The twentieth century is called the century of towns because large and small towns spread throughout the world. Most of them have a planned lay-out. Among them the grid pattern with straight streets intersecting others at right angles is the most widespread. Dan STANISLAWSKI, starting from the Spanish towns of the New World, traced the origin of this design back to Mohenjo-daro in the Indus Plains. A. SMAILES also considered the development of the grid pattern in the Orient, Greece, Rome, and Europe in the Middle Ages. However, both studies explained little about the plan of Mohenjo-daro itself. This author restored the original grid plan of Mohenjo-daro, and further took up another grid town, the ancient Chinese towns in the Huanghe Plains, which occurred not much later than the Indus towns, for not only their plans but also their measurement standards resemble each other very much. The author proposes that the basic grid of Mohenjo-daro would have been about 180 meters (which correspond to 100 danda in ancient India) square and subdivided into 16 sections of about 40 meters (120 Mohenjo-daro feet) square; and that, starting with the Shang capital at Zhengzhou, the Chinese grid of one li (about 400 meters) square could have begun. Thus the idealized capital plan written in Zhouli could have been modeled after this Shang capital, but it was not followed fully from the beginning. Later, however, it was realized with some modifications during the Northern Wei, Sui and Tang Dynasties. The grid pattern town planning of the Indus towns was adopted in the Indian subcontinent, namely not only India but also in Nepal, Ceylon, Burma, and as far as Cambodia. On the other hand, it spread via the Orient to Europe and finally to the New World. At the same time the Huanghe town plan was practised in Korea and Vietnam, and inspired the capital planning in Japan, i. e. Nara and Kyoto were modeled after Luoyang and Chang'an in the Sui and Tang Dynasties.
This study sets out to examine the structural and spatial changes in the relative status of the hundred largest urban centers in Japan from 1880 to 1980 in relation to the development of railway network; it charts the structural changes by means of rank-size curves and rank correlation coefficients, and the spatial changes by means of descriptive maps. We can date the structural and spatial patterns of the present urban system as well as the geographical features of the present transport network back to 1908, to a large extent to 1880, even to the era before 1868. Urban development after 1880 was not the emergence of a brand new system but rather a process of readjustment of the already existing urban system. The major changes after 1880 were the increasing metropolitan dominance and the relative growth of the centers along the main spine of the Pacific coast. There were greater structural changes in the periods before 1908, in contrast to the periods of structural stability, with further reinforcement after 1908 of the dominance of the metropolitan areas and the main spine. Although the periods before 1908 are characterized by the major development of the railway network, it is difficult to establish a clear spatial relationship between the newly opened lines and the urban development beside them. The railway network rather reinforced metropolitan dominance and the further growth of the centers along the main spine. If we are allowed to extend the definitions of a metropolis and main spine, to take into account of the more rapid growth of the regional centers than the three metropolitan areas after the mid-1970s and the development of the new transport networks of super highways, Shinkansen lines and air services after the 1960s, the basic trend will not change in the near future.
One of the problems arising in the use of grid squares as spatial units is the possible difference of areal information obtainable from two identical but displaced grid systems. The uncertainty of patterns on binary grid maps is examined as such a problem, and statistically calculated for some simple cases. It is concluded that the use of a grid map is undesirable if the map pattern is to be observed at the scale of a unit cell.
The present sea level in Miyako Island is the highest recorded during the Holocene, determined by systematic investigation of marine notches and beachrocks occurring along the coast. The elevations of the retreat points of the notches are consistent with the present tidal range and exposure. All beachrocks which are known to be due solely to tidal conditions are situated at an intertidal or almost intertidal zone. Near-intertidal beachrocks have been dated 425±70 and 1520±60 y. B. P. (Loc. 13), and 2120± 75 y. B. P. (Loc. 31), showing a relative stability of sea level during at least the past 2100 years. This situation clearly differs from the Late Holocene emergence prevailing in most of the major axial islands in the Ryukyu Frontal Arc. Such a difference can be ascribed to recent crustal movements.
With reference to the IASH's classification in World Glacier Inventory (UNESCO/IASH, 1970) and the revised classification by MtiLLER et at. (1977), the author has proposed a type of mountain glacier, the sub-alpine niche glacier formed by abnormally heavy snowfalls and the drifted snow, but not including any kind of avalanche, and named it the “Mount Chokai type glacier” (TsucHIYA, 1978b). During the period 1972-1981, yearly field surveys on the “Kaigata Glacieret, ” one of these types of glaciers, were carried out and the results showed several peculiar features of a very small glacier. Kaigata Glacieret was formed at about 1400m a. s. l., some 2000m below the so called climatic snowline, and covered about 0.04 km2 during a one-year period of development, and then diminished to form tiny ice patches during the course of a few years. The estimated maximum snow depth at the site of this glacier at the end of a snowfall season was about 45m or more. The ablation rate is very high and the observed maximum rate during a warm rainstorm was 1.4cm/h. The rate of glacier ice development was very fast. The density of the remnant snowpack on this glacier coninued to increase almost up to the stage of glacier ice after only one ablation season. The flow rate of the Kaigata Glacieret showed some variation. In the ablation years following accumulation, such as 1975 and 1979, the flow was more rapid. Similar features are recognized on other smaller perennial snow patches (including ice bodies) in the Japanese snowy mountains, but there had been no observational evidences on the flow before 1975. A comparison with similar small glaciers such as the Whakapapanui in New Zealand showed some common features, such as the site of formation being far below the climatic snowline, westerly prevailing winds in the snowfall season, leeward slope fronts at low latitudes, snowdrifts as the main nourishment sources, and large interannual fluctuations.
The purpose of this paper is to clarify the regional and seasonal features of cold fronts in Japan and the surrounding area. The main materials used in this study are visual pictures of the Geostationary Meteorological Satellite (GMS). First of all, the monthly overlapping locations of cold fronts, the other fronts and squall lines have been examined in a wide area including Japan (Figure 2). Cold fronts pass over the main island (Honshu) of Japan most frequently in April, September and October. In May fronts lie, on an average, at 22°-23°N; in winter cold fronts often move around 20°-25°N, and sometimes move toward about 10°N. Second, cloud continuity diagrams along the two cross sections of 128°E and 140°E have been illustrated by means of the daily cloud pictures (Figures 4, 5 and 6). Concerning 25°-40°N, the cold frontal cloud bands at 140°E tend to become longer and narrower, and higher in brightness than at 128°E.
Climate and agriculture on the Hainan Island is described based on the literature published in the recent years and on the author's study trip in January and February, 1984. First, the synoptic climatological background such as the frontal zones, cloud distribution, stream lines at the 850 mb level, typhoon frequencies and fog distribution are described. The southwesterly winds and northeasterly winds prevail at the surface level to the 850 mb level and the typhoon plays an important role. Secondly, distributions of air temperature, rainfall, and durations of rainy and dry seasons are presented using the climatological records. Climatic division by Ha Daizhang is introduced to show the eight regions. Thirdly, agriculture in the Hainan Island is reviewed with special reference to climatic conditions. Rubber, rice and tea cultivations are dealt with in detail. In particular, the five areas of rubber cultivation were described in terms of cold and wind damages. The upper limit of rubber cultivation can be 300-350m a. s. l. in the northern part of the island, but 500m in the southern part. Of special interest is the recent development of hybridized rice cultivation. They grow in winter, ear and flower normally in middle and later March and mature in middle and later April in the southern part of the island. Then they are brought to the home regions in the other part of China and sown there in the normal growing season. Fourthly, typhoon damage and cold damage are described and damage classes are introduced. Especially, early rice and late rice cultivation and cold damages on rubber trees are dealt with, because they are most striking in the development of tropical crop agriculture on the Hainan Island. Lastly, a perspective is given for further development.