2004 年 69 巻 5 号 p. 553-564
The urban graveyard theory has emphasized that mortality is higher in urban than in rural areas. Osamu SAITO has reviewed research on urban and rural mortality in Tokugawa Japan (1603〜1867), and concludes that the theory has not yet been confirmed but seems likely if Farr's law is applied. This conclusion seems to have been widely accepted. In this paper, I examine whether it is reasonable to conclude that the urban mortality rate was high by using the data for 1883 to test whether Farr's law holds in the Japanese case. The results of my analysis are as follows: First, the mortality rate depends on the distributional patterns of population density in urban areas. Second, Farr's law should only hold for areas with similar patterns. Third, since urban areas in Japan showed various patterns at the beginning of the Meiji period (1868〜1912), only large areas with similar patterns follow Farr's law. Fourth, the existence of the different patterns should lead us to expect a wider range of urban mortality.