Abstract
Since the early 1900's studies in the factorial ecologies have made a tremendous progress not only in Western cities but also in non-Western cities. Fortunately, in Japan also, the census tract data of larger cities with over 200, 000 inhabitants were first published in Popu-lation Census of 1970. The aim of this paper is to clarify, through the analysis of the census data, the characters of the factorial ecologies in the cities of Hiroshima and Fukuoka. The writer hopes that such a comparative study may contribute to the general understand-ings of urban social geography in Japanese cities.
Though manufacturing activities have developed more in Hiroshima (population, 541, 998) than in Fukuoka (pop., 853, 270), both cities are important regional capitals in Chugoku and Kyushu districts respectively.
With forty-two common variables selected from the census data as shown in Table 2, the writer extracted some factors by attempting the factor analysis (principal axis method) in both cities and drew some maps on the basis of the distribution of standardized factor scores. In Fukuoka city the author attempted two analyses by using study areas of different boundaries.
The results of this analysis are summarized as follows:
1) By analyzing the structure of factor loadins, four main factors in each city were labelled as shown in Tables 3a, 3b and 3c. Family status and socio-economic status occur-red as the two main factors of each city. They correspond to the general factor structure of Western cities, though the weight of the two factors is contrary to most of American cities.
2) High relationship between the same labelled factors like family status and socio-economic status is generally recognized in each city as shown in Table 4, which indicates correlation matrix between each factor. However, it does not always mean the same structure of factor loa.dings. The third and fourth factors in one city are quite different from those in another city; while in Hiroshima blue-collar workers appear as the third and mobility status as the fourth factor; in the built-up area of Fukuoka city, the third factor is named student boarding area factor. Thus it seems that with lower weight in the factors the difference in urban characters of both cities stands out conspicuously.
3) The difference in the spatial size of the study areas is related with the characters of factors concerned. As the total city area of Fukuoka expands outwards from the built-up area to the rural areas, the spatial variation between built-up area and rural area is more obvious than the variation between urban core and peripheral residential area. Contrary to it, because of the smaller size of the total city area in Hiroshima, the spatial variation between urban core and peripheral residential area is obvious, especially in the case of family status factor with tonal pattern. Consequently, for the comparative study of fac-torial ecologies, it is desirable to unify the study areas with the geographic city areas as an areal unit, as Sweetser had asserted.
4) The spatial pattern of each factor in both cities are shown in Figures 2a-d and 3a-d. By the analysis of variance (Table 5) it became clear in each city that family status factor displayed a tonal pattern while the factor of socio-economic status had a sectoral distribu-tion. This result also corresponds to the spatial pattern of both factors in Western cities. Furthermore, in Hiroshima the distribution of blue-collar workers as the third factor is closely related to that of transportation facilities and industrial establishments, while the areas of high mobility are located in recently developed residential areas and in the con-centrated areas of residence for public officers. On the other hand, in Fukuoka city, the distribution of student boarding areas are concentrated in the vicinity of universities and colleges.
