Origin of Eight-Year Generating Cycle 87 



variable solar output; indeed, these barometric vari- 

 ations do closely follow the variations in the sun- 

 spot frequency and the other products of the sun's 

 variable activity. From 1873 to 1890 there is a 

 remarkable agreement between these two classes 

 of phenomena, hut if the pressure period is 8 years 

 and the sunspot period 1 1 years that connection cannot 

 he complete in its nature. It is evident that we shall 

 be obliged to prosecute this study yet further by 

 constructing similar barometric tables for other 

 countries, in each hemisphere, and also that it will 

 be well to include in our research some account of the 

 temperature, the vapor tension, and the magnetic 

 and electric fields of the atmosphere before attempting 

 any further remarks on the subject. . . . 



" Finally, it is evident that for the United States 

 we can now correlate the years which have similar 

 secular variations, as 1878, 1885, and 1893, and 

 study them clmiatologically to see if there are 

 any marked and prevailing features which charac- 

 terize them. It is clear that this discussion opens 

 to us for the first time the prospect of a scientific hasis 

 for seasonal forecasts, certainly so if it should prove 

 to be the case that the maximum-pressure years 

 differ distinctly from the minimum-pressure years 

 in their seasonal character, for in the event of our 

 happening on a maximimi year we could then pre- 

 dict the feature due to a falling barometer, or, if 

 we are in a minimmn year, to a rising barometer, at 

 least for one or two years ahead." ^ 



1 Report of the Chief of the WeaUier Bureau, 1900, 1901, vol. ii, pp. 

 1004, 1005. 



