Origin of Eight-Year Generating Cycle 81 



records of North Dakota and South Dakota, for May 

 and June, from 1882 to 1918 were searched for evidence 

 of periodicity between the hmits of three and twelve 

 years, and the result indicated that, if there is a real 

 cycle of rainfall for May and June in these two repre- 

 sentative states its most, probable value is in the neigh- 

 borhood of eight years. When an eight-year cycle is 

 fitted to the rainfall data, the maxima occur, approxi- 

 mately, at 1882, 1890, 1898, 1906, 1914, and are prac- 

 tically synchronous with the dates of the maxima of 

 rainfall of Prescott, Arizona, the growth of pines in 

 Arizona, and the maxmia of the international crop 

 cycles in the United States, in the United Kingdom, 

 and in France, The graph of the Dakota rainfall is 

 given in Figure 23. 



Starting from the hypothesis that American economic 

 activity is largely dependent in its ebb and flow upon 

 the prosperity of the agricultural Middle West, I sought, 

 in an early study ^ of meteorological cycles, to discover 

 a periodicity in the rainfall of the Mississippi Valley. 

 As the meteorological records were much longer for the 

 Ohio Valley than for the states further west, the rainfall 

 of the Ohio Valley was made the basis of the first ex- 

 amination. The method employed was a test of the 

 possible presence of cycles between the limits of three 

 and thirty-three years, and the evidence was unmistak- 

 able. If there is a real cycle of rainfall in the Ohio 

 Valley its most probable length is about eight years 

 and its maxima are approximately synchronous with 

 the maxima of the May and June rainfall of the Dakotas, 

 the annual rainfall of Prescott, Arizona, the growth 



1 Economic Cycles: Their Law and Cause, 1914. 



