1903.] C. Little — On two remarkable rain-hursts in Bengal. 



55 



certain, but the character of the isobars and of the pressure changes, 

 would appear to indicate the possibility of its advancing in a north- 

 westerly direction." 



July 9th. — " The cyclonic storm in Guzerat has been almost station- 

 ary during the past twenty-four hours." 



July \Qth. — " The storm in Kathiawar is filling up but will pro- 

 bably continue to give rain in Guzerat, during the next twenty-four 

 hours." 



On July llth no reference is made to the storm which was no 

 longer shown by the ground observations and rainfall in India was 

 confined to restricted areas. 



The following Table, gives the rainfall in Guzerat, between the 6th 

 and llth July, and the amount which from the normal Tables was likely 

 to fall during those days : — 





Actual rainfall bet- 



Normal rainfall bet- 





ween 6th and 



ween 6th and 





llth July. 



llth July. 



Snrat 



2-17 



311 



Ahmednagar ... 



0'70 



191 



Bhavnagar 



65 



71 



Veraval 



605 



0-7G 



Kajkot 



0*70 



2-28 



Bhuj 



100 



118 



Deeaa 



032 



1-35 



It will be seen that Veraval alone received excess rainfall, and that 

 at three other stations, Ahmednagar, Rajkot, and Deesa, rainfall was 

 much in defect. 



I have pointed out that the two storms from the Bay in period (B) 

 recurved towards the Kumaou Hills, and that the storm from the 

 Arabian Sea filled up in Guzerat after causing rainfall near the coast. 

 In fact, that storm was very similar in its behaviour, to the first storm 

 in period (D), which filled up in south-west Bengal, and it is not unlikely, 

 that the filling up was due to similar causes, if these were only known. 

 It may, therefore, be assumed that throughout period (B) there was some 

 influence which prevented the advance of cyclonic storms, whether 

 from the Bay of Bengal or from the Arabian Sea, into the north-west 



