1904.] L. Rogers — Special Report on Fever in Dinajpur Dist, 39 



having been 49 per thousand, and it has been pointed out that tin's is an 

 unusual state of affairs. In line 4 of table VI are shown the birth-rates 

 of the tlianas in which I worked, and it will be seen that they are very 

 much lower than those of the previous year. The explanation seems 

 to be simply that as the highest death-rate due to fevers take place in 

 the autumn, its effect in reducing the birth-rate will not be evident 

 until the figures of the succeeding year are available, so that the lower 

 birth-rate of 1903, corresponds with the high death-rate of 1902, as 

 might have been expected. The large number of infants which die 

 within a few days of death in the malarial season, their mothers having 

 suffered from malaria before delivery, which has been already pointed 

 out, also go to swell both the birth and death-rates of the Dinajpur dis- 

 trict and partially account for the relatively high birth-rate. No definite 

 relationship between the birth and death-rates of the different thanas 

 can be made out from the figures given in table II, the birth-rates not 

 having varied greatly in different parts of the district. The fact that 

 both the birth and the death-rates are lowest in the two southern 

 thanas may possibly be due to less efficient registra.tion in that part of 

 the district. The rates for all the villages visited have also been ob- 

 tained for me by the District Superintendent of Police, Mr. F. L. Peters 

 (to whom I am greatly indebted for all the trouble he has taken over my 

 enquiry), but the figures show such extreme variations that it is clear 

 that the populations of individual villasfes are too small for such a 

 study to be of any value. Taking several villages too^ether I find that 

 the birth-rates were unusually low in most of the villages of the very 

 unhealthy Ranisankail circle, wliere they averaged only about 30 per 

 thousand, but more than that cannot be said. 



Vaeiations in the Mortality from Fevers in Individual Villages. 

 Next we have to consider the variations in the surroundings of differ- 

 ent villages in relationship to the fever mortalities in them. Here we 

 have to be specially cautious in drawing conclusions for several reasons. 

 In the first place the villages are mostly very small, containing two or 

 three hundred inhabitants only, so that small and possibly accidental 

 variations in the number of deaths in a single year will make great 

 differences in the rate per thousand of population. In the second place 

 the subdivision of a circle into different villages for the purposes of mor- 

 tality returns in this part of Bengal is a very artifieal one, for villages 

 in the sense of a collection of considerable number of houses close to- 

 gether are rarely met with, each so-called village really consisting of a 

 number of scattered homesteads dotted at irregular intervals over a con- 

 siderable area, which may amount to one or more square miles, so that the 



