1904.] C. Little— Oi/cZowe in the Bay of Bengal 37 



is associated with exceptioual distribution of rainfall, frequently with 

 heavy flooding. 



The only remedy is investigation of the upper strata of the atmos- 

 phere, because ground level observation fail to display the causes, and 

 therefore fail to indicate its occurrence beforehand. 



It may be said tliat with the introduction of wireless telegraphy 

 earlier and better information will be available. The experience of 

 these two ships in the cyclone of November 14th shows how difficult it 

 is for ships at sea to find out what is happening within a few miles of 

 them, and it may be inferred that, even with an efficient system of wire- 

 less telegraphy, guaranteed to continue working in all kinds of weather, 

 and more especially to be independent of thunderstorms, it is doubtful 

 whether much information will be procurable from the shipping in the 

 Bay at the time of the storm. 



The Andaman connection might have given some information on 

 the ]3fch which would have shown that a storm was approaching the 

 north of the Bay and moving towards Gopalpur, but on the point of 

 greatest importance, viz., whether the storm would strike the mouth of 

 the Hooghly or Chittagong or Akyab, information from the Andamans 

 would have been of no assistance. 



When we come to consider that wireless telegraphy fails when 

 thunderstorms are occurring, and that thunderstorms are constantly 

 occurring in the early stages of cyclonic weather, the probability of 

 assistance in forecasting weather by that means becomes practically 

 nil. Wherever wireless telegraphy has been tried so far, its working 

 has been temporarily abandoned during thunderstorms ; and this is so 

 great a drawback in storm warning work that the United States 

 Weather Bureau, after a lengthy trial of wireless telegraphy, have put it 

 aside for the present, and are continuing the laying of submarine cables 

 over the short distances between their mainland and neighbouring 

 islands. If wireless telegraphy fails for that reason as an aid in storm 

 warning in the temperate region, it is much more likely to fail in the 

 tropics where lightning is an almost constant accompaniment of even a 

 slight atmospheric disturbance. Those who doubt the interference of 

 lightning with wireless telegraphic work may refer to the latest Adminis- 

 tration Report of the Telegraph Department in India. 



It is practically certain that no improvement in storm warning 

 will follow the introduction of wireless telegraphy. A cable to the 

 Andamans would, undoubtedly, give valuable information, but only 

 ground level information which is not sufficient to settle the important 

 question of recurving. There remains the investigation of the upper 

 strata suggested by Mr. H. F. Blanford more than twenty years ago 



