Geological Survey 12th Annual Report (Part I) The Eruptive Rocks of Electric Peak and Sepulchre Mountain, Yellowstone National Park

Having described the geological structure of Electric Peak and of Sepulchre Mountain and the occurrence and character of the igneous rocks in each locality, it remains to point out the relationship of the two groups of rocks to each other, and the petrological deductions which may be drawn from their investigation.

To arrive at the relationship of the volcanic rocks of Sepulchre Mountain to the intrusive rocks of Electric Peak it is necessary to observe, in review of the facts already presented, that the latter cut through Cretaceous shales and sandstones and have imparted sufficient heat to them to metamorphose them for a great distance, indicating the passage of large quantities of molten magma through the fissures; while the lavas of Sepulchre Mountain rest on Cretaceous strata and also carry large blocks of black shale inclosed within them. They plainly show by their crushed and dragged portions that a profound fault has separated the block of Sepulchre Mountain from that of Electric Peak, dropping the former down considerably more than 4,000 feet. Consequently the volcanic rocks of Sepulchre Mountain once occupied a higher elevation than the present summit of Electric Peak and its bodies of intrusive rock.

In Electric Peak there is a system of fissures that radiates outward toward the south and southwest, as shown by the dikes of porphyrite. At the west base of Sepulchre Mountain there is a system of dikes and intruded bodies that radiates outward toward the north and northeast. These fissures antedate the great faulting just mentioned and represent the east and west halves of a system of fissures trending from north and south around to northeast and southwest which crossed one another at the point where the broadest body of intruded rock is now found. The axis of this system appears to have been inclined toward the east, that is, to have dipped toward the west, and was cut across by the great fault which dropped Sepulchre Mountain.

The igneous rocks that broke through the strata of Electric Peak consist of a series of porphyrites, occurring in sheets between the strata, and another series of diorites and porphyrites that were erupted through the vertical fissures just alluded to. The central fissure or fissures became the conduit through which the molten magmas followed one another at successive intervals of time. In the outlying narrow fissures the magmas solidified as dikes of porphyrite, while within the heated conduit they consolidated into coarse grained diorites of various kinds. The magmas of this series of eruptions became more and more siliceous. Their succession is indicated in the accompanying table.

TABLE XVI.—Order of eruption of the rocks at Electric Peak and Sepulchre Mountain.

The igneous rocks that formed the breccias and lava flows of Sepulchre Mountain with their dikes and larger intruded bodies constitute a series of andesites, basalts and dacites, which reach a degree of crystallization that places part of them among the porphyrites. They commenced with an andesitic breccia that is filled with Archean fragments, which must have been thrown from some neighboring center of eruption located in an Archean area. Such a center exists a few miles to the north. This was followed by a series of magmas that were at first somewhat basic and became more and more siliceous. The series is represented in the right hand column of Table XVI. From this it is seen that the succession of eruption in each locality was the same, after the first period, A, in which the magmas evidently came from different sources. Each series of the second period began with basic magmas and ended with acidic ones. Their division in the table into four groups is not intended to convey the idea that they belong to four distinct periods of eruption. The whole series in each case is more correctly a single, irregularly interrupted succession of outbursts of magma that gradually changed its composition and character. Upon comparing the rocks which have resulted from the corresponding phases of these series of eruptions, the similarity of the porphyritic forms is immediately recognized. The nature and distribution of the phenocrysts in the different varieties of andesite and dacite, which determine their macroscopical habit, have their exact counterpart in the different varieties of porphyrites. The microscopical characters of the phenocrysts in the corresponding varieties of porphyrites and of the intruded andesites and dacites are identical. The character of the various groundmasses, however, is different in the two groups, being more highly crystalline in the porphyrites—many of the andesites being glassy. Many of the finer grained diorites have a habit, derived from the distribution of the ferromagnesian silicates and larger feldspars, which resembles that of some of the andesites and dacites which correspond to them chemically.

Finally, the study of the chemical composition of the intrusive rocks of Electric Peak and of the volcanic rocks of Sepulchre Mountain proves that these two groups of rocks have identical chemical compositions, for the varieties that have been analyzed are but a few of the many mineralogical and structural modifications assumed by these magmas on cooling. The analyses serve as indications of the range of the chemical variability of these magmas.

From the geological structure of the region, then; from the correspondence between the order of eruption of the two series of rocks; from the resemblance of a large part of the rocks of both series, macroscopically and microscopically, and from the chemical identity of all the rocks of both groups, it is conclusively demonstrated that:

I. The volcanic rocks of Sepulchre Mountain and the intrusive rocks of Electric Peak were originally continuous geological bodies.

II. The former were forced through the conduit at Electric Peak during a series of more or less interrupted eruption.

III. The great amount of heat imparted to the surrounding rocks was due to the frequent passage of molten lava through this conduit.

We have, then, in this region the remnant of a volcano, which has been fractured across its conduit, has been faulted and considerably eroded; and which presents for investigation on the one hand, the lower portion of its accumulated debris of lavas, with a part of the upper end of the conduit filled with the final intrusions; and on the other hand, a section of the conduit within the sedimentary strata upon which the volcano was built.

CORRELATION OF THE ROCKS ON A CHEMICAL BASIS.

Correlating the two groups of rocks according to their chemical composition and arranging them as in Table XVII, we see that the hornblende-mica-andesites, Nos. 95 and 102, are the equivalents of the quartz-mica-diorites, Nos. 215, 213, 205, 227, and 223, and of the quartz-pyroxene-mica-diorite, No. 211. The dacites, Nos. 129, 131, are the equivalents of the quartz-mica-diorite-porphyrites, Nos. 233 and 230. The hornblende-pyroxene-andesites and the pyroxene-andesites, Nos. 33, 80, 20, 2, and 21, are the equivalents of the coarse grained pyroxene-mica-diorite, No. 197, with variable percentage of quartz, and of the fine grained diorites, Nos. 176 and 177, and of a fine grained facies, No. 171.

The dacites and hornblende-mica-andesites included within this correlation are intruded bodies within the breccia of Sepulchre Mountain, and have the same mineral composition as the corresponding porphyrites and diorites of Electric Peak. They differ from them in structure and degree of crystallization the details of which have already been described in earlier parts of this paper.

The glassy andesite with pyroxene and hornblende phenocrysts, however, present the utmost contrast to the chemically equivalent, coarsely crystalline diorites. In the former the hypersthene, augite, hornblende and plagioclase are sharply defined, idiomorphic crystals in a groundmass of glass, which is crowded with microlites of plagioclase and pyroxene, besides grains of magnetite. The hornblende is brown, occasionally red, and the other phenocrysts have all the microscopical characters which distinguish their occurrence in glassy rocks. In the diorite the hornblende is green, in some cases brown; and the hypersthene, augite and hornblende are accompanied by biotite, and are all intergrown in the most intricate manner, with evidence that they commenced to crystallize in the order just given. The labradorite is often clouded with minute opaque particles, which are characteristic of its occurrence in many diorites; it is surrounded by a shell of more alkaline plagioclase, which with occasional individuals of orthoclase and considerable quartz, closed the crystallization of the magma. Magnetite, apatite and zircon are the accessory minerals. The quartz contains fluid inclusions, which complete the correspondence of this diorite with typical diorites of other regions.

From the structure of this region, which has been so finely exposed by faulting and erosion, it is evident that of the different magmas erupted a part found their way into vertical fissures and took the form of dikes; part reached the surface and became lava flows and breccias, while other portions remained in the conduit. Therefore the various portions of the magmas solidified under a variety of physical conditions, imposed by the different geological environment of each the most strongly contrasted of which were the rapid cooling of the surface flows under very slight pressure, and the extremely slow cooling of the magmas remaining within the conduits under somewhat greater pressure.

TABLE XVII.—Correlation of the two groups of rocks upon a chemical basis.

TABLE XVIII.—Correlation of the grades of crystallization of the rocks from Sepulchre Mountain and Electric Peak.

(click on image for a PDF version)

The effect of this diversity of conditions upon the degree of crystallization of the various portions of these rocks is well shown in the accompanying Table XVIII, which has been derived from Tables VIII and XIV.

In this table are presented all of the specimens from Sepulchre Mountain and Electric Peak. They are arranged in four principal divisions: First, the breccias and lava flows; second, dikes and larger bodies intruded in these breccias; third, dikes in the Cretaceous strata of Electric Peak; fourth, the main stock and its immediate apophyses. These groups are still further subdivided into columns which correspond to mineralogical differences in the rocks and bear the same letters as the mineralogical subdivisions in Tables III, VIII, XII, and XIII. Consequently each of the four principal groups has the most basic members at the extreme left and the most acidic ones at the extreme right. The mineralogical range is, therefore, repeated four times. The table illustrates a number of facts. It exhibits the relative degree of crystallization of the breccias, lava flows; dikes, and stock rocks, and shows that a great number of intermediate steps can be recognized between the most glassy andesite and the coarsest diorite. It shows that the dike rocks furnish the connecting link between these two extremes and that the dike rocks of Electric Peak have the same range of grain as the majority of those of Sepulchre Mountain. But many of those at Sepulchre Mountain are still finer grained and some are glassy, being vesicular also. Between these rocks there is the closest possible resemblance macroscopically, and the two groups might have been described conjointly so far as their petrographical characters were concerned. The variation of grain within each of the four principal divisions is very significant when taken in connection with the geological occurrence of the different rocks. The limited range of variation in the first group is in accord with the fact that all of these rocks are surface ejectamenta. The range in the third group from more crystalline basic rocks to less crystalline acid rocks, as already pointed out on page 621, shows the greater tendency of the basic rocks to crystallize. And since the dikes here represented are nearly the same size, this variation of grain corresponds to differences in the chemical composition of the rocks. On the contrary the variations in the second group indicate a slightly greater crystallization of the acid rocks. This, however, is due to the fact that the basic rocks in this group, with a few exceptions, occur in small dikes, while the acid rocks for the most part form broad intruded bodies a number of hundred feet wide. In these cases the size of the mass has had more influence on the degree of crystallization than the chemical composition of the magma has had. In the fourth group the basic rocks exhibit a wider range of grain than the acidic, being much coarser and also considerably finer grained than the latter. This arises from the fact that the basic rocks form a much larger mass and exhibit great variation of grain, having fine grained facies that have been fully discussed in an earlier part of this paper.

These Tertiary diorites and others that cut the volcanic lavas in several localities in this region correspond to the andesdiorites and andesgranites of Stelzner who described stocks of granular rocks penetrating the andesitic tuffs in the Argentine Republic. The study of these Tertiary granular rocks led him to the conclusion that the degree of crystallization of eruptive rocks is in no way dependent on their age, but depends on the physical conditions under which the mineralogical differentiation and the cooling of the magma took place.¹

---

¹Alfred Stelzner: Beiträge zur Geologie und Paleontologie der Argentinischen Republik. Cassel and Berlin, 1805, p. 207.

"Sie (die Andengesteine) wird uns, wie ich meinerseits glaube, immer mehr und mehr erkennen lassen, dass die grössere oder geringere Krystallinltät eruptiver Gesteine keineswegs, wie man so lange und so hartnäckig behauptet hat, von dem Alter der letzteren abhängig ist, sondern lediglich von den physikalischen Umständen, unter denen die mineralische Differenzirung und Erkaltung der gluthflüssigen Magmen vor sich ging."

---

From the study and comparison of the chemical analyses of the two groups of rocks under investigation it is demonstrated that the magmas that reached the surface of the earth in this place had exactly the same chemical composition as those which remained inclosed within the sedimentary strata. It proves with equal clearness that the different conditions attending the filial consolidation of the ejected and of the intruded magmas affected not only their crystalline structure, but their essential mineral composition. The most marked illustration of this is in the occurrence of biotite in the two series. In the volcanic rocks of this locality biotite is an essential constituent of the more siliceous varieties, and is only rarely found as an accessory constituent of the varieties with less than 61 per cent of silica. In the intrusive rocks it is an essential constituent of all the coarse grained varieties, even the most basic. In the finer grained porphyritic forms it is a constituent of the groundmass to a variable extent. The second most noticeable difference is the presence of considerable quartz in the coarse grained forms of the basic magma and its absence from the volcanic forms of the same magmas.

From these observations, then, we see that in this region there are chemically identical rocks which have distinctly different mineral compositions, but which were once parts of a continuous body of molten magma. We are led therefore to the conclusion that—

The molecules in a chemically homogeneous fluid magma combine in various ways, and form quite different associations of silicate minerals, producing mineralogically different rocks.²

---

²This conclusion is the same as that stated by Justus Roth:

"Es können mineralogisch ganz verschiedene Gesteine in dieselbe Gruppe gehören, denn feurigflüssige Massen von gleicher oder sehr nahe gleicher chemischer Zusammensetzung können in verschiedene Mineralien ausinanderfallen. Die Ursachen, welche diese Erscheinung bedingen, lassen sich höchstens muthmassen und mögen in Unterschieden des Druckes, der Temperatur, des umgebenden Mediums der Unterlage u. s. w. gesucht werden." Die Gesteins-Analysen in tabellarischer Übersicht und mit kritischen Erläuterungen, Berlin, 1861. p. VVI.

---

The bearing of these facts upon the question of the classification of igneous rocks is that, since different portions of a large body of a chemically uniform magma may assume a variety of geological forms within the earth's crust or upon its surface and may crystallize into rocks with different mineral composition, it is more proper to consider intrusive and effusive rocks that have like chemical composition as corresponding or equivalent rocks than those forms of the two series that have similar mineral composition. Thus we would not say that certain volcanic rocks which are the equivalents of certain intrusive ones differ from them chemically by such and such variations among the oxides, for the term equivalent would then simply refer to their mineralogical character, and we might be comparing portions of totally different magmas that had no geological connection with one another. Used in the other sense, we should say that certain volcanic rocks differ from their corresponding or equivalent intrusive rocks by the presence or absence of certain minerals, and in this way we would be grouping together the extrusive and intrusive portions of the same body of magma. The classification would then rest on a common geological and chemical basis.

In this region of Electric Peak and Sepulchre Mountain the greatest mineralogical differences accompany the greatest differences in structure or degree of crystallization; hence we may assume that the causes leading to each are coexistent. The source of these causes must be sought in the differences of geological environment, and these affect the rate at which the heat escapes from the magmas and the pressure they experience during crystallization.

It is to be remarked that the most essential mineralogical difference between the intruded rocks and their chemically equivalent extrusive forms is the much greater development of biotite and quartz in the intruded rocks; these minerals being abundant even in the basic intrusions and absent from their basic volcanic equivalents. That their simultaneous development is naturally to be expected in many cases is evident from a consideration of the character of their chemical molecules and that of other minerals common to these rocks. For if we assume that biotite is made up of two molecules, K and M corresponding respectively to K₆Al₆Si₆O₂₄ and R₁₂Si₆O₂₄, and compare these with the molecules of orthoclase, K₂A1₂Si₆O₁₆, of olivine, R₂SiO₄, and of hypersthene RSiO₃, we see that molecules which under some conditions might have taken fine form of olivine or hypersthene and potash-feldspar, which latter may have entered into combination with lime-soda feldspar molecules to form somewhat alkaline feldspars, may under other conditions combine as biotite with the separation of free silica or quartz; in which case also the feldspars of the rock would be less alkaline.

Another mineralogical difference between the two groups of rocks just mentioned is the greater development of hornblende in the intruded rocks in place of augite, which is chemically its equivalent, though it has not been determined whether in this case the hornblende of the diorite has precisely the same composition as the augite of the andesite. The probability is that there are slight differences between them.

EFFECT OF MINERALIZING AGENTS.

The crystallization of quartz, biotite, and hornblende in fused magmas, according to our present knowledge, requires the assistance of a mineralizing agent; for it has been demonstrated by synthetical research that these minerals will not crystallize into the forms they assume in igneous rocks when their chemical constituents are fused and simply allowed to cool under ordinary atmospheric conditions. But they have been produced artificially with the aid of the mineralizing action of water and other vapors. Now there is ample evidence both in the ejected lavas and in the coarsely crystallized rocks in the conduit that water vapor was uniformly and generally distributed through the whole series of molten magmas, and there is no evidence that there existed in the magmas which stopped within the conduit any more or different vapors than those which existed in the magmas that reached the surface. Hence we conclude that:

The efficacy of these absorbed vapors as mineralizing agents was increased by the conditions attending the solidification of the magmas within the conduit.

Moreover, if it is necessary, as advocated by the French geologists, MM. Michel Lévy,¹ de Lapparent² and others, to refer the crystallization of certain minerals, as quartz, to the mineralizing influence of absorbed vapors, it is evident that the required mineralizing agent is universally present in sufficient quantities, since there are no instances where a magma of the requisite chemical composition has failed to crystallize completely with the development of quartz when subjected to the proper physical conditions.

---

¹"Structures et Classification des Roches Éruptives." Paris. 1889, pp. 5 and 12.

²Revue des Questions Scientifiques. Paris, 1888, p. 26.

---

However, it is probable that differences in the amount or in the kind of mineralizing agents produce differences in the degree or nature of the crystallization of similar magmas which have solidified with the same geological environment.

It has been suggested by Dr. H. J. Johnston-Lavis³ that the nature of the rocks surrounding a conduit through which molten magmas pass materially affects the amount and character of the vapors introduced into these magmas, which will vary as the surrounding rocks are more or less porous and are saturated with different kinds of waters. The effect of these vapors on the structure and composition of igneous rocks is also discussed by the same writer.

---

³"The Relationship of the Structure of Rocks to the Conditions of their Formation." Sci. Proc. of the Royal Dublin Soc., vol. 5 (m. s.), part 3, July, 1886, pp. 113 to 155.

---

The effect of differences in the amount of the mineralizer in a single magma is well illustrated in the structure of the obsidian at Obsidian Cliff, Yellowstone National Park,⁴ where the alternating layers of holocrystalline and glassy rock appear to be unquestionably due to the irregular distribution through the magma of vapors, which in the upper portion of the flow have produced alternating layers of pumice and compact glass. The mineralizing agent was present, however, in the alternate glassy layers as well as in the crystallized or in the pumiceous ones, for in the highest portion of the flow the whole mass is pumiceous but in different degrees, and the presence of absorbed vapors may be detected chemically and physically in the compact layers. Its amount, however, was not sufficient to produce complete crystallization under the attendant physical conditions. Its effectiveness in this case was controlled by the geological occurrence of the magma.

---

⁴Obsidian Cliff, Yellowstone National Park, by J. P. Iddings. Seventh Annual Report of the Director of the U. S. Geological survey, Washington, D, C., 1888, p. 287.

---

It is to be observed, in addition, that whatever the mineralizing vapors in acidic magmas may be, there is the same evidence of their existence in intermediate and in basic magmas, whether we investigate them chemically or physically, or study the phenomena of their geological occurrence. There are even indications of their greater abundance in the basic lavas, many of whose glasses contain a high percentage of water, and the highly vesicular character of whose lava-flows is universal. Nor are the geological evidences less conclusive that demonstrate the existence of abundant explosive agents in the basaltic and andesitic magmas that have hurled their shattered masses over broad areas of country, and have piled vast accumulations of basaltic breccia throughout our western territory.

Nevertheless, with all these evidences of the universal presence of mineralizing agents in basic magmas, we do not recognize their influence upon the microstructure or crystallization of basic lavas. We may assume, then, that in the majority of these cases they have no influence.

But when the basic magmas become coarsely crystalline, and separate into minerals, the crystallization of some of which we have already referred to the action of mineralizing vapors, we may logically assume that in these cases the absorbed vapors have influenced the crystallization of the magmas.

If this reasoning is correct, then the action of mineralizers upon basic magmas is controlled by the physical conditions under which they solidify.

Finally, if mineralizing agents are universally present in igneous magmas, and if their action, so far as we can observe it is controlled by the physical conditions imposed by the geological history of each eruption, we should not regard the presence or absence of certain minerals, relegated to the influence of mineralizing agents, as evidence of the presence or absence of these agents in the molten magma; but we should see in it the evidence of special conditions controlling the solidification of the magma, and should seek the fundamental causes of the mineralogical and structural variations of a rock in the geological history of its particular eruption.

APPLICATION TO THE CLASSIFICATION OF IGNEOUS ROCKS.

The facts brought out by the study of this occurrence of igneous rocks seem to the writer to have a direct application to the problem of the general classification and description of igneous rocks. For while this occurrence cannot be regarded as a representative of all others, still it typifies to a very great extent the relations that exist between intruded magmas and their extrusive forms.

We have observed that in this locality a series of molten magmas was erupted through a common conduit during a succession of fracturings of the sedimentary strata. These magmas not only differed among themselves chemically, but varied somewhat in different portions of one and the same body, producing chemical facies of the main body of a particular rock mass.

When we consider the variations in the chemical composition and structure, and mineral constitution of a continuous geological body, such as may occur along an irregularly shaped crevice or system of fissures from their narrow and remote terminations toward their wider junctions with the main conduit, as well as the interpenetration and welding of older and newer portions of the magmas filling the conduit, with their consequent transitions in some places, and sharply marked intersections or contacts in others, we see that the resulting mass of igneous rocks presents a geological body whose complexity exceeds that of the most intricate web of vegetable organism.

Chemically considered there is a wide range of composition embracing the middle of the whole series of igneous rocks of the surrounding region. In percentage of silica they range from 53 per cent to 69 per cent; and if certain contemporaneous intruded rocks in the immediate neighborhood be included, the range of variation in the intrusive rocks is about the same as that of the volcanic rocks, from 48 per cent to 74 per cent.

Structurally, there are all forms from coarsely granular to porphyritic glassy, including all possible intermediate structures.

Mineralogically, there are all the combinations existing in this region, from that of quartz, alkali-feldspar, and mica, to that of basic lime-soda-feldspar and pyroxene, with a little olivine.

Hence the rocks include granite, granite-porphyry, quartz-porphyry, and rhyolite; diorite, quartz-mica-diorite, diorite-porphyrite, pyroxene-porphyrite, hornblende-mica-andesite, hornblende-andesite, pyroxene-andesite, dacite, and basalt. The glassy form of the granite-porphyry or of the quartz-diorite-porphyrite is not found in the immediate vicinity of Sepulchre Mountain, but occurs in the region south as a modification of the rhyolite at the Upper Geyser Basin. The still more siliceous rhyolite of Sepulchre Mountain is represented by a facies of the microgranite at Echo Peak, a point 12 miles south of Electric Peak.

Notwithstanding the range of structural variations within the mineralogical groups just mentioned, it is not possible to trace in exposure any one group through this series of structural variations. It becomes evident that while a perfectly continuous body may, and undoubtedly does in some instances, connect the glassy form of a consolidated magma with a coarsely granular form through intermediate stages of crystalline structure, yet the connected occurrence of all these forms is not a necessity, and in fact does not always exist. For if we consider the course of eruption of a magma that varies in its chemical composition, or the successive outbursts of a series of magmas that differ chemically from one another, we see that if a basic magma which has reached the surface of the earth and has produced glassy rocks—andesites—and has filled the disrupted strata with intruded sheets and dikes of porphyrite, and stands in the conduit under conditions which would eventually produce coarse grained diorite—if a basic magma in this stage of solidification be followed through the same conduit by a more siliceous magma, then the viscous body within the conduit would be forced out on the surface and its place occupied by the later magma, which would thus sever the connection between the intruded sheets or dikes and the surface lavas, and would deprive both of a coarse grained equivalent. Moreover, it is well known that in volcanic regions it usually happens that the lava that flows from a cone severs its connection with the molten magma in the crater, which often descends again within the conduit.

In the case of a great body of magma which varied in composition during a prolonged eruption, so that the first portion of it differed considerably from the last portion, the surface flows and earliest intrusions, if continuously connected with the deep-seated portion, would grade into it not only through a variety of structural modifications, but through a series of chemical and mineralogical variations, so that their actual geological connection would be with a coarse grained rock of a different type.

Furthermore, the magmas, which can be recognized at this locality as having constituted independent eruptions, not only differ in their chemical composition from one another, but vary to such an extent within their own mass that the chemical facies of one body correspond to the main portion of another. Hence the members of the series overlap one another in composition. Consequently a classification or consideration of the various forms of rocks of the same chemical composition involves in this case the artificial grouping of parts and facies of different geological bodies.

In the study and discussion of the igneous rocks of this region it has been found that the natural and most intimate grouping of the rocks brings together varieties of the surface or extrusive rocks which differ chemically, mineralogically, and to a certain extent structurally. In another group it brings together varieties of coarse grained rocks which vary chemically, mineralogically, and to a certain extent structurally. And in another group it presents a collection of intruded sheets and dikes, with similar chemical and mineralogical variations, and another range of structural variations. The distinction between these groups is the range of the structural variations in each, which is coupled with their mode of occurrence. But here also is an overlapping of the groups, there being no sharp line between the first and second, or between. the second and third. This, however, is not so much of an objection to the treatment of the subject as that which would follow a grouping upon a chemical basis, for the latter would still heave unreconciled the mineralogical variations that are dependent on the mode of occurrence. It is this complicated relationship which has rendered a clear and comprehensive description of the occurrences so difficult.

Since this complication of relationships between all varieties of igneous rocks exists universally, as it has been shown to exist at Electric Peak and Sepulchre Mountain; and since the classification of igneous rocks along any single line of relationship can not be a simple and at the same time a natural one, it seems to the writer that the most satisfactory treatment of the subject brings together into groups for purposes of description rocks of similar or allied structures, but of various mineral and chemical compositions.

This grouping appears the more rational when it is considered that the chemical variability of rock magmas which leads to the formation of local modifications of rocks or to their chemical facies is, as the writer believes and hopes to be able to demonstrate at another time, the underlying principle which gives rise to the chemical differences among the rocks themselves. In other words, the chemical differences of igneous rocks are the result of a chemical differentiation of a general magma. And in a very special manner all of the igneous rocks of any locality are so intimately related to one another chemically that there is far more reason for considering them as a complex chemical unit than as a number of independent, well defined magmas.

It is to be remarked, moreover, that if, as demonstrated in this paper, the conditions attending or controlling the crystallization of igneous magmas, whether affecting simply the rate of cooling, or acting through the medium of a mineralizing agent within the magma itself—if these conditions determine the species and character of the minerals developed, as well as the crystalline structure of the rock, then the grouping together of rocks of allied structures unites those rocks in which the mineralogical characteristics bear a certain relation to the chemical composition, which relation is different from that which exists in rocks that have crystallized under different conditions. There is, therefore, in such a grouping more than the similarity of structure or the geological association of the rocks in the field.

While the grouping of igneous rocks on a basis of crystalline structure, which would bring together coarse grained forms, medium grained forms, and extremely fine grained and glassy ones, is in a very large measure equivalent to classifying them on a geological basis, still the precise connection between the crystalline structure and geological occurrence of all igneous rocks is not so uniform that it can be expressed in simple terms.¹ It is not, in fact, the particular mode of occurrence of a rock, geologically considered, that determines its structure, but the physical conditions attending its eruption and solidification. And since these physical conditions may be occasioned by somewhat different geological circumstances, the resulting similar structures may be found with different geological environment. That is, a large mass of magma deep within the earth's crust may attain a crystalline character through the cooling of so large an inclosed mass, which may be more closely related, if not identical, to the crystallization of a much smaller mass that has solidified within highly heated rock walls, than it is to the structure of an equally large mass that has been chilled by being forced a longer distance through colder rocks, or that has solidified on the surface of the earth. As another example, narrow bodies of magma which have solidified at very much the same distance from the surface of the earth differ widely in their crystalline structure, according to the temperature of the rocks surrounding them at the time of their consolidation.

---

¹Compare in this connection the conclusions of M. Michel Lévy: "Ainsi, en résumé, les conditions de gisement nous paraissent en relations trop complexes avec les facteurs de la cristallisation pour ponvoir être substituées, comme entrée de classification, à ma notion plus précise et toujours présente de la structure des roches."—Structures et Classification des Roches Éruptives. Paris, 1889, p. 10.

---

Recognizing, then, the intricacies of these geological and physical relations, it seems to the writer advisable to base the classification of igneous rocks on that character which may be determined with certainty from the rocks themselves, namely, the crystalline structure, and which, at the same time, is to so high a degree an exponent both of the chemical composition of the magmas and of the physical and geological conditions attending their solidification.