140 Economic Cycles: Their Law and Cause 



In regard to the use of statistical generalizations in 

 meteorology, we have the cautious opinion of Lord 

 Kelvin: "I cannot say whether anything with reference 

 to Terrestrial Meteorology is done once for all. I 

 think probably the work will never be done." There 

 is always need of checking up statistical conclusions in 

 the light of new data, and this necessity applies to the 

 generalization that in the Mississippi Valley the annual 

 rainfall passes through a double cycle of thirty-three 

 years and eight years. This conclusion is undoubtedly 

 warranted by the data that lie at the basis of the in- 

 vestigation, but it would be a grave fault, indeed, to 

 hold that the cycles do not alter with the flow of time. 

 Whether they change or retain their characteristics can 

 be determined only by accumulating more data than 

 are at present available. 



We come now to the second part of our general 

 problem, namely, to the consideration of the relation 

 between rainfall and the yield of the crops, and again 

 the questions of data and method must be settled. 

 In choosing the data, the prime consideration was to 

 make sure that the crops selected should be representa- 

 tive of the conditions of crop-producing in the Middle 

 West. The five principal crops in the Middle West are 

 corn, hay, wheat, oats, and potatoes, and of these five 

 all except wheat were taken to serve as representative 

 crops. Wheat was omitted because of technical dif- 

 ficulties: First, it is impossible, except for recent years, 

 to separate in the published statistics the yield per acre 



