EPIDEMICS. 



65 



The temperatures in 1903 (Q) and 1905 (S) were also subnormal 

 during the critical period, but with much greater irregularity and not 

 to such an extent as in 1904. 



To recapitulate, it is seen that altliough the general average 

 of temperatures for 

 the whole area for 

 the 7-month and 3- 

 month })eriods in 

 1903, 1904, and 1905 

 were not much dif- 

 ferent, still in those 

 States where the rust 

 attack was most se- 

 vere in 1904, namely. 

 North Dakota, South 

 Dakota, ^linnesota, 

 Iowa, and Wisconsin, 

 temperatures for the 

 7-month period aver- 

 aged generally much 

 lower in 1904 than 

 in 1905, and for the 

 3-montli period aver- 

 aged about the same 

 as in 1905. During 

 the 1-month period 

 temperatures were 

 consistently subnor- 

 mal in 1904, averag- 

 ing 2.67 degrees be- 

 low normal for the 

 whole region ; tem- 

 peratures were 3^ de- 

 grees below normal 

 over South Dakota, 

 North Dakota, ^Mhi- 

 nesota, and Wiscon- 

 sin, this average be- 

 ing considerably 

 lower than that of 

 either 1903 or 1905. 

 It is seen, then, that 

 the unusuall}' low temperature over tliis region was a very 

 important factor, if not the determining factor, for the prevalence 

 of rust in 1904. Low temperatures made the crop as a whole late. 

 88550°— Bull. 216—11 5 



Fig. 2.— Temperature chart, showiug average monthly departure 

 from normal in several States in 1903, 1904, and 1905. X, Normal 

 line for a 7-month period preceding the maturity of the grain: X', 

 normal line for a 3-month period preceding the maturity of the 

 grain; X", normal line for 1 month including the heading period; 

 K, N, and Q, lines for 1903; L, O, and R, for 1904; and M, P, and 

 S, for 190.5. The lines show departure of temperature from normal 

 during the periods indicated, respectively, the distance between two 

 adjacent horizontal lines representing 1 degree F. 



