1884.] of the Bay of Bengal in 1883. 179 



the Briton and Runnymede troop ships. Both vessels were dismasted 



and thrown on the Andamans. 

 1850 — November 17th — 19th. In the Andaman Sea. Passed east of 



Port Blair and travelled N. N. W. 

 1854 — April 21st — 23rd. A violent hurricane in the Gulf of Martaban 



and Rangoon. 

 1858 — April 9th — 10th. A storm from the Andamans to Cape Negrais. 



Much destruction of property occurred at Henzada and Rangoon, 



between which the centre passed. 



There is no direct evidence in this list that any of these storms 

 passed from the Martaban Gulf into the Bay of Bengal. It is, however, 

 probable that the third storm in the list did so. It appears to be parallel 

 in time of occurrence, and line of advance, with the storm under discussion. 

 There is no apparent theoretical reason in support of the opinion that 

 a storm generated in the Gulf of Martaban should not pass into the Bay. 

 Experience certainly appears to indicate that such a line of motion is 

 very rare. The great majority of the cyclonic storms in the Gulf of 

 Martaban are generated to the east of the Andamans and north-east of 

 the Nicobars, and advance in a general northerly direction across the 

 south coast of Burmah, when they rapidly break up. 



Hence, although experience is doubtless valuable in indicating the 

 probabilities of the occurrence of cyclones, and their line of motion, it 

 should be most carefully borne in mind, that they are mere probabilities 

 based, at the present time, on very limited experience, and that it 

 would almost certainly be misleading and dangerous to dogmatize our 

 limited experience into rules or laws, which might fail on their first 

 appHcation. 



Another feature deserving notice was the short period of its exist- 

 ence. Favourable conditions, according to the condensation theory, were 

 present for some days previous to the 10th. The log of the Kwang Tung 

 for the 7th proves that there was no cyclonic vortex in existence in the 

 Andaman Sea on that day. The various observations of the 8th and 9th 

 indicate that cyclonic motion on a considerable scale had not commenced 

 on either of these days. The observations of the 10th, on the other 

 hand, establish the existence of a small depression on that day which 

 rapidly developed into a large atmospheric whirl. Hence the existence 

 of the cyclonic vortex dates from the evening of the 9th, or morning of 

 the 10th. The circulation intensified and developed rapidly on the 

 evening of the 10th and morning of the 11th, so that there was a large 

 barometric depression and cyclonic circulation on that day, to the north- 

 east of the Andamans. The cyclone was then moving north-westwards. 



