120 H. F. Blanford— Rainfall of the Camatic. [April, 



eleven years, and computing out the hypothetical rainfall of each year 

 of the cycle, on this assumption, I find that the periodical variation so 

 determined is at the time of the minimum 6*7 inches below the average 

 of 34'3 inches, and at the time of maximum 7'3 inches above it, the 

 total mean oscillation being 14 inches, or more than two-fifths of the whole. 

 The recorded rainfall of any single year deviates on an average from 

 the hypothetical value by only 3'6 inches. The result is far greater and 

 more striking than I, or I believe any one else, would have ventured to 

 anticipate. 



But now comes a curious point. No other province of Southern 

 India, as far as I have yet ascertained, exhibits anything approaching 

 to such regularity. The rainfall of Mysore and Bellary, a region more 

 subject to famine than Madras itself, shews scarcely an appreciable 

 oscillation coinciding with the sun-spot cycle, except that it exhibits a 

 very distinct minimum in 1876. The province of Malabar has indeed 

 two very distinct minima in 1866-67 and 1875-76, but there was another 

 greater than either in 1881 and the maxima were very irregular, occur* 

 ring in 1871, 1874, 1878 and 1882, and the Deccan and Hyderabad are 

 not more concordant. At present therefore only the Carnatic can be 

 said to have its rainfall varying in an eleven-year cycle, but in this case 

 the coincidence is very marked indeed. 



On the question of the periodical recurrence of famines, I have 

 nothing to add to the facts detailed in the Report of the Famine 

 Commissioners. 



The conclusions drawn by Mr. Hill that the winter rainfall and 

 summer rainfall of the N.-W. Provinces both shew a periodical varia- 

 tion, but of a different character in the two cases, and that the summer 

 rainfall tends to vary in the opposite direction to the winter rainfall 

 have also been critically examined. 



Taking the rainfall of the N. W. Himalayan stations alone, from 

 November to May inclusive, and tabulating it year by year for about 26 

 stations, I find that it shows amid some irregularities a very much more 

 distinct cyclical variation than that obtained by Mr, Hill from the 

 registers of twenty stations, half on the hills and half on the plains. 

 It seems to have been at a maximum in the winter and spring of 

 1864-1865, to have decreased with some irregular variations, but on the 

 whole pretty steadily, till 1872-73, to have again increased in like manner 

 till 1877-78, after which it became more irregular, dropping to a 

 minimum in 1878-79, then increasing again till 1880-81 and decreasing 

 to 1883-84. Notwithstanding these irregularities, when the probable 

 oscillation in the eleven-year cycle is computed out it is found to be not 

 less striking than that of the Carnatic rainfall. It amounts when at its 



