1904.] 0. Little — The Simalayan summer storm of Sept. 2dthj 1903. 1^3- 



this condition and the persistence of the dry winds in those 

 years. But there is evidently some other condition which 

 operates in producing these winds, and in retarding oi' interrupt- 

 ing the rainfall in years such as 1883, when the atmospheric 

 pressure does not exceed the average and may even he below it. 

 The dry winds in such seasons are less lasting, but for a time 

 they are as strongl)-- marked as in famine years. I have given 

 elsewhere my reasons for believing that in such cases they are 

 due to an unusual extent to the thickness of snow on the North- 

 Western Himalayas. 



* * * * * .*.... 



But there were indications that througliout this rainy interval the 



influence of the snows was operative. On the JSTorfch- Western 



Himalayas, the rain of the first half of July was accompanied 



with frequent thunderstorms and hail storms (always a sign of 



an unsteady monsoon), and in clear intervals, the upper clouds. 



could be seen drifting steadily from the North-West, proving the 



existence, at no great lieight, of the dry current which about the 



19th descended and replaced the monsoon, prevailing over the 



whole of Northern India and Rajputana, and influencing the 



winds down to the Deccan on the south and Bihar on the east." 



Further reference is made in the report for 1884, to the connection 



between the North- Westerly winds, the snowfall in the Himalayas and 



the rainfall distribution in Northern India ; and in the report for; 1885, 



the last which Mr. Blanford wrote before retirement, the following 



remark occurs : — 



'* The year under review affords a very striking instance, in 

 confirmation of the hypothesis, on which these latter forecasts have 

 been principally based, viz.,ih.Sit the persistence of dry North- 

 westerly winds in Western India and Westerly winds in Nor- 

 thern India is greatly influenced by the depth and extent of the 

 snow accumulation on the North- West Himalayas." 

 It will, I think, be clear from the above extracts that fin important 

 factor, in Mr. Blanford's opinion, in estimating the character of the 

 Weather in Northern India, not only for the monsoon season as a whole, 

 but for brief periods ahead, was connected with the upper atmospheric 

 strata. 



He repeatedly expresses his conviction of the need of some in- 

 fluence from the Noi'th, " some condition which operates in retarding or 

 interrupting the rainfall," and shows that the pressure variations or 

 pressure anomalies fail to account for rainfall failure or distribution. 

 What that condition is he merely indicates as an important subject for 

 J. II. 21 



