i) Metamorphism

a) The interaction between deformation and recrystallization

In zones IIb and III, grain displacement is essential of the deformation of rocks (flow fold), whereas in zones I and IIa, block movement is so (lens fold). Corresponding to this contrast, modal amounts of relic augite show an abrupt decrease from zone IIa to zone IIb. In zones I and IIa, relic augite is almost perfectly preserved during metamorphism, while in zones IIb and III, it is completely altered. The stratigraphical thickness of the basic volcanic sediments in zones IIa and IIb is up to 100 meters. Therefore, we exclude the the effect of layered structure of rocks on the deformation style. It is concluded that grain displacement becomes dominant with proceeding of the recrystallization. The rock deformation performed with the grain displacement is represented by the viscous flow, and that with block movement is by the plastic deformation. The interaction between the deformation and recrystallization reveals in the rocks deformed with viscous flow of zones IIb and III.

As discussed in the chapter of texture, the mechanical accretion, which results the formation of albite spot, is controlled by the collision and annealing of grains. The collision is induced by the rock compression, and the anneal proceeds in the case that the collided minerals are compositionally homogeneous and they are under the stress and at medium to high temperature. Therefore, zoned minerals do not grow in the mechanical accretion process such as the albite spot. Table VI-1 lists the unzoned and zoned minerals in the basic rocks of zone IIb and III. Epidote and amphiboles show the continuous clear zoning, and this compositional zoning has been formed with rising temperature, maintaining the surface equilibrium as discussed in the chapter of mineralogy. On the other hand, chlorite has no compositional zoning and in one thin section its chemical composition shows usually the same value at every portion. This chemical homogeneity of chlorite may be explained by following two alternatives: one is body diffusion, and the other is grain boundary diffusion models. According to KURATA (1972), chemical heterogenity of chlorite grain up to 1 mm in length may be formed at the expence of garnet retrogressively. This fact may indicate that the body diffusion model is not appropriate at low temperature, and thus we must explain the homogeneity of chlorite in terms of grain boundary diffusion.

Chlorite forms the aggregates of fine-grained chlorite and it is possible, if not probable, that individual grain in the aggregates is composed ofsubmicroscopically disorientated grains. The grain boundaries of this submicroscopical aggregates offer an effective diffusion path. If so the overall diffusion becomes significantly easier as compared to homogeneous chlorite. The dominant diffusion path is along the grain boundaries and/or subgrainboundaries.

Chemical heterogeity of chlorite appears along the cleavages (grain boundary), thereby showing that it is brought about by the grain boundary diffusion in chlorite reported by Kurata. Therefore, it is concluded that the chemical homogeneity of chlorite is caused by the grain boundary diffusion, and then the individual chlorite grains, being surrounded by the grain boundaries, and microscopical subgrain boundaries, may be sized submicroscopically.

Nevertheless, compositionally homogeneous chlorite aggregate has not been annealed as in the case of albite aggregate. This fact may be explained by the following two reasons. The cellular reconstruction of chlorite have the large activation energy, and then its rate is very small at the low temperature. Another is that if chlorite grew porphyroblastic, it is crushed by rather weak stress and forms the polycrystalline aggregate. Kink bands and other deformation bands of chlorite are common in the metamorphic rocks, thereby showing that chlorite is deformed by the weak stress during metamorphism. Although coarse grains of chlorite are crushed by the such stress, the fine grained or submicroscopical fine-grained chlorite may not maintain the internal strain by which chlorite is ruptured to the polycrystalline aggregate.Even weak stress arising on the polycrystalline aggregates gives rise to grain boundary slip as the elementary process of the creep deformation or the viscous flow of chlorite aggregate. Therefore, chlorite has maintained the fine grained aggregate and the chemical homogeneity throughout metamorphism.

It follows that all of chlorite has maintained the chemical equilibrium with surface of amphibole, epidote and other metamorphic minerals at every stage of progressive metamorphism. This conclusion is supported by the Al-enrichment in chlorite with rising temperature.

Composition bands: Unzoned albite, calcite and quartz are annealed to form crystallographically homogeneous grains, if they are collided with each other in the stress condition. Therefore, as is discussed in a previous section, albite, calcite and quartz are segregated to form the compositional banding by the uniaxial stress during metamorphism. Nevertheless, the compositional banding may show the mono-mineralic bands, as is shown in albite spots. From the segregation of these minerals, actinolite, chlorite, epidote and sphene form another composition bands. Therefore, the schist under the uniaxial compression may be composed of actinolite +-chlorite + epidote and albite + calcite + quartz bands. Further, in the calcic bands, albite, calcite and quartz grow porphyroblasitcally through the mechanical accretion and anneal processes. In deed, it seems that the size of albite, quartz and calcite in the mafic bands is smaller than that in the calcic bands. It is likely that the ratio of thickness of the mafic and calcic bands have the similar value in some rock specimens. This model may be the stress-induced ordering in the sense of rheology.

Reaction zone: The segregation process might have been taken place throughout the metamorphism. This segregation of albite quartz and calcite produces the compositional gradient in rock. The epidote+chlorite bands are undersaturated in SiO₂, and the albite + quartz + calcite bands are oversaturated. Therefore, the following reactions may occur along the boundary between those composition bands;

2NaAlSi₃O₈ + Mg₃Si₂O₅(OH)₄ = Na₂Mg₃Al₂Si₈O₂₂(OH)₂ + H₂O,
6Ga₂Al₃Si₃O₁₂(OH) + Mg₁₇Al₈Si₁₀(OH)₂₈ + 7SiO₂
= 6Ga₂Al₄Mg₃Si₆O₂₂(OH)₂ + 11H₂O.

Usually, glaucophane occurs along the boundary in the basic banded schists, and in the rocks of zone III barroisitic amphibole rich in Ts-component is common along the boundary between two compositional bands. MIYASHIRO (1958) also has found the reaction zone consisting clinopyroxene and calcic plagioclase with small amount of quartz and sphene between the grandite+ quartz or calcite+quartz and hornblende bands. He also discussed that these bands were not derived from impure calcareous sediments, because they differ from such sediments in chemical composition, especially in total Fe/Mg ratio. Therefore, the segregation induced by stress was followed by the chemical reaction, during the progressive metamorphism, because of the formation of compositional gradient in rock. These reaction along the boundary between bands may advance the viscous flow because of the production of fluid phase.

Therefore, the Sambagawa metamorphism may contain the recrystallization and the mechanical segregation induced by the stress in rocks of zones IIb and III, while in zones I and IIa basic rocks, metamorphism may have not been accompanied by the mineral redistribution. These two processes of chemical reaction and mechanical redistribution are independent with each other, but because of forma tion of the compositional gradient in rock, the chemical reaction is controlled by the mechanical process.

b) Thermal History of Metamorphism

Growth Model of Amphibole Composite Grains

As discussed in the chapter of mineralogy, actinolite alkali amphibole interfaces of the composite grains maintain the various compositional gaps, corresponding to various temperatures. From the composition profiles across the interfaces (Fig. VI-1), it is confirmed that the composite grain grew under the rising temperature without effective diffusion. Fig. VI-2 indicates the growth model for the composite amphibole grain. This model explains the composition profile across the interface and the different compositional gaps at different interfacies, on the assumptions that growth surface coincides with the interface between actinolite and alkali amphibole throughout the recrystallization and the activity of the growth surface ceases at different time at the different interface. The duration of activity of the growth surface may be due to the supply of reactants into the interface and the surface state controlled by the strain at the interface. These factors may vary from interface to inter face. The actinolite and alkali amphibole compositional gap can be approximated by the binode of simple mixture as shown in the right hand side of Fig. VI-2. Fig. VI-2 indicates the model of amphibole composite grain growth in these condition.

At stage of (1), infinitesimal amount of actinolite and alkali amphibole nucleates along the parting of host alkali amphibole at the expence of chlorite, pumpellyite and quartz at the temperature of T₁. At this stage, the chemical compositions of actinolite and alkali amphibole are represented as points on the binodal line as indicated on the right-hand side. With increasing temperature, actinolite and alkali amphibole grow on the newly formed boundary planes between growing actinolite and alkali amphibole. At the stage of (2), when the temperature reaches T₂, the supply of amphibole forming reactants into one boundary ceases, whereas that into another boundary continues, because of the variation of surface condition induced by the stress. Then, as far as the boundary surface is incoherent and the reactants are supplyed into that boundary surface by the grain boundary diffusion, composite growth of actinolite and alkali amphibole may continue on this surface and it makes the new grain boundary between them.

At the stage of (3), when the temperature reaches T₃, the recrystallization ceases, and the compositional gap at one boundary reaches the value corresponding to the two phase region at T₃ in the binary system of actinolite and alkali amphibole as shown in Fig. VI-2. At this stage, the composition gap at another boundary maintains the same value at T₂, because diffusion does not occur. Therefore, the compositional gaps may vary from interface to interface, and thus the compositional zoning of actinolite and alkali amphibole may appear asymmetric. Moreover, the zonings in actinolite and alkali amphibole show an increase of Na and Ca from core to interface, respectively.

Thermal Trends of Progressive Metamorphism

The actinolite-alkali amphibole solid solution series can be approximated by simple mixture, and the ratio (Tc-T)/Tc at many interfacies can be obtained from the following equations;

1 - T/Tc = 1 - 2y/ln[(1+y)/(1-y)],
y = 1 - 2X₁,

where Tc and X₁ are the critical temperature and actinolite-content in alkali amphibole, respectively.

As the composite growth is followed by rising temperature, the range of rising temperature for composite growth can be estimated by the range of (Tc-T)/Tc ratio at the interfacies of many composite grains in the rocks of zones IIa and IIb. This ratio ranges from 0.04 to 0.24 in zone IIb, while in zone IIa it does from 0,18 to 0.24 as shown in Fig. VI-3 and Table VI-2. If we assume Tc=600°K, the ranges of formation temperature are about 120 and 40°C in zones IIb and IIa, respectively. The lowest temperature for glaucophane schist is consistent with that of the lawsonite zone condition in New Caledonian metamorphic belt reported by TAY- LOR and COLEMAN (1968).

The (T-P) condition of regional metamorphism can be obtained from the phase transition of some minerals, i.e., jadeite + quartz = albite, aragonite = calcite, and by the oxygen isotope partition between lawsonite and glaucophane, and calcite and quartz, and others. Otherwise, these geothermometer and geobarometer may not be applied for the thermal history of regional metamorphism, i.e., the time-dependent regional metamorphism, because these transitions occur univariantly on a (P-T) plane. Oxygen isotope thermometry, however, makes it possible to obtain the range of formative temperature of some minerals. Nevertheless, most of metamorphic minerals show the compositional zoning owing to rising temperature, and the results of oxygen thermometry for magnetite + quartz, calcite + quartz and lawsonite + glaucophane assemblages does not indicates the strict range of formative temperature in the regional metamorphism. It is neccesary that the measurement of isotope ratio must be done on the boundary between two minerals.

On the other hand, the Mg-Fe partition ofchlorite-garnet pair and biotite-garnet one, etc., may be possible to determine the equilibrium temperature at the boundary between them. However, above mentioned grain displacement during metamorphism may indicate that the biotite-garnet, chlorite-garnet and other pairs are not in equilibrium with each other at the interfacies even if they are crystallized in equilibrium with each other.

Therefore, the range of formation temperature in the progressive metamorphism could be estimated by the phase relation as a continuous function of temperature, if the mineral pair governed chemically by such a phase relation maintains the equilibrium state at every stage of metamorphism. In other words, if some minerals governed by such a phase relation show the parallel growth in equilibrium with each other at their boundary, and the compositional gap varies from interface to interface, the range of progressive temperature could be estimated. The compositional gap is also uniquely determined by the temperature, because the actinolite-magne-sioriebeckite solid solution can be approximated by simple mixture. Therefore, the range of formative temperature in progressive metamorphism can be obtained by these amphibole composite grains in each grade.

It is noteworthy that the temperature range of zone IIb is larger than that of zone IIa. From the compositional zoning of barroisitic amphibole, the temperature range of zone III may be also larger than that of zone IIb. Therefore, we also obtain the difference of temperature between zones IIa and IIb as around 70°C. On the other hand, the maximum difference of pressure between these zones can be obtained from the total thickness of the Kashiwagi, and Upper Sambagawa formations, if the thickness of overburden sediments is constant throughout the metamorphic terrane. The total thickness of their formations measures about 2 km, and this value is approximated as 0.7 Kb. It follows that the ratio ΔT/ΔP is approximated as 100°C/Kb. It is evident that this ratio represents the (T-P) trends of the metamorphic fades series proposed by MIYASHIR.O (1961) (Fig. VI-4).

On the other hand, amphibole composite grains have grown under the condition of the glaucophane schist facies, judging from the stable mineral assemblage of ctinolite + glaucophane + chlorite + epidote + quartz in the basic composition. Therefore, the amphibole composite grain growth may have performed in the high pressure condition. Rough estimation of pressure can be obtained by the content of jadeite component in omphacite coexisting with quartz and albite in zone IIb (see ESSEN and FYFE, 1967). If we assume 200°C as the formative temperature of omphacite, the pressure around 6 to 7 kb is obtained from the jadeite content in omphacite. It is a problem whether pressure may increase or decrease with rising temperature in a range from 200 to 300°C as discussed above. The three cases of pressure trends with rising temperature are illustrated in Fig. VI-5. The case 3), where pressure decreases with rising temperature, formative (T-P) trend in zone IIa passes the univariant line of laumontite decomposition reaction. Therefore, the case 3) may not be acceptable for the Sambagawa metamorphism in the Kanto Mountains, because of absence of laumontite in zone IIa and I. The cases 1) and 2), however, may be suitable for the conditions of disappearance of aragonite, and prehnite in zones IIa and IIb and the compositional zoning of alkali pyroxene showing an increase of diopside and acmite components toward the outer part of grains, if the range of pressure increase is around 1 Kb. Therefore, even in the case 2), where pressure increases with rising temperature, as the range of pressure increase is small, the trends of (T-P) in zones IIa and IIb may be approximated by the case 1), that is, the pressure is nearly constant during temperature rise.

Fig. VI-4 indicates the (T-P) trends of zones IIa, IIb and III during progressive metamorphism. In this figure, the geothermal gradient at the inner trench slope (b) estimated by Oxyburgh and Turcotte (1971) passes the metamorphic conditions of the Franciscan and Sambagawa metamorphic belts (B, and x-y-z in Fig. VI-4). The (T-P) trends of the Sambagawa metamorphism may indicate the deviation from the geothermal gradient at the inner trench slope toward that beneath the inter arc basin (geothermal gradient a in Fig, VI-4).

c) Age of Sambagawa Metamorphism

From geological consideration, it is concluded that regional stratigraphical gap exists between the Sanchu and Kamiyoshida formations in the southern marginal area, whereas in the central area of the mapped district the gap lies between the Ato-kura and Upper Sambagawa formations. The Sanchu and Atokura groups are correlated to the Lower to Upper and the Upper Cretaceous ages, respectively. On the other hand, the Kamiyoshida and Upper Sambagawa formations are also correlated to the Middle to Upper Triassic and Lower Permian to Upper Carboniferous, respectively. Therefore, the time gap between the Atokura and Upper Sambagawa formations is larger than that between the Sanchu and Kamiyoshida formations. The Atokura and Sanchu groups are exposed parallel to the general trends of the Sambagawa and Chichibu belts in the border zone between the areas exposed the Chichibu and Sambagawa groups, and in the southern marginal zone of the mapped district respectively. Fig. VI-6 shows the time gap versus the distance from the area exposed the Sanchu group.

Judging from the lithology of the Permo-Triassic sedimentary rocks, they have deposited in a basin shallower than some kilometers in depth. Further, the Cretaceous rocks overling unconformably the Permo-Triassic rocks contain conglomerate, thereby showing the depositions near the coast line. Accordingly, the metamorphism had taken place at certain periods between the Jurassic and Early Cretaceous (180-110 my bp). Fig. VI-7 illustrates the change of depth of the Mamba and Upper Sambagawa formations. The above mentioned temperature rise in the high P/T type regional metamorphism may require for the duration more than 10 mys, on the basis of thermal conduction accompanied with the thermal recovery (refer following section). Further, the duration of overall metamorphic event may be at least 10 mys. Therefore, the sinking and uprift of metamorphic terrane must have been performed in the duration of 60 million years. According to the isotope age determination by YAMAGUCHI and YANAGI (1970) the Sambagawa metamorphic age is about 110 mybp, and thus the sinking of metamorphic terrane succeeded from 180 to 140 mybp, and in the period of 140 to 110 my the thermal recovery had taken place. Then, the uprift of metamorphic terrane may have been rapidly performed. In this model, the rates of the sinking and uprift are 0.5 and 1 mm/year. The estimated metamorpbic age is older than the K-Ar and Rb-Sr ages by BANNO and MILLER (1961).

The depth of metamorphism can be estimated by its pressure condition, because the so-called tectonic overpressure induced by stress is less than 1 kb in the rock which behaves as the viscous fluid under the stress through the grain displacement process. The depth around 20 to 15 km can be obtained from the assumption of density of rock, = 3 gr/cm³. It suggests whether enormous amount of the Jurassic sediments cover the Permo-Triassic rocks at certain period between the Jurassic and Early Cretaceous, or the Permo-Triassic rocks have squeezed into some thick beds. The latter case contains the idea that the Permo-Triassic beds were manifolded by major overturned folding. From this idea, the physical conditions of the deep-seated Permo-Triassic rocks may satisfy that of the glaucophane schist fades. On the other hand, in the former, we must accept the views that the vast amounts of the Jurassic and Cretaceous sediments had been erroded out before the deposition of the Ato-kura and Sanchu groups.

These two ideas may be experimented by the lithic fragments and its primary ages in the sedimentary rocks of the Shimanto and Torinosu groups in the Shikoku, and the Kobotoke group in the Kanto Mountains. The lithology of the Torinosu group and Sakamoto formation (Upper Jurassic) may not satisfy the former idea that vast amount of Jurassic sediments covered the Sambagawa and Chichibu groups.

ii) Genesis of the Sambagawa Metamorphic Terrane

The high P/T type regional metamorphic terrane is widely exposed in the fossil and survived suture zone of oceanic and continental plates or two continental plates and it is accompanied by the low P/T type metamorphic belt situated on its continental side (MIYASHIRO, 1961). These paired metamorphic belts and low geo-thermal gradient estimated from low heat flow along the inner trench slope (HASEBE et al., 1970) confirm that the high P/T type metamorphism occurs beneath the inner trench slope proposed by MIYASHIRO (1961). Further, the high P/T type regional metamorphism occurred after the Paleozoic age. Recently, ERNST (1972) summarized the results of isotope age determination in the world, and obtained the same conclusions as MIYASHIRO'S suggestion.

LARSON and CHASE (1972), and LARSON and PITTMAN (1972) succeeded to clas-sifing the geomagnetic lineations of the Phenix, Hawaii and east of Japan Arc. According to their results, the geomagnetic lineations of these areas were formed from the Jurassic and Early Cretaceous. Therefore, it is confirmed that the trench model of the high P/T type regional metamorphism can be applied for the Mesozoicage. Moreover, Larson and Chase suggested that the Kura-Pacific ridge which produced the Kura and Pacific plates had gone beneath the Japan Arc, at the period around 80 my bp. From this fact, UYEDA and MIYASHIRO (1973) pointed out that the high P/T type regional metamorphism ceased at 30 mys before the collision between the Japan Arc and ridge, by comparing the Sambagawa metamorphism with the Franciscan ones. They also suggested that the cease of metamorphism may be due to the temperature rise derived from the descending hot oceanic plate.

The Sambagawa metamorphism has been characterized by the temperature rise maintaining the constant pressure as discussed in the previous sections. The temperature rise shows that the geotherm of the deep seated metamorphic terranes changes from that corresponding to the area of low heat flow to that the area of high heat flow (UYEDA and HORAI, 1964). The low geothermal gradient may be explained by the cooling of the inner trench slope derived from the descending of the cool oceanic plate at the trench. If the cool oceanic plate stops descending, the oceanic plate and the continental edge may be heated by the thermal conduction. On the other hand, the deep seated metamorphic belt may be heated by the descending of hot oceanic plate in the neighbourhood of the survivied ridge as suggested by UYEDA and MIYASHIRO (1973). The temperature rise might be called as "thermal recovery" for the above mentioned reasons. This thermal recovery may be followed by the uprift of the metamorphic belt. The uprift of the metamorphic belt may be caused by the same reason as the thermal recovery of the metamorphism. This reason is whether the hot oceanic plate descends beneath the continent or the island arc or the oceanic plate stops descending. The uprift may be due to both of these reasons.