224 On the Cyclones of the Bay of Bengal. 



hypothesis for the explanation of the cyclones of the southern 

 Indian Ocean, is a very remarkable one. It may also be added, 

 in further confirmation of it, that it affords some explanation 

 of the connexion between cyclone-frequency, rainfall, and sun- 

 spot-frequency. Mr. Meldrum's investigations have estab- 

 lished that the frequency and energy of cyclones in the 

 Indian Ocean and the rainfall of the globe vary directly with 

 the number of sun-spots, or are subject to the same law of 

 periodicity. This relation is almost self-evident, if it be 

 admitted that massive terrestrial atmospheric phenomena are 

 due solely or primarily to the action of solar radiant heat, 

 and that any marked periodicity in the cause (solar heat and 

 action) must impress a corresponding periodicity on the 



more important phenomena of the earth's atmosphere 



If it be granted that rainfall varies with change of the sun's 

 photosphere, as indicated by the number of sun-spots, and 

 that cyclones are due to rainfall, it is probable that cyclone- 

 frequency will be subject to the same law of periodicity 

 which affects the distribution of rainfall, and, consequently, 

 that the occurrence of cyclones will vary with the number of 

 sun-spots. This probable mutual connexion between phe- 

 nomena apparently unconnected appears to give additional 

 weight and force to the condensation theory." 



In calling attention to severe cyclones Mr. Eliot remarks, 

 " That severe cyclones at the head of the Bay during the 

 month of May have generally occurred about the period of 

 minimum sun-spots." He also finds the same connexion 

 characteristic of the opposite transition-period, viz. October: — 

 " The three cyclones which have occurred in the month of 

 October during the present century, happened at or near 

 minimum sun-spot years." He embodies the results in two 

 general propositions as follows : — 



"First. That severe cyclones, L e. cyclonic storms ex- 

 tending over a large area and of considerable barometric 

 depression, and therefore attended by a destructive storm- 

 wave, occur at or near the minimum sun-spot period, and 

 that the year of minimum solar spots, or the preceding year, 

 appears to be the most probable. 



" Second. That, during the months of November and De- 

 cember, only ten severe cyclones have occurred during the 

 present century ; or, taking every eleventh year as defining 

 the minimum solar-spot period, they occurred as follows : — 

 1822, three storms, viz. in 1822, 23, and 24 ; 1833, three 

 storms, viz. in 1831, 32, and 33 ; 1844, one storm in 1842 ; 

 1855, one storm in 1852 ; 1866, one storm in 1864 ; 1877, 

 one storm in 1876." " Hence," Mr. Eliot says, "there is 



