160 0. Little— The Himalayan summer storm of Sept. Sdth^ 1903. [No. 4, 



practically no monsoon either in the Arabian Sea or in the Bay of 

 Bengal, if it be understood that a smooth sea and comparatively light 

 winds indicate the absence of monsoon ; and, as the Pioneer states, the 

 skies had become cloudless in Upper India. We have also the evi- 

 dence of the meteorologists of St. Xavier's College, Calcutta, that in 

 their opinion the rains would terminate early — an opinion given quite 

 independently of the Weather Bureau. Assuming that the chances 

 were 10 to 1 against a disturbance developing in the Bay of Bengal or 

 the Arabian Sea during the latter half of September or first week of 

 October, and 10 to 1 against that storm moving north-westward, if it 

 happened to form, the chances were 100 to 1 against rainfall in the 

 United Provinces or the Punjaub. The above chances are not at all 

 unlikely if we remember that the monsoon had become exceedingly 

 weak in the Indian sea area, and that the sky had become cloudless in 

 Upper India : or, in other words, that atmospheric conditions in the 

 ^orth of India and the Bay of Bengal were such as in ordinary years 

 are experienced in the early part of October. Experience has shown 

 that under such circumstances no cyclonic storm has ever advanced 

 from the Bay of Bengal or the Arabian Sea in the direction of North- 

 west India. 



The conclusion therefore is that the subordinates of the Weather 

 Bureau had two alternatives. They had either to stand by the systesn 

 and come to the conclusion that rainfall had ceased for the year in 

 North-West India, or they had to put aside the system and strike out a 

 new course for themselves. That the former course was adopted is in 

 my opinion not only natural ; the adoption of any other was impossible. 



One of my reasons for making the above newspaper quotations is 

 to emphasise, as strongly as it is possible to do, the fact that weather in 

 Northern India just after the middle of September was of a kind to 

 suggest an early termination of the rains in Northern India, and that 

 before the end of the first week of October rainy weather of an intense 

 character had become established over a great part of Northern India. 



A change so noticeable, as that most undoubtedly was, must have 

 been due to well-marked local causes, which ought to become apparent 

 on subsequent enquiry. Now whether these causes can be brought to 

 the surface from the meteorological records or not, there can be no 

 question that the events of the early days of October were not foreseen, 

 and in my opinion could not, under the present system, have been fore- 

 seen, for reasons which I have been trying to make clear in previous 

 papers read before this Society. Until the atmospheric processes are 

 better understood than they are at present, it is a mere truism to state 

 that even approximg-tely accurate forecasting is impossible. 



