254 



Journal of Agricultural Research 



Vol. XXII, No. s 



season. It seemed possible, therefore, that by making successive plant- 

 ings a period might be found for this latitude when the mean soil tem- 

 perature is sufficiently high to materially check or completely inhibit 

 onion smut infection. 



In order to test this out, a series of plantings at intervals of from 8 to 

 14 days was begun on June 18, 1920, at Madison, Wis. Onion seed 

 was sown in smut-free soil in an open trench about i inch deep and was 

 then covered with naturally infested soil similar to that used in certain 

 of the laboratory experiments. Two varieties. Red Globe and Yellow 

 Bermuda, were used, one lo-foot row of each variety being put in at 



S/ 5 



Fig. 2. — Graph showing the daily mean soil temperature at a depth of i to 2 indies as it occurred in the 

 "successive planting" plots. Since the weather continued rather cool, one bed was covered with glass 

 to insure a higher temperature. The temperature of the uncovered bed is shown by the solid line, 

 the temperature of the glass-covered bed by the broken line. For further details see Table XII and 

 and accompanying text. 



each planting. Temperatures of the soil at a depth of i to 2 inches were 

 obtained by means of a self-recording thermograph. The hourly mean 

 temperature for each day was then secured by adding temperatures as 

 recorded for each hour and dividing the sum by 24. These computa- 

 tions are represented graphically in figure 2. Since the weather in 

 July was unusually cool, a higher mean soil temperature was secured for 

 some of the plots by covering them with an ordinary glass cold frame. 

 Inasmuch as the dry weather and high temperature would cause a rapid 

 desiccation of the surface layer of soil, the plots were watered thoroughly 

 on alternate days or oftener during the early growth of the plants. The 



