The purpose of this paper is to reveal the regional structure of Japanese agriculture in terms of part-time farming. For this purpose, the writer made up a matrix of the characteristics of part-time farming by 1, 140 unit areas (shi and gun) in Japan. And then factor analysis is applied to this. The mapping of the factor scores has revealed the regional structure of part-time farming. The same methods are also applied to both the characteristics of agricultural management and that of non-agricultural employment.
As a result, four regional distribution patterns were distinguished with regard to the regional structure of part-time farming (Figs. 1_??_4), eight regional distribution patterns to the agricultural management structure (Figs. 5 and 6), and three regional distribution patterns to the non-agricultural employment structure (Fig. 7).
Next, the relationships among those spatial patterns were examined by canonical analysis. Four statistically significant relationships were recognized between the regional structure of part-time farming and that of agricultural management. The major findings are as follows;
1. The regions with more productive farming show larger percentage of full-time farmers, fewer aged agricultural laborers, smaller percentage of female farmers, and smaller percentage of part-time farmers.
2. The regions prosperous for traditional farming on dry fields usually cultivate wheat, barley, potatoes, miscellaneous cereals, pulses and vegetables, or operate horticulture. In those regions, a larger percentage of the people are self-employed or engaged in side business of fishery as well as being continuously employed in other businesses.
3. The agricultural regions with larger percentages of aged laborers, female laborers, or side business types of farming (whether their income is mainly earned from farming or from other activities, or from jobs in other regions) have less prosperous cropping with vegetables, less active horticulture, and lower level of land productivity, while they practice more active traditional farming on dry fields.
4. The regions growing industrial crops and/or operating sericulture have high percentages of farmers engaged in non-agricultural jobs, day laborers, farmers going to other regions to work, or farmers with self-employed enterprises.
Moreover, three statistically significant relationships are found between the regional structure of part-time farming and that of non-agricultural employment (Tables 4 and 5)
1. Side-businesses of farms such as farmers with subsidiary jobs, day laborers, or farmers going to other regions to work are more strikingly found both in the regions with meager activities in tertiary industries excluding public service, and also in the regions with active fishery and public service.
2. Farms operating self-employed non-agricultural enterprises and part-time fishery are distributed especially in the regions with active tertiary industries on the one hand. On the other hand, farmers continuously employed in other businesses are more dominantly found in the regions with active manufacturing.
3. The regions characterized by aged and female laborers in agriculture have more activities such as mining, manufacturing, and energy-supplying enterprises.
According to those results, it is noted that the influences of part-time farming on the regional structure of agriculture are recognized to be as important as agricultural management and non agricultural employment. And it can safely be said that non-agricultural employment of farmers has closely been related to the spatial distribution of tertiary industries except public service which is dominant in metropolitan or major local centers.

View full abstract

There are a number of countries and areas in the world consisted of central place systems in which periodic markets still play an important role. However, comparatively few studies have dealt with them despite the recent development of periodic market studies. For example, J. H. Stine presented, through the adaptation of central place theory, a locational theory of periodic markets based on his research in post-war Korea. But it was different from the real conditions of periodic markets and central place system in Korea so that it was not so appreciated, although his approach to the locational theory was valuable as a first attempt. The recent research conducted by S. Park also could not sufficiently throw light on the function and role of periodic markets in the central place system.
Thus, in order to understand the existing conditions of periodic markets and characteristics of a central place system which includes periodic markets, it is important and effective to analyse the actual functions and structure of periodic markets within the whole system of central places. Such study will contribute not only to fill the gap in the central place studies but also to understand the locational bases and existing conditions of periodic markets. Thus, the writers discuss the central place system including periodic markets in the surrounding areas of Kongju in Chung Cheong Nam Do.
The main results obtained are summarized as follows:
1. Although periodic markets in this area have increased in number during the chaotic period after the Korean War, recently they tend to decline due to the loss of patrons, especially in the vicinity of higher-order central palces like Kongju with the population of 42, 500. Therefore, they are distributed densely in the areas remote from a higher-order centre, and most of them are located together with other central establishments in the administrative centres of municipalities (myeon).
2. The hierarchy of periodic markets in terms of their turnovers are found not only at present time but also in the colonial period. The relationship between the turnovers of periodic markets and the centrality represented by ‘functional index’ of W. K. Davies, however, is stronger in lower-order central places (Fig. 4). It means that the periodic markets play a more important role in lower-order central places than in higher-order ones, and that higher-order places have more various central functions than expected from the turnovers of the periodic markets.
3. Since many of central places classified by the analysis of service areas do not expand their service areas even for the supply of higher-order goods, we can not evaluate the concepts of ‘inner range of a good’ or ‘threshold population’ considered till now generally as basic principles of central place location.
4. For the service areas of central places, the purchasing areas of fish (Fig. 7) and the collecting areas of rice (Fig. 10) correspond closely to the spheres of periodic markets, whereas mag-geoli (a kind of Korean liquor made of rice) are sold in each district of primary school (Fig. 6). Though the purchasing areas of men's suits (Fig, 8) are slightly larger than those of fish, they have not drastically expanded in spite of the improvement of transportation conditions in the last ten years, and are generally narrower than the service areas of dentists and doctors. It is a different phenomenon from the case in Japan. The reason for this phenomenon is that in addition to the shortage of dentists, the close relationship between periodic markets and the surrounding rural areas has been maintained. Periodic markets are supported by peasant farmers from surrounding areas, who sell their agricultural products, buy their neccessaries of life, and stay for a long time at the periodic markets which they patronize.

View full abstract

Meteorological satellite pictures show the cloud lines which appear around the Japan Islands under a winter monsoon situation. The cloud lines can be classified into four types. The first type is the cloud lines from the open sea surrounded by cold land or sea ice areas. The second type is the cloud lines over the Pacific Ocean, starting from the cloud area over the Japan Sea, and going through the lowlands between the mountain ranges of the Japan Islands. The third type is the cloud lines which start i n the Japan Sea and continue to the western coast of the Japan Islands. From there no cloud lines are seen until reaching the eastern coast to the lee of the gaps in the mountain ranges of the Japan Islands, and continue out over the Pacific Ocean. The fourth type is the cloud lines to the lee of the mountains.
The purposes of the present study are to describe the forms of the cloud lines and also to compare the atmospheric condition in relation to the appearance of the cloud lines. The satellite data used in this study are visible photographs taken from the polar-orbit satellite, NOAA-5. The visible information was used to determine the existence and location of cloud lines. The periods of this study are December, 1976 to March, 1977 and December, 1977 to March, 1978.
From the visible information, we can determine eight major areas of cloud line appearance. Most of the cloud lines appear under the winter monsoon situation. A circle with its center at a point on the cloud line and with radius of 400 km can be drawn. If a center of low pressure or a center of high pressure appears within the circle, a cloud line does not appear.
The cloud lines to the east of Sendai, in northern part of Japan for example, are described by the relationship between the directions of the lower wind and the cloud line direction. The cloud line to the east of Sendai belongs to the third type of cloud line. The reason why these cloud lines were selected is that the aerological station is located at Sendai, upward of the cloud lines. It is easy to compare the cloud line direction with the wind direction at a lower level. The mean value and standard deviation of cloud line direction is 295° and 15°, ranging from 280° to 320°. On the other hand, the mean values and standard deviations of wind direction at the 900, 850 and 800 mb levels are 295°, 295° and 290°, and 10°, 10° and 10°, respectively. The mean values of the cloud line direction are equal to those of the wind direction, but the coefficients of correlation for each case are very low. However, the scatter diagrams showing the relationship between the cloud line direction and the wind directions show close correlations, except for the cases having 280° of cloud line direction. The regression equations are shown as follows:
D₉₀-290=1.2 (Dir-290)-10.0 (r=0.73)
D₈₅-290=1.0 (Dir-290)-5.0 (r=0.69)
D₈₀-290=0.8 (Dir-290)-5.0 (r=0.59) where D₉₀, D₈₅ and D₈₀ represent the wind directions at the 900, 850 and 800mb levels, respectively. “Dir” represents the cloud line direction. The coefficient of correlation is shown by the “r” in parenthesis.
The cloud free distance means the distance from the coastal line to the upward appearance point of the cloud line along the cloud line direction. It is considered that the longer the cloud free distance is, the stabler the atmosphere around the cloud line is. The mean value of the cloud free distance to the east of Sendai and its standard deviation are 75km and 50km, ranging mainly from 50km to 100km. As an index of atmospheric stability, the temperature difference between the 850mb level and the mean sea surface temperature is used. Cloud lines do not appear under conditions when the temperature difference is lower than 21°C.

View full abstract