CEREAL EXPERIMENTS AT THE WILLISTON SUBSTATION. 27 



MIDSEASON VARIETIES. 



Seven of the eleven varieties under discussion may be classed as 

 medium in maturity. The average date of ripening of these varieties 

 for the seven years ranges from August 2 to 5, 101 to 103 days being 

 required from seeding to maturity. Six of these varieties have 

 yielded more than any of the early or late oats, while the seventh, 

 Swedish Select, has yielded less than the two stocks of White Russian. 

 All are open-panicled varieties and all have white hulls except Prob- 

 steier, which is yellowish white. 



The farmer in the locality of Williston who grows any of these 

 varieties, with the possible exception of Swedish Select, will make no 

 mistake. The average yields here reported show that there is little 

 choice between them, though Abundance, Lincoln, Siberian, and 

 Silvermine have slightly exceeded the others in average yield, due in 

 part to their extremely high yields in 1914. The weight per bushel 

 of the Silvermine is higher than that of any of the other varieties in 

 the test. 



LATE VARIETIES. 



The White Russian and Tartarian are very similar, if not identical, 

 varieties of late oats. They mature about August 6 to 9 at Williston, 

 requiring from four to six days more from seeding to maturity than 

 the midseason varieties just discussed. The heads are long, com- 

 pact, and turned to one side (side, or horse-mane, oats). The kernels 

 are white, long, and slender. In favorable seasons they usually 

 yield well, but in ordinary or particularly unfavorable years they 

 usually fall far below the earlier varieties. 



The 7-year average yield of the best White Russian (C. I. No. 732) 

 is 7\ bushels lower than that of Abundance, the best open-panicled 

 variety. Both stocks of White Russian have yielded more than the 

 Tartarian, though the difference is largely due to the low yield of the 

 Latter in 1914. The yield of the Tartarian was low in 1914 because of 

 poor germination and the poor stand which resulted. 



K cepl where large acreages of oats are grown, so that it is desirable 

 to have the maturity of the crop extend over a considerable period 

 in prevenl losses in harvesting, the growing of side oats in this dis- 

 trict is not recommended. 



BATE-OF-SEEDING HOST WITH OATS. 



A rate-of-seeding lest with oats has been conducted for four years 

 (1911 to 1914) at Williston.. The Swedish Select variety lias been 

 used in this test, in which the rates of seeding were 2, 4, 6, 8, and 10 

 peck-.. The LO-peci rate of seeding was discontinued in 1911. The 

 re ul1 of this experiment are shown in Table XIV. 



