The Guanyindong site is situated at Shajing Commune, Guizhou Province, some 152km northwest of Guiyang (106°E, 27°N). First led by PEI, W. C. and then by LI Yanxian and WEN Benheng, four excavations under the supervision of the Institute of Vertebrate Paleontology and Paleoanthropology, Academia Sinica, were carried out during the years 1964-1973. According to the observation of LI, W. X. the stratigraphy of the cave is composed of nine layers. From the top to bottom: Layer 1; black sandy clay, 15-70cm thick. Layer 2; red clay, 40-240cm thick. Stone implements and mammalian remains were found. The lower part of this layer is lighter in color than the upper, and it contains a great deal of rock debris. Layer 3; loose sandy clay, brown-yellow and gray-yellow in color, 30-160cm thick. A few atrifacts and mammalian remains were found. This layer contains limestone blocks and pebbles. Layer 4; sandy clay, brown-yellow and red-yellow in color, 90-150cm thick. Stone artifacts and mammalian remains were found. This layer contains limestone blocks and pebbles. It is capped with a layer of hard stalagmitic crust. Layer 5; gray sandy clay, 10-15cm thick. Rich in stone artifacts and mammalian remains. Limestone pebbles were included in this layer. Layer 6; soils similar to those in Layer 4, except for the absence of the stalagmitic crust and lacking limestone blocks. This layer is less cemented. Stone artifacts and mammalian remains were found. The thickness of this layer is 20-60cm. Layer 7; variegated sandy clay, 15-70cm in thickness. The soil color is gray-yellow. Stone artifacts and mammalian remains were found. Layer 8; variegated coarse sand, 15-50cm thick. The soil color is yellow. Poor in stone artifacts and mammalian remains. Pebbles and limestone blocks were found. Layer 9; a barren layer, more than 450cm in thickness. Coarse sand, pebbles and clay were interbedded. No excavation was carried out on the bottom. Except for the first layer, the remaining layers were divided into three groups: Layer 2 belongs to group A, Layers 3-8 belong to group B, and Layer 9 belongs to group C. There are 23 kinds of fossil mammals found in Guanyindong cave; all belong to the Ailuropoda-Stegodon fauna complex. LI, W. X. suggested that group A may be referred to the Middle Pleistocene and group B to the early Middle Pleistocent, for it bears the marks of Tertiary survival. Accordingly, the stone artifacts from both groups belong to the Lower Paleolithic. The author believes as follows: First, as suggested by PEI, W. C., the sediments appearing under Layer 3 were deposited by streams. Second, on the basis of taphonomy, humans are unlikely to have occupied the cave. The discovered mammalian remains might have been brought in by Rodentia and Carnivora. Third, the Uranium Series dating done at Peking University shows that the date of Layer 2 is 57±3ka; those of Layer 5 are 76±4ka, 84±5ka and 104±6ka; and that of Layer 8 is 115±7/6ka. Accordingly, Layers 2-8 should belong to the Late Pleistocene. In July 1988, the author visited the IVPP and made some observations of several stone artifacts from the Guanyindong site. Those stone implements are mainly flake tools including canted points, flake points, notched flakes, awls, becs, backed knives and a handaxe. Among them the marks of early blade industries were indentified, including crested blades, bladecores, bladecore platform rejuvenated flakes and prepared platform blades. These blade industries are perhaps the earliest known in mainland China. Thus, the stone implement assemblage found from the Guanyindong site represents the most important finding of Middle Paleolithic tools in China.
Lake Nojiri is situated in the northern part of Nagano Prefecture, and Tategahana Site is located on the west coast of the lake. This site has been excavated ten times during the period from 1962 to 1987, and these excavations have recovered many Paleolithic artifacts associated with fossil bones of Palaeoloxodon naumanni (Naumann's elephant) and Sinomegaceros yabei (Yabe's giant deer) from the Upper Pleistocene Nojiri-ko Formation. Paleolithic artifacts and big game fossils were continuously recovered from the Upper Nojiri-ko Member I down to near the base of the Lower Nojiri-ko Member III. The paleolithic artifacts, which coexist with big game fossils, are made of stone, bone, and wood. There are two toosl-making techniques within this range. One of them, belonging to the Paleolithic tradition, is characterized by bone tools and trapezoid flakes laid down during the depositional epoch from the Lower Nojiri-ko Member III up to the Middle Nojiri-ko Member III. The other is characterized by lengthwise flakes laid down during the depositional epoch from the Middle Nojiri-ko Member III up to the Upper Nojiri-ko Member I (about 24, 000y.B.P.). We propose to name these two industries the “Nojiri-ko Culture”. This culture is characterized by tools made of big game bones. In addition, the “Kill-Site” hypothesis suggested for this site in 1984, is verified on circumstantial evidence: the skull of a Naumann's elephant was uncovered together with a spear-form wooden tool, and ribs of a Naumann's elephant together with a bone cleaver and other bone tools.
A very important problem for Quaternary research is determining when human beings first settled the Japanese islands. Recently many artifacts of the Early Paleolithic age, dating from before 30, 000 years ago, have been discovered in the northern part of Sendai plain, North Japan. The age of these artifacts has been determined mainly by radiometric dating methods. On the other hand, tephrochronology is an effective technique for establishing Quaternary stratigraphy in the Japanese islands and their surrounding area. The author investigated the age of horizons bearing artifacts on representative sites of that age from the viewpoint of tephrochronology. The stratigraphy of proximal tephra layers is indicated in Fig. 1. Useful widespread tephra layers for chronological study of this area are Toya (ca. 90, 000-100, 000y.B.P.), On-Pm I (ca. 80, 000y.B.P.), Aso-4 (ca. 70, 000y.B.P.) and AT (ca. 21, 000-22, 000y.B.P.). Toya, On-Pm I, Aso-4 and AT were discovered from the horizons between IcP and KtA, IcP and KtA, N-N and N-Y, and N-Y and NK-U, respectively. At Babadan A site, artifacts of the Early Paleolithic age were excavated from horizons in Ando soil found below IwP, just below and above IcP, between IcP and KtA, just above N-Y, and between N-Y and NK-U. Consequently, the artifacts excavated from the horizons below KtA belong to the Early Paleolithic age. At Zazaragi site, another representative site belonging to the Early Paleolithic age, the artifacts were excavated from the horizons of orange-colored volcanic ash (the lower part of Yasuzawa volcanic ash). During the excavation, this deposit was regarded as an airfall tephra layer. But the author interpreted this deposit as a weathered part of a pyroclastic flow deposit (N-Y3) on the basis of the following observations: (1) The boundary of the lower part of Yasuzawa volcanic ash and lower pyroclastic flow deposit is irregular and not clear in some place. (2) It is massive and not layered. There is no intercalation of soil. (3) It is not well sorted. (4) It includes charcoals. (5) It has segregation pipes. Some of them develop at the boundary of the lower part between the Yasuzawa volcanic ash and the lower pyroclastic flow deposit. The author believes that the artifacts were incorporated into pyroclastic flow and carried to their present site. The horizons from which artifacts of the Early Paleolithic age in Japan have been excavated are plotted in Fig. 8. It has been confirmed that some widespread tephra layers originated from gigantic eruptions and covered the area surrounding the Sea of Japan (cf. MACHIDA and ARAI, 1933). The author believes that common chronological studies of Paleolithics in this area are made feasible by tephrochronology.
Since 1980, 38 early and middle Paleolithic sites have been found in Miyagi Prefecture, northeastern Honshu, Japan. In particular along the Eai River, stratigraphic sequences can be well observed on terraces of tephra. Until now 9 of them, Yamada-Uenodai, Zazaragi, Kitamae, Nakamine C, Shibiki, Babadan A, Nagakuki, Aobayama B, and Kashiwagi had been excavated systematically. Industries from these sites are divided into three groups by typology, lithic reduction systems, raw materials and tephrochronology of the layers including artifacts. The oldest group A which is found from the sites located on the highest terraces is characterized by the small pieces of shattered vein quartz. The small size and crudity of the industry reflect probably only the raw material. Vein quartz is very intractable, but it is broken into fragments with useful, razor-sharp edges. Tools made on these blanks are typologically less specialized. There were also a few andesite artifacts, which gave rather massive and crude impressions. The group A has been found from Babadan A strata 20, 32, 33, Nakamine C stratum VII and Aobayama B stratum 11d. They are older than 110, 000y.B.P. Group B is 110, 000-45, 000y.B.P. As compared with the group A, small tools of quartz are less dominant and tools of siliceous shale are increased. Various kinds of scrapers and points are characteristic of this group. Crude and massive tools such as bifaces, pics, choppings and ovades are also observed. This group was included in Yamada-Uenodai lower, Nakamine C stratum VII and Shibiki stratum 9. The industry of Kashiwagi stratum 7 has a character common to both A and B groups. From stratigraphy and production techniques, industries from Kitamae lower and Shibiki stratum 8 are probably comparable to the later part of this group. Group C is 45, 000-35, 000y.B.P. Major raw material is siliceous shale. Reduction systems and variations of lithic tools are the same as in the group B, however, axe-shaped tools are highly characteristic. Crude tools of andesite are thoroughly disappeared. Babadan A strata 6, 7, Zazaragi stratum 13, Nakamine C stratum III and Shibiki stratum 6 are included this group. Besides the industries found in Miyagi Prefecture, those of Souzudai and Fujiyama may be included in the group B, and Gongenyama II, Irinosawa and Tama New Town No. 471 in the groupes B or C, and Kiribara in the final stage of the group C. From these findings many problems arose as to the migration of Homo erectus, and the transition from early to middle or from middle to late Paleolithic in eastern Asia. Since the beginning of our studies, interdisciplinary researches have been carried out for solution of various questions. Such kind of cooperations shall be maintained and promoted for further investigation.
The Tomizawa site comprises more than 27 layers within a depth of 5m which include superposed archaeological horizons ranging from the Paleolithic to modern ages. The site is situated on the coastal plain in the southeastern part of Sendai City, northeast Honshu, Japan. The location is on the backmarsh surrounded by natural levees along the Natori and Hirose Rivers flowing east into the Pacific. The 33 archaeological horizons include modern laid earth and arable soil (in layer 1); paddy fields with irrigation ditches of modern age (2), from modern to middle age (3), and middle age (4-5); deposits of the Heian Period (6-7), of the early and late Kofun Periods and of the late Yayoi Period (8), and of the early Yayoi Period (9); traces of Jomon Period (16) and of Paleolithic age (25-27). The Paleolithic horizon dates from about 23, 000 years before present in 14C age. With the rise in sea level in the Late Pleistocene to Holocene, the Paleolithic horizon was covered by fluviatile deposits over which the later horizons were superposed successively. More than 100 Paleolithic implements include 2 knives, 1 notch, 1 retouched flake, 1 hammer stone, 2-3 cores and many flakes. In layers 25-26, buried forest materials, dung of deer and sites of open-air fires were found. The flora consists of Picea glehnii, Pinus koraiensis, Abies sp., Tsuga sp., Betula ermanni, Alnus hirsuta, Larix kamtschatica, Myriophyllum sp. and others. The site was found by excavation preceding the construction of a primary school; however, realizing the significance of this site, the municipal authorities relocated the school and designated the site for a field museum of archaeology and natural history.
Middle-Upper Pleistocene strata yielding abundant macroscopic plant and pollen fossils are distributed here and there in the coastal plain and inland basin areas in Northeast Honshu, Japan. Among them, those in the southern part of Tohoku district are investigated in full detail. On the basis of stratigraphical distribution of the taxa of woody plants obtained from these strata, the local assemblage zones are recognized as shown in Table 1, namely, TODERA 1-2, TODERA 3, TODERA 4, TODERA 5-6, TODERA 7, TODERA 8, LIG, LG III, LG II and LG I fossil assemblage zones, in ascending order. LIG and LG I fossil assemblage zones can be subdivided into two subzones: the former is subdivided into LIGa and LIGb fossil assemblage subzones, and the latter into LG Ia and LG Ib subzones. The LG I fossil assemblage zone is assigned to the later cold time (about 35, 000-10, 000y.B.P. based on radiocarbon dating); the LG II fossil assemblage zone may be from the middle temperate time (about 50, 000-35, 000y.B.P. based partly on radiocarbon dating), and the LG III fossil assemblage zone may be the early cold time of the period corresponding to the Last Glacial Age in the glaciated areas. The LIG fossil assemblage zone is assigned to the Last Interglacial Age. Moreover, the fossil assemblage zones of TODERA 8, TODERA 7, TODERA 5-6, TODERA 4, TODERA 3 and TODERA 1-2 may date from the middle to early ages of the Middle Pleistocene. The fossil assemblages belonging to the LG II, TODERA 7 and TODERA 1-2 zones and LIGb subzone chiefly consist of the taxa commonly found in the temperate deciduous broadleaf forest in the Japanese Islands. On the other hand, the fossil assemblages belonging to the LG Ia subzone and TODERA 8 zone are chiefly composed of the taxa distributed in the subalpine to montane or the subarctic to boreal mixed conifer and deciduous broadleaf forests in the Japanese Islands, Kurile Islands and Sakhalin. The fossil assemblages belonging to the TODERA 5-6, TODERA 4, TODERA 3 and LG III zones and to the LIGa and LG Ib subzones consist of a mixture of the taxa in the temperate deciduous broadleaf forest and the taxa of needleleaf and broadleaf trees which are distributed in the subalpine to montane or the subarctic to boreal mixed conifer and deciduous broadleaf forests in Japan. The fossil assemblages belonging to the TODERA 4 zone and LIGa subzone include more taxa distributed in the temperate deciduous broadleaf forest than those in the subalpine to montane or in the subarctic to boreal mixed conifer and deciduous broadleaf forest. The fossil assemblages belonging to the TODERA 3, TODERA 5-6 and LG III zones and LG Ib subzone are composed of taxa consisting in opposite ratitos to those belonging to the TODERA 4 zone and the LIGa subzone. Although stratigraphcal gaps are often found in the Middle to Upper Pleistocene strata in Northeast Honshu, it is clearly recognized that the plant fossil assemblage composition periodically changed through the Middle to Late Pleistocene. In Northeast Honshu, the periodic oscillation of vegetation, which shifted between two forests represented by the present ones in the Japanese Islands, the temperate deciduous broadleaf forest and the boreal mixed conifer and deciduous broadleaf forest, became more remarkable with time during the Middle to Late Pleistocene.
The Middle and Late Pleistocene mammalian faunas of Japan are described with new opinions on their succession and relation to the continental faunas. Although fossil materials assignable to early Middle Pleistocene are seemingly scarce in Japan, the fauna of that time is considered to have been transitional between the Early and Middle Pleistocene ones. On the other hand, fossil records which are younger than early Middle Pleistocene are abundant from the mainlands of Japan; viz. the Honshu-Shikoku-Kyushu area. In the middle Middle Pleistocene, the fauna of this area contained a considerable number of taxa which are extant today in the area (about 50%). It was also characterized by a high proportion of endemic species and the predominance of temperate forest elements. From this time to the late Middle Pleistocene, several species disappeared from the fauna; at the same time, immigrants from the continent were scarce. The faunal characters of the late Middle Pleistocene were basically identical with those of the preceding time. In the early Late Pleistocene, no mammal seems to have immigrated from the neighboring continent, and faunal composition was almost consistent with that of the late Middle Pleistocene. The elements of that fauna still persisted in the late Late Pleistocene, apart from the extinction of a few forms. In addition to the fact mentioned above, immigration from the northern part of the continent was recognized in the late Late Pleistocene, although it was restricted to a few large herbivore forms and to a short time duration. The introduction of the continental faunas to the mainlands of Japan during Middle and Late Pleistocene times was not so remarkable as previously inferred. Therefore it becomes doubtful that the faunas of the area were drastically replaced by the immigration of the Choukoutien, Wanhsien and Loess faunas of China during those times.