- 著者
-
梶川 勇作
- 出版者
- The Association of Japanese Geographers
- 雑誌
- 地理学評論 (ISSN:00167444)
- 巻号頁・発行日
- vol.46, no.3, pp.211-215, 1973-03-01 (Released:2008-12-24)
- 参考文献数
- 4
- 被引用文献数
-
1
Christaller and Lösch had used the hexagon as the modular unit in their models. One of the commonest criticisms for the hexagonal framework is that it is for too rigid and abstract. But Haggett was trying to test in his book whether hexagonal arrangement does in fact exist. One of his results was that one-third of Brazilian counties (municipios) had exactly six neghbours and the mean contact number (number of adjacent territories) was 5.71. Thus, he concluded that criticism for the hexagonal system as over-theoretical may have been too hasty. In this paper author is testing the contact number of some administrative areas in Japan. These mean contact numbers are 5.5 for the prefecture (fig. 1), 5.9 for the old state (fig. 2), 5.8 for the county (tab. 1), 5.5 for the communal unit and 5.7 for the section (fig. 4), Moreover, it is suggested that the frequency of the contact number (fig. 3 and 5) have the order of 6, 5, 4, 7 and 8. These results show that hexagon is the modular unit even in the administrative areas of Japan. It is also suggested that the frequency curve of the contact number tends to be identified with the stochastic one and is similar to Bunge's experimental result of the floating-magnet in the tab.