The thermal structure of the Sambagawa belt, which is reflected by minerals produced during the highest temperature of metamorphism, generally is characterized by the occurrence of the highest grade schists in the middle of the structural pile. The origin of this structure has been analysed in the upper horizon of the Sambagawa schist sequence in Central Shikoku as an example for the structure of the whole belt. The upper horizon of the Sambagawa schist sequence in Central Shikoku consists of three nappes, the Saruta nappe 11, the Saruta nappe I and the Fuyunose nappe in descending order of structural level. The Saruta nappe Il shows a downward increase of metamorphic grade from the garnet zone, through the albite-biotite zone, to the oligoclase-biotite zone. The Saruta nappe I consists mainly of rocks in the albite-biotite zone and partly in the oligoclase-biotite zone, but the direction of the increase of metamorphic grade is not clear. The Fuyunose nappe shows an upward increase in metamorphic grade which changes from the glaucophane zone to the barroisite zone. It is concluded that the Saruta nappe I1 and Saruta nappe I were overturned and then mechanically coupled with the Fuyunose nappe. The Sambagawa metamorphic field, which is of the highest temperature phase of metamorphism, appears to have had an inverted thermal gradient and a thermal structure comparable with that expected in the deeper parts of a subduction complex.

Quartz c-axis fabrics of the Sambagawa schists produced along a late Mesozoic convergent plate margin were analysed so that their tectono-metamorphic history could be clarified. It has been noted by many authors that quartz fabrics produced by earlier phase deformation are easily modified by strain increment during later phase deformation. This paper attempts to elucidate the high-temperature phases of prograde metamorphism (Sim-Bim phase) and of retrograde metamorphism (Sbl phase and Sb2-1 phase) from quartz grains included in garnet and plagioclase porphyroblasts. Quartz c-axis fabrics for all these phases are explained in terms of a type I crossed girdle, without (only rarely with) higher concentration in the principal axis of strain Y (X>Y>Z), that must have been produced by the activity of a dominant slip system such as rhomb 500°C).