PART I,] 



Influence of Atmospheric Temperature. 



217 



It will be observed that in both table and diagram the months are arranged 

 starting from November, in place of from the commencement of the calendar year. 

 The same arrangement has been followed in the subsequent tables and diagrams 

 in connection with Calcutta. The arrangement was adopted because November, as 

 the month of average prevalence, forms a good starting point for comparison ; but 

 other advantages also attend it, one of which is that, so far as the phenomena of 

 rainfall are concerned, October makes a more natural termination to the year than 

 December. 



Atmospheric-pressure curve [9 years] ... 



Cholera deaths [3S years] 



Diagram 1. — Atmospheric Pressure and Cholera-prevalence in Calcutta. 



The result of the comparison shows that the season of minimum prevalence is 

 characterised by low atmospheric pressure. Farther than this, however, the coincidence 

 ceases. There is no indication of the existence of any definite relation between 

 degree of atmospheric pressure and prevalence of cholera. December and January, 

 the months of maximum pressure, show less prevalence than November on the one 

 hand, and much less than February, March, April, and May on the other. So 

 again June shows lower atmospheric pressure, but much higher prevalence than 

 August and September. Atmospheric pressure, considered in the light of these 

 data, cannot be regarded as exerting any direct influence on the prevalence of 

 cholera. The coincidence of low atmospheric pressure with minimum prevalence 

 must be regarded as such only, or if any influence be exerted by the pressure, 

 it must act through some intermediate agency. 



(b) Atmospheric Temperature. 



A mere glance at the table and diagram below renders it evident that temperature, 

 if it exert any influence on the variations in prevalence of cholera in Calcutta, does so 

 only in a very subordinate way. We find periods of maximum, minimum and medium 

 prevalence occurring with an almost unaltered temperature. For example, the average 

 temperatures of April and July only differ by 1°"2, whilst the former month is one 

 of maximum, the latter one of minimum prevalence. Again, the temperatures of March 



