Origin of Eight-Year Generating Cycle 83 



for a period of 41 years, from 1870 to 1910. The corre- 

 lation between the rainfall in the Ohio Valley and the 

 rainfall of Illinois was r = .60, and the eight-year cycle 

 in the Ohio Valley rainfall was repeated in an eight-year 

 cycle in the rainfall of Illinois. The graph is likewise 

 given in Figure 23. 



In October, 1920, H. W. Clough of the United States 

 Weather Bureau published in the Monthly Weather 

 Review an article on ''An Approximate Seven- Year 

 Period in Terrestrial Weather, with Solar Correlation." 

 From this paper I make use of two facts: the first re- 

 fers to atmospheric pressure at Toronto, in Canada, 

 and at St. Louis, in the United States. In Mr. Clough 's 

 paper a graph is given of atmospheric pressure at To- 

 ronto from about 1843 to 1911, and a similar graph is 

 given for the pressure at St. Louis from about 1874 to 

 about 1917. An inspection of these shows (1) that for 

 the interval covered by the two graphs, they are approx- 

 imately congruent ; (2) that the maxima occur, approxi- 

 mately at 1850, 1858, 1866, 1874, 1882, 1890, 1898, 

 1914, and are therefore synchronous with the eco- 

 nomic and meteorological cycles already examined. 



The other fact that I take from Mr. Clough 's paper 

 refers to atmospheric pressure in Europe. "Mauer 

 {Archives des Sci. Phys. Nat., Geneva, May, 1918) 

 called attention to a well-marked periodicity of seven 

 to eight years in the winter pressure in Central Europe, 

 with maxima as follows: 1818-19, 1827-28, 1835-36, 

 1843-44, 1850-51, 1857-58, 1865-66, 1873-74, 1881-82, 

 1889-90, 1897-98, 1904-05, 1912-13." 



If now the dates of the maxima in Mauer's European 



