Yu. A. Kuznetsov

ORIGIN OF IGNEOUS ROCKS

“It is a great thing to reach with the mind into the depths of the earth, where nature forbids hands and eyes to extend; to wander in thought in the underworld, to penetrate by reasoning through narrow fissures, and to bring forth into the clarity of sunlight things and deeds darkened by eternal night.”

M. V. Lomonosov. On the Strata of the Earth.

Preface

The problem of the origin of granites is at present the most acute problem for petrology. The problem of the origin of basic and ultrabasic rocks is apparently considered clearer and has scarcely been discussed in the literature in recent years. It seems to me that the question of the origin of granites cannot be resolved apart from the question of the origin of other rocks of igneous aspect; therefore, below, the question of the origin of igneous rocks—or, more precisely, rocks of igneous aspect in general—will be considered.

In the first part of the work, the existing (modern) views on the question of the origin of igneous rocks and the principal disagreements on this question will be set forth as briefly as possible; moreover, I shall refrain from presenting its history, considering it known in its main outlines to most readers.

In the second part, factual material is presented in the minimum amount, constituting the initial data for constructing a general hypothesis of the formation of igneous rocks.

In the third, I have attempted to formulate my own views on the question of the origin of igneous rocks—views based partly on my own observations and partly of a deductive character.

Review of Views on the Origin of Igneous Rocks

Definition of Certain Concepts

Before proceeding to expound the essence of the question, I consider it necessary to define the content of certain terms with which it will be necessary to operate further.

“Magma.”—Different authors invest this term with very different meanings, but all agree that magma is a material from the solidification of which rocks are formed. The disagreements concern the aggregate state of magma (a homogeneous melt—solution, or a mixture of crystals and liquid; P. I. Lebedev’s “gas magma” [25]), as well as the boundaries between the concept of “magma” and the concept of “hydrothermal solution.”

It seems to me that the concept of “magma” should be defined as follows: magma is a natural silicate melt that periodically arises in the lithosphere, containing in solution a certain quantity of volatile components; it is physically homogeneous or, more often, represents a mixture of liquid and crystals that are relics of the melted (dissolved) substrate or products of crystallization of the liquid. The leading sign of magma is the flow of the mass without simultaneous deformation of the crystals, if they are present in the magma, which is achieved at approximately 25% liquid in the mixture. One may speak of the birth of magma when, as a result of slow quantitative changes (increase in temperature, change in composition during metasomatism, etc., and, as a consequence, an increase in the quantity of pore fluids), a given part of the lithosphere acquires a new quality—fluidity, the capacity for intrusion or eruption. Another qualitative leap corresponds to the transition of magma into a magmatic rock.

According to this definition, a melt impregnating a rock, or a solution filtering through it, cannot be called magma; but, according to the same definition, the growth of a metamorphic process into a magmatic one is entirely natural.

The term “migmatite,” which has become widely used in the literature and has been applied to designate a mush of magmatic liquid and crystals, is, according to our definition of the concept “magma,” superfluous.

“Magmatic rock” is the product of solidification of magma. A synonym is “igneous rock,” since magma that has not moved in space is inconceivable. Since, as a result of certain metamorphic processes (granitization, basification), rocks may be obtained that are indistinguishable in composition and structure from magmatic ones, it is appropriate to use the nongenetic term “rock of magmatic aspect.”

“Metamorphism” is any natural process occurring under conditions of elevated temperatures and pressure and expressed in a change in the chemical and mineralogical composition, structure, and texture of the primary rock, separately or together; moreover, during metamorphism a molecularly mobile phase is present at every given moment in a small quantity, mainly in the form of pore fluids. Any metamorphism expressed in a change of chemical or mineralogical composition, or in processes of recrystallization, is accompanied by transfer of matter and metasomatism. In the case of simple recrystallization of the primary rock, metasomatism has a diffusion character. In the case of a change in chemical composition, it may be diffusive or infiltrational.

“Metasomatosis” or “metasomatism” is any process as a result of which the primary substance changes its chemical composition and which is accompanied by the introduction of some substances and the removal of others. Metasomatosis may proceed as a physical process of displacement or as an exchange reaction.

Metasomatosis is not a specifically metamorphic process, since it also occurs in magmatic systems and in the processes of weathering, sedimentation, and diagenesis (for example, dolomitization of a known silt); but a rare metamorphic process is not accompanied by metasomatosis. However, only those cases of metamorphic transformations should be called “metasomatic metamorphism” in which there is a significant influx of matter and a substantial change in the chemical composition of the primary rock, which is accomplished mainly during infiltrational metasomatosis.

“Granitization” is a metasomatic process accompanied by the introduction of alkalies and silica (and the removal of Mg, Fe, Ca), as a result of which a solid rock (without passing through a magmatic stage) becomes more similar to granite than it was before; in the limit it acquires

attains a granitic composition. The transition from granitization to magma formation is a common phenomenon. The products of granitization may be various feldspathized rocks, gneisses, and also certain granites and granodiorites.

“Magmatization”—in the original sense of Sederholm [37]—is the mechanical mixing of magmatic and nonmagmatic material, more often as a result of injection along bedding. Later this term was extended to cases of metasomatic granitization. It would be more correct to use this term in its original sense. Hence “migmatite” is a mixed rock in which magmatic and nonmagmatic material is clearly distinguished.

“Basification” is a metasomatic process accompanied by the introduction of Ca, Mg, Fe, and alkalies and the removal of silica, as a result of which a solid rock becomes more basic than it was before. Basification is usually closely associated with granitization, but the same phenomenon of basification is also characteristic of contact metasomatism connected with medium-depth acid intrusions. The products of basification are biotite and cordierite hornfelses, amphibolites, certain diorites, etc.

“Rheomorphism” is the partial or complete melting of preexisting rocks as a result of their heating and metasomatic introduction of material. The formation of granitic magma as a result of granitization is a particular case of rheomorphism.

The Fundamental Nature of the Disagreements between the “Magmatists” and the “Transformists” on the Question of the Genesis of Granitoids

As already noted, disagreements on the question of the genesis of granitoids are especially great.

There are exceedingly many shades of opinion on this question, but quite clearly all researchers who have written on the genesis of granites may be divided into two groups—“magmatists” and “transformists,” or “metamorphists,” who believe that in the formation of large granite intrusions the leading role is played by the purely metasomatic process of granitization. It must be said that at present there are no magmatists who would deny the reality of the process of granitization, and there are very few transformists who would deny the formation of at least some granites by a purely magmatic route. Thus, at first glance it seems that the differences in the views of magmatists and transformists are only quantitative and consist in a different assessment of the role of granitization processes. However, these differences in views are much deeper and, as N. G. Sudovikov [45] emphasized, are fundamental. The point is that, according to the ideas of the magmatists, the process of granitization is only one of the manifestations of the emanational activity of intrusive granitic magma, which is conceived as “primary,” or derived from another, equally “primary,” basaltic magma. According to the ideas of the “transformists,” metasomatic granitization is a completely independent process, not only not caused by the presence of granitic magma, but, on the contrary, in some cases causing its appearance. Granitic magma, according to these ideas, is a product of granitization. It should be stipulated that most transformists, though not all, conceive the relationship between granitization and granitic magma in approximately this way. The “extreme transformists” consider all intrusive rocks to be products of granitization.

Views of the “Magmatists”

The views of the “magmatists” on the origin of granitoids are well known, familiar to everyone, and there is no need to dwell on them at length. As an example, I shall briefly set forth the ideas of several major scholars. The most

In his recent works P. Niggli [35, 36] has proved to be an orthodox magmatist; he holds that granitic magma is the product of a complex crystallization differentiation of basaltic magma (according to which the crystals that separate out first, sinking into more heated zones, dissolve again, thus serving as agents for the transport of molecules). At the same time, judging from certain statements, P. Niggli allows for the possibility of the formation of granitic magma by the remelting of ancient granites and rocks close to them. He regards granitization as a real process, but one connected with the emanational activity of cooling granitic magma.

N. Bowen [10, 11] believes that the majority of granites formed from granitic magma, which is a product of the crystallization differentiation of basaltic magma, which in turn arises during the fractional melting of the crystalline peridotite of the sialic shell of the Earth. But at the same time he admits that other paths of origin are possible for granites and granitic magma, for example: the formation of granitic magma as a result of the remelting of geosynclinal sediments or of the granitic shell of the Earth; as a result of the differentiation of a synthetic magma formed by the dissolution of acid material in basaltic magma; he also does not exclude the possibility of the formation of granitoid rocks by metasomatic means.

F. Yu. Levinson-Lessing, in his latest articles [29, 30, 31] devoted to the problem of the genesis of igneous rocks, continued to defend the idea of the origin of granites from a special ancestral granitic magma. True, in his genetic classification of igneous rocks [31] he provided for such types as “anatexites” (remelted rocks) and “metasomatites” (rocks produced as a result of the metasomatic alteration of sedimentary rock by magma or by some fluid magmatic emanations), but, judging from his sharp criticism of all hypotheses of the metamorphic origin of granites and even of hypotheses of the origin of granitic magmas as a result of the remelting of rocks of the sial, one must think that Levinson-Lessing attached very little significance to these genetic groups.

In recently published polemical articles, N. A. Eliseev [13] and G. D. Afanas’ev [5] came forward as “magmatists,” the latter attaching great significance to the phenomena of granitization, but linking the latter with the emanational activity of acid intrusions, while on the whole remaining entirely on the positions of F. Yu. Levinson-Lessing.

Views of the Transformists

I shall dwell somewhat more fully on the new conceptions of the transformists, still little known among us. The essence of these conceptions comes down to the fact that granites and granitoid rocks in general are recognized not as magmatic proper, but as essentially metamorphic products. They are not products of the crystallization of granitic magma, but arise through the metasomatic reworking of primary rocks of any origin and almost any composition that previously occupied the volume of the granitic body. This metasomatic granitization is generally accompanied by the introduction of K, Na, Si and the removal (in some cases) of Mg, Fe, Ca, while the amount of liquid phase during the process of granitization is very small or, according to some conceptions, entirely absent. Some transformists believe that in the mass undergoing granitization the amount of liquid phase gradually increases and, finally, it becomes fluid. Quantitative changes lead to a new quality—granitic magma arises, capable of intrusions.

But it must be noted that among the transformists there are very many shades in the understanding of the processes of granitization and even fundamental disagreements

on a number of questions, chiefly on the question of the mode of transfer of matter during granitization. Various investigators conceive this transfer as being accomplished by:

a) infiltration of granitizing solutions, liquid or gaseous;
b) diffusion of ions in a solid medium;
c) diffusion of ions in a liquid medium (immobile).

The hypothesis of infiltration granitization is the oldest and was developed as early as the beginning of this century by Termier [58, 59], then by Sederholm [57], Eskola [52], and others, and is now supported by D. S. Korzhinsky [17]. According to this hypothesis, the introduction of alkalis and SiO₂ is effected by solutions, gaseous or liquid, filtering through the pores of the rock. Such filtration is quite real under conditions of shallow depths, but is hardly possible at great depth and under conditions of very high pressures, where the processes of granitization are chiefly manifested. Therefore the majority of modern “transformists” assign the principal role in the transfer of matter during granitization to diffusion in a solid or liquid medium, the latter being conceived as intergranular films.

The hypothesis of granitization by diffusion in a solid medium (“dry granitization”) is the most fashionable among foreign scientists. It is based on well-studied and quite real phenomena of diffusion of atoms through the crystal lattice, which account, in particular, for recrystallization phenomena and chemical reactions in the solid state, without a solvent. The most ardent defenders of the hypothesis of “dry granitization”—Perrin and Roubault [37], Ramberg, Bugge¹)—regard the latter as the only mode of formation of all granitoids and decisively object not only to the possibility of the occurrence of granitic magmas, but also to all other conceivable modes of granitization.

The hypothesis of granitization by diffusion of ions in intergranular liquid films enjoys no less popularity than the hypothesis of “dry granitization.” This hypothesis is adhered to mainly by Wegmann, Kranck, Backlund, Read, and in our country by N. G. Sudovikov [45].

Concepts of N. G. Sudovikov

The views of N. G. Sudovikov [45] on the question of the origin of granites may briefly be set forth as follows.

All metasomatic granites are associated with orogenic zones, which makes it possible genetically to link granitization and folding, but the character of this connection remains unknown. In time, granitization is confined to the folding phase and develops as a slowly increasing and slowly fading process. Most often granitization begins to develop at the stage of the strongest movements and ends after their complete cessation, under conditions of a quiet tectonic setting. Granitization is a metasomatic process, often developing jointly and simultaneously with migmatization—the free injection of liquid material. The process of granitization is predominantly, but not necessarily, a deep-seated phenomenon. This process is also possible at very shallow depths, if one takes into account that in places the value of the geothermal gradient decreases to 6–7 meters and that complete melting of the rock is not necessary for granitization. But the products of granitization at shallow and great depths differ substantially. For deep zones the predominant process is potassium metasomatism; for shallow zones, sodium metasomatism, expressed in the processes of albitization and the formation of albitic granites. In general, in the process of metasomatic granitization, alkalis, SiO₂, Fe, Mg, and Ca migrate in large quantities. The composition of the migrating elements

¹) According to the paper by Bowen [11].

depends on the composition of the substrate rocks in which granitization develops. In this process, for the most part, an introduction of K, Na, and SiO₂ and a removal of Fe, Mg are observed. Opponents of granitization often ask: why are granites, and not other rocks, the products of granitization? This question is based on a misunderstanding, since other rocks also arise. For example, during granitization basic rocks are transformed first into quartz-feldspar amphibolites, then into amphibole gneisses, and finally into granites, and the process may stop at any stage. But the granitic composition is the limiting one for granitization because, upon reaching this stage, the granitized substrate becomes plastic (more correctly—“fluid.” Yu. K.) and leaves the zone of granitization, forming an intrusion. Thus a process increasing quantitatively leads to a new quality. Slowly developing granitization and the gradual increase in the amount of liquid phase lead to the emergence of a new quality—the ability to be injected. Masses arise that are similar to magmatic ones (in my view, “magma” arises!). As a result of discussing possible modes of transport of migrating substances, N. G. Sudovikov comes to the conclusion that infiltration of solutions is unlikely and possible only at shallow depths. Diffusion in a solid medium is also unlikely. Apparently, the chief role belongs to diffusion in intergranular liquid, which is rather not an aqueous solution but a melt, arising as a result of differential melting at the expense of the resources of the granitizing rock itself. Whence comes the flux of alkalis and SiO₂ necessary for granitization is unknown; the removal of Mg, Fe during granitization creates a front of basification advancing ahead of the granitization front. Likewise, the cause (primary cause) of granitization remains unknown. The immediate impulse for the development of regional metasomatism, including granitization, is apparently an increase in temperature, which causes the appearance of intergranular liquid, providing the possibility of diffusion. The causes of the temperature increase may be various. Examining the question of the genesis of granitic bodies, N. G. Sudovikov concludes that among them there may be:

1) magmatic intrusions (products of crystallization of pure melts of rheomorphic origin);

2) nonmagmatic intrusions (injection of relatively viscous but plastic masses with a small amount of liquid phase);

3) unmixed, nonintrusive bodies occurring at the site of their formation (widely developed in regions of regional granitization).

Batholiths, according to N. G. Sudovikov, are mainly nonintrusive bodies in which, however, in places there may occur a transition of granitized rocks into the rheomorphic state, and, consequently, also the development of flow structures, which may be taken as signs of the magmatic origin of batholiths. The connection of granitic intrusions with ores, according to N. G. Sudovikov, is not genetic but paragenetic: both come from regions of granitization, and he regards mineralization in the upper zones of the orogen as one of the manifestations of the front of basification.

Critique of N. G. Sudovikov’s Views

This, in general, very interesting article by N. G. Sudovikov was subjected to criticism by G. D. Afanas’ev [5] and N. A. Eliseev [15]. The critical remarks of G. D. Afanas’ev amount essentially to the fact that certain ideas of N. G. Sudovikov seem doubtful to G. D. Afanas’ev and that he stands with F. Yu. Levinson-Lessing on the question of the origin of granites.

Of much greater interest is the sharply polemical article by N. A. Eliseev. The basic propositions of this article, in their briefest form, may be formulated as follows:

a) the author points out that N. G. Sudovikov has not worked in post-Precambrian folded zones and has not seen true granites, whence his obvious underestimation of the role of “intrusive granites.” At the same time N. A. Eliseev does not deny the metasomatic origin of certain granites;

b) N. A. Eliseev believes that N. G. Sudovikov makes a mistake in speaking only of metasomatism during granitization and in failing to note that granitization begins with melting, with selective melting capable of developing into complete melting. N. A. Eliseev holds, moreover, that the melts arising in small quantities during granitization in the form of an intergranular film may be regarded as magma. In other words, according to N. A. Eliseev, granitization is not metasomatism, but the selective melting-out of granitic melts. (This, from my point of view, is a misunderstanding or an obvious unwillingness to understand N. G. Sudovikov’s principal conclusion, although I agree with N. A. Eliseev that one of the results of granitization is the appearance of granitic melts. Melting during granitization, with a progressive increase in the liquid phase, is the result not only of a rise in temperature, but also of a change in the composition of the substrate during metasomatism. Moreover, melting is only one side of the process; the other is a change in the composition of the solid phase. In addition, from my point of view, the intergranular liquid appearing during granitization cannot be called “magma,” even if it is a silicate melt);

c) N. A. Eliseev quite correctly notes that N. G. Sudovikov uses the term “plasticity” incorrectly when he says that the final product of granitization, as a result of an increase in the amount of liquid phase, acquires plasticity and becomes capable of intrusions. In this case it is better to speak of a liquid-flowing, and not a plastic, state;

d) in his critical article N. A. Eliseev expresses doubts on a whole series of questions. Thus, for him the composition of migrating substances, the reasons for the granitic composition of the final products of granitization, the possibility of the manifestation of basification, etc., are doubtful. N. A. Eliseev, in doing so, ascribes to the transformists and, in particular, to N. G. Sudovikov, ideas that were expressed by no one. For example, nowhere does N. G. Sudovikov say that ultrabasic rocks can be formed as a result of basification;

e) N. G. Sudovikov’s most interesting thought—that “the process of gradual granitization... may be regarded as a quantitatively increasing process leading to a new quality. Slowly developing granitization and the gradual increase in the amount of liquid phase lead to the emergence of new qualities, namely, the capacity for injection intrusion. The masses arising in this way in many respects become similar to magmatic ones...” N. A. Eliseev did not understand, or does not wish to understand, and by this revealed his complete philosophical helplessness. Yet this thought is undoubtedly correct, and for its sake alone one may forgive all the other sins of N. G. Sudovikov. I would only, instead of the last sentence in the cited quotation, write: “Thus granitic magma is born”;

f) in the conclusion of his article N. A. Eliseev discusses the question: are the ideas of the transformists progressive? And, naturally, he arrives at a negative conclusion, accusing the transformists of neglecting the data of experimental petrography, the doctrine of facies, the existing theory of metasomatism, etc.

He formulates his assessment of the ideas of transformism as follows: “Transformism took shape in isolation from reality. For it

is characterized by agnosticism, elements of mysticism, by ascribing to the phenomena of granitization and metasomatism something mysterious, not amenable to study by modern methods of investigation.”

Are the Ideas of the Transformists Progressive?

Indeed, in the article by N. G. Sudovikov, and also in the works of many other transformists, there are many extremes; many conclusions are in no way substantiated, and many objections can be raised against these views. But would it be correct to carry out a blanket disparagement of these new ideas and to deny completely their progressive role in petrology?

These views are unquestionably progressive, despite all the exaggerations and errors of the transformists and despite all the criticism of them by N. A. Eliseev and other magmatists.

They are progressive, first of all, because it was precisely the transformists who drew attention to the fact that rocks of magmatic aspect may also have another, non-magmatic origin. As a result of the investigations of the transformists, the old customary notion has been destroyed that the massive-crystalline structure of a rock is an indisputable and self-evident proof of its magmatic origin. Therefore at the present time every geologist is already obliged to take into account the various possible modes of formation of granitoids (and of other rocks of magmatic aspect) and to seek and cite in descriptions evidence not only of metasomatic, but also of magmatic origin. In connection with the development of the ideas of the transformists, the need has matured to develop genetic criteria for rocks of magmatic aspect. In other words, the new posing of the question of the genesis of granites stimulates the development of petrological science in general.

The views of the transformists are also progressive because only the transformists (true, only some of them), in the course of studying metasomatic granites and developing hypotheses of granitization, came close to the question of the causes and conditions of the birth of granitic magma and made more or less successful attempts in this direction. For the “magmatists,” by contrast, it is characteristic that almost all of them deal only with questions of the origin of rocks and carefully avoid, or simply do not consider it necessary to discuss, the question of the origin and development of the magmas themselves, regarding them as eternally existing, primordial (“parental” magmas of F. Yu. Levinson-Lessing), omnipresent (in Niggli’s expression), thereby revealing their metaphysical essence.

But here it is necessary to draw attention to one harmful and anti-scientific tendency inherent both in the extreme metamorphists (Perrin and Roubault, Backlund, partly Sudovikov) and in the extreme magmatists (Niggli, partly Bowen)—a tendency expressed in the desire to create a universal petrogenetic theory; I emphasize, not a “single” but a “universal” one, in the sense that in these “universal” theories the entire diversity of phenomena is explained by a single universal cause. For example, in Bowen and Niggli such a universal theory is the theory of crystallization differentiation of basaltic magma, by means of which they try to explain the origin of all magmatic rocks. Perrin and Roubault in general do not recognize a magmatic origin for intrusive rocks, considering all of them to be products of metasomatism in the solid state. It is quite obvious that any such “universal” theory, which does not take into account all the complexity of geological phenomena and their variability in space and time, is doomed in advance to failure, being metaphysical. V. I. Lenin, speaking of the dialectical laws of development, considered it necessary to emphasize that “phenomenon is richer than law,” whereas our metaphysical “universalists” try to force nature into an artificial scheme extrapolated to a greater extent...

part of a single special case, where it can be, and actually is, valid.

In this connection one cannot fail to mention Levinson-Lessing, who throughout his life fought against Bowen’s universal theory of the crystallization differentiation of basaltic magma and was among the first to draw attention to the circumstance that rocks of similar chemical and mineralogical composition, usually combined into a single classificatory group, may have different origins, i.e., may be heterogeneous. Recognition of the widespread occurrence of heterogeneous formations was reflected also in his genetic classification of igneous rocks, published in one of his last articles [31].

Survey of Views on the Origin of Basic and Ultrabasic Rocks

There is considerably less disagreement on the question of the genesis of basic and ultrabasic rocks. The fundamental question of their magmatogenic or metamorphogenic origin is resolved almost unequivocally by nearly all investigators, and almost no one doubts their magmatic genesis. The only exceptions are certain extreme transformists, who deny in general the magmatic origin of all intrusive rocks and connect the formation of basic and ultrabasic intrusions with processes of basification.

Substantial disagreements exist on the question of the independence of ultrabasic magma. Many, and perhaps even the majority of investigators (Bowen [10], Niggli [35], Levinson-Lessing [29]) consider that all ultrabasic rocks are products of the crystallization differentiation of basaltic magma. Along with them Holmes [56], H. Hess [54], Aleshkov [1], and D. S. Belyankin [7] incline to the same view, and adduce sufficiently convincing evidence that ultrabasic rocks are formed at the expense of a special ultrabasic magma arising during the melting of the peridotite substrate of the sima. H. Hess [54] moreover assumes that in nature there are two principal genetic types of hyperbasites—one of them is the product of the crystallization of ultrabasic magma, the other the product of crystallization differentiation—of basaltic magma.

There are also disagreements concerning the nature of the substrate at whose expense basic and ultrabasic magmas are formed. Bowen [10, 11] considers basaltic magma to be a product of fractional melting of the peridotite belt of the lithosphere. Most investigators acknowledge the existence of a basaltic belt, situated above the peridotitic one and serving as the source of basaltic magmas. At the same time, during the discussion on questions of cosmogony [46], held in April 1951, Chetverikov reported the results of the study of an artificial rock formed during the melting of marls in the underground combustion (gasification) of coal. It had an ophitic structure and consisted of anorthite, hedenbergite, and ferruginous olivine, i.e., the composition of a basic gabbro and an entirely “magmatic” appearance. Citing data from physical chemistry, he comes to the conclusion that basic magmas may be products of the selective melting of any basic material and that recognition of the existence of a “basaltic shell” of the lithosphere is not obligatory. V. N. Lodochnikov [34] had earlier arrived at the same conclusion on other grounds.

Finally, there are disagreements on the question of the causes of the diversity of basic and ultrabasic rocks. Some investigators see the cause of the diversity in phenomena of crystallization differentiation, others assign the chief role to assimilation, etc.; but we shall not dwell further on these questions here.

Initial Data for Constructing a General Hypothesis of the Formation of Igneous Rocks

Grounds for Distinguishing Three Genetic Complexes

At the first conference on questions of cosmogony, held in April 1951 at the Geophysical Institute of the Academy of Sciences of the USSR, the report by Academician O. Yu. Schmidt, “The Problem of the Origin of the Earth and the Planets” [46], was subjected to comprehensive discussion. O. Yu. Schmidt’s hypothesis was sharply criticized by a number of astronomers, but was warmly supported by the overwhelming majority of geophysicists and geologists. Indeed, this hypothesis accords better than others with the basic data of geology and geophysics and gives the geologist a good instrument for developing new geological, in particular petrological, hypotheses and theories. I shall briefly recall its basic propositions concerning the state and direction of development of our planet. The Earth arose through the accumulation of cosmic dust and meteoritic matter and was never in a completely molten state, being, on the contrary, a cold body in the first moments of its existence. The heating of the Earth as a result of radioactive decay began long ago, is taking place now, and will continue for many billions of years. This heating is also a powerful source of energy for mountain building. Another source of energy lies in the gradual differentiation (gravitational) of the Earth’s matter. Gravitational differentiation in a plastic, but viscous, heated medium proceeds very slowly, amounts to the rise of light particles and the vertical displacement of matter in general, and is the cause, for example, of deep-focus earthquakes. According to O. Yu. Schmidt’s calculations, this differentiation of matter can release as much energy as radioactive decay.

Subsequently, the state and structure of the Earth, as well as the principal forms of its motion, are considered from the point of view of O. Yu. Schmidt’s hypothesis.

It has long been established that the most widespread types of igneous rocks are, on the one hand, basalts and andesites, and on the other—granites and granodiorites.

Élie de Beaumont, as early as 1847, concluded that there exist two fundamentally opposed groups of igneous rocks: the first—a basic group, predominantly volcanic, but also including traps and serpentinites; and the second—acidic, predominantly plutonic and granitic. Each group is accompanied by its own suite of characteristic mineral deposits, and there are fundamental differences in the mode of emplacement between granite and other igneous rocks. Later it was established that granitic and granodioritic intrusions are spatially confined to folded zones, and in time—to phases of folding, whereas massive outpourings of lavas (predominantly basic) occur either before folding—in the stage of subsidence of a geosynclinal zone—or after folding, in connection with the emergence of deep faults. A. A. Polkanov [39, 40] clearly established fundamentally different forms of manifestation of magmatic activity for areas of the cratogen and areas of the orogen. The former are characterized by fissure intrusions and outpourings of basic magma, which are accompanied by sheet intrusions, lopoliths and laccoliths, as well as various ring and central intrusions, predominantly of the same basic, and rarely alkaline, magma. In the zones of the orogen there are clearly distinguished: a) pre-orogenic magmatism, manifested in the form of submarine (often deep-water) extrusions, dikes, and sheet intrusions of basic magma, which prove to be deformed together with the enclosing geosynclinal strata and appear in the greenstone

phase in the form of characteristic spilite-keratophyre and diabase series; b) synorogenic intrusions of acid magma of the batholithic type, the root parts of which constitute zones of migmatization and granitization; c) postorogenic, predominantly fissure intrusions and extrusions, mainly of andesitic lavas. That is, preorogenic and postorogenic volcanism of orogen zones has much in common with the volcanism of cratonic regions, but synorogenic volcanism is sharply distinguished by the preferential development of acid—granitoid—intrusions. In my work on the facies of igneous rocks [20] I emphasized that the extrusive and hypabyssal facies are characterized by the predominant development of rocks of basic composition or by products of differentiation and assimilation of basic magma, whereas the facies of intermediate depths and the abyssal facies are represented by granitoid intrusions.

All these indisputable regularities and generalizations of modern geology, as well as a number of other facts, make it possible to draw the conclusion that in nature there exist, first of all, two genetically independent complexes of igneous rocks: the granitoid complex, manifested in the abyssal and intermediate-depth facies, and the basaltoid complex, manifested in the hypabyssal and extrusive facies, i.e., to return to F. Yu. Levinson-Lessing’s conception of granite and basalt magmas, but on a new basis. These conceptions have also spread abroad. For example, Andersen and Kennedy [42] subdivide magmatic formations into two genetic groups: “volcanic” and “plutonic” associations, including in the first the products of basalt magma, represented by lavas and the small intrusive bodies accompanying them, and in the second—granitoid intrusions and the migmatites and granitized rocks associated with them. Read, as a result of his “ruminations on granite” [42], arrives at the conclusion that there exist three large classes of rocks: “Neptunian,” “volcanic,” completely corresponding to the volcanic association of Andersen and Kennedy, and “plutonic,” in which he combines granitoids and all metamorphic rocks—something with which, of course, one cannot agree at all.

Thus, there can be no doubt as to the reality in nature of independent basaltoid and granitoid genetic complexes of igneous rocks. But it seems to me that alongside them one may distinguish a third, special genetic complex of hyperbasite intrusions, which, however, is very limited in distribution.

Let us examine these three genetic complexes in more detail.

A. The Granitoid Complex

It has long been quite firmly established that acid igneous rocks occur predominantly in intrusive form; moreover, granitoid intrusions, like areas of granitization, are always located in folded zones, being also, in time, confined to the phases of the most intense folding movements that complete the formation of the given structural zone. Acid effusives of liparitic and dacitic composition in most cases are products of postorogenic volcanism (more rarely preorogenic), and the question of their genetic relationships with one or another magma will evidently be resolved differently in different cases.

Abyssal granitoids [20], occurring only among Archean and Lower Proterozoic crystalline schists, in the overwhelming majority of cases appear either as concordant akmolithic bodies accompanied by massive arteritic injection (migmatization), or else as zones of continuous metasomatic granitization. In both cases sharp contacts are absent, and gradual transitions to the enclosing rocks are present, as are replacement structures and inherited gneissic textures. In general, these intrusions in the overwhelming majority of cases

bear all the signs of metamorphogenic—metasomatic origin. Cross-cutting intrusions of granitoids in the abyssal facies are of subordinate origin. Abyssal granitoids are for the most part characterized by a uniform granitic composition, are represented predominantly by potassic (microcline) granites, and the appearance of more basic varieties is rare.

Medium-depth granitoids [20], in contrast to abyssal ones, in most cases form cross-cutting “intrusive” bodies, occurring with a sharp contact among folded, normally sedimentary strata (more rarely among crystalline schists) that have undergone, at the contact, hornfels recrystallization (superimposed in the case of contact with crystalline schists). Elements of independent internal prototectonics are often observed; linear textures are common, but they are not inherited. Medium-depth intrusions have mainly a granitic or granodioritic composition, but in the marginal zones syenitic, dioritic, and even gabbroic and hornblenditic rock types are common, which is connected with the considerable scale of assimilation phenomena. These intrusions, in general, give the impression of magmatic formations and do not reveal such clear signs of metasomatic origin as do the abyssal granitoids.

There is no doubt that it is precisely the disregard of the facies features of intrusives that is the cause of the sharp disagreements between the “magmatists” and the “transformists,” since most of the leading “transformists” have worked all their lives among Precambrian metamorphic strata, whereas most “magmatists” know well the intrusions of medium depths and have never dealt with the ancient Precambrian.

In my work “A Scheme for the Classification of Facies of Magmatic Rocks” [20] I distinguished, in addition to the abyssal facies of magmatic rocks, an even deeper—ultra-abyssal (charnockitic)—facies. The granitoids of this facies are very distinctive (hypersthene-garnet granites, granosyenites and granodiorites, hypersthene quartz diorites) and evidently arise under quite specific conditions, always being emplaced within a complex of crystalline schists of the hypersthene-gneiss facies. In the southern part of the Yenisei Ridge, the ultra-abyssal granitoids (“charnockites”) are characterized by massive occurrence, the appearance of igneous rocks, the presence of xenoliths of the enclosing rocks, and in some cases by cross-cutting contacts, with a complete, however, absence of manifestations of contact metamorphism. At the same time, however, the charnockites are characterized by the same mineral associations as the enclosing crystalline schists, and consist of variable quantities of quartz, orthoclase, plagioclase, hypersthene, monoclinic pyroxene, and garnet, differing only in a somewhat elevated content of potash feldspar. The structures of the charnockites are granoblastic, with a weak manifestation of hypidiomorphism. Among the charnockites, as also among the enclosing garnet and hypersthene gneisses, interbeds and lenses of coarse-grained material, enriched in quartz and orthoclase and sometimes possessing graphic structure, are frequently observed. At one time I came to the conclusion that these pegmatite-like formations were of lateral-secretion origin, and that the charnockitic bodies formed as a result of ultrametamorphism (palingenesis) without the participation of volatile substances in this process; I clearly emphasized the fundamental differences of the charnockitic formations from the later abyssal granites (the Tarak and Posolnaya intrusions), which arose as a result of metasomatic granitization and migmatite injection, carried out with a substantial participation of volatiles.

The well-known features of the conditions of occurrence, composition, and structure of granitoids of orogenic zones, partly set forth above, make it possible to draw the following general conclusions:

1. Granitoid bodies under different facies conditions manifest themselves differently and are formed by different pathways.

a) Ultra-abyssal granitoids (charnockites) are the result of “dry” ultrametamorphism, accompanied by some introduction of alkalies (enrichment in potassium feldspar, sometimes quartz). In the event that, as a result of such ultrametamorphism, significant volumes of rocks of granosyenitic composition are formed, they acquire mobility and become capable of intrusion; but such a “palingenetic” magma does not possess a sufficient reserve of energy, cannot cause contact metamorphism, and cannot travel far.

b) Abyssal granitoids—in a considerable part of their mass—are a product of granitization in situ, but of “wet” granitization, since the participation of volatiles in this process is beyond doubt and is proved by the metasomatic development of grid microcline, the biotitization of garnet and hypersthene, etc. Some abyssal granitoid bodies apparently represent a product of layered injection of magma, accompanied by phenomena of migmatization. More rarely, these are discordant bodies.

c) Mesabyssal granitoids are mainly discordant intrusive bodies formed by the emplacement of a mush-like magma, predominantly of granitic and granodioritic composition. Apparently, under conditions of mesabyssal facies, phenomena of granitization are also widespread. In any case, special investigations are needed to resolve the question of the mechanism of formation of mesabyssal granitoid plutons.

d) Hypabyssal granitoids (and acid effusives) represent the product of crystallization (and solidification in the form of glass), predominantly of homogeneous liquid magma. The question of the origin of this magma in different cases is apparently resolved in different ways. Part of the acid lavas and granitoids, judging by the composition of the magmatic complexes with which they are associated, almost certainly have a “basaltoid” origin. Another part, probably, has a “granitoid” origin. However, the ascent of granitoid magmas into the very upper layers of the lithosphere and their outpouring onto the Earth’s surface is a highly improbable phenomenon and, if it occurs, then rarely. In any case, mesabyssal granitoids in most cases formed from mush-like magmas, and hypabyssal ones from homogeneous magmas. The ascent into the upper layers of the lithosphere of a mush of liquid and crystals should lead above all to the loss of volatiles, an increase in viscosity, and crystallization of the magma, but in no way to its complete homogenization.

1. The diversity of the petrographic composition of granitoid bodies is also due to different causes under conditions of different facies.

a) In the ultra-abyssal facies, the petrographic diversity of “autochthonous” charnockite bodies is explained by the composition of the substrate that underwent ultrametamorphism, and partly by the amount of introduced potassium. The constancy of composition (hypersthene-garnet granosyenites) of “allochthonous” charnockites is probably due to the fact that these rocks have an anatectic composition; moreover, it is possible that precisely the granosyenitic, and not granitic, composition of the eutectic is caused by a shift in the position of the eutectic point toward the feldspars under conditions of very high pressures¹.

b) Abyssal bodies of granitoids are generally characterized by a very uniform, predominantly granitic, anatectic composition. A certain diversity, expressed mainly in changes in the content of biotite and microcline, is due to the degree of metasomatic feldspathization (granitization), and in other cases to assimilation and migmatization, i.e., to the whole aggregate of contact phenomena.

c) The petrographic diversity of mesabyssal granitoid intrusions is almost entirely due to contact phenomena (reaction assimilation and exocontact metasomatism); “gas” differentiation plays some role (granite-aplites of the apical parts of the granite intrusions of Altai). Other types of differentiation play no role in the formation of granitoid massifs. The structural and, in part, mineralogical diversity of abyssal and mesabyssal granitoid bodies may also be due to autometamorphic recrystallization and cumulative crystallization, as well as to autometamorphic and allometamorphic (superimposed) metasomatism (microclinization of granodiorites with the formation of porphyritic microcline granite, according to G. D. Afanas'ev [3, 4]).

d) Acidic hypabyssal intrusions of “granitoid” origin have not been distinguished, have not been studied, and are unlikely to be widely distributed.

1. Consequently, granitoids of various depth facies are either products of metamorphism and ultrametamorphism in situ, or products of the solidification of intruded masses—“magmas”—generated during these processes, as an expression of the emergence of a new qualitative state under the slow increase of quantitative changes in the thermodynamic conditions and in the composition of the substrate during metamorphism.

B. Basaltoid Complex

Naturally, the most reliable material for identifying the individual features of the basaltoid complex can be provided by Quaternary volcanism, whose products have been studied sufficiently well and for which generalizing works by Wolf [60] and A. N. Zavaritskii [14] are available. A comparative study of the products of Quaternary volcanism makes it possible, on the basis of a number of features of their chemistry and mineralogical composition, to distinguish among them above all two characteristic types, which have long been called the “Pacific” and “Atlantic” types. Intermediate between them is the “Mediterranean” type of rocks.

The Pacific type of Quaternary volcanic rocks is spatially associated with the Pacific and Alpine folded zones. It is characterized by a normal alkaline-earth series of magmatic rocks; moreover, andesitic, basaltic, and dacitic lavas are the most widespread, rhyolites are rare, and alkaline types are very rare. Naturally, lavas could not always break through to the surface and in part solidified at some depth, forming rocks of intrusive aspect. In any case, in deeply eroded volcanic edifices, dikes, stocks, and laccoliths composed of rocks of the gabbro, diorite, and peridotite types have been established; monzonite and essexite are encountered rarely. It may be regarded as beyond doubt that these intrusive rock bodies are genetically connected with the products of subaerial volcanism, being its intrusive (hypabyssal) facies. This is indicated sufficiently convincingly by their geological position and by the commonality of the chemistry and mineralogical composition of the lavas and the corresponding intrusive rocks.

The Atlantic type of rocks is characteristic of volcanic islands of the Atlantic, Pacific, and Indian Oceans, and is also widely distributed within the foreland of the orogen, being confined to major faults already within the platform. In this type, lavas of basaltic composition (including olivine-basaltic and trachybasaltic) are sharply predominant, and trachytes, phonolites, and various alkaline basaltoid rocks are characteristic, though they occur in small amounts. In deeply eroded volcanic edifices, intrusive bodies of gabbro, essexite, and other alkaline gabbroids, monzonite, nepheline syenite, and alkaline feldspar-free rocks have been found. Isolated occurrences of alkaline syenite and alkaline granite are known.

It should be borne in mind that, in the overwhelming majority of cases, Quaternary volcanism manifested itself in the form of volcanoes of the central type; moreover, this volcanism is either post-orogenic (Pacific type) or the volcanism of regions of a craton in the stage of uplift. As for mass fissure eruptions, accompanied by dikes and sheet intrusions, they are characteristic of regions and periods of subsidence and occur, on the one hand, in craton zones (trap formations), and on the other—in orogen zones at the stage of formation of geosynclinal sediments (spilite-keratophyre formations).

Trap formations have long been known, are widely distributed, and have been well studied. They are characterized, in general, by a very uniform composition of effusive and intrusive representatives (basalts and gabbrodiabases or dolerites) and by generally insignificant differentiation, although in individual cases (the Siberian trap formation, the Deccan Traps) the appearance of small quantities of alkaline rocks is established. In the roof of the dolerite intrusions of the Karroo, zones of granophyric rocks are not uncommon in the upper contact of dolerite sills, as are veins of granophyres intruding the dolerite. Walker and Poldervaart [47] present convincing evidence of their rheomorphic origin.

In spilite-keratophyre formations we usually encounter a more diverse complex of rocks, whose individual features, however, are always strongly obscured by phenomena of autometamorphism (albitization). At the same time, the usual sharp predominance of spilitic rocks over keratophyres is established quite distinctly, and suppositions are put forward that spilite-keratophyre formations fully correspond to the normal alkalic-earth (Pacific) type, while their individual features are conditioned by the geological circumstances of formation (submarine eruptions).

Everything that is known concerning the principal features of the entire basaltoid (volcanic) complex permits the following conclusions to be drawn:

1. It is beyond doubt that all the rocks of the complex, both intrusive and extrusive, are products of the crystallization or solidification of magma, i.e., are magmatic in the full sense of the word.

2. Since basalts and dolerites (gabbrodiabases) are the sharply predominant quantitative type of rocks in this complex, and since, it would seem, one cannot name a single volcano that did not pour out basaltic lavas along with others, the conclusion is natural that the magma ancestral to the whole complex was basaltic in composition. This conclusion has long since become customary for everyone and is confirmed by all new data.

3. The chemical-mineralogical diversity of the rocks of the basaltoid complex is conditioned by processes of differentiation and assimilation. At the same time, the diversity observed in the Atlantic type, and also in trap formations, and expressed in the formation of rocks of the alkaline series, is almost certainly due to processes of differentiation, chiefly crystallizational, whereas the diversity of rocks of the Pacific type (the normal alkalic-earth series) is probably conditioned not only by differentiation but also by assimilation of a considerable quantity of sialic acidic rocks.

4. The wide distribution of granophyres of rheomorphic origin in the roof of dolerite sills of the Karroo, as well as the long-known cases of remelting of granites in contact with diabases, suggest that the intrusion of hot basaltic magma into sialic rocks may lead not only to assimilation but also to remelting of the latter. The possibility is not excluded that liparites and rhyolites associated with basalts in the absence of intermediate varieties are products of the eruption of such acidic magmas of rheomorphic origin.

1. “Basaltoid” volcanism is manifested not only by surface eruptions, but also by intrusive bodies accompanying the latter, emplaced chiefly in a hypabyssal setting; i.e., the products of “basaltoid” volcanism may have both an effusive and an intrusive aspect. The spectrum of rocks of intrusive aspect that arise as a result of the differentiation and assimilation of basaltic magma is varied, but what is characteristic is that normal granites and granodiorites are practically absent among them. They are represented chiefly by rocks of the gabbro and dolerite type and by alkaline gabbroids, more rarely by diorites, monzonites, hyperbasites, nepheline syenites, and very rarely by alkaline syenites and alkaline granites. Examples of intrusions genetically connected with effusive complexes and having a “basaltoid” origin are: the post-Devonian alkaline intrusions of the Aldan region; the trap intrusions of the Siberian Platform; the Stolbovaya intrusion of nordmarkite in the vicinity of Krasnoyarsk, the Abaza albite rocks, possibly the Maina “granodiorite” intrusion, etc.

B. The hyperbasite complex

Among the ultrabasic rocks at least two genetic types are quite clearly indicated. One of them is represented by weakly serpentinized pyroxenites, dunites, and peridotites, associated with gabbro-norite rocks and linked with the latter by gradual transitions. Ultrabasic rocks of this genetic type either participate in the structure of large lopoliths (for example, the Bushveld), or form small independent bodies known to me, for example, from the Kuznetsk Alatau and the Yenisei Range and described in a number of other regions. Their genetic connection with gabbro and norites is entirely obvious, and it is beyond doubt that ultrabasic rocks of this genetic type are products of the differentiation of basaltic magma and, consequently, are a constituent part of the “basaltoid” complex.

We are now interested in another genetic type of ultrabasic rocks—the type represented by serpentinite apoharzburgite intrusions of “serpentinite (ophiolite) belts.”

This type of serpentinite intrusion, comprehensively studied by V. N. Lodochnikov [32], is usual only for folded regions. The intrusions are characterized by a lens-shaped form; by the vertical occurrence of the serpentinite lenses; by an exceptionally weak contact effect on the country rocks; by a very uniform serpentinite composition, with the serpentinites in the overwhelming majority of cases developing after harzburgites, much more rarely after dunites and pyroxenites; unserpentinized varieties of these rocks are very rare; equally rare are feldspathic rocks associated with the serpentinites, except for plagioclase veins, which are rather hydrothermal formations.

As a result of discussing the question of the origin of serpentinite intrusions, V. N. Lodochnikov [32, 33] came to the quite categorical assertion that in no case can they be products of the crystallization differentiation of basaltic magma. Holmes [56], Aleshkov [1; 2], and Hess [54] insisted on the independence of ultrabasic magma, with Hess attempting also to explain the mechanism of intrusions of hyperbasite magma arising as a result of selective melting of the peridotite belt of the lithosphere. The idea of the independence of ultrabasic magma is supported by D. S. Belyankin [7]. It seems to me that these considerations concerning independent hyperbasite magma deserve every attention and should be accepted.

An Attempt to Construct a General Hypothesis of the Formation of Igneous Rocks

Igneous rocks—heterogeneous formations

Thus, it may now be considered firmly established that basaltoid, granitoid, and hyperbasite complexes are formed on the surface and within the lithosphere separately in space and time, being connected with various forms of geotectonic movements. Series of rocks connected with granitoid, basaltoid, and hyperbasite magmatism are very distinctive. At the same time, however, it is necessary once again to emphasize that convergence of features in derivative basaltoid and granitoid magmas, in basaltoid and hyperbasite magmas, and in igneous proper and certain metamorphogenic formations is a quite ordinary matter; moreover, in many cases rocks very close in their chemical-mineralogical composition and structural features arise by entirely different paths.

Consequently, it may now be considered firmly established that many, if not the majority, of the groups and families, and even species, of “igneous” rocks distinguished in petrographic classifications may arise in different ways, i.e., may be derivatives of granite, basalt, and hyperbasite magmas, and also products of granitization or basification, i.e., essentially metamorphic formations. The diversity of igneous rocks arising from each separate type of initial magma may be due to the phenomena of assimilation and differentiation. In general, by the present time quite enough data have been accumulated to speak of the convergence of features in genetically heterogeneous igneous rocks as a perfectly ordinary and normal phenomenon, and also to suppose that the entire aggregate of rocks that it is customary to call igneous constitutes a complex of very heterogeneous formations. At the same time, phenomena of heterogeneity are found not only in syenitic and dioritic types, which was established already by F. Yu. Levinson-Lessing [29], but are spread much more widely, being found in all types; moreover, the number of examples of heterogeneity is continually increasing. Thus, heterogeneity among igneous rocks is a regularity, and not an accident, as it seemed earlier. Theoretically this regularity is justified by the fact that the mineral composition of crystalline (igneous and metamorphic) rocks is wholly determined by the values of the coordinates “x,” “t,” “p” during the process of rock formation and does not depend on the origin of the initial substance; the structure of these rocks, however, is determined not only (and perhaps even “not so much”) by the mode of formation as by the properties of the mineral components composing the rock. From this the following conclusion should be drawn: 1) all attempts to create a “universal” theory of the formation of igneous rocks have been, and will be, entirely fruitless (see p. 20); 2) the question of the genesis of igneous rocks must be resolved for each type separately and with consideration of the entire geological setting and the whole history of its development.

A. Origin of the granitoid complex

On the relationship between magmatism and metamorphism

We have already had occasion to convince ourselves that there exist very different points of view on the origin of granitoids; moreover, some investigators, for example, consider them igneous formations, while another part consider them metamorphic. What has been set forth above permits one to assert that not ...

others are right and wrong, and that granitoids may have both a magmatic and a metamorphic origin. Closely connected with the question of the origin of granitoids is another, broader and more general question—the question of the relation between magmatism and metamorphism, which is likewise resolved differently by different investigators and which it is absolutely necessary to clarify if we wish to approach a solution of the question of the origin of granites.

Among most modern investigators there is a definite tendency to connect with magmatic activity not only contact metamorphism proper, but also the formation of crystalline schists. For example, D. S. Korzhinsky particularly emphasizes that “all metamorphism of silicate rocks is connected with magmatic phenomena and takes place under the direct influence of magma or postmagmatic solutions” [16]. But alongside these prevailing conceptions there are also others, according to which regional metamorphism and the formation of crystalline schists are processes completely independent of magmatic activity. Harker, for example, states outright that “it seems more natural to regard intrusions as an ordinary particular instance of regional metamorphism than to accept them as the cause of the latter” [49].

It seems to me that, on this question of the relation between acid magma and metamorphism as well, the defenders of both opposing points of view are right, but right not in general, only in specific particular cases. From my point of view, the cause of the disagreements on the question of the relation between magmatism and metamorphism, just as on the question of the origin of granites, lies in the metaphysical limitation of the defenders of extreme and opposing views; in the fact that both magmatism and metamorphism are considered in frozen forms, without regard to their historical development and their interaction with the surrounding environment; in the fact that particular regularities are taken for general ones.

This question of the relation between metamorphism and volcanism was discussed by me in sufficient detail in a work published as early as 1941. The principal conclusions on this question may be formulated as follows:

1. It may be considered proven that in the root parts of geosynclinal zones, in connection with their folding, metamorphism develops, expressed in the development of crystalline schists and independent of magmatic activity.

2. This metamorphism, at great depths, in zones that are especially strongly heated, may grow over into melting, i.e., magma formation, which is an expression of the transition of quantitative changes into qualitative ones.

3. Magma born at depth acquires the capacity for independent movement, and the process grows over into its opposite... Magma born at great depth and arriving at shallow depths comes into contradiction with the surrounding rocks. The magma–continent system proves to be nonequilibrium, which serves as the cause of contact phenomena expressed in assimilation and contact metamorphism.

I consider these conclusions correct even now. The question of the relation between metamorphism and volcanism must be resolved not in general, but with regard to the development of the process and the given concrete setting. Regional metamorphism of great depths and acid magma in the stage of its origin are generated by a common cause, are paragenetically connected with one another, and both the magma and the silicate masses from which it arose are in equilibrium with one another. But magma that has moved from the place of its birth into the upper zones of the lithosphere inevitably comes into contradiction with the surrounding

medium, which is expressed in the active contact effect of the magma on the surrounding rocks, i.e., here the magma already causes metamorphism.

The question of the origin of granitoids must be resolved in an entirely analogous way. Magmatic granitoids undoubtedly exist—these are granites of the upper structural levels, having an obviously intrusive mode of occurrence. But at the same time, granitic magmas that arise at depth in the lower structural levels must have their own prehistory, and before magma appears—a qualitatively new formation—the substrate at whose expense it is born must undergo gradual quantitative changes, expressed not only in its heating, but also in the gradual approach of its composition to that of granite. This “premagmatic” stage of development may also be called the stage of granitization, since during it the substrate gradually acquires the composition of granite. Consequently, in nature there exist both magmatic and metamorphic granites, but they take shape at different stages in the development of the process and in different geological settings. Such is the dialectic of magmatic phenomena.

Let us consider the process of granite formation in its development.

Stage of Granitization

Above we defined granitization as a metasomatic process accompanied by the introduction of some elements and the removal of others, as a result of which a solid rock (without passing through a magmatic stage) becomes more similar to granite than it was before and, ultimately, acquires a granitic composition. Consequently, granitization necessarily requires an influx of matter and, apparently, an influx of energy.

The question of the source of the energy required for granitization is common to the question of the source of the energy of tectonic processes, which was considered by P. N. Kropotkin [18] and a number of other investigators. The cosmogonic theory of O. Yu. Schmidt provides for the phenomena of radioactive decay and gravitational differentiation of the Earth’s matter as the main sources of energy for tectonic processes (and consequently also for granitization). However, in the present case this question seems to me secondary.

Far more important is the question of the influx of matter. The composition of the substances migrating during granitization has been considered by many investigators, and recently by N. G. Sudovikov [45]. One may agree with him that, in the general case, granitization is accompanied by the introduction of alkalies and silica and by the removal of iron, magnesium, and calcium. This question of the composition of the migrating substances is resolved in each individual case relatively simply by comparing the compositions of the original rocks and the products of granitization. Much more complex, and still in no case resolved, is another question—where do the alkalies and other substances necessary for granitization come from? Some investigators (Chetverikov, Belousov) seek the source of the granitizing substances in the central parts of the planet, linking their migration upward with the process of its differentiation that is still continuing, according to the hypothesis of O. Yu. Schmidt [40]. Others try to find local sources of these substances. It seems to me that this question, too, must be resolved differently in individual cases.

For example, in the southern part of the Yenisei Ridge, granitization phenomena appeared during the formation of the Lower Archeozoic Kansk metamorphic complex and the charnockites associated with it, and also during the formation of the Posolnaya and Tarak granite intrusions.

The Kansk metamorphic complex is composed mainly of garnet and hypersthene gneisses, which are products of the recrystallization of sedimentary rocks. Among the garnet gneisses, consisting of quartz, ande-

sillimanite, garnet, and a small amount of biotite, garnet–orthoclase gneisses are very often encountered, occurring as small lenses and streaks, veinlets, and irregular segregations with gradual transitions to the enclosing garnet–plagioclase gneisses. In some cases entire outcrops were observed composed of massive garnet–orthoclase rocks, among which, however, there are always blurred bands and lenses of garnet–plagioclase gneisses. I at one time regarded these garnet–orthoclase rocks as products of metamorphic differentiation (i.e., as lateral-secretion formations), accompanied by migration of substances, the metamorphic differentiation having developed into palingenesis, the product of which was the various charnockites (see above). These observations and the conclusions drawn from them 15 years ago may be translated into modern terms as follows: 1) these lens-shaped and vein-shaped masses of garnet–orthoclase gneisses, as well as all charnockitic rocks, are products of granitization, and the granitization reached the stage of formation of granitoid rocks; upon reaching this stage the granitized material acquired mobility and could form intrusions; 2) granitization was accompanied by an influx of potassium and a small amount of silica; 3) the source of the potassium was undoubtedly the enclosing masses of garnet–plagioclase and hypersthene–plagioclase gneisses, now completely devoid of potassic feldspar; i.e., granitization here is a particular case of metamorphic differentiation.

Incidentally, a number of features of the structure and mineralogical composition also make it possible to speak about the mode of transfer. The complete absence of replacement phenomena, the preservation in the granitized rocks of the same garnets and hypersthenes that characterize the enclosing rocks, the absence of zoned plagioclases, the orthoclase character of the potassic feldspar, the ideal freshness of all components of the rock, and the absence of “secondary” minerals—all this compels one to speak of sterile conditions of metamorphism and granitization and completely excludes any possibility of infiltration of solutions. The indicated features of the granitized rocks compel one to think of diffusion of potassium, and of diffusion rather in the solid state.

Granitization manifested itself quite differently, for example, in the formation of the “Posolnaya intrusion,” which is composed of very uniform microcline granites and is accompanied by a broad field of rocks feldspathized to varying degrees; moreover, the granites themselves have a clearly metasomatic character, expressed, in particular, in inherited textures, widely developed phenomena of replacement of plagioclase by microcline and of amphibole by biotite, etc. The granites and feldspathized rocks develop at the expense of various primary substrates—partly at the expense of Archean crystalline schists, partly at the expense of weakly metamorphosed sedimentary rocks of the Lower Proterozoic; moreover, the zones of granitization can be traced for hundreds of kilometers, being distinctly confined to regionally developed tectonic zones. In this case granitization was also accompanied by an influx of alkalis (chiefly potassium, less sodium) and insignificant amounts of silica, as well as by the removal of calcium, magnesium, and iron. But the mechanism of granitization was undoubtedly different. Judging by the developed structures of metasomatic replacement, by the phenomena of biotitization of garnet and hypersthene in the rocks of the Kansky metamorphic complex during their granitization, etc., one may think that granitization here proceeded with a significant participation of volatile substances and most likely had an infiltration character, the source of the granitizing solutions being sought somewhere at great depth; in any case it is not local.

Thus, the process of granitization proceeds differently in different cases.

Magmatic Stage

It seems to me that N. G. Sudovikov is quite right when he says: “...the entire gradual process of granitization, with the subsequent transition into a rheomorphic and intrusive state, may be regarded as a certain quantitatively increasing process leading to a new quality. Slowly developing granitization and the gradual increase of the liquid phase lead to the emergence of new qualities, namely—the capacity for injection, intrusion” [43, 96]. But the sentence following this quotation: “The masses arising in this process become in many respects similar to magmatic ones,” already gives rise to serious objections, since precisely these mobile masses arising during granitization should be called “granite magma.”

Are other modes of origin of granite magma possible? Certainly they are possible, but it is unlikely that they play any major role in the formation of granitoids. For example, cases of remelting of granitoids at contacts with diabases have repeatedly been described; rheomorphic granophyres formed at the expense of argillites at contacts with the dolerites of the Karroo have been described [47]; and the idea has been expressed of the rheomorphic origin of the red granophyric granites of the Bushveld complex [53]. In general, the formation of granite melts and granite rocks under the influence of the thermal and emanational action of basaltic intrusive masses on the surrounding rocks, close to granite or granitoid composition, is apparently quite possible, but this mode can hardly have any major significance, above all because some special conditions are required under which the country rocks undergo remelting or rheomorphism, but phenomena of assimilation of the substance of these country rocks by basaltic magma are absent.

The ideas of the origin of granite magmas as a result of selective melting of sialic material, first advanced by Eskola [52], and supported, for example, by Eliseev [13] and Afanas’ev [4], and formerly shared by me, enjoy wide popularity. Theoretically, the selective melting-out of the most fusible mixtures upon simple heating of any multicomponent system is a wholly natural process. When sialic material of the earth’s crust is heated, the first portions of the liquid phase should have a granite composition, since granites may be considered an anchi-eutectic system. The amount of this granite liquid depends on the composition of the initial material and will be large only when the initial material has a granite composition. In the general case, however, this eutectic liquid will quickly cease to be eutectic (i.e., will cease to be granitic) as soon as, in the progressing process, components begin to melt that are in excess relative to the composition of the eutectic mixture.

Thus, the process of selective melting may give different results: 1) when ancient granites are melted, a homogeneous, in any case purely granite magma may arise. But this process will not be a process of selective melting, since by the latter is always meant the melting-out of fusible eutectic alloys from mixtures of non-eutectic composition; 2) during selective melting of non-granitic initial material, various amounts of granite melt may be obtained. In the event that the amount of liquid is sufficient, the entire mass of partially melted material may acquire mobility and become capable of intrusion, but this mass will not have the composition of granite—it cannot differ substantially from the composition of the original rock. Obviously, this theoretical case is not suitable for explaining the origin of granite intrusions. In the same case, if the amount of liquid is small, it likewise cannot give rise to granite “magma.” First of all, this

intergranular liquid, according to our definition, is not yet “magma.” In order for this intergranular liquid to be able to give rise to “magma,” its concentration is necessary, i.e., its displacement in space. But such displacement of the melt by mechanical infiltration is impossible so long as the liquid is in a film state and, probably, is greatly impeded in liquids filling capillaries. The chemical action of this newly born liquid upon the surrounding crystals is also excluded, since at the moment of its birth the liquid is in equilibrium with the surrounding medium. In general, these ideas concerning the possibility of the birth of granitic magmas as a result of selective melting of non-granitic material, without its preliminary granitization, seem to me improbable.

Thus, the most important and most typical case of the birth of granitic magma may be represented as follows.

In the root parts of an orogen (a geosynclinal zone), in connection with intensive tectonic movements and as a result of a rise in temperature (and, possibly, also of the simultaneously beginning migration of matter), phenomena of regional metamorphism begin to develop, expressed in recrystallization with the formation of mineral associations stable under the conditions of the given temperature and pressure. Recrystallization is necessarily accompanied by metamorphic differentiation, which manifests itself with particular intensity in the deepest zones. The essence of the process of metamorphic differentiation is: dissolution, counter-diffusion of the most mobile components (primarily alkalies and alkaline earths), diffusion metasomatism with the development of porphyroblasts, concretions, and interbeds, whereby further enrichment takes place in a portion of the original rock initially relatively enriched in the given component. One manifestation of metamorphic differentiation is quartz-feldspar “lateral-secretion” veins, lens-shaped and sheet-like segregations of granitoid rocks among crystalline schists of non-granitic composition, and similar formations described by many investigators, beginning with Holmquist [55], including the present author. Since in such metamorphic differentiation the selective enrichment of certain portions of the rock in alkalies and depletion in alkaline earths is very common, as a result of which the composition of these portions approaches that of granite, it may be asserted that granitization is a special case of metamorphic differentiation, but a special case of particular significance and sometimes manifested on a colossal scale, probably precisely because the most mobile components take part in this form of metamorphic differentiation.

Attention has already been drawn above to the fact that the process of granitization may proceed in different ways. In the case where the migration of matter was provided only by the diffusion of atoms in a solid medium or in film liquids, and where volatile components did not participate in the migration, the result of granitization proves insignificant because diffusion can manifest itself only over limited distances; lateral-secretion veins, granitoid rocks unmoved or displaced over small distances, and charnockite masses arise (see p. 24). In the same case, if not only alkalies and alkaline earths but also considerable masses of the most mobile volatile components were involved in the migration of matter in the process of metamorphic differentiation, there was an opportunity for not only diffusion of atoms to manifest itself, but also infiltration of gases or very dilute solutions. The second condition for the possibility of infiltration is probably the presence of weakened tectonic zones. It is apparently no accident that granitoid intrusions, zones of granitization, and zones of particularly intensive metamorphism are associated with deep-seated faults. Under these conditions the migration of matter may already proceed as

over very great distances, and metamorphic differentiation leads to the creation of sometimes enormous masses of rocks granitized to varying degrees and of “metasomatic” granites. At the same time, I believe that the difference between veins of granitic composition in lateral-secretion veins and massive metasomatic granite is quantitative rather than qualitative.

At a certain stage of the process of metasomatic granitization, when, owing to the local concentration of a number of components, including volatiles, the composition of the granitized material approaches that of granite or, otherwise, the composition of the ternary eutectic—quartz + potassium feldspar + acid plagioclase—melting begins under continuing metasomatism. When the volume of the liquid phase reaches approximately 25% of the total volume, the entire system—crystals + liquid—acquires the ability to flow as a whole. This moment may be regarded as the moment of birth of granitic magma, possessing, for the time being, only the potential capacity to form granitic intrusions.

Newly born granitic magma that has not been displaced in space is in equilibrium with the surrounding medium and, naturally, is incapable of any contact action or phenomena of assimilation. Its aggregate state may vary. It may be thought that in a tectonically quiet setting the magma, even before its intrusion, may completely homogenize and then intrude in the form of a homogeneous melt. In a tectonically disturbed setting the magmatic mass begins to move early, before complete homogenization; a mush of crystals and liquid intrudes. In almost all cases granitic magmas concentrate within themselves large quantities of volatile components brought into the zone of granitization during the stage of metasomatism. Therefore, having a comparatively low temperature, they possess high chemical activity, which, however, manifests itself only when the magma comes into motion and comes into contact with systems not in equilibrium with it. Once intrusion of granitic magma has begun, it subsequently develops by itself, sustained by the energy accumulated in the magma during the stage of granitization and manifested chiefly in the chemical activity of the magma. In this process the upward advance of the magma can apparently occur in two ways: as mechanical injection and by gradual advance upward and to the sides of the front of granitization and magma formation, i.e., by “magmatic replacement” [17].

The birth of acid granitic magmas in time is apparently confined to the most intense phases of geosynclinal folding. Upon completion of folding, magma situated at depth under colossal pressure gains the possibility of independent ascent into the upper zones of the lithosphere. It is quite evident that by the time of arrival of the intrusion the sedimentary strata of the upper zones have already undergone their folding, foliation, and sometimes regional metamorphism. And thus magma, born at great depth and arriving at shallow depths, finds itself in contradiction with the surrounding medium. The system magma + country rocks here proves to be fundamentally nonequilibrium, and the degree of disequilibrium is determined essentially by the difference between the temperatures of the magma and of the country rocks and by the difference in their chemistry, including the amount and composition of volatiles. In striving to create a new equilibrium appropriate to the given setting, it assimilates the wall rocks, at the same time giving the latter an excess of heat and its own constituent parts, i.e., causing contact metamorphisms of various types.

Causes of the Diversity in the Composition of Granitoid Intrusions of Intermediate Depths

The considerable petrographic diversity of granitoid intrusions of intermediate depths may be due to various causes: 1) above all

in all, even in the zone of origin, acid magmas of somewhat different composition may arise, chiefly granitic and granodioritic; 2) during its ascent, granite—or, more correctly, granitoid—magma inevitably assimilates considerable quantities of foreign material; 3) assimilation continues also at the site of the final emplacement of the intrusive body, and this process leads to the appearance, in the marginal zones and schlieren masses, of a greater part of material of a more basic composition than granite; 4) in particular cases an important role is played by gaseous differentiation, which accounts, for example, for the appearance of aplite-like granites in the apical parts of the Altai granite intrusions; 5) under conditions of medium-depth intrusions, processes of metasomatism are of great importance, in particular with the formation of rocks of magmatic aspect and even of granites, i.e., processes of granitization are characteristic also of intermediate depths, but here granitization is already caused by magmatic emanations released by the granite magma during its cooling and crystallization. Examples of metasomatic rocks of magmatic aspect may be the metasomatic diorites from the Central Caucasus described by G. D. Afanas’ev, as well as the porphyroid microcline granites that arose as a result of metasomatic microclinization of granodiorites [4, 5]; 6) finally, one must take into account recrystallization and accumulative crystallization, which apparently proceed on a large scale in the postmagmatic stage of the life of acid intrusions, as A. N. Zavaritskii recently pointed out [14]. As a result of this process, not only pegmatites may arise, but the appearance of the entire mass of granite may be substantially altered. For example, the constant confinement of quartz veins to bodies of aplites occurring among the granites of some Altai intrusions is probably explained by recrystallization of the granite with the development of aplite along a fissure through which solutions circulated, depositing quartz and other vein minerals.

As for the processes of crystallization differentiation, they apparently play no role in the formation of acid intrusions.

It remains unclear to this day whether the outpouring of granite magma onto the surface and the formation of acid extrusions and hypabyssal intrusions of granitic composition are possible. In the composition of Quaternary lavas, liparitic lavas play a wholly subordinate role, occurring chiefly together with intermediate, andesitic lavas, which indicates their probable connection with basaltic magma. But in more ancient volcanic series, massive outpourings of predominantly acid lavas, with a wholly subordinate role of basic ones, are not uncommon. An example may be the Middle Devonian keratophyre-spilite series of Rudnyi and Central Altai, in which quartz keratophyres and quartz albitophyres are the decidedly predominant type. It is difficult to explain the compositional features of this series otherwise than by the outpouring of a special acid magma.

B. Origin of the Basaltoid Complex

The complete lack of development of the question of the origin of basaltic magmas

Strange as it may seem, despite the long dominance in petrology of the hypothesis of a single parental basaltic magma, the question of how and by what means this magma is formed remains, to this day, wholly undeveloped; and, for example, the mechanism of formation of granitic magma is now more understandable than the mechanism of formation of basaltic magma. Evidently, reference to an eternal and omnipresent parental basaltic magma fully satisfied the majority of investigators. At the same time, the question of the origin of basaltic magmas is far from as simple as it appears at first glance, and it breaks down into a number of particular questions.

The question of the substratum at whose expense basaltic magmas are formed. The usual notion that basaltic magma arises at the expense of a special basaltic shell or belt of the lithosphere is by no means shared by everyone. For example, V. N. Lodochnikov [34] decisively denied the possibility of the existence of such a shell. Bowen believed that basaltic magma arises as a result of selective melting of a peridotite substratum. Recently new data have appeared which also argue against a basaltic belt. Among these are new data on the connection of volcanism with deep faults and deep-focus earthquakes, which make it possible to speak of the inception of magmatic foci at least at depths of 100–120 km, i.e., deeper than the lower boundary of the distribution of the hypothetical basaltic belt and within the limits of a less hypothetical simatic shell. All this compels one to approach very cautiously the determination of the character of the substratum at whose expense basaltic magmas are born—magmas which, incidentally, are by no means so uniform in their composition as was formerly thought.

In exactly the same way, the question of the mechanism of magma formation remains completely unclear; this question is connected with the problem of the aggregate state of the hypothetical basaltic substratum. In the case of its glassy state the matter is resolved simply. Magma formation is the lowering of the viscosity of the glass in connection with a rise in temperature or a lowering of pressure. But the existence of the glassy basaltic belt itself cannot be reconciled with present-day ideas about the structure of the Earth. If, however, the basaltic belt has a crystalline structure, it is difficult to admit the possibility of its melting as a whole without the preliminary melting-out of albitic or oligoclase eutectic magmas, the products of crystallization of which would have to be encountered far more often than is actually the case. Therefore the more probable supposition is that basaltic magmas arise through selective melting of “sima,” presumably having a peridotitic composition.

Finally, the question of the composition of the basaltic magma itself also remains unclear. At present the existence of at least two types of basaltic magmas is established quite distinctly—the Atlantic (olivine-basaltic or trachybasaltic) and the Pacific (tholeiitic), each of which yields its own complex of differentiates. But it remains unclear whether both types are independent, arising in different geotectonic settings and perhaps at the expense of different substrata; or whether both types are products of differentiation of some third type of basaltic magma; or, finally, whether one of these types is derived from the other. The regular confinement of the Atlantic type to oceanic islands and of the Pacific type to continents and especially to folded zones makes very tempting the supposition that tholeiitic magma is a product of contamination of the primary olivine-basaltic magma by sialic material.

Likewise, the cause of the origin of basaltic magmas remains unclear, and only the connection of manifestations of basaltic volcanism with fracture tectonics and with areas of predominant subsidence is clear. One might therefore think that the immediate cause of melting is a lowering of pressure in the substratum; but this is contradicted by the connection of modern volcanism of the Pacific belt with deep faults dipping beneath the continent and therefore having, rather, an overthrust character.

All these questions have in essence still scarcely been discussed and are very far from being resolved. At the same time, the periodic appearance in the lithosphere of basaltic magmas in enormous quantities and of very constant composition remains a fact with which we must reckon.

Causes of the Diversity of the Petrographic Composition of the Basaltoid Complex

Clearer is the question of the subsequent fate of basaltic magmas and of the composition of the rocks of the basaltoid complex, which has already been partly discussed above. It may be considered established that basaltic magmas are capable both of assimilating foreign material and of differentiating.

Differentiation in its pure form is manifested in the Atlantic (olivine-basalt) type and leads to the formation of rocks of the alkaline series—alkaline basaltoids, trachytes, phonolites, and intrusive rocks of corresponding composition, predominantly hypabyssal. In genetic association with such a basaltoid complex there may even appear alkaline syenites and alkaline granites, but the latter have been encountered in such regions (the post-Jurassic intrusions of Aldan, the trap formation of the Deccan) where phenomena of assimilation cannot be regarded as excluded. Since such alkaline complexes of basaltoid origin are confined chiefly to the inner parts of oceanic areas and to platforms, i.e., they develop in anorogenic regions with an absent or thin sial layer, it may be thought that precisely these conditions are favorable for the manifestation of differentiation in pure form with the formation of alkaline types.

The diversity of rocks arising from magmas of the Pacific (tholeiitic) type is more probably connected with the combined manifestation of assimilation and differentiation, as a result of which basaltic magma gives rise to the normal calc-alkaline series: basalt—andesite—dacite—rhyolite, and to intrusive rocks of corresponding composition. Evidence for the manifestation of assimilation by basaltic magma of the material of the country rocks, with the appearance of more acid varieties, up to rocks of the quartz-diorite type, is abundant; moreover, such assimilation is expressed very effectively even in diabase dikes, for example, in the diabases of the southern part of the Yenisei Ridge or in the diabases of the Tomsk region. One might suppose that assimilation is the sole cause of the appearance of intermediate and acid rocks of basaltoid complexes. However, the fact that the tuffs ejected from the vent of a volcano when its activity is renewed always have a more acid composition than the lavas subsequently flowing out, and also comparison of the composition of lavas that erupted simultaneously in 1937 from the Klyuchevskaya Sopka volcano from parasitic craters at the base of the volcano and at its summit [15], show that in the vents of central-type volcanoes gravitational differentiation is realized comparatively easily and rapidly, leading to the accumulation in the upper part of the magmatic column of relatively acid and light material. Volcanic complexes of the Pacific type are known to be confined to zones of folding, being postorogenic formations, and consequently are formed in a tectonically unstable setting and in zones with sharply thickened sial, which, of course, should favor the possibility of assimilation of acid material and the corresponding change in magma composition.

Assimilation by basaltic magma of foreign material, at least judging from observations in diabase dikes, is expressed almost exclusively in the simple remelting or dissolution of foreign material entering the still liquid-magmatic stage, as is evidenced by the frequently occurring grains of melted xenogenic quartz or microcline in dike diabases or basalts.

The processes of differentiation of basaltic magma proceed differently in different cases. Edwards’ investigations [51] suggest that differentiation processes lead to substantially different results when manifested in closed (sill, lopolith) or “through” (dike) chambers. In the first case, according to Edwards, differentiation leads only to the accumulation of iron in the residual melts; in the second—to the appearance of acid rocks of the nor-

of the alkaline-earth series. A comparison of the chemistry of five successive cycles of eruptions of Vesuvius, each of which began with the ejection of trachytic or phonolitic pumice and ended with the outpouring of alkaline basaltic lavas, shows that the lavas of each cycle are not a simple repetition of the lavas of the preceding cycle, but have their own peculiarities of chemistry; moreover, the lines connecting the initial points of the vectors that depict the composition of the rocks of each cycle on Zavaritskii’s diagram and thus reflect the chemical peculiarities of each cycle do not coincide with one another, but are arranged in parallel [15, fig. 62]. This makes it possible to say that the character of the differentiation taking place in the volcanic conduit during the period of quiescence differs substantially from the character of differentiation in a deep volcanic chamber.

The mechanism of differentiation of basaltic magma, for a long time, while Bowen’s hypothesis reigned almost without division in petrology, seemed simple and comprehensible; moreover, the entire variety of rocks, at any rate those belonging to the basaltoid complex, was explained by most investigators as the process of crystallization differentiation, although immediately after the appearance of Bowen’s hypothesis voices were raised in protest against it by Levinson-Lessing, Lodochnikov, Fenner, and others. At present it may be considered proven that the process of crystallization differentiation plays, on the whole, a very modest role—not only in trap formations, where it is manifested only in changes in the composition of femic minerals [44, 47], but even in layered concordant intrusions of the Shonkin-Sag type, which were formerly cited as a classic example of the manifestation of crystallization differentiation in place. More detailed studies have shown that all such layered intrusions are multiphase and that, consequently, the process of differentiation did not proceed within the intrusive body, but in a “deep-seated chamber.” As for the mechanism of differentiation in “deep-seated chambers,” we essentially know nothing; but crystallization differentiation can hardly proceed there, if by “deep-seated chamber” one understands the zone of origin of basaltic magmas. Therefore the different composition of successive “phase” injections of magma may be due either to differentiation in the feeder channel during the ascent of the magma, or to the fact that these “phase” magmas were originally of different composition.

Yu. A. Bilibin [9] came to the conclusion that crystallization differentiation alone cannot explain the entire diversity of rocks of the alkaline complex of the Aldan region, and that, alongside crystallization differentiation, magmatic (diffusional) differentiation also took place there. This proceeded in the direction of increasing the content of \(K_2O\) at the expense of a uniform decrease in the content of all the remaining components, whereas crystallization differentiation led to an increase in the content of \(SiO_2\) in the successively separating differentiates. This scheme seems to me sufficiently artificial, especially since all processes of both magmatic and crystallization differentiation are assigned by Bilibin to a “deep-seated chamber,” while the mechanism of diffusional differentiation is in no way explained. At the same time, it is quite obvious that crystallization differentiation is not the only kind of differentiation responsible for the diversity of rocks of the basaltoid complex. Alongside it, there is undoubtedly some kind of magmatic differentiation, leading, in particular, to the “settling out” of light portions of magma in the upper parts of a magmatic body, manifested especially effectively in the conduits of central-type volcanoes. The mechanism of this differentiation is still quite unclear, but the possibility is not excluded that what occurs here is stratification of the magmatic solution according to specific gravity, analogous to the stratification of aqueous salt solutions established by the experiments of Zemyatchenskii and Kuzmin [22].

B. Origin of the Hyperbasite Complex

Above, grounds were given for distinguishing a special hyperbasite magmatic complex, manifested in the form of intrusions of “serpentinite” or “ophiolite” belts and representing the product of an independent hyperbasite magma. The question of the origin of the hyperbasite complex is resolved much more simply than the question of the origin of the basaltoid complex, although only in the most general form.

The substrate at the expense of which hyperbasite magmas arise may be regarded as the peridotite–simatic shell, whose reality is demonstrated by the presence of a continuous layer with a density of 3.2, lying at a depth of about 60 km within the Eurasian continent, 20 km in the Atlantic Ocean, and almost on the very bottom of the Pacific Ocean. Of course, the peridotitic composition of the simatic layer is a hypothesis, but a very probable one. Hyperbasite magma, born within and at the expense of the sima, probably has a harzburgitic composition, but is considerably enriched in water and other volatiles, which lower its melting temperature. The mechanism of magma formation is selective melting of the substrate with concentration in the liquid phase primarily of the volatiles present in small quantities in the substrate. The cause of melting is unclear, but it is beyond doubt that the conditions under which hyperbasite magmas appeared occurred in the history of the Earth relatively rarely, usually only once in the history of the development of each given geosynclinal structure. The birth of hyperbasite magma is connected with deep-seated faults developing along the boundaries of the geosynclinal trough, and specifically with the early stages of development of deep-seated faults. Hyperbasite complexes show a certain differentiation, with dunites separated out on the one hand and pyroxenites on the other, while the average harzburgitic type sharply predominates. Judging from the fact that dunites usually occur only in large, deeply eroded massifs, whereas pyroxenites are confined to small and apical parts of hyperbasite massifs, one may think that the differentiation had a gravitational character; but whether this differentiation took place in a homogeneous or heterogeneous medium, i.e., whether it was magmatic or crystallization differentiation, is difficult to say. The usual conception of crystallization differentiation is contradicted by the facts of the injection of dunites into peridotites. It is curious to emphasize that hyperbasite magma apparently possesses absolutely no capacity for assimilation of the country rocks, and its contact effect on the country rocks is very weak and has mainly a hydrothermal character. All this indicates a small reserve of free energy in hyperbasite magmas and, above all, their relatively low temperature.

G. Origin of the Principal Types of Magmatic (and Magmatic-Appearing) Rocks

It was indicated above and, I hope, demonstrated, that most groups and families, and even species, of igneous rocks distinguished in petrographic classifications may arise in various ways and are heterogeneous formations. The limits of diversity of petrographic types that may arise at the expense of granitic, basaltic, and hyperbasite magmas, as well as as a result of granitization, were outlined above. Here an attempt will be made to give an overview, even if only the briefest, of the possible modes of formation of the principal types included in the principal groups of the mineralogical classification of rocks customary to us.

Groups of Ultramafic Rocks

Among ultramafic rocks two genetic types are distinguished quite clearly: derivatives of basaltic magma and derivatives of hyperbasitic magma.

The latter are represented by harzburgites poor in iron (the ratio \(Mg : Fe > 7.5\)), calcium, and alumina, more rarely by lherzolites, dunites, and pyroxenites, forming lens-shaped bodies oriented along zones of deep-seated faults that border large folded structures. For these intrusions, intensive autometamorphic serpentinization and weak contact metamorphism are highly characteristic.

Ultramafic derivatives of basaltic magma are predominantly gortonolite dunites and peridotites relatively rich in iron (the ratio \(Mg : Fe > 3.5—7.5\)), as well as augite pyroxenites, giving gradual transitions through gabbro-pyroxenites and gabbro-peridotites to the gabbro type. In these rocks primary banding is common; serpentinization is not characteristic. Association with gabbro and norites in concordant central intrusions, as well as in dikes, is usual, and their predominant distribution is in cratogenic regions. Kimberlites, meimechites, and picrites as effusive representatives should also be assigned to this same type.

Sometimes a third genetic type of ultramafic rocks is also encountered: these are hornblendites, more rarely pyroxenites, arising as a result of assimilation and contact metasomatism of carbonate rocks by granitoid magmas. Pyroxenites of this type are represented by diopside varieties.

Groups of Gabbro-Basalt

The overwhelming majority of gabbros, norites, anorthosites, all gabbro-diabases (dolerites) and basalts, all basaltic porphyrites, diabases, and spilites are naturally direct products of crystallization of basaltic magma or its derivatives. The diversity of chemical-mineralogical composition within this group is explained partly by the different character of the initial basaltic magmas (olivine-basaltic or tholeiitic), and mainly by the phenomena of assimilation and differentiation. But it must be borne in mind that rocks of the gabbro type are sometimes encountered in the basic marginal zones of granitic intrusions, and in such cases they no longer have any relation to basaltic magma, being typical hybrids of granitic magma with limestones. Such gabbros are usually characterized by light-colored diopside pyroxene and are poor in ore admixtures. It should also be noted that some rocks described as anorthosites, gabbros, and norites are in reality crystalline schists of the deepest metamorphic facies—the facies of hypersthene gneisses.

Groups of Granite-Liparite and Granodiorite-Dacite

Within this group one may distinguish a number of genetic types, united together only by the formal feature of a certain commonality of chemical-mineralogical composition. These are granitoids of metasomatic and rheomorphic origin, granitoids and corresponding effusives—products of granitic magma, and, finally, granitoids and acid effusives—products of basaltic magma.

Metamorphogenic granitoids are naturally represented only by rocks of intrusive aspect. To these should be assigned, first of all, the metasomatic granites proper of zones of granitization, always microcline-bearing, often porphyroid, sometimes tending toward granosyenitic and even alkaline types (metasomatic granodiorites, common in the Sudo-

granites (at any rate, I have never observed them in Karelia). Characteristic of them are the absence of cross-cutting contacts and gradual transitions in the degree of granitization to the enclosing rocks, the presence of unreplaced remnants, a layered inherited texture, an imperfect idiomorphic, often granoblastic, usually porphyroid or porphyroblastic structure, phenomena of replacement of plagioclase by microcline and of other minerals, sieve-like structures in microcline, etc. Metasomatic granites are distributed chiefly in the Precambrian and occur among crystalline schists; they are typical representatives of the abyssal facies, although phenomena of granitization, on a not especially large scale, have been recorded, for example, in the Altai as well, in Paleozoic formations, and here developed at lesser depths.

The charnockites of the Yenisei Ridge described by the author, which are products of “dry” granitization and form autochthonous and allochthonous bodies, also belong to the metamorphogenic granitoids. They differ in composition, consisting of quartz, orthoclase, plagioclase, garnet, hypersthene, and biotite in variable amounts, and are very inconstant in composition, approaching now the type of granite, now granosyenite, quartz diorite, and quartz norite. Characteristic is the absence of phenomena of contact metamorphism and autometamorphism. Charnockites occur only among crystalline schists of the hypersthene-gneiss facies and are representatives of the ultra-abyssal facies of igneous rocks.

Pyroxene granophyres and granophyric granites, associated with trappean and gabbronorite intrusions of sill and lopolithic form and occurring in the roofs of the latter, should be assigned to the metamorphogenic granitoids. Most recent investigators are inclined in favor of their metamorphic or rheomorphic origin [47].

Granitoids of magmatic origin (products of granitic magma) are widely distributed both in the Precambrian and in post-Precambrian folded structures. They are represented by granites, adamellites, granodiorites, more rarely trondhjemites and alkaline granites, giving gradual transitions into one another. Their characteristic features are cross-cutting or partly cross-cutting mode of occurrence, sharp contacts with the enclosing rocks and contact metamorphism of the latter, the presence of chilled zones expressed in a decrease of grain size with approach to the contact, the presence of flow textures and the independence of the latter from the texture of the enclosing rocks, normal hypidiomorphic structures, and clearly manifested hydrothermal autometamorphism. The petrographic composition may be variegated; varieties often show gradual transitions into one another, but not into the enclosing rocks. It is characteristic that granitic intrusions proper are very uniform in composition (granites, granosyenites, aplites, and pegmatites), whereas granodioritic intrusions have a very variegated composition (granodiorites, adamellites, sometimes granites, plagiogranites, quartz diorites and diorites, tonalites, etc.). Evidently, granodioritic intrusive complexes develop in cases where granitic magma possessed great chemical activity and a great capacity for assimilation. In the opposite case, granitic complexes of simpler composition took shape.

Effusive and hypabyssal intrusions owing their origin to granitic magma that broke through to the surface apparently exist (Altai effusive and intrusive quartz porphyries and quartz albite porphyries?), but it does not appear possible to distinguish them from acid effusives and hypabyssal intrusions of basaltic origin.

The granite–liparite and granodiorite–dacite group quite evidently includes acid derivatives of basaltic magmas as well. In any case, there can hardly be any doubt that dacitic and liparitic lavas of volcanoes that erupted predominantly basalts and andesites are products...

by parts of basaltic magma contaminated by acidic material of sial and having undergone gravitational differentiation. Likewise there can be no doubt that such acidic melts of basaltoid origin could give rise also to hypabyssal intrusions, represented chiefly by alkaline granites and granosyenites, as well as quartz albitites and quartz oligoclasites (the Semeitau magmatic complex of NE Kazakhstan, the quartz albitites of the Abakan deposit, perhaps the Maina granodiorite).

Groups of diorite-andesite and syenite-trachyte

The heterogeneity of these two groups of igneous rocks was quite distinctly established by Levinson-Lessing [29], who pointed out that their intrusive and effusive representatives are not even chemical equivalents and have different origins.

The effusive representatives of these groups undoubtedly and entirely have a basaltic origin, judging from their regular association with basaltic rocks in modern volcanoes and ancient volcanic complexes. Andesites give completely gradual transitions to tholeiitic basalts and are, apparently, products of contamination of basaltic magma by acidic material, perhaps accompanied by differentiation. Trachytes are linked through trachybasalts with olivine basalts and are products of differentiation of olivine-basaltic magma.

The matter is more complicated with the intrusive representatives of these groups. The overwhelming majority of diorites and syenites (quartz diorites and diorites, syenites and granosyenites), associated with granites and granodiorites in granite and granodiorite intrusive complexes and forming marginal zones in large plutons or small satellite bodies, undoubtedly are hybrids of granitic magma with the enclosing rocks. But alongside them dioritic and syenitic rocks of basaltoid origin are fairly widely developed; these are intrusive, predominantly hypabyssal equivalents of andesitic and trachytic lavas, which, of course, could not only be poured out onto the surface but also form intrusive bodies. Unfortunately, this type of dioritic and syenitic rocks is difficult to distinguish from the granitoid type, although its representatives should differ rather sharply, above all in chemical composition, being complete equivalents of the corresponding effusives. Apparently, many augite diorites and diorite-diabase types of dike and “small” intrusions of various ages, widely developed in the Salair and the Kuznetsk Alatau, belong to this genetic type. The Middle Devonian intrusions of quartz alkaline syenites, widely developed in the Eastern Sayan and partly in the Kuznetsk Alatau, should also be assigned here.

Finally, it should be noted that diorites and syenites may also have a metasomatic origin; moreover, metasomatic diorites and metasomatic syenites have been well studied and described [5].

Groups of alkaline rocks (nepheline syenite-phonolite and alkaline gabbroids and basaltoids)

The genetic connection of alkaline gabbroids and basaltoids with basaltic (olivine-basaltic) magma is undoubted. Likewise, evidently, the question of phonolites is easily resolved: being regularly associated in volcanic complexes with basalts, they are evidently also genetically connected with them. Nepheline syenites are often associated with these same volcanic basaltoid complexes; examples of this were given by A. N. Zavaritskii [15]. Yu. A. Bilibin [8; 9] comes to the same conclusion about the genetic connection of the alkaline intrusions of the Aldan region with basaltic magma, and his arguments appear quite convincing. There are grounds to think that, in general, the overwhelming majority of nepheline-syenitic

intrusions of craton regions and the latest postorogenic alkaline intrusions of folded zones have the same basaltic origin. However, alongside this predominant genetic type there really exists another, genetically connected with granitic magmas and owing its origin to processes of assimilation of carbonate rocks. Nepheline syenites of the Lake Bulankul’ area may serve as an example; they constitute the marginal zone of a granosyenite intrusion at the contact with limestones.

Conclusions

The principal conclusions that may be drawn from this very brief review of the possible modes of formation of the chief types of magmatic (and magmatic-looking) rocks suggest themselves and may be formulated briefly:

1. The groups of rocks in the classical mineralogical classifications include very heterogeneous formations, and the classifications themselves are formal; they are based only on the feature of chemical-mineralogical composition and take no account whatever of the origin of magmatic rocks.

2. The need has become ripe to develop a new genetic classification of magmatic rocks, the first variants of which have already been proposed by F. Yu. Levinson-Lessing [31] and by me [21]. But beforehand an enormous amount of work must be carried out on questions of the genesis of magmatic rocks, in many cases still entirely unclear, and on questions of establishing criteria for the genesis of magmatic rocks.

Conclusion

In petrology, and especially in the question of the origin of magmatic rocks, disagreements are exceptionally great on a number of questions; there are especially many hypotheses claiming “universality,” and at the same time here, alongside the newest achievements of science, hopelessly obsolete concepts peacefully coexist. Apparently, every scientist periodically feels the need to systematize the knowledge and ideas in that field of science in which he works, and to develop some more or less coherent world view on the basic, genetic, nodal questions of his science. What has been set forth above constitutes my working views on the question of the origin of magmatic rocks, which, it seems to me, I have succeeded in bringing into a more or less coherent system corresponding to geological facts and methodologically sound. At the same time I clearly realize that everything set forth above is still only a scheme and the first outline of a general theory of the magmatic process. Here, probably, there are many insufficiently substantiated and perhaps even premature conclusions. Very much in the problem of the origin of magmatic rocks remains unclear. Where and how is basaltic magma born? If it is born at the expense of “sima,” then why in some cases does basaltic magma appear, and in others hyperbasitic magma? How is the different energy level of granitic, basaltic, and hyperbasitic magmas to be explained? The answer suggests itself that the energetics of magma is determined not by thermal but by chemical energy. But in that case, how is one to explain the usual very high energy potential of granitic magma? The hypothesis set forth above does not explain this. Nor, in general, can this question apparently be explained by the laws of classical physics and physical chemistry. Evidently, in the processes of magma formation phenomena of another order come onto the stage, for example those connected with the liberation of intra-atomic energy. These and similar questions still await their resolution.

In the course of developing the topic, the work I undertook at times seemed to me premature, but I was inspired and justified by the remarkable words of M. V. Lomonosov, which I took as the epigraph to my article.

References

1. Aleshkov A. N.—Dunite–peridotite massifs of the Polar Urals. Mat. Kom. eksp. vyp. 18, Academy of Sciences of the USSR, 1929.

2. The same—On the question of “primordial” magma. Publ. Acad. Sci., geological series, issue 5, 1940.

3. Afanas’ev G. D.—Phenomena of microclinization in granodiorite intrusions of batholithic type in Transcaucasia. Izv. AN, geological series, No. 3, 1949.

4. The same—Granitoids of ancient intrusive complexes of the Northwestern Caucasus. Proceedings of the Institute of Geological Sciences, issue 69, petrographic series (28). Academy of Sciences of the USSR, 1950.

5. The same—On the role of granitization in the formation of granitoid massifs of certain folded regions. Izv. AN SSSR, geological series, No. 4, 1951.

6. Backlund H.—The problem of granitization. Collection “Problems of the Formation of Granites,” 1, 1949.

7. Belyankin D. S.—On the question of the present state and prospects of the study of magmas and magmatic rocks. Izv. AN, geological series, No. 5, 1947.

8. Bilibin Yu. A.—Post-Precambrian intrusions of the Aldan region. Petrography, USSR, series 1, regional petrography, issue 10. Petrography of the Aldan. Izv. AN SSSR, 1941.

9. The same—Petrology of the Ylymakh intrusion. Gosgeolizdat, 1947.

10. Bowen N.—Magmas. Collection “Problems of the Formation of Granites,” 1, 1949.

11. The same—The granite problem and the method of repeated prejudices. Collection “Problems of the Formation of Granites,” II, 1950.

12. Wager K. A.—Paths of differentiation in the Deccan traps. Translation. Collection “Geology and Petrography of Trap Formations.” Foreign Literature Publishing House, 1950.

13. Eliseev N. A.—Granitization and metasomatic granites. Scientific Notes of Leningrad University, No. 2, 1951.

14. Zavaritskii A. N.—On pegmatites as formations intermediate between igneous rocks and ore veins. Zap. Vseros. miner. ob-va, No. 1, 1947.

15. The same—Introduction to the petrochemistry of igneous rocks. Publ. Acad. Sci. USSR, 1950.

16. Korzhinskii D. S.—Factors of mineral equilibria and mineralogical facies of depth. Proceedings of the Institute of Geological Sciences, issue 12, Acad. Sci., 1940.

17. The same—Granitization as magmatic replacement. Izv. AN, geological series, No. 2, 1952.

18. Kropotkin P. N.—On the origin of granites. Soviet Geology, No. 9, 1940.

19. The same—O. Yu. Schmidt’s cosmogonic theory and the structure of the Earth. Izv. AN SSSR, geographical and geophysical series, No. 1, 1950.

20. Kuznetsov Yu. A.—A scheme for the classification of facies of magmatic rocks. Proceedings of the Mining-Geological Institute of the West Siberian Branch of the Academy of Sciences, issue 5, 1949.

21. The same—On the origin, nomenclature, and classification of magmatic rocks. Izv. AN SSSR, geological series, No. 6, 1951.

22. Kuz’min A. M.—Mass crystallization at many levels. Izv. Tomsk Polytechnic Institute, vol. 65, issue 2.

23. Kupletskii B. M.—A review of contemporary views on the origin of granites. Izv. AN, No. 3, 1942.

24. Lebedev P. I.—The problem of studying basic magma. Problems of Soviet Geology, No. 1, 1936.

25. The same—On the question of the nature of silicate and ore magmas.—Collection “Questions of Mineralogy, Geochemistry and Petrography.” Publ. Acad. Sci. USSR, 1946.

26. Levinson-Lessing F. Yu.—On the main problems of petrogenesis. Izv. SPb Polytechnic Institute, Department of Technology, Natural Sciences and Mathematics, 14, issue 1, 1920.

27. The same—A few thoughts on differentiation and the nature of magma.—Ibid., issue 2, 1915.

28. The same—On the question of the genesis of igneous rocks. Proceedings of the Mining Museum, Academy of Sciences, issue 3, 1929.

29. Levinson-Lessing F. Yu.—The problem of the genesis of magmatic rocks and paths toward its solution. Publ. Acad. Sci., 1934.

30. The same—Problems of magma. Scientific Notes of Leningrad State University, 3, No. 17, 1937.

31. The same—Problems of magma. Izv. AN, geological series, No. 1, 1939.

32. Lodochnikov V. N.—Serpentines and serpentinites, the Il’chur ones and others, and petrological questions connected with them. Proceedings of TsNIGRI, issue 38, 1936.

33. The same—Serpentines and serpentinites and petrological questions connected with them. Problems of Soviet Geology, No. 5, 1933.

34. The same—Certain general questions connected with magma that gives rise to basaltic rocks. Zap. miner. ob-va, part 68, No. 2, 1939.

35. Niggli P.—The problem of the formation of granite. Collection “The Problem of the Formation of Granites,” 1, 1949.

1. Onge—Leucocratic, trondhjemitic, and leucosyenitic magmas and anatexis. The Problem of Granite Formation, II, 1950.

2. Perrin R. and Roubault M.—On the granite problem. Coll. “The Problem of Granite Formation,” II, 1950.

3. Polkanov A. A.—On the question of the heterogeneity of phonolites. Izv. AN SSSR, No. 5, 1944.

4. Onge—Basic principles of the genetic systematics of intrusive bodies. Proceedings of the Jubilee Session of Leningrad State University, 1946.

5. Onge—Genetic systematics of intrusive bodies of the platform—craton. Izv. AN SSSR, geological series, No. 6, 1946.

6. Reynolds D. A.—The sequence of geochemical changes leading to granitization. Coll. “The Problem of Granite Formation,” II, 1950.

7. Read H.—Meditations on granite. Coll. “The Problem of Granite Formation,” I, 1949.

8. Onge—Granites and granites. Coll. “The Problem of Granite Formation,” II, 1950.

9. Sobolev V.—Petrology of the traps of the Siberian Platform. Proceedings of the Arctic Institute, No. 43, 1936.

10. Sudovikov N. G.—Metasomatic granites. Bulletin of Leningrad University, No. 10, 1950.

11. Proceedings of the First Conference on Questions of Cosmogony. Publishing House of the Academy of Sciences of the USSR, 1951.

12. Walker F. and Poldervaart A.—Dolerites of the Karroo of the Union of South Africa. Translated in: Coll. “Geology and Petrography of Trap Formations.” Foreign Literature Publishing House, 1950.

13. Usov M. A.—Geology of magmatic rocks. Problems of the Geology of Siberia, vol. 1, 1945.

14. Harker A.—Metamorphism. ONTI, 1937.

15. Hess H. H.—Island arcs, gravity anomalies, and serpentinite intrusions. Proceedings of the XVII International Geological Congress, vol. II, 1937.

16. Edwards A. B.—Differentiation in the dolerites of Tasmania. Translated in: Coll. “Geology and Petrography of Trap Formations.” Foreign Literature Publishing House, 1950.

17. Eskola P.—On the origin of granitic magmas—Miner. u. Petr. Mitteilungen, Bd 42, teil 5–6, 1932.

18. Fenner C.—A view of magmatic differentiation.—Journal of Geology, 1937, No. 2, pp. 156–163.

19. Hess H.—A orimary peridotite magma. Amer. Journ. of Sc 1936, XIXV, No. 209.

20. Holmquist P.—Typen und nomenklature der Adergesteine. Geöl. Fören Forhand Bd, 43, 1921.

21. Holms. A.—The origin of igneous rocks. Geol Magas. 1932, 59, 543.

22. Sederholm I.—On migmatites and assoc. precambr. of Southwestern Finland. Bull. Com. geol. Finland, 58, 1923; 107, 1934.

23. Termier P.—Les Schistes cristallins des Alpes Occident, C.R. IX Congres geol. Intern, Wienna 1903 1990).

24. Onge—Sur la genese des terrains cristallophyllions, C. R. XI Congros. geol. Intern. Stockholm 1912.

25. Volf F.—Der Vulkanismus, Bd II, Stuttgart 1923.

26. Wegmann—Zur Deutung der Migmatite.—Geol. Rundschau 28, 1928; 26, 27, 1935.

1. The influence of pressure change on the melting temperature is manifested most strongly precisely in quartz. Hence, it is theoretically possible to determine the depth of formation of granitoids from the quartz content in anatectic systems. ↩