[Vol. 10 



146 



ANNALS OF THE MISSOURI BOTANICAL GARDEN 



pod and stem blight of soybean are correlated with the rainfall 

 of the growing and ripening season. The curves of fig. 8, founded 

 on data collected at the Raleigh station of the United States 

 Weather Bureau, represent normal and total precipitation at 

 Raleigh during the months of June, July, August, and September, 

 of the 3 above-mentioned years. The curve for 1920 shows 



Fig. 9. 



that a normal amount of rain fell during June and July 

 amount decreased markedly during August and appr 



The 

 ?hed 



normality again in September. In 1920 the disease was first 

 observed about August 1, and it spread rapidly from diseased to 

 adjacent healthy plants during the greater part of this month, 

 but with the continuance of dry weather in September a marked 

 falling-off of new infections was observed. The curve for 1£ 

 lies, for the greater part of its course, far below that representing 

 normal rainfall. Not until September, the month having a low 

 normal precipitation, do the two curves approach each other. 

 In this season very few diseased plants were found and on these 

 the disease usually involved the stem only near the ground level. 

 In 1922, the amount of rainfall decreased from near normal in 

 June to subnormal in July. It arose above normal in August 

 but sank to practically nothing in September. During this 

 season the disease appeared later and was notably less severe 

 than in 1920, but very much more prevalent than in 1921. 

 Figure 9 is intended to illustrate the estimated relative preval- 

 ence of this disease during these 3 summers. The monthly 

 mean temperatures for the 3 seasons lie very close together. 

 Apparently this factor cannot account for the prevalence of the 

 disease noted above. 



