41 



temperature variation is less, and the seasons are determined 

 by the prevailing winds and rainfall, the epidemic season is 

 also well defined, but varies in different localities. 



The seasonal incidence of plague is not due to any 

 periodicity in the breeding of rats. In India, in localities 

 where the latter was investigated, it bore no relation to the 

 season when plague occurred. It was found by the Com- 

 mission (Reports, 1908, p. 266, and 1910, pp. 446 and 483), 

 however, that the epidemic season coincided with the 

 period of greatest flea prevalence. In Bombay, Belgaum, 

 Poona, and two Punjab villages the Commission undertook 

 a census of rat fleas upon captured rats, in the course of 

 which the fleas on 150,000 rats were recorded. In each 

 case the observations extended over more than one year. 

 The analysis of the figures showed a seasonal variation in 

 the number of rat fleas in all the localities. The average 

 number per rat varied in Bombay between 3 and 7, in Poona 

 between 2 and 11, in the Punjab between 2 and 12, and in 

 Belgaum between 1 and 17. The interesting point is that in 

 all the places examined plague is epidemic when the average 

 number of fleas is well above the mean, and the height of 

 the epidemic corresponds fairly closely with the season of 

 maximal flea prevalence. A similar seasonal variation 

 in the prevalence of X. cheopis and correspondence between 

 the maximum of these fleas and the epidemic period 

 has been observed by Kitasato (1909) in Japan, Tidswell 

 (1910) in Sydney, and by Gauthier and Raybaud (1910-11) 

 at Marseilles, and by Andrew (1911) in Tongschang in 

 Northern China. The incidence of the epidemic period in 

 the season of greatest flea prevalence under such different 

 climatic conditions as obtain in Bombay, the Punjab, 

 Poona, Tongschang, and Marseilles is in itself suggestive, 

 and, in view of the results of animal experiments already 

 referred to, one is justified in believing that the most 

 favourable time for the epidemic is largely determined by 

 the seasonal prevalence of rat fleas. 



(5) The rat-flea hypothesis also affords an interpretation 

 of the fact that epidemics decline when the mean daily 

 temperature passes 85 F. A mean temperature of 85 F. 

 in -India means also dry weather. Fleas, whether in the 

 larval, pupal, or imago state, are rapidly killed off at this 

 temperature unless associated with a high degree of 

 humidity. 



The Commission for the Investigation of Plague in India 

 found that in Bombay the proportion of fleas which become 

 infected, the time they remain infected, and the number of 

 successful experiments on flea .transmission was greatest 

 during the winter months (Reports, 1907, pp. 197-204, 1908, 

 pp. 510-15 and 529-38). An analysis of their results 

 showed 67 per cent, successful transmissions when the 



