YIELD OF SPRING WHEAT 5 



years, the general trend of results, and exceptional or unusual behav- 

 ior in certain years. Detailed data for each year at each station are 

 given in the appendix. 



Assinniboine 



The northern Montana branch station, herein called the Assinni- 

 boine station in keeping with previous publications, is located near 

 Havre, Mont. The precipitation is the lowest of any of the stations 

 studied, and the soil has a medium to high water-holding capacity. 

 Data are available for 22 years during the period 1916-38. Data for 

 1934 are not used, because the crop was seriously damaged by hail. 

 In 1916 the plots had uniform preparation, all being on land broken 

 from prairie sod in June 1915. All plots were wet to a depth of 2 

 feet at seeding time, and the yields were abnormally high, largely as a 

 result of precipitation nearly double the normal in May, June, and 

 July. If this year were excluded, the showing from 2 feet of wet soil 

 would be much less favorable than it is. The only really good yield 

 obtained from seeding in soil that was not wet to a depth of more than 

 1 foot was on plot A, spring-plowed, in 1932. This yield ''was at- 

 tributed to favorable quantities and distribution of rainfall in April 

 and May, unusually heavy rainfall (4.50 inches) in June, and to free- 

 dom from diseases and insects" (2, p. 7). The other two plots were 

 each wet to a depth of 2 feet. The yields on them differed by only 



1 bushel per acre but averaged 5 bushels more than the plot that 

 had but 1 foot of moist soil in the spring. 



On both spring-plowed and fall-plowed land continuously cropped 

 to wheat the average yield and the expectancy of a fair crop were 

 materially higher when the soil was wet to a depth of 2 feet than when 

 the depth of moist soil was 1 foot or less. The fall-plowed plot was 

 wet to a depth of 2 feet much less frequently than the spring-plowed 

 plot. The spring-plowed plot was wet to a depth of 3 feet or more 

 only three times in its history and the fall-plowed plot only twice. 

 A good yield was produced from this condition only in 1927 on both 

 plots. 



There was only 1 year when the fallowed plot did not have more 

 than 1 foot of wet soil in the spring. In that year the condition 

 was common to all methods, and the yields were low. In 8 years 

 the soil was wet to a depth of 2 feet, and in 13 years it was wet to 3 

 feet or more. 



The data on the depth to which the soil is wet and the yields 

 show why the growth of small grain following small grain is not a 

 recommended practice at this station (1). 



WlLLISTON 



Crops were produced at Williston from 1909 to 1920, inclusive, in 

 the methods of cultivation and rotation project, but soil-moisture 

 determinations, from which the depth to which the soil was wet in 

 the spring may be estimated, are available for the 7 years 1910-16 

 only. Near failures were experienced in the 2 years when available 

 water in the continuously cropped plots was limited to the first foot, 

 and good yields in the 5 years when the soil was wet to a depth of 



2 or 3 feet. In 1910 and 1911 when available water in the contin- 

 uouslv croDped plots was limited to the first foot, the fallowed plot 



