岩石学の論理構造 : 特に花崗岩問題に寄せて

Logical Structure of Petrology: With Special Reference to the Granite Problem

KOJIMA George

廣島大學地學研究報告

Geological report of the Hiroshima University

The science of Japan is characterized by reading papers, interpretation, and smuggling foreign ideas, as disclosed by one of the most ingenious reviewers of Japan, Takaaki YOSHIMOTO. If that be the situation, where is to seek the originality of scientist?
Among them, problems related to the interpretation in petrology are discussed from the standpoint of the logical structure of petrology. The petrology is situated between descriptive geology and theoretical physical sciences. In the petrology are used various technical terms belonging either to descriptive geology or physical sciences. For instance, the term "magma" is a descriptive one, denoting a kind of flowing rock-forming mass which gives rise to a volcanic rock through solidification or cooling, and it must not be identical with "liquid", a kind of phase of material. When the term "magma" is equalized to "liquid", abstraction based on an assumption is implied. The problem is not that of assumption, but of the fact.
The most typical case of interpretation in geology has been provided by the Old Testament, the tale of the Deluge. It has been the most effective public fantasy of the Western people, as the Tennô (the Emperor of Japan) fantasy stands for the Japanese. The Testamental interpretation can be traced even up to E. SUESS: "Das Antlitz der Erde". The interpretation in the present geological science can be divided into two categories; that is, the abstractive and the inductive interpretations. In the abstractive interpretation, one side of things and affairs is picked up, interpreted after some models, and generalized. For example, if the assemblage of jadeitic pyroxene and quartz is found in a rock, the rock is interpreted to have been formed about 30 km below the surface. On the other hand, in the inductive interpretation, one starts from experiments or principles, forms models, then picks up one side of facts and affairs as the informations fitted to the models. Examples are afforded by papers of geophysicists.
In this connection, the inductive method is discussed from the standpoint not of J. S. MILL, but of E. HUSSERL and M. MERLEAU-PONTY: the induction is to read the essentials of things and affairs directly. The induction is not "power", but "light" clarifying the future.
As the most prominent work of petrology, "The Evolution of the Igneous Rocks" (N. L. BOWEN, 1928) is analysed with respect to its logical structure. BOWEN'S inductive chain of inference is as follows: petrographic province→rock association→community of origin (consanguinity) →differentiation of a single original magma (= basaltic magma)→fractional crystallization of basaltic liquid. The arrow in this series represents the inductive leap, not the logical consequence. In this chain structure, the term begins from geological one to that of physical science. Therefore, the inductive chain represents the sequence of abstraction. Geological sciences, including petrology, are constructed of many domains of different level (order) of abstraction, that being the fundamental difference from physics, so to speak, the idealizing fiction. About the role of the experiment BOWEN'S words should be noticed; that is, "... since the investigated systems are always much simpler than magmas, it is not possible to use directly the actual quantitative values of concentrations, temperatures, etc., of the experimental results. The principal service of these must be rather to point the road."
The present writer names the cpigones of BOWEN the bowenists. One of problems in the works of bowenists is the transgression of border between branches of science. When they infer the mechanism of magmatic differentiation as gravitational sinking of crystals within a magma chamber or squeezing of magmatic liquid through a narrow passage in the Crust, the problem enters into the field of dynamic side of geology, in which it must be discussed dynamically based mainly on the fabric in the nature.
The latter half of the paper is devoted to the granite problem. In the beginning, it is noticed that naming of granite based solely on the mineralogy and texture of the rock is impossible. The classification of natural things and affairs, including rocks, should be based on the form of phenomenon, as exemplified by the natural water, which is classified as river water, lake water, groundwater, etc., according to its form of phenomenon. In this connection, the granite problem should be one of field geology, as pointed out by H. H. READ. The granite problem must begin with clarifying the genesis or the history of formation of individual granitic mass. From this standpoint, the history of the granite
problem is reviewed briefly.
Then, experiments on granitic substances are reviewed. Important results are summarized as follows:
1) Under the condition of H2O saturation, the minimum temperaturc of liquidus-solidus becomes lower with increase of PH2O, and the rate of the lowering is distinct up to PH2O = ca 1 kb.
2) In the case of H2O undersaturation, the temperature interval between liquidus and solidus is conspicuously larger than in the case of H2O saturation. This suggests that, in the granite magma of moderate temperature, much crystal grains of refractory minerals, such as pyroxene, intermediate to calcic plagioclase, hornblende, would be suspended in the liquid magma.
3) In the peraluminous haplogranitic system, excess alumina can melt into haplogranitic liquid only in a limited amount, that excluding the possibity of peraluminous granitic liquid magma. On the other hand, in the peralkaline haplogranitic system (mole ratio alkali: alumina > 1), peralkalinc hydrous solution persists down to such lower temperature as about 300°C, and the H2O content attains to about 50 wt. percent in the solution. The gas phase in equilibrium with this solution becomes gradually enriched in silicate contents, and the immiscible gap between the liquid and the gas disappears at PH2O= 1.25 kb, 320°C. These results are very significant to chang our view on the nature; that is, from the discontinuity view to the continuity one.
History of formation is reflected on the fabric of granitic rocks. It is characterized by the "Kristalloblastese". The granitic texture is no more than the granoblastic one. Several characteristics which appear to be significant to clarify the genesis of granitic rocks are listed as follows:
1) Plagioclase having An-rich core with normal zoning is occasionally included. Magmatic type of twinning is observable in the core. It represents the crystal of refractory character in experiment.
2) Glomeroporphyroblastic megacrystals of plagioclase, consisting of several crystals of commonly andesine-oligoclase arranged subparallel to each other, are often found. Zoning and turbid An-rich core are only rarely observed. This type of plagioclase is often found in metamorphic rocks, representing an intermediate stage of growth of large porphyroblast.
3) Plagioclase forms sometimes porphyroblastic crystals including granoblastic crystals of pyroxene, hornblende; and biotite, which resemble porphyroblasts in amphibolite, often in the course of feldspathization or anatexis. The host plagioclase appears hetero-geneous in the extinction position, that suggesting the absorption of included plagioclase which had formed granoblastic aggregate with hornblende.
4) Alkali-feldspars occur ① as phenocrystic large crystals, ② as porphyroblastic mega-crystals of irregular shape, ③ forming granoblastic aggregate with plagioclase and quartz, ④ filling interstices, and ⑤ as patches in plagioclase crystals (not always anti-perthite). At least, some of these features must be attributed to the metasomatism or replacement at rather lower temperatures, probably by the action of hydrous alkaline solutions.
5) Quartz occurs, as is the case for alkali-feldspars, ① as possible phenocrysts, ② as porphyroblast-like crystals, ③ as granoblastic aggregate with oligoclase and alkali-feldspar, and ④ intersticially. The formation may have lasted from the early-magmatic to the later hydrothermal stages.
6) K-feldspar, Na-feldspar, and quartz form such textures as ① graphic — micropeg-matitic — granophyric — spherulitic intergrowth, ② myrmekite, ③ perthite, and ④ intergranular albite. Some of these textures may be related to metasomatism.
7) There is a tendency of mafic and felsic mineral groups to collect into separate melan-ocratic and leucoratic parts, forming gneissic or network structure. Mafic minerals form lenses, seams, and clots. These features are comparable to those of metamorphic differen-iation; that is, the contrast between melanosome～restite～palaeosome and leucosome～mobilizate~neosome.
8) Deformation structures are observed selectively among mineral species. Deformation structures of minerals are commonly found in more basic varieties of granitic rocks, such as granodiorite, quartz-diorite, quartz monzonite, and tonalite. Minerals of earlier stages or inherited from the origin before granitization, namely, those having refractory behaviour show deformation structures, whereas those of later formation, such as alkali-feldspar and quartz show no traces of deformation. Especially, plagioclase megacrystals have been fractured to form glomeroporphyritic aggregates, and the trace of twinning is often distorted. Such features of deformation can well be explained by the hypothesis that, before the crystallization (probably recrystallization) of K-feldspar, Ab-rich plagioclase, and quartz, the crystal mush mainly consisting of hornblende and plagioclase, presumably in some cases with alkali-feldspar, quartz, and biotite, flowed to the upper part of the Crust. The presence of intergranular fluid phase is not excluded, but it is needed that there prevails enough level of stress to give rise to intragranular flow of crystals consisting the crystal mush. The flowing material having these characteristics must be quite different from the so-called "magma", the molten volcanic material. Therefore, there cannot always exists any correlation between the rhyolite magma and the flowing granitic material. In this respect, is significant H. RAMBERG'S experiments, which concluded 1~103 poises of viscosity difference between the country rocks and the flowing granitic material, which forms dome-shaped or batholithic plutons. To cause fracturing and fragmentation of country rocks along the magma channel is needed large values of the viscosity difference.
J. J. SEDERHOLM and afterwards E. WEGMANN has stressed the significance of metamorphic basic dykes in granitic and gneissic masses (the "Sederholm effect" named by P. ESKOLA). The role of the metamorphic dykes is twofold; that is, the "vertical synchronization" and "die Stockwerkanalyse". The inference is based upon the assumption that the intrusion of the dyke occurs when the country rocks arc situated under the "Oberbau" conditions — lower temperature and pressure — , while the ductile deformation and metamorphism take place under the "Unterbau" or "Zwischenzone" conditions — higher or intermediate temperature and pressure. The assumption cannot be held in the presence of recent knowledges about the visco-elastic behaviour of rock materials. Even the hot molten magma can be fractured by a shock. Therefore, it can safely be said that the flowing granitic material can be fractured, and intruded by basaltic magma. The basaltic magma may be quenched down to the temperature of the granitic material. Afterwards, the dyke may be deformed — sheared and boudinized — , and metamorphosed by the surrounding granitic material, including anatexis. While, the granitic rocks surrounding the basic dyke may be partially melted (transfusion, mobilization, rhemorphism) to form acid dykes and veins, which penetrate the dyke itself. It is suggested that the activity of granitic material would be related to the ascent of basic magma into the Crust.
Lastly, it is stressed that there exists a spectrum from liquid magma to solid flow through all variations of crystal mush. On the other hand, with respect to the nature of inter-granular fluid, there exists also a spectrum from magmatic liquid to hydrothermal solution. There should not be any discontinuity between flowing "magma" and rigid rock mass. Generally speaking, the trend of view on the nature is on the road to the continuity view, as shown in the history of sciences.

jpn

廣島大學理學部地學教室

[ISSN] 0073-2303

[NCID] AN00213676