80 Generating Economic Cycles 



the Douglass 7.3-year cycle in the Arizona pines. The 

 two graphs, considering the errors in the estimated 

 lengths of the cycles, are practically congruent, and 

 from 1867 to 1910 the Douglass 7.3-year cycles in tree 

 growth at Flagstaff run well with the eight-year cycles 

 in the rainfall at Prescott. 



The close relation between meteorological cycles and 

 crop cycles is infonningly illustrated in the crops 

 of the spring-wheat area in the United States. In 

 1916, among the states of the United States, North 

 Dakota ranked first in the production of spring wheat, 

 second in the production of barley, and seventh in the 

 production of oats; and to these three crops, in 1918, 

 seventy per cent of its crop area was devoted. The 

 semi-arid region of the United States begins to the west 

 of the one-hundredth meridian, and this meridian 

 divides nearly in two the state of North Dakota. The 

 average precipitation does not greatly exceed the 

 minimum essential for vegetation, and, consequently, 

 we find a close correlation between the crop yield and 

 the rainfall of the critical months, which are May and 

 June. The correlation coefficients ^ between the rain- 

 fall of May and June and the yield of the crops are, for 

 wheat, r =.66; for oats, r =.79; and for barley, r =.73. 



With this close relation between rainfall and crop 

 yields established for this very important, semi-arid 

 agricultural region, one is again led to inquire whether 

 the rainfall itself is not subject to law.^ The rainfall 



^ "Forecasting the Crops of the Dakotas," Political Science Quarterly, 

 June, 1920, p. 219. 



^ This question I have treated in a paper on "Forecasting the Crops 

 of the Dakotas," Political Science Quarterly, June, 1920. 



