In Japan, up to now, historical geographers have usually tried to be restored the geographical landscape and the other historical aspects with every time of the past to the original state by using ancient documents, ancient maps, and cadastral maps. And then the geographically restored landscapes at every time have been projected one after another on a map which is used in the present time. This is known in terms of “successive cross-section method”. The long and the short of it is that the geographical land features (natural aspects) had been of little importance in their investigations. So, on the contrary, the author has been made a full discussion of the above-mentioned subject from a stand-point of the natural changes of micro-landform which is inseparably bound up with the process of land occupation by man in the historical age. The Fukuchiyama-Ayabe, Kaibara, Sasayama basins surrounded by mountains on all sides but one, and the Ohno and Katsuyama basins situated in the Ohuetsu mountain regions were selected as basins for the research and investigation.
Taken altogether it has been definitely shown by the following three works that, land development considered, our research and investigation goes a long way toward a better under-standing of the manor structure in the mountain basins: that is, (A) to group the land sur-face of their basins under several landform types, (B) to search after the earth mark allotted according to the grid pattern of Jon-system in the Nara a era that underlie the history of land development in the basins, (C) to clear up the relation between the allotment of grounds and the landform.
The results obtained are as follows:
1) As a rule, the land surfaces in basin are formed from about thirteen varieties of diluvial upland, alluvial fan, river terrace, valley plain, flood plain, natural levee, and the like. However, we are able to classify them into two groups; one is basin formed from depositional phases as the main constituents and the other is basin come into being from erosional phases along the river courses which have worn channels through the soft depositional phases.
2) In the Kaibara basin, depositional phases are observed and on the land surface we could not be recognized the existence of the erosional phases. On the other hand, in the Ohno basin, the erosional phases are gradually forming on the landsurface of depositional phases. Among the river terraces, in the Sasayama basin, flood plains have well-developed out of the landform which does not show as yet the erosional phase. And in the Fukuchiyama-Ayabe basin, the natural levees have highly grown among the river terraces.
3) A difference of micro-landform has brought diversity on the distribution of allotment of land performed according to the grid pattern of Jon-system, namely, in a case of the Kaibara basin the earth marks of land allotted in accordance with the Jon-system are found out on the flood plains of depositional phase. In the area formed from the braided rivers, we can not recognized the earth mark. In witness where of the braided rivers, it did not develop into a clear formation of erosional phase there. Ohno basin where the growth of depositional phase is conspicuous is a far cry from the Kaibara. On the left-over areas (back swamp) where were no filled up by a river, the earth marks allotted according to the grid pattern of Jon-system in the Nara era have observed distinctly. However, there is no allotment of land on the dormant alluvial fans. In the Sasayama and Katsuyam.a basins, on the other hand, they spread out neatly on the land surface of river terraces.
On the contrary, in the Fukuchiyama basin the allotments of land performed in accordance with the Jon-system locate on the flood plains among river terraces.

View full abstract

The writer intended to make clear the distribution area of evergreen broad-leaved forests in Kanto District, which are the climax forests of lowland vegetation. By surveying remnant forests in certain parts of mountain areas and the forests belonging to Shinto shrines and Buddhist temples, the latter are well conserved in the original condition, the inland limits of distribution were studied in relation to the climatic conditions. The species composition and structure of the forests were also dealt with in comparison with those in south-western Japan, the native ground of the forests.
(1) Using the available data on the flora or the vegetational reports, the remnant stands of natural evergreen broad-leaved forest are plotted on a distribution map (Fig. 1). From the map the following facts must be noticed. The remnant stands appear on the coastal hilly terrain in Kanagawa, Chiba and Ibaraki Prefectures and on the foot of mountains in Kanagawa, Tokyo, Saitama, Tochigi and Ibaraki Prefectures. It is an important fact that they are small in number in Gunma Prefecture, whilst they are few in the central parts of the Kanto Plain because of destruction by human activities.
(2) Next, thermal condition controlling the distribution of evergreen broad-leaved forests was surveyed as follows : The warmth index (W.I.) was estimated by monthly air temperature with a lapse rate of 0.61°C per 100m. Topographical map (1:1, 000, 000) was used for base map. The areal mean values of the heights with various warmth indicies within an area of 30km×30km were calculated for each square using the estimated temperature at the 159 climatological stations in this district. The distribution of the warmth index is shown in Fig. 2. By the same method the distribution map of coldness index (C.I.) was drawn as given in Fig. 3. If we superimpose Fig. 2 on Fig. 3, it can be observed that the isolines, running along the foot of the mountainous area in Kanto District, are C.I. -5°C, W.I. 105°C, W.I. 90°C etc. from the Pacific Ocean to the inland and the isolines of W.I. and C.I. run roughly in parallel with alternal pattern. Comparing the distribution map of the remnant forests (Fig. 1) with them (Figs. 2, 3), the boundary of the remnant forests are found between the lines of W.I. 105°C and 90°C and it coincides with the lines of C.I. -10°C in Kanto District. For the actual heights of the upper limits of vertical distribution of evergreen broad-leaved forests, mostly the Cyclobalanopsis forests, the warmth and the coldness indices in Kanto District are calculated (Table 1), showing that they are approximately represented by C.I. -10°C. Therefore, the inland boundary of the evergreen broad-leaved forests of this district is expressed by C.I. -10°C as the same for Tohoku District.
(3) Field observations were tried at 15 stations in this district as show in Fig. 1. Quantitative data by quadrat method was taken at places as given in Table 2. Size of quadrat was 100m² - 400m². All trees with a diameter more than 3cm at breast height were checked in a quadrat as well as stratification, forest floor vegetation, and the number of species and of individuals. Basal areas at breast height were calculated from the trees with a diameter more than 10cm (A group), 3-10cm (B group), as shown in Table 2. Three types of evergreen broad-leaved forests are recognized from the dominant species on the table: Cyclobalanopsis type (with Abies firma or Torreya nucif era) on the slope of Mt. Takao Mt. Nishikanasa, Mt. Sengen A, Mt. Dokko, and at Iko Shrine. Shiia type on the slopes of Mt. Sengen B, Mt. Takadate, and Mt. Butchozan. Machilus type at Kozaki Shrine.
Basal area at breast height of the Cyclobalanopsis spp.

View full abstract

The moving observations in and around a New-Town at Hibariga-Oka (Lark Hill) have been carried out to measure a specimen of urban temperature distribution. This New-Town is located at the western outskirts of Tokyo and the distance between the centers of them is about 20 km and 1 hr's ride by the suburban and subway trains (Fig. 1). The town has been consisted of 182 residential buildings in the area of 335, 000m². Most of them are four storied and they have been lived by mostly commuters to Tokyo. And its total population reachs almost 10, 000.
The temperature distribution has been measured several times during the second half of 1966, using a thermister thermometer mounted on a bicycle or an automobile about at 1m above the ground. The observation route and observation points have been selected beforehand as were indicated on Fig. 2.
The observed temperatures at various points were converted into the relative values. This values were represented by the differences between the observed values and those of the central points (B) of the town respectively.
When they were calm and not overcasted, heat islands were identified clearly as on Sept. 20, Oct. 8 and 13 (Figs. 5 a, b and c) . And almost every occasion, the lowest temperature was recorded at the western outside of the town. However, when the strong wind prevailed as on Nov. 3 (Fig. 5 d) or completely overcasted as on Nov. 30 (Fig. 5 e), heat islands were not observed.
Every two-hourly observations have been scheduled particularly from 1800 to 0600 on Dec. 10-11, and a process of heat island formation has been recognized.
At the beginning of the observation at 1800, there was no appearance of heat island (Fig. 6 a). However, heat island began to appear about at 2200 and then its formation became remarkable more and more as the time passed to midnight. The full formation of heat island on this occasion reached at the early morning at 0400 (Fig. 6 b-g). No appearence of heat island at the early evening might be influenced by the heat production caused by traffic congestion at the eastern-side of the town. And the full formation at 0400 might be attributed to the sparsity of the traffic. The comparison of the hourly mean temperature change of out-side (A) and inside (B) of the town has shown no significant increase of the temperature difference between these two points as the time passed (Fig. 7).
The vertical temperature profiles up to 25m has been also observed from 1700 to 2400 on Dec. 10 at the central part of the town. The temperatures at the height of 25m, 18m, 12m, 5m and 1m have been recorded by electric thermometers installed on a water-tank tower. The profiles showed strong surface inversions all through the observation period. The largest temperature difference between 25m and 1m was 2.5°C at 1800. Then the difference began to decrease to 0.5°C at 2200 (Figs. 8 and 9).
Checking the sequences of the vertical temperature profiles and the wind situations (ob-served at the nearest weather observation station of 5km away from the town) during the night of Dec. 10-11, it has been distinguished that the existance of fairly remarkable correlations between increasing of the wind speed and weakening of the surface inversion and sudden rise of surface temperature, as were recorded at 1820, 2020, 2120 and 2230 (Fig. 8). The surface temperature differences between inside and outside of the town have been shown on Fig. 10. The remarkable is the increasing of the difference just after the weakening of the surface inversion in the town as it were at 1810 to 1835 on Fig. 10.
The heat production by human activities in the town has been roughly estimated about 6.4×10⁹ car/hr. This amount is evaluated almost equivalent with the total heat need to warm up the town air by 1°C. However, the amount of produced heat by the town might not be sufficient to create heat islands in temperature distribution.

View full abstract