Fragments of earthenwares of the Earliest Jomon age are found on a sea-cliff at Shimotakabora on the west coast of O-shima Island, Izu Islands (34°44.3′N., 139°21.6′E.). They are unearthed from the upper brown weathered ash of a unit layer constituting pyroclastic fall deposits exposed there. Associated with them, are found obsidian flakes, an angular block of biotite rhyolite pumice, several round pebbles of compact basalt, and a small amount of charred wood and bone fragments. All the earthenware fragments are identified to be of the Hirasaka type of the Earliest Jomon age whose type locality is in the Miura Peninsula, South Kanto. Two samples of the obsidian flakes have been determined by fission track method to have the same uranium content and age as obsidian exposed on Kozu-shima Island about 60km south-southwest of O-shima. The biotite rhyolite pumice block is, in petrographic characters, similar to the biotite rhyolite from any one of Nii-jima, Shikine-jima or Kozu-shima Islands, all of which lie to the south-southwest. These obsidian and rhyolite pumice were brought by the Earliest Jomon (Hirasaka) men to O-shima Island for making living tools. On the sea-cliffs at Onoue and Tatsunokuchi to the south of Shimotakabora, earthenware fragments of the Kayama, Kijima plus Sekiyama, Moroiso (?) and Odoriba types have been unearthed from the particular stratigraphic levels of superimposed pyroclastic deposits. The earthenware fragments of the Hirasaka type are found at the layer lying nine fall units below the level which contains the earthenware fragment of the Kayama type about seven thousand years old. As the time represented by a unit pyroclastic layer is thought to be a hundred and several tens of years, the earthenwares of the Hirasaka type may be brought to O-shima eight to nine thousands of years ago. This estimated age well coincides with that given so far to the type.
The Myoko volcano is situated in the northern part of the Fossa Magna, Central Japan, and constructs the Myoko volcanoes together with the Niigata-Yakeyama, Kurohime, Iizuna, and other volcanoes. The Myoko volcano in general sense, is divided into the older Myoko volcano and Myoko volcano, and the growth-history of the latter is further subdivided into the Ist, IInd, IIIrd, and IVth stages. The eruptives of the each stage, except for those of the Ist stage in which no basalt occurs, consist petrographically of basalt-andesite association. The strata formed during the Myoko volcano-IIIrd stage are collectively called the Mitaharayama group, and consist of lava flows, pyroclastic (scoria, bomb, and others) fall deposits, pyroclastic flow deposits and volcanic mud flow deposits. They are widely and thickly distributed around the caldera (Fig. 1). The rocks of the essential eruptives are basalt, pyroxene-olivine andesite, and hornblende-andesite. They are characterized by a large amount of basalt to basic andesite and by a subordinate amount of acidic andesite (Fig. 2). During the activity of the IIIrd stage of the Myoko volcano, the rocks changed from basalt, through pyroxene-olivine andesite, to hornblende andesite, and the mode of the eruption also changed from the eruption of the scoria fall by the Strombolian eruption, through the alternated eruption of the lava flow and pyroclastics such as bomb by the Vulcanian eruption, to that of the pyroclastic flow by the Peléan eruption (Table 1). A conical stratovolcanic cone with the elevation of 2, 800 to 3, 000m may have been formed at a certain period during this IIIrd stage (Fig. 4). The age of the IIIrd stage-activity falls in the late Quaternary, and is older than about 30, 000 years ago.
The morphology of plant opals in the humic horizons of present and buried volcanic ash soils distributed in Tohoku, northern Kanto, eastern Tokai and Kyushu districts were investigated. Plant opals in leaves of some grasses were also examined. The results obtained were summarized as follows; 1) The morphology of plant opals in leaves of Graminaceous grasses is closely related to taxomic groups of grasses, which well confirms that of the previous reports. Dumbbel shaped plant opals, Panicoids, are characterized by subfamily Panicoideae and saddle shaped plant opals, Chloridoids, are contained specifically by Eragrostoideae. 2) The sizes of Sasaoid type plant opals vary with taxomic groups within subfamily Bambusoides Their median of size distribution is 24μ in genus Sasa and is 18μ in genus Pleioblastus. 3) The plant opals separated from the leaves of Miscanthus sinensis (Susuki in Japanese) are more soluble than those of Pleioblastus sp. (Nezasa in Japanese) when treated with 1%-Na2CO3 solution for an hour at 60°C. Microscopic observation of the insoluble residues confirms the above fact. 4) The geographical distributions and mutual relations in the various types of plant opals in the humic horizons of volcanic ash soils are as follows. Sasaoid types are recognized commonly in all sites studied, while Panicoid types are found mainly in samples from Tohoku, northern Kanto and a part of Kyushu districts. Festucoid and Point shape types show high frequency in samples from Hokkaido. Fan shape and Elongate shape types as well as Sasaoid types are recognized extensively. Chloridoid are not common types. The amounts of Point shape types and the ratios of Fan shape types vs. Elongate shape types are closely related to the amounts of Festucoid types.
Examination of dune sand columns in the coastal region of North Kyushu, Japan, and the recognition of paleosols allowed the division of the dune sand formation into four members, which are Ashiya dune sand (the youngest, Holocene), Awaya dune sand, Enokizaka dune sand and Onizu dune sand (the oldest). Some members of dune sand occur on each of two terrace surfaces. The upper surface occurs on the Ashiya beach deposits at approximately 8 metres above sea level and the lower surface occurs on the Mitoma beach deposits at about 1 metre above sea level. The upper surface is overlain four dune sand members with three layers of well developed paleosols; the lower surface is overlain three dune sand members with two layers of well developed paleosols. The upper surface predates the Würm Glacial and the lower surface correlates the WI/WII (Gettweiger) interstadial age. Paleosols do not have a complete paleo-solum. The break between a paleosol and overlying dune sand of the succeeding deposits is associated with discontinuous gravel and peat. It would appear that the upper part of the solum of the paleosols had been stripped off prior to the deposition of the succeeding dune sand layer. Peat at the base of the Awaya sand member immediately above the second paleosol gives a radiocarbon data (GaK-5157) of 22, 020±740 years B.P. A hypothesis is presented suggesting that the dune sand members were accumulated by the prevailing wind of the directions ranging N to N10°W during substages of the Würm Glacial. The periods of soil development indicated by the paleosols were at interstadials in the Würm Glacial age.