minerals are formed; and by the absence of mineral segregation due to difference in specific gravity. There is still, however, much difference of opinion as to whether certain tabular-shaped masses should be classed as veins or as dikes. Some of the earlier investigators have also advocated the theory that veins are contemporaneous in origin with the inclosing rock, while others have believed the vein minerals to represent a recrystallization or transformation of the inclosing rock; but these views have now been generally discarded by geologists, and, since they are unsupported by evidence, it is not necessary to discuss them here.

Methods of Vein Formation

There are six different ways by which it is conceivable that mineral matter transported in solution may reach its resting place in a vein:

1. It may enter, along an open fissure, and be deposited on the walls until the fissure is more or less filled.
2. It may enter along a fracture, bedding plane, or similar passage, and, as it is deposited, force the wall rock apart, thus making room for the growing vein.
3. It may be introduced into an open fissure through small openings in the walls, and thus more or less fill the fissure.
4. If may be introduced through the walls of a narrow passage or incipient fracture, and, as it is deposited, force the walls apart.
5. It may enter along a fracture or other passage, and replace part or all of the minerals of the wall rock.
6. It may be derived from the surrounding material and be concentrated by locally replacing part or all of the minerals in the country rock.

Before estimating the relative importance of these methods of vein formation, it is necessary to establish criteria by which the origin of a given vein can be recognized. In studying the larger metalliferous veins having commercial value, it is sometimes difficult to discover evidence bearing on the mechanics of their formation, for only the final results are available for examination, and evidence that probably existed during the early stages of vein formation may have been largely obliterated by replacement or later alterations. The problem is rendered more difficult by the fact that many veins are the result of two or more methods of ore deposition.

In searching for evidence bearing on the mechanics of vein formation, the writer has not confined his attention to the large complex metalliferous veins, but has also made, a detailed study of the small and relatively simple veinlets of a region of unaltered sedimentary rocks.1 Natural deposits of mineral matter found elsewhere than in veins were also examined, and several different types of veins were successfully produced in the laboratory, where the origin and development of the more important vein structures could be observed in detail. A discussion of the six ways in which it is possible for veins to be formed, together with such evidence as has been thus far collected, is given below. The present paper is preliminary in nature, and some vein phenomena are not discussed. The investigation is being continued, but it is only through the collaboration of many observers that such problems can be solved.

The hypothesis that veins have been deposited from, solutions circulating along open fissures until filling was more or less complete is extremely old, and is to be found, in one form or another, in writings on ore deposits from the time of Agricola to the present day. Drusy cavities or vugs, and comb-structure, banding or crustification, are the evidences commonly cited in recent text-books as proof that veins were deposited in open fissures.

Drusy cavities are most common in veins recently formed at shallow depths, and, if we eliminate the openings due to solution in the belt of weathering, they are very rare, or entirely absent in veins formed at great depth. The gold-quartz veins of the Southern Appalachian region are believed to have been formed at depths of three miles or more; and, so far as the writer has been able to ascertain, they contain no vugs. If veins were formed through the filling of open fissures, there is no reason why vugs should not be as common in depth as in higher levels.

The close similarity in structure between drusy cavities and geodes, as emphasized by both Posepny and Shaler, is proof that open fissures are not essential for the development of the former. Shaler observed that in the process of enlargement, which geodes undergo, they sometimes condense and deform the inclosing rock strata. The texture of the vein matter surrounding vugs is often such as to indicate that the crystals lining the cavities have been deposited, as in geodes, from infiltrating solutions rather than from solutions circulating along an open fissure. In many instances the vugs are lined in part with minerals that are secondary in origin.

It has been demonstrated experimentally that pre-existing fissures are not essential for the development of drusy cavities; such openings are commonly present in veins, produced in the laboratory, which have made room for themselves by pushing their walls apart. A careful consideration of all the facts leads to the conclusion that the crystal lining of druses was deposited on the walls of open cavities, and that druses may be formed in veins that were deposited in open fissures, but the mere presence of these druses in a vein does not prove that the vein as a whole was thus deposited.

Werner seems to have been the first to call attention to banding and to explain it as being due to differences in the composition of the solutions from which the minerals were deposited; but Fournet observed that banding was not always parallel to the walls of a vein, since fragments of the adjacent rock were sometimes surrounded by layers of vein minerals, as, for example, in the “ring-ore” of the Hartz. De la Beche discussed the formation of comb-structure and of vugs or druses, and cited these phenomena as tending to prove that some veins owe their width to repeated reopenings of the fissures.

Posepny, in his treatise on “The Genesis of Ore Deposits,” devoted many pages to “the filling of open spaces,” and referred to banding or “crustification” as the characteristic feature of cavity-filling; but he acknowledged that the genesis of the non-crustified deposits was more difficult of determination, and concluded that “they were not deposited in pre-existing spaces.”

Veins of the latter class are, however, very much more common than those showing crustification. Sometimes banding is very imperfect, and again one part of a vein may show distinct crustification, while in other parts this structure is entirely absent. The crustified portion of some veins “is surrounded by solid quartz possessing no banded or comb-structure.” Banding, as well as vugs, may result from the filling of open fissures; it would then necessarily be symmetrical with respect to the center, but in many veins it is asymmetrical. Banding can be formed in other ways, however, and therefore it can not be regarded as proof of the manner of vein deposition. In veins produced in the laboratory, which were not deposited in open spaces, both symmetrical and asymmetrical banding were obtained by changing the composition of the vein-forming solutions. A banded structure may also result from replacement or from metamorphic processes.

When crystals are slowly enlarged by additions of material from supersaturated solutions, they develop crystal faces on their exposed surfaces; prismatic crystals, such as quartz, are commonly normal to some supporting surface from which they project, and therefore often have a parallel orientation. This is well illustrated by the crystals lining geodes, drusy cavities in veins, and similar openings in rocks; but in vein quartz crystal faces are relatively rare or entirely absent except in the immediate vicinity of drusy cavities, and microscopic examination proves that parallel orientation is likewise uncommon in massive quartz. There is unquestionably some recrystallization of vein minerals contemporaneous with their deposition, as a result of which certain crystals may be enlarged at the expense of others, but this would not change the crystallographic orientation. Crystal faces and parallel orientation are most common in veins formed at very shallow depths, and seem to be practically absent from the deeper veins. Veins formed at shallow depths occasionally show a spheriodal texture with prismatic crystals of quartz radiating outward from numerous central nuclei. Vein quartz usually has a granular texture similar to that shown by many igneous rocks, being due to mutual interference of the growing crystals. The tendency to assume a regular polyhedral form is stronger in pyrite than in any of the other minerals commonly present in veins, and yet pyrite crystals seldom show idiomorphic outlines except where they are embedded in older minerals which they have partly replaced.

If veins were formed through deposition of minerals on the walls of open fissures, there should be a suture line near the center of the veins, marked by occasional vugs where filling was incomplete; but most veins show no trace of such a structure; in many of the smaller veins individual crystals extend from wall to wall, and vugs, when present, are not always confined to the central portions of veins. Rickard has described and figured such a cavity which extended for several feet along the foot-wall of the Jumbo vein in the Enterprise mine of Rico, Colorado.

A soft gouge or selvage is often found along one or both walls of a vein, and, if this were present before the deposition of the vein minerals, it could never have remained in place on the walls of an open fissure even while supporting no additional load of vein minerals. Occasionally the gouge matter is found in the center of a vein.

The change in mineral composition when a vein passes from one rock to another has been attributed in several instances to substances present in the wall rock that act as precipitants; but this action could hardly occur if the vein-forming solutions were separated from the country rock by layers of impermeable mineral matter deposited on the walls of the fissure. Most veins are accompanied by some replacement and mineralization of the wall rock, but there is no evidence that the process of replacement is completed before fissure-filling begins. The mineral composition of some veins varies greatly along their strike; this is difficult of explanation under the assumption that the minerals were deposited from solutions filling continuous open fissures.

It is inconceivable that large fissures could have remained open at great depth below the surface during the long period of time required for their filling by mineral deposition. A a result of his ingenious experiments, Adams concludes that small cavities may exist in granite to a depth of at least 11 miles, but he distinctly points out that the walls of larger, openings would collapse under much less pressure. The latter fact has apparently been overlooked by some geologists. Data obtained from tests on small specimens of stone free from fractures and flaws can not be used in computing the depth at which fissures would be closed by collapse of their walls. In the vicinity of many veins the country rock shows evidence of having been greatly shattered prior to the entrance of vein-forming solutions, for branching veinlets sometimes extend out from the main vein and ramify in all directions. It is especially difficult to understand how open fissures could be maintained in rock that has been shattered to the extent shown by the “ horse-tail” structure of veins at Butte and in the Basin District, Montana.

The hydrostatic pressure of solutions occupying an open fissure would of course help to support the walls, but, on the other hand, there are factors that would aid in closing them. Deep-seated veins are formed under high-temperature conditions, which, together with the presence of mineralizers, would reduce the differential stress necessary for rock flowage. Veins are often far from vertical, and some are nearly horizontal.

It is only with great difficulty that openings are maintained in some of the deeper mines during the extraction of ores, and yet veins have been formed at very much greater depth than any reached in mining. Most of the deeper mines are practically dry in their lower levels, which indicates that even the smaller openings in rocks are largely confined to relatively shallow depths.

Most veins are found in mountain regions or in other areas that have been subjected to orogenic stresses, where, because of lateral pressure, gaping fissures are least likely to form; veins are relatively rare in regions of epirogenic movements where normal faults and monoclinal folds prevail. Moreover, the ore deposits that do occur in regions of important normal faulting, as, for example, at Clifton, Arizona, are not, as a rule, found occupying fault fissures. The presence of slickensides and stria on the walls of veins formed along faults are evidence indicative of closed rather than of open fissures.

The walls of many veins contain angular irregularities of such a nature that the opposite walls would fit perfectly into one another. Such veins evidently do not occupy fault fissures, nor can they be explained by any process of replacement. Many veins also split up and finger out into the country rock in such a way as to prove that the spaces were not formed by faulting.

Yawning chasms in any, way comparable in magnitude with the many barren and metalliferous veins are not known today at any place on the surface of the earth, and they are less likely to occur at the depths where most veins have formed; there is no reason, therefore, for assuming their existence in the past, for veins are formed in rock strata of all ages. Neither are partly filled fissures to be found, if the veins containing drusy cavities are excepted. Some veins are enormous in size; according to Suess, the Great Pfahl in the Bavarian Forest is somewhat over 150 km. in length and averages 70 to 115 m. in width. He states that this great quartz vein represents the infilling of a long cleft produced by dislocation. When we consider the multitude of veins, barren as well as metalliferous and the extremely rare occurrence of open fissures of any kind, it is difficult to believe that many veins have been formed through the filling of such openings.

In the Bendigo gold fields of Victoria, Australia, and also in Nova Scotia, large masses of auriferous quartz, known as saddle-reefs, occur in the crests of anticlinal folds, but open spaces, with the exception of solution caverns in limestone, have never been observed in anticlinal folds, although folded rocks in all stages of dissection are found in many parts of the earth. Simple filling of open fissures of tectonic origin does not explain a change in the width of veins on passing from one wall rock to another, such as occurs in the veins of the San Juan Mountains, Colorado.

The lenticular form of veins is also evidence opposing the hypothesis of deposition in open fissures. Most veins are more or less lens shaped; this form is often accentuated, and in extreme cases the lenses are bulbous or almost spherical. Sonie veins alternately swell and pinch; some consist of small separate lenses strung out along the same line of strike and connected by an almost imperceptible seam; others are made up of small and apparently disconnected parallel lenses irregularly distributed along a narrow belt or zone. Sometimes the lenses are remarkably symmetrical (Fig. 5); often they are irregular. The lenticular form is usually best developed in veins formed at depth, or where the wall rock is a highly laminated schist, slate, or shale. The walls are usually sharply defined and there is often other evidence that replacement has not been an important factor in the formation of the lenses.

De la Beche showed that the relative displacement of the walls of an irregular fracture could result in the formation of open spaces which thin out where the opposite walls are in contact, and this has frequently been advanced as an explanation of lenticular veins; but no method of faulting can account for the space occupied by apparently disconnected lenses, and, in some instances, there is positive evidence that the walls have been separated without other displacement. Billingsley and Grimes have recently described short lenticular veins in granite which show no evidence of faulting, “such as shearing in the intervening country rock or displacement of intersected dikes of aplite.” They attribute the drawing apart of the walls to tension resulting from the cooling and contraction of the granite; but the distribution of the veins is not such as to relieve tensile stresses in all directions in a shrinking mass, many granites show no evidence of lenticular openings either filled or unfilled, and lens-shaped veins are more common in sedimentary and metamorphic rocks than in igneous rocks.

The bedding planes of sedimentary rocks and the folia of schists and slates, where in contact with lenticular veins, are commonly curved to conform with the curvature of the lenses. The later origin of the veins is indicated by the fact that their occasionally transect the rock structure and contain included fragments of the wall rock. Such displacement of structural features antedating the formation of the vein can be explained only on the theory that the walls have been pushed outward by some force acting from within, for the effects of tectonic forces acting parallel to the rock structure could not be limited so as to confine the formation of lenticular openings to a single plane.

Lenticular veins were believed to be igneous intrusives by some of the earlier geologists, but true dikes do not show such extreme forms, and the theory is also open to many other objections previously cited. Graton concluded that the lenticular gold-quartz veins of North and South Carolina were deposited from solutions which were under such great pressure as they ascended toward the surface that they were able to force apart the walls of small fractures and thus form the lenticular openings.

This explanation was advocated by Lindgren who opposed the theory that the veins were formed through the “force of crystallization.” If the enlargement of vein spaces were due to the pressure of fluids, very flat lens- or tabular-shaped openings would be formed, and not a series of thick lenses more or less isolated and disconnected; moreover, the pressure of the fluid would have to be uniformly maintained until the openings were filled, or the walls would collapse. It is obvious that this

theory is inapplicable to many lenticular veins, such as those of calcite or gypsum, formed near the surface in unaltered sedimentary strata. The pressure of fluids cannot account for the formation of lenses that are subdivided by thin partitions of included schist detached from the walls, as illustrated in Fig. 1. In view of all the facts, it must be concluded that the typical lenticular veins were not deposited in preexisting openings.

The peculiar banded structure, sometimes referred to as “book structure” or “ribbon-ore,” resulting from the alternation of narrow veinlets of ore with parallel sheets of slate or schist, is difficult or impossible of explanation under any theory of ore deposition in preexisting openings. Becker and Day have called attention to the fact that in some cases “it can be shown conclusively that the distribution of the slate is not due to faulting.” This structure is well developed at many places along the Mother Lode of California, and also in some of the gold and copper veins of the Southern Appalachians. Usually the veinlets, consisting of quartz or calcite, are lenticular in form, and pinch out to be replaced by others a little to one side or farther along the strike; sometimes layers of slate are wholly included within the veins.

The commonest evidence indicating that veins were not entirely deposited in open fissures is furnished by the presence in them of angular inclusions of the country rock. These fragments often show by their shape that they have been detached from an adjacent wall and displaced laterally without falling. Where the inclusions are unaltered by replacement, they could refit, perfectly into the positions which they formerly occupied. Inclusions that have suffered alteration are usually more or less rounded or subangular. Posepny states that upon search he always found points of contact between such fragments; but, in many instances, the writer, by making parallel sections, proved that the inclusions were not in contact with the walls or with other fragments.

Inclusions are rare in some veins and plentiful in others, in some cases making up the larger portion of the ore mined. The inclusions also vary in size, up to masses so large that it is a question whether they should be classed as “horses” or as country rock inclosed between two branches of a vein. There is no essential difference between some breccia veins and certain ore deposits that consist of a network of anastomosing veinlets which ramify in all directions through the country rock. In veins that are horizontal or dip at flat angles, the inclusions are no more plentiful along the foot-wall than near the hanging-wall.

Deposition of mineral matter in opening filters with gas is practically limited to the belt of weathering. Such deposits are, as a rule, easily recognized; they are often finely banded, and usually exhibit stalactitic, mammillary or analogous forms. As they are almost exclusively secondary in origin and obviously of little importance in unaltered veins of primary ores, further discussion is unnecessary here.

The evidence outlined above leads to the conclusion that deposition of mineral matter in widely open fissures is of minor importance in the formation of veins. Minerals deposited in open spaces are present in many veins, especially those formed at shallow depths, though in most cases they make up a relatively small proportion of the total vein mass. Some veins have possibly been formed entirely through the simple filling of open fissures, but they are believed to be extremely rare.

The hypothesis that veins may be deposited from solutions entering along extremely, small passages and that the growing veins have made room for themselves by forcing apart the inclosing walls has been suggested by investigators from time to time; but it,has never received general acceptance, and is practically ignored in text-books treating of the origin of ore deposits. Comparatively little evidence in support of the hypothesis has been published, and the mechanics of the process by which growing crystals exert pressure has not been clearly understood. According to Andree, who has compiled an excellent bibliography of the subject, von Weissenbach was the first to assume a “force of crystallization ” in explaining geologic phenomena. Lavalle, however, appears to have been the first to record experimental proof of pressure exerted during crystal growth.

It has been demonstrated beyond question that growing crystal may exert force and actually grow in directions in which their growth is opposed by adjacent solid bodies, but some geologists have doubted that this force is sufficient to separate the walls of veins at great depths. In replying to this objection, attention is called to the fact that all crystals have been formed under pressure, and many mineral crystals must have developed under enormous pressure. The effect of pressure or of strain is to increase slightly the solubility of most substances, but the change is very small as compared with that due to difference in temperature. So long as a crystal surface is in contact with a solution supersaturated with respect to that surface, growth will continue no matter how great the pressure or strain may be. The supersaturation of a solution in contact with a growing crystal is maintained by circulation and diffusion through the solution; and, when a crystal grows in a direction in which growth is opposed by an adjacent solid, it is due to the fact that the material necessary for growth is able to diffuse between the crystal and the other solid.

The film of solution separating the crystal from the other solid is not expelled by the pressure developed; otherwise, growth would cease. If two perfectly smooth parallel surfaces were separated by a thin layer of liquid which wets them both, capillarity would tend to bring the two surfaces close together, and reduce the thickness of the separating film to a minimum; but it is improbable that this film could be completely expelled by capillarity or even by external force that does not rupture the solids. When a crystal, in growing, approaches another solid, the surfaces that, are closest together tend to become parallel, because deposition is most rapid where diffusion is least restricted.

There is abundant evidence that solutions, especially when under great pressure and high temperature, can penetrate even the densest rocks. Penetration takes place between the mineral crystals and along cleavages and microscopic fractures in the minerals themselves. It has been suggested that diffusion may take place directly through individual crystals, but this is doubtful, and certainly has not been proved. Evidence of the enormous quantities of mineral matter which may diffuse for considerable distances through small capillary and sub-capillary spaces is furnished by the size of replacement deposits. The presence of ore minerals having idiomorphic surfaces, where in contact with older minerals which they have partly replaced, proves that in such instances the solution and removal of the replaced minerals has resulted from the growth of the ore minerals, and therefore did not precede the precipitation of the latter so as to form open spaces for their reception.

The pressure effects accompanying the growth of some crystals are evidently, in part at least, due to the tendency to develop crystal faces, but this “linear force of growing crystals” is believed by the writer to be of minor importance. In previous papers he has expressed his belief that these pressure effects are to be attributed chiefly to the molecular forces associated with the separation of solids from solution.

Laboratory experiments prove that the pressure effects are independent of the cause of precipitation; for they have been obtained where precipitation was induced by cooling, by evaporation, by reactions between liquids, by reactions between liquids and solids, and by reactions between gases and solids. When a single molecule of the solid passes by diffusion into the thin film of solution separating a growing crystal from a foreign body, and attaches itself to the surface of the crystal, a slight displacement of the entire crystal relative to the foreign body results. According to this conception, the mechanism of the process is somewhat analogous to that of a hydraulic jack. It is inconceivable that the addition of a single molecule could displace the crystal in any other way, and there is no evidence of periodicity in the growth of most crystals. The diffusion of a solid through a solution is ascribed to osmotic pressure, and its separation therefrom to the relation between osmotic pressure and solution pressure. The writer thinks it preferable not to use the term “force of crystallization” until it has been definitely established that the pressure effects accompanying the separation of solids from solution are limited exclusively to crystalline solids. There is some evidence that the deposition of non-crystalline substances, such as limonite, may also result in the development of presssure under favorable conditions. This question is now being investigated experimentally, but as yet with only negative results.

Lindgren thinks that a platy or schistose structure due to development under one-sided pressure would necessarily characterize veins which have made room for themselves by forcing their walls apart, and that the absence of this structure is evidence indicating that the veins were not so formed. There is, however, much evidence opposing this view. A crystal developing at depth below the surface, completely surrounded by interlocking minerals, is probably under nearly uniform pressure in all directions. Pseudophenocrysts of biotite which have developed in highly schistose rocks under, mass static conditions do not show parallel orientation, although in growing they have forced apart the folia of the schist. Calcareous concretions which develop in shales and push apart the bedding-planes, as shown in Fig. 4, may be even granular in texture. The writer has examined, microscopically, thin sections cut normal to the surface of these concretions and others cut parallel to the surface without observing any difference in the appearance of the calcite grains. In some concretions that are relatively free from impurities the calcite crystals are elongated normal to the surface.

Even in cases where growing crystals are subjected to much greater pressure in one direction than in others, they are not necessarily flattened normal to the direction of greatest pressure, for the shape of crystals is often determined by the relative accessibility, in different directions, of the material for growth. This has been proved experimentally by growing slender columnar, and even fibrous, crystals which were elongated in the direction of greatest pressure.

In laboratory experiments conducted during the last four years, the writer has been successful in growing veins which have made room for themselves by pushing apart their walls; and in these veins such structures as drusy cavities and banding have been obtained. The drusy cavities were due to a local deficiency of the material for vein growth, resulting either from insufficient concentration or from relative inaccessibility; and evidence has been elsewhere cited in support of the hypothesis that at least some of the drusy cavities occurring in mineral veins are similar in origin.

Banding was produced in laboratory veins by merely changing the composition of the solutions during the growth of the veins. Mineral veins showing asymmetrical banding could have been formed in open fissures only through successive reopenings of the fissure, and usually there is no evidence supporting such an assumption. Asymmetrical banding, as demonstrated by laboratory experiments, may develop in growing veins that make room for themselves whenever the circulation of the vein-forming solutions is limited to a single wall. There is much evidence indicating that the banding found in metalliferous veins is often due to deposition from solutions circulating along the walls of the growing veins, and that the minerals deposited last are sometimes concentrated near the walls rather than in the center of the veins, as would be necessary on the assumption of ore deposition in open fissures.

Many large veins show a regular arrangement of quartz in the center and chalcedony near the walls. On the assumption that such veins are referable to the activity of hot-spring waters, Beyschlag, Vogt and krusch attribute this regular arrangement to the cooling effect of the walls on the ore-bearing solutions, for, as is well known, quartz usually separates from solutions at a higher temperature than chalcedony. It is just as logical, however, to assume that the vein-forming solutions gradually became cooler because of the slow cooling of the igneous rocks from which the heat was derived.

The expiring stages of igneous activity are often marked by the formation of veins through deposition from hot ascending solutions, and in such veins the commonest gangue minerals are quartz and calcite. Similar deposits are formed about hot springs where the waters reach the surface. Calcite is usually deposited later than quartz, but in some instances this order is reversed, and then there is more or less replacement of calcite by quartz. According to Spurr, the calcite veins of Ararat Mountain in the Tonopah District are in places beautifully banded, with quartz in the center and carbonates near the walls. These veins are locally as much as 20 ft. thick, but are exceedingly irregular and non-persistent. The rhyolite forming the walls is in places silicified by the vein-forming solutions, but there is no mention of replacement of calcite, and in many cases the carbonates were observed occupying cavities in the silicified rhyolite, thus indicating a later deposition for the carbonates. Spurr states that “these are fine examples of veins which have filled open fissures,” but the facts cited above would indicate that the veins had made room for themselves by forcing their walls apart, and this hypothesis is also supported by the presence of numerous angular rhyolite fragments within the veins.

Small roughly banded veins with calcite in the center and later pyrite along the walls have been described by the present writer. The pyrite is partly in the form of idiomorphic crystals which replace both the earlier vein calcite and the wall rock.

An irregular and more or less interrupted banding, as in specimens of ore from Silverton, Colorado, is much more common in veins than an evenly banded structure such as is usually found in travertine, agate, and other deposits known to have been formed in open spaces. Very imperfect banding or absence of banding may result when mineral deposition, instead of being confined to the walls, takes place partly or entirely within the growing vein, the later minerals making room for themselves by replacing or mechanically displacing the earlier minerals. The later minerals may be localized, through deposition along well-defined channels of circulation, or they may be deposited throughout the vein mass. Where the latter occurs, the material for vein growth must be supplied, at least partly, by diffusion between the earlier minerals. There is much evidence that this method of mineral deposition is of the utmost importance in the formation of ore deposits.

Many veins are intersected by numerous small but well-defined veinlets, consisting chiefly of the later minerals, though often the growth of these veinlets has been practically contemporaneous with the formation of the vein as a whole. Minute veinlets are often revealed by microscopic examination of ore when their presence would not be suspected from a megascopic examination. These microscopic veinlets frequently cut directly through the older minerals, and separate the fragments without other displacement. Occasionally the veinlets may be traced considerable distances, but usually they are non-persistent, and often they are confined to the limits of a single crystalline individual. When thus limited, the veinlets sometimes follow crystallographic directions in the older minerals.

The deposition of mineral matter within a growing vein is not, however, limited to intersecting veinlets, and there is also evidence of more or less contemporaneous recrystallization of some minerals during the enlargement of veins. When very small veinlets are examined microscopically, it may be observed that the individual mineral crystals, usually extending from wall to wall, tend to vary in size with any variation in the thickness of the veinlets. This can be explained only on the assumption that the enlargement of the veinlets has been accompanied by an enlargement of some of the vein minerals and the elimination of others. In other words, the crystals which are less stable because of smaller size, or

unfavorable orientation or location, tend to redissolve, and thus furnish additional material for the growth of their more fortunate neighbors.

Evidence of mineral deposition within growing veins is furnished by the separation of included fragments of the country rock from one another as well as from the walls. In such cases the fragments often show by their shape that they would fit perfectly together. Inclusions are occasionally found in most veins formed by simple deposition unaccompanied by appreciable replacement, and in some veins the inclusions make up a large percentage of the vein mass. Where the inclusions are abundant, they are sometimes rather uniformly distributed (Fig. 2), as in the so-called “bird’s-eye” quartz found in gold veins of the Swauk district of central Washington. The angular inclusions are of black shale, much silicified, while the matrix is of quartz and calcite. The quartz is partly in the form of small prisms which radiate outward from the separated fragments. At their ends some of these prisms mutually interfere with those radiating from neighboring inclusions, while others project into drusy cavities. According to Smith, the wall rock is in places much shattered, with many small quartz veins traversing it in all directions, so that it is difficult to draw any limits to the vein itself, and there is a transition from this shattered wall rock into the typical “bird’s-eye” quartz. Russell described these veins, and suggested that the vein minerals “in crystallizing have exerted a force analogous to the expansion of water in freezing, which crowded the rock fragments asunder.”

Von Cotta discussed breccia structure in veins, and stated that in soine cases it seems to be due to a special force of crystallization of the principal vein mass, while in others it is to be explained by the opening of vein fissures, more than once. The latter explanation, however, is obviously inapplicable to such veins as those, described above in which the spehrulitic texture is well developed with quartz prisms radiating in all directions from the nucleal fragments. In some veins having a spherulitic texture the nuclei of the spheroids consist of earlier vein minerals instead of fragments of the country rock. This is well illustrated by some of the veins in the Telluride and Silverton districts. Less frequently the nuclei are surrounded by several layers or concentric shells differing in mineral composition. This texture, variously known as “cockade ore,” “ring ore” and “concentric ore,” was described by von Weissenbach as early as 1836. He noted that there was no contact between the original fragments, and attributed their separation to the crystallizing force of the later minerals. The spherulitic textures seem to be limited to veins formed at comparatively shallow depths.

The laboratory veins, which made room for themselves by forcing their walls apart, were sometimes branching, and frequently contained inclusions of the wall material—a structure which could not be duplicated experimentally by any method of vein deposition in open spaces.

In a manner similar to that described above, certain vein minerals of early deposition are commonly ruptured, and the fragments separated by the growth of later minerals. This is well illustrated by many tourmaline crystals found in quartz veins, and also by pyrite and other sulphide minerals present in ore deposits. Pyrite is usually one of the first sulphide minerals deposited, and microscopic examination shows that the pyrite crystals are commonly ruptured and the fragments separated by later minerals, such as chalcopyrite, bornite, etc. Graton and Murdoch, after an extensive investigation, state that

“this is the characteristic structure of the average primary copper ore. In such ores there can be no doubt that the pyrite has undergone actual mechanical deformation, for in some instances it is found that somewhat separated fragments would fit perfectly together, even to the minor irregularities, and would collectively form a compact grain if the intervening cement were removed. Such an occurrence gives the appearance of an exploding bomb. The cause of this crushing and distortion is not yet clear, but it is plain that it took place in an early stage of crystallization of the ore, because no trace of such a thing is commonly to be found in the later sulphides or in the gangue, in the latter of which it would be expected that crushing or at least strain-shadows would be seen in thin sections if these materials had been subjected to stress.”

Emmons discussed this structure of sulphide ore deposits, and attributed it to regional metamorphism, but it is also a peculiarity of deposits found in non-metamorphic areas, and this peculiar structure is not characteristic of any type of metamorphic rock. The ore deposits described by Emmons are, it is true, situated in a belt of intensely dynamo-metamorphosed rocks, but the orebodies, while roughly lenticular in form and elongated parallel to the rock structure, are very wide in comparison with their length. Emmons, in commenting on this fact, states that “as a class they seem to be wider than deposits not metamorphosed. This is just the opposite of what should be expected, for, from internal evidence, one gets the impression that they have been squeezed thin.”

The separation of fragments of pyrite and other early minerals is evidently due to the growth of the later minerals; but the manner in which the fractures were first formed is not always clear. In some cases the fracturing of crystals appears to have resulted from external pressure, often due to the growth of later minerals, and in other cases it is probably caused by the development of later crystals in the small cavities present in most crystals or along, planes of incipient cleavage or parting. The latter view is supported by the presence in some crystals of dots and stringers of the later minerals arranged chiefly in parallel lines along crystallographic directions.

While some minerals found in veins are commonly fractured and the fragments separated, this structure appears to be rare in other minerals, such as galena, having equal or greater porosity and cleavability. This difference can not always be attributed to later crystallization. In some cases it is possibly due to a difference in the surface tension existing between different minerals and the ore-bearing solutions, but no definite conclusion can be drawn until more data are at hand.

Minerals of early crystallization are occasionally bent as well as fractured, and then the evidence of deformation by external forces is conclusive. Emerson has described large crystals of spodumene apparently bent, and in some cases broken, by, the growth of later minerals, for the vein in which they occur is not crushed or sheared. The forces that bent the large spodumene crystals through angles of 45° and 90° without fracturing must have been applied gradually through a considerable period of time while the crystals were imbedded in a matrix of other minerals. Bent and broken crystals of covellite; the intervening spaces being filled with “later, but nevertheless primary, bornite,” have been described by Graton and Murdoch. The flakes of mica and chlorite found in some veins are often bent and shredded by the growth of sulphides.

The hypothesis that veins have made room for themselves by forcing their walls apart furnishes a ready explanation of lenticular form and of “book structure” or “ribbon ore,” features characteristic of veins formed in highly laminated rocks, and these features are difficult or impossible of explanation under any other theory of vein formation. According to this view, “book structure” has resulted where the ore-bearing solutions have entered along closely spaced parallel planes, and deposited mineral matter in the form of veinlets, which, in growing, have gradually separated the laminae of the country rock. Becker and Day were the first to suggest this origin for the so-called “ribbon-ore” of the Mother Lode district in California. The separation of the silicified shale fragments in the ore breccia forming the Enterprise blanket of the Rico district, Colorado, should probably be attributed to a similar process. Ransome has reproduced a photograph of a beautiful specimen of this breccia from the New-man Hill mine.

Dunn states that the saddle-reefs and leg-reefs of the Bendigo district, Australia, are often laminated by the inclusion of thin films of slate and occasionally of “thick substantial flakes.” The laminations are farther apart where the reefs are widest, and closest together where the reefs are narrowest. The laminations are also more closely spaced near the walls. Dunn cites these facts as evidence that the quartz has forcibly made room for its accumulation, and concludes that the growth of the quartz layers was continuous during the period of reef formation.

Lenticular form is characteristic of most aggregations of mineral matter that develop in highly laminated rocks, such as shales or schists, and displace rather than replace the older minerals. Eye-like lenses may result from the growth of a single crystal, in which case the outer cir-

cumference of the lenses is filled in with other minerals, as in Fig. 3. The single crystal may be of such mineral as garnet, biotite, pyrite, galena, or sphalerite, while the remainder of the lens is filled, with quartz or calcite. Larger lenses are formed by the growth of crystalline aggregates. The calcareous concretions shown in Fig. 4 are essentially similar to the gold-quartz lenses shown in Fig. 5, and in both cases the form has obviously resulted from similar processes. In a previous publication, the writer has cited the lenticular form of certain auriferous-quartz veins as evidence that the vein minerals had made room for themselves by forcing the walls apart. One of these veins, exposed in the Tellurium mine, Virginia, consisted of lenses which were in some cases wonderfully symmetrical in form. In laboratory experiments, lenticular veins were obtained through the slow growth of crystalline masses of salts between layers of heavy cardboard coated with paraffine.

According to this hypothesis, the veins have been deposited from solutions which have penetrated slowly along lines of least resistance, such as faults, fractured zones, planes of bedding or schistosity, or through porous strata; where appreciable openings existed, these have been filled, but the

veins have chiefly made room for themselves by forcing apart the inclosing country rock. The form assumed by the growing deposits is determined by the nature of the opening along which the solutions enter, by the accessibility of the ore-bearing solutions during the growth of the ore-body, and by the forces resisting enlargement. When the material for growth is everywhere equally available, the shape produced is that which requires the least expenditure of energy. This furnishes an explanation of the swelling of some veins where they pass into less rigid rocks, and of the growth of “saddle-reefs ” in the crests of anticlinal folds where the rigidity of arching strata relieves the underlying beds of part of their load.

After carefully considering all the facts, there seems to be no escape from the conclusion that it is possible for growing veins to make room for themselves by forcing their walls apart; and the evidence indicates that many veins have been formed in this manner.

That veins have been formed of material leached from the inclosing rock, transported in solution and introduced into open fissures through their walls is the old lateral-secretion hypothesis advocated by Bischof, Sandberger, and others. It was vigorously opposed by Posepny, and has been abandoned, in its original form at least, by practically all geologists. Much evidence against the hypothesis has been recorded by others and need not be repeated here. It is also open to the same objections as all hypotheses postulating gapping fissures as a prerequisite for the formation of veins.

In its narrower form this hypothesis confines the derivation of the vein minerals to the wall rock in immediate contact with the deposits, and, therefore, is especially applicable to such veins as the calcite veins in limestone and calcareous shales, gypsum veins in gypsiferous strata, quartz veins in siliceous rocks, and chrysotile veins in serpentine. The veins belonging to this class are commonly horizontal or nearly so, and frequently ramify in all directions through the country rock so as to make up a large proportion of the total mass; hence the difficulty of explaining the formation and maintenance of open fissures is even greater for such veins than for others.

Laboratory investigations indicate that two different types of deposits may be formed in open spaces when mineral matter is introduced through small openings in the walls. One type is formed when mineral matter is deposited from solutions subsequent to their entrance into cavities, in which case the deposits are similar to those formed in open spaces by circulating solutions. The minerals deposited on the inside of geodes, and also in many of the drusy cavities of veins, have evidently been deposited in this way, though there is no evidence that large veins have had such an origin. In the belt of weathering, the deposits usually show stalactitic or analogous forms.

The other type is due to the deposition of mineral matter from super-saturated solutions which do not enter the cavities; In this case additions are made to growing crystals only at their bases, where they are attached to the walls, thus making them columnar, acicular, or fibrous. Because of inequalities in the supply of material, the resulting fibers are unequal in length and in places entirely absent. Where the fibers are in close contact with one another, they are usually parallel and normal to the surface from which they grow; where the fibers are isolated or in small groups, they are commonly curved, and sometimes form tangled coils. Thin crusts are often formed on the walls and later pushed outward by the growing crystals, which also occasionally detach fragments of the wall material and displace them in a similar manner. Mineral deposits of this type may sometimes be seen forming on the walls of caverns and in other favorable places. Vugs and smaller openings found in veins often contain minerals, usually secondary in origin, which exhibit this structure; but there is no evidence of vein fissures having been filled with such deposits. The filiform varieties of the native metals have probably developed in this way, and some intermediate forms may be due to a modification of wire-like varieties by solutions which have occupied the cavities.

That the material for vein formation has been supplied through small, closely spaced openings in the walls, as the growing veins have made room for themselves by forcing their walls apart, is the hypothesis advocated by the present writer, in previous papers, as offering an explanation of cross-fiber veins. Most of the evidence supporting this view has been published, and therefore it need not be discussed in detail here.

A somewhat similar explanation of the origin of certain fibrous veins was suggested by von Weissenbach in 1850, but this was not known to the writer until after the publication of the papers referred to above. Von Weissenbach described the growth of fibrous gypsum crystals on the walls of old buildings and of the needle ice that forms on clayey soils, comparing it with the formation of the fibrous veins. He thought that vein walls of soft material might be pushed back, or that the growth of the needle-like crystals was simultaneous with the widening of the vein fissures which resulted from shrinkage due to drying. As evidence in support of this hypothesis, he states that fibrous veins are never persistent and never show banding or drusy cavities; but this, statement is not altogether true, for color banding parallel to the walls is sometimes present in veins of fibrous celestite, and drusy cavities have been observed in similar veins of calcite and of gypsum.

Cross-fiber veins are composed essentially of a single mineral, which is also usually an important constituent and often the dominant mineral in the wall rock. For example, there are fibrous veins of quartz in sandstone, of calcite in limestone, and of chrysotile in serpentine. In rare instances the vein mineral is not found in the wall rock, but in such cases it is a secondary mineral which has evidently been derived from minerals that are present in the wall rock. Veins of fibrous chalcanthite an inch in thickness have been found in the Yavapai mine at Morenci, Arizona.

All cross-fiber veins, irrespective of mineral composition and kind of wall rock, are characterized by the same structural peculiarities. The mineral fibers are parallel to one another, and in many instances there is direct evidence that they extend in the direction in which the walls have been displaced during the process of separation. In most veins the fibers are approximately normal to the walls, but frequently they are oblique, and, if the course of the vein is not straight, the fibers may be normal to the walls at one place and oblique at another. This is well illustrated in specimens from Hoboken, New Jersey, showing nemalite veins in serpentine. The fibers are usually straight, especially when the veins are small, parallel, and not very numerous; often they are curved or abruptly bent, and this may be observed most frequently where the veins branch, intersect, and ramify in all directions. Sometimes there are several bends in the fibers, thus giving the veins a banded appearance due to unequal reflection of light where the fibers run in different directions, and ordinarily this pseudo-banding is symmetrical with respect to the central portion of the veins. A central parting is present in some veins, several partings are present in others, and in many veins the fibers extend from wall to wall without a break. The partings are often very irregular, appearing as though displaced parallel to the fibers, and similar irregularities may occasionally be noted in the bands that run parallel to the vein walls. Angular inclusions are common, especially along lines of parting, and in many instances it is evident that these fragments have been merely detached from the adjacent walls and displaced laterally. Fibrous veins as a class are non-persistent, and usually have a marked lenticular form.

The structural features here given are common to all kinds of cross-fiber veins; and, with the exception of lenticular form, all of these features have been duplicated in veins grown in the laboratory. All of the features may be satisfactorily explained under the hypothesis outlined above, while many of them are inexplicable under any other hypothesis so far advanced.

The formation of veins through lateral secretion is not necessarily confined to the zone of vadose circulation, as argued by Posepny and Raymond, for it has been proved that the vein minerals may be transported to the walls of the growing vein by diffusion through solutions occupying minute spaces, instead of by circulation.

The fibrous structure has been attributed by the writer to a mechanical limitation of crystal growth through the addition of new material in only one direction. In veins of the asbestiform minerals, the fibrous structure is accentuated by a normal prismatic habit and cleavage. In studying cross-fiber veins of calcite and of gypsum, it was found that the diameter of the fibers varied with the spacing of the openings through which the material for growth was supplied, and that the veins were non-fibrous where they passed through relatively impervious chert nodules which forced the vein-forming material to diffuse between the walls. Therefore, if material reached a growing vein through openings in the walls, spaced sufficiently far apart, it would be impossible to determine from the structure whether the vein had been formed through a process of lateral secretion or not. A gradation is, therefore, possible from fibrous veins to non-fibrous veins of class II. The fibrous structure of some veins may be made coarser or even completely obliterated as a result of recrystallization, since, for many minerals, a fibrous form is not the most stable.

It has been suggested elsewhere that minimum resistance to growth may often be the factor which has determined the location of chrysotile veins in serpentine, and there is considerable evidence that cross-fiber veins in general tend to develop in those places where the least expenditure of energy is required in making room for them. The cross-fiber veins in stratified rocks are commonly parallel to the bedding, especially where the beds have not been greatly disturbed; and in folded strata the veins in places may be largely confined to the crests of anticlines. These facts are well illustrated by the crocidolite veins of West Griqualand, South Africa, which are briefly described below. For this description the writer is indebted to Dr. A. W. Rogers, Acting Director of the Geological Survey of the Union of South Africa.

The cross-fiber crocidolite is interbedded with cherty and magnetite layers, as a rule, but it breaks across these beds where the latter are bent and broken, though the amount of cross-fiber mineral obtained from these cross-fractures or veins is quite insignificant. Where the beds are sharply folded the cross-fiber mineral thickens very much in the arches and troughs, and it may disappear completely in the connecting limbs The folding and fracturing of the beds was completed before the crocidolite assumed the cross-fiber form; I have not seen the fiber disturbed by these processes, and it appears that its formation took place very soon after, or during, the time when the beds were disturbed Minute fibers of crocidolite are distributed, more or less abundantly, through the sedimentary beds lying between the layers of cross-fiber mineral.

The formation, of veins through replacement of country rock by ore minerals, introduced in solution through small cracks, was suggested by Charpentier as early as 1778. Little evidence in favor of the hypothesis could be cited at that time, and therefore it received almost no attention for nearly a century. Facts learned from the study of pseudomorphs were gradually applied to the formation of ore deposits, and, as evidence accumulated, it became evident that metasomatism or replacement is an important factor in the formation of ore deposits. In 1910 Lindgren published his essay, “Metasomatic Processes in Fissure Veins,” in which he described in detail the various changes that take place in wall rock as a result of vein formation, and suggested a classification . of veins based on characteristic metasomatic processes. It is now generally recognized that while the wall rock of many veins shows absolutely no evidence of metasomatism, most of the larger metalliferous veins, on the other hand, are bordered by altered zones, varying in width and intensity of alteration, and some veins have been formed almost entirely through processes of replacement.

Criteria for the recognition of replacement have been given by Lindgren, Irving, and others, and therefore a discussion of this phase of the subject may be omitted here. In regard to the mechanics of replacement, the views of the present writer are slightly at variance with those of the authorities just cited, and it is this question that is discussed in succeeding pages.

The changes which take place in wall rocks as a result of vein-forming processes are often complex, for most rocks are composed of several minerals differing in chemical composition and physical properties, while the ore-bearing solutions may also contain many constituents. The problem can be simplified somewhat, however, by considering separately the various ways by which changes may take place in each individual mineral of which the rock is composed.

One mineral may take the place of another in at least three different ways: (1) by solution of the original mineral and subsequent deposition in the space vacated; (2) by solution of the original mineral and simultaneous deposition of the replacing mineral without chemical reaction between them; (3) by a chemical reaction involving the addition, subtraction, or interchange of one or more elements.

(1) While minerals are sometimes deposited in spaces of dissolution, this process is believed to be of little importance in the formation of veins, and, strictly speaking, it is not a process of metasomatism. With such a method of replacement it would be impossible for the replacing material to inherit and preserve details of internal structure, regular arrangement of inclusions or the structure of organic remains. It would also be impossible for the replacing mineral to show idiomorphic boundaries, where in close contact with the partly replaced mineral. Microscopic examination of minerals that are partly altered or replaced always shows the replacing mineral to be in close contact with the earlier mineral, and, even under the highest magnification, no trace of intervening space is observable.

It has been suggested that mechanical deposition may follow so closely after dissolution as to make the two processes appear as one, but it is highly improbable that two processes could be independent and at the same time so closely synchronized. Such results might be obtained, however, if solution and deposition were interdependent, and therefore simultaneous; i.e., if solution of the replaced mineral induced the deposition of the replacing mineral, or if deposition of the latter forced the former into solution.

(2) Mineral crystals in contact with supersaturated solutions must grow, and, if surrounded by other minerals, the latter will be mechanically displaced, or, if they are rendered more soluble by pressure, they may be removed in solution and deposited elsewhere. The mechanical displacement of individual minerals in a compact rock would necessitate enormous forces; the pressure per unit area would have to be very much greater than that required in separating the walls of a growing vein. The solubility of nearly all substances, so far as known, is increased by pressure and strain, and this appears to be especially true of quartz and calcite, minerals that are common in rocks subject to replacement. Therefore, when crystals develop in compact rocks, the necessary space is usually obtained through solution rather than mechanical displacement.

In his paper on “The Nature of Replacement,” Lindgren draws a distinction between replacement in rigid rocks and crystallization in yielding substances. He acknowledges that in soft material “the stresses produced by crystallization may become of direct importance,” but he does not believe that these forces are sufficient to make room for new crystals “in a rigid medium like quartz.” In support of this view he states that “the general absence of fracturing or breaking of the host mineral is evidence enough; even optical anomalies and undulous extinction are uncommon around crystals developing in comparatively rigid minerals.” Under his summary of previous views, Lindgren refers to the occurrence of clear gypsum crystals in clay as proof of a “force of crystallization,” and then on the other hand cites “instances where the growing crystal has failed to exert this mechanical force, as in the well known calcite crystals in the sand of Fontainbleau, which include so much foreign material as to appear like crystals of sand.”

Now these apparently contradictory phenomena are susceptible of direct experimental investigation; and the laboratory experiments of the writer demonstrate that mechanical resistance to crystal growth is only one of several factors determining whether foreign material be included or excluded by a growing crystal. The size of the pore spaces in the inclosing material is usually the controlling factor; and, in some cases perhaps, the surface tension between the solids and the solution is also of importance.

When a growing crystal is surrounded by foreign material in which the pore spaces are large, as for example in sand, the pressure is unevenly distributed over the crystal surface, growth is relatively rapid on the exposed areas while there is insufficient time for diffusion to supply additional material to the surfaces that are in contact with the foreign bodies. Therefore, the latter are gradually surrounded and included in the growing crystal. On the other hand, when the pore spaces are small, as in clay, pressure is more uniformly distributed over the crystal surface, growth is slow, and new material may reach the entire surface of the crystal by diffusion. Therefore, the foreign bodies are mechanically displaced by the growing crystal. In the so-called rigid rocks, the pore spaces are small and the growth of new crystals extremely slow; therefore, new crystals usually exclude most foreign material.

The absence of fracturing and of optical anomalies in partly replaced minerals is not, in the writer’s opinion, to be regarded as proof that the growth of the new minerals was unopposed by mechanical force. Distortion and fracturing of crystals can be produced only by a differential stress exceeding the elastic limit of the substance. Pressure resulting from the growth of new minerals is developed very slowly, and in general it is transmitted through a thin layer of solution which supplies the material for mineral growth; Each mineral grain, in compact rocks is surrounded by others, which increases its resistance to distortion. Moreover, strained crystals are less stable than those that are not in a state of strain; and, therefore, the replacing solutions tend to dissolve such crystals in preference to others.

New minerals that make room for themselves by exerting pressure on the surrounding material must have begun their growth in small openings filled with gas or liquid. The replacing minerals often appear as microscopic grains, completely surrounded by the older minerals. Sometimes minute fractures, through which the solutions probably entered, may be recognized, but often they are invisible. In many such cases, however, the presence of sub-microscopic passageways may be inferred from the fact that dots and stringers of the new mineral develop along cleavage directions in the older crystals or along the planes of fluid inclusions so common in quartz. The enlargement of new crystals is sometimes accompanied by a breaking up of the older inclosing crystals, and the substitution therefor of interlocking aggregates of smaller grains. This fact has been recognized by Lindgren, though he does not attribute the fracturing to pressure from the growing crystal.

In some rocks the development of new minerals is accompanied by solution and removal of the less stable minerals and mechanical displacement of those that are more stable. Rocks in which replacement has barely begun commonly contain disseminated crystals of metasomatic pyrite, and, when these crystals are examined microscopically, it may sometimes be observed that in their immediate vicinity the more soluble minerals of the country rock, such as quartz, are relatively scarce, while the less soluble minerals, such as muscovite and chlorite, are crowded together, and in some cases bent or broken (Fig. 3).

It is difficult, if not impossible, to explain well developed crystal form in metasomatic minerals under any hypothesis except that the crystals in growing have exerted pressure on the surrounding material. The isolated pyrite crystals that develop in wall rocks usually show idiomorphic outlines; but, where pyrite has completely replaced the older minerals, crystal form is much less common, because the growing crystals have mutually interfered with the growth of one another.

The mechanical replacement of one mineral by another, in the manner outlined above, would be, necessarily, volume for volume, and not molecule for molecule. Pseudomorphs might be formed through the replacement of crystals imbedded in a matrix of more stable minerals, but it is improbable that the more minute details of internal structure could be preserved.

It is conceivable that mechanical replacement may also result when solutions dissolve the mineral that is being replaced, and thus lower the pressure on the replacing mineral sufficiently to permit additions of new material. This method of replacement may be applicable to some deposits formed in the belt of weathering, but it is improbable that primary ores in veins have been formed in this way. Primary vein minerals are deposited from solutions which, as a whole, are moving toward regions of lower temperature and less pressure, and for this reason the solutions are dominantly agents of deposition rather than of solution. This assumption is borne out by the general absence of solution cavities in primary replacement deposits below the belt of weathering. Even under the microscope, it is impossible to detect any intervening space between a partly replaced mineral and the replacing mineral, or other evidence that solution is dominant over deposition.

(3) When one mineral replaces another as a result of a chemical reaction, a definite number of atoms is added, subtracted, or interchanged, the number depending on the quantity of mineral replaced and the nature of the reaction. But the number of molecules that are deposited and left behind in solid form upon the completion of the reaction depends also on the solubility of the new mineral under the conditions existing at the time of its formation. In some cases the original substance is entirely removed, as when copper is replaced by silver from a solution of silver sulphate.

The new mineral formed by the reaction may be practically equal to the old one in volume, but usually the change is accompanied by either a decrease or an increase in volume. In case of a decrease, the replacing mineral is commonly porous or cellular in texture, and sometimes shows drusy cavities lined with small crystals. This is well illustrated in pseudomorphs of native silver after argentite and of smithsonite after calcite. The voids may be filled, however, through deposition of other minerals, especially in primary ores deposited by ascending solutions. When the volume of the new mineral is greater than that of the old, the surrounding material may be mechanically displaced in order to make room for the additional volume, or, more commonly, the pressure resulting from expansion may force into solution either a part of the inclosing rock or a part of the new mineral, and thus remove enough material to compensate for the increase in volume. In many cases adjustment probably results from two or more processes, and the problem may be further complicated by contemporaneous reactions involving neighboring minerals.

Irving and Lindgren have referred to the absence of optical anomalies in adjacent mineral grains and the lack of distorted structural and textural lines as evidence that pressure, due to increase in volume, is of little importance in the process of replacement, but with this view the writer does not agree. Pressure resulting from chemical change in the minerals of a compact rock is developed slowly, the resistance to stress is great and the solubility of most minerals is increased by pressure and strain. Therefore, it is only when relatively insoluble minerals are present that evidence of mechanical displacement or distortion of mineral crystals may be expected. Most of the arguments given above in support of the view that the growth of new minerals in rocks may result in the development of pressure are equally applicable here.

It has been demonstrated experimentally that chemical reactions, similar to those producing changes in the minerals of rocks, may be accompanied by the development of pressures due to expansion in volume; and that these pressures may be many times the crushing strength of the material as ordinarily determined. An experiment has been described in which a porous porcelain cup containing anhydrous cupric chloride was ruptured through the formation of the hydrous salt. Similarly, a glass bottle of 150 c.c. capacity, having walls with a minimum thickness of 1.5 mm., was fractured by the expansive force resulting from the hydration of zinc, nitrate. Although such reactions are accompanied by an increase in the volume of the solid, there is usually a decrease in the volume of the system (the anhydrous salt plus water); and, where this is true, pressure can have no direct effect in preventing the reaction. The pressure developed is therefore limited solely by the resistance offered to expansion.

In many metamorphic rocks, quartz, has disappeared from places subjected to maximum stress, and recrystallized where the stresses were least, while at the same time minerals, such as mica, which are less soluble, have been bent and displaced. Evidence has been cited elsewhere in support of the view that chrysotile veins may be due to the solution of serpentine under pressure resulting from its formation and the deposition of the material as chrysotile at places where the forces opposing expansion were sufficiently feeble. The general absence of vugs in chrysotile veins may be due to the pressure under which they are formed, since drusy cavities are not uncommon in fibrous veins of other minerals, such as calcite and gypsum.

The metasomatic replacement of a rock is often the result of extremely complicated processes, partly mechanical and partly chemical.

In the case of individual minerals it may be impossible to determine whether replacement was the result of mechanical processes, or of chemical interchange, or, perhaps, of both combined. It is realized that such processes can never be classed as wholly chemical or wholly physical, for replacement of any kind involves the removal or addition of atoms; but, nevertheless, it is believed that the distinction outlined above is important. Silica is pseudomorphic after calcite, and commonly replaces fossil shells in such a way as to preserve the minutest details of internal structure; but, on the other hand, perfectly formed quartz crystals sometimes develop in crystalline limestone. In both cases calcite is replaced by silica, though the two processes are evidently different. It is suggested that the first is primarily a chemical process and due to chemical interchange, while the second is chiefly mechanical.

The segregation of mineral matter originally dispersed through the surrounding material and its deposition in relatively concentrated masses through some process of replacement is of common occurrence. In this way nodules and layers of chert or flint are formed in calcareous rocks, and pyrite is segregated as single crystals, or as larger crystalline aggregates. It is conceivable that some bedded veins and veins paralleling contacts might be formed in this manner, but it is highly improbable that well defined tabular deposits could develop transverse to the rock structure. Such veins have been recognized by LeConte, and classed under the heading of “veins of segregation.” They can be distinguished from the replacement veins described above only by determining that the vein minerals were derived from the inclosing rock and were not introduced along fractures or similar channels. For most veins the evidence is overwhelmingly against the hypothesis that the ore minerals were derived from the immediate wall rock. Many of the arguments used by Posepny and others in combating the old lateral secretion theory of Sandberger are applicable here, and need not be repeated. Most veins are also formed along fractures or other structural planes which have facilitated the circulation of ore-bearing solutions, and thus controlled the form of the deposit. The irregular and nodular form of some of the magnesite veins occurring in serpentine suggests that they may have been formed partly in this way.

Some tabular and lenticular ore deposits, inclosed in igneous rock, and commonly regarded as magmatic in origin, should possibly be classed as veins, and included in the present discussion; for there is no essential difference between (1) the transfer of ore minerals through the agency of mineralizers set free during the final stages in the solidification of an igneous magma, and (2) the solution and transportation of ore minerals originally disseminated through a rock mass. In both cases the ore minerals may be deposited through the partial or complete replacement of earlier minerals. Through their comprehensive investigations, Tolman and Rogers have established the general law that sulphide minerals are among the final products of magmatic differentiation rather than the first.

Summary and Conclusions

It is obvious that all veins have not been formed in the same way, and that many veins, particularly the larger metalliferous ones, are the result of more than one method of deposition. There are gradations between the different types of veins and also between veins and other types of epigenetic ore deposits. It is the opinion of the writer that the methods of ore deposition herein discussed are applicable to all such deposits.

All minerals deposited in veins or other orebodies occupy space previously filled by other material, either gaseous, liquid, or solid. Minerals are deposited in openings filled with gas or liquid, where such spaces are available, but only very small open spaces are believed to exist at the depths where most ore deposits have been formed. It has been demonstrated that yawning fissures are not necessary for the formation of veins, and the evidence indicates that very few veins have been formed wholly by deposition in pre-existing openings. All minerals are formed under pressure, and no pressure could prevent the growth of a mineral crystal the surfaces of which were in contact with a supersaturated solution. Mineral-bearing solutions, especially under conditions of high temperature and great pressure, are able to enter the densest rocks by penetrating between adjacent crystals and even along mineral cleavages and sub-microscopic fractures. The introduction of mineral matter through very small openings is probably due chiefly to diffusion through thin films of solution. The circulation of solutions is greatly facilitated by fault fractures, planes of bedding or schistosity and by porous strata; therefore, most veins are formed along such passages.

Growing veins may make room for themselves either by mechanically displacing or by replacing the wall rock with which they are in immediate contact, and there is every gradation between these two extremes. When the minerals of the country rock in contact with a vein are forced aside, other adjacent minerals are likewise displaced, and final adjustment is attained either wholly or in part through condensation of porous material, rock flowage to places of less pressure, solution and removal of unstable minerals, or recrystallization to minerals of higher density.

All veins having cross-fiber structure are lateral secretion veins which, in growing, have forced apart their own walls, but all lateral secretion veins are not necessarily fibrous. The fibrous structure is a direct result of the manner of deposition, which has mechanically limited crystal growth to a single direction.

The replacement of one mineral by another may be primarily a mechanical or a chemical process. When the process is mechanical, the solution and removal of the old mineral is induced by pressure resulting from the growth of the new and, therefore, the volume of the replacing mineral equals the volume of the mineral replaced. When replacement results from a chemical reaction, the ratio between the number of atoms added and the number removed is constant for a given set of conditions, and the change may be accompanied by either an increase or a decrease in volume.

The importance of the fact that crystals may grow in directions in which growth is opposed by adjacent solid bodies has not been generally recognized by geologists, but it is a fact that must be taken into consideration in framing an adequate theory of ore genesis. . The hypothesis that ore minerals have made room for themselves by mechanically displacing older minerals and. rocks, elucidates many phenomena which otherwise are contradictory. It apparently solves some of the questions which have, resulted in such widely divergent views as to the origin of certain sulphide orebodies.

Inclusions of wall rock, often angular in shape, are common in most sulphide orebodies, including those which have been classed as magmatic in origin. The fragments vary in size from microscopic specks up to blocks weighing thousands of tons. In some instances their position suggests that they have been pried off the adjacent wall. Occasionally the inclusions are bent, twisted or split into smaller fragments. The high fluidity of the ore-bearing solutions (or magmas ?) is indicated by the manner in which the minerals have penetrated along minute fractures in the inclusions and in the wall rock, and yet there appears to be no tendency for the fragments to segregate under the influence of gravity. Usually there is evidence of considerable replacement, but often replacement is of minor importance. These facts are all in accord with the hypothesis that the concentration and deposition of the ore minerals has been brought about through the agency of mineralizers, and that the ore minerals have made room for themselves, in part at least, by mechanical displacement of older minerals. The facts, are apparently irreconcilable under a hypothesis that attributes the introduction of the ores to magmatic injection, magmatic sloping, or any similar process.

As evidence accumulates, it is becoming more and more certain that the formation of ore deposits is essentially a process of orderly and progressive concentration through the agency of mineralizers, and that the different types of deposits are due largely to difference in the original magmas. The writer believes that the ores make room for themselves, in most cases, by mechanically displacing or by replacing the older minerals and rocks.