1887.] H. F. Blanford— Rainfall of the Garnatic. 117 



best known is Dr. now Sir, W. W. Hunter who in a pamphlet written 

 in his well known brilliant style of exposition, set forth these now 

 familar facts of the rainfall, basing them, however, on the much more ex- 

 tensive data of sixty-four years' registers, together with the further im- 

 portant observation, which I believe originated with Mr. Pogson, viz. 

 that the famines of Southern India recurred at intervals of about 

 11 years, coinciding with the recurrence of the sun-spot minimum. It 

 fell to me, as an official duty, to review the evidence adduced in Dr. 

 Hunter's pamphlet and other writing, on the subject, the economic bear- 

 ings of which were obviously of the highest importance, and I had to 

 point out that while the rainfall registers of the Madras Observatory, 

 undeniably showed a certain amount of fluctuation of the kind pointed 

 out by Mr. Norman Lockyer and Dr. W. W. Hunter, any other stations 

 in Southern India, the rainfall registers of which were among the oldest 

 extant, to wit Bangalore and Mysore and other stations in the peninsula 

 and other parts of India, viz., Bombay, Calcutta, Nagpur and Jubbul- 

 pore, with a partial and doubtful exception in the case of Nagpur, failed 

 to shew any trace of such a cyclical variation. Bnt it was notorious 

 that the famines the periodicity of which was considered corroborative 

 of the rainfall oscillation had sometimes affected the Carnatic, sometimes 

 Mysore and Bellary, sometimes Hyderabad and the Deccan, and sometimes 

 the Northern Circars and Orissa ; so that the two classes of facts must be 

 regarded as to a great extent independent of each other. 



In a lecture delivered on the 18th May 1877 at the Royal Institu- 

 tion in Albemarle Street, General R. Strachey also pointed out the in- 

 sufficiency of the evidence adduced to bear out the conclusions drawn 

 from them. His criticism is so much to the point, and the objection 

 raised, notwithstanding its obvious validity, is so frequently disregarded 

 by persons who enter on enquiries of this kind, that it is worth quoting 

 in its original form of expression. Speaking of Dr. Hunter's results he 

 gays, " He has inferred from what must be held to be altogether in- 

 sufficient numerical data that sure indications of periodicity exist. He 

 arrives by an arithmetical process at certain figures, which he regards 

 as the probable mean amount of rainfall in the successive years of the 

 11 year's cycle, and finding a maximum and minimum among them, he 

 infers that this is a proof of a true periodical variation. But such a 

 result alone proves nothing. To test its value, it is necessary to com- 

 pare the calculated quantities of rain for the several years with the 

 quantities actually observed, and then to consider whether the differ- 

 ences are of a character to justify the belief that the calculated quantities 

 afford a reliable approximation to the truth and what sort of approxi- 

 mation * * * The only conclusion that seems possible from such 



