29a THE CEREALS IN AMERICA 



slightly decreased.^ On clay soils heavy rains are more like!); 

 to pack the soil unfavorably on rolled than upon unrolled land. 

 On such soil a light harrowing after rolling may prove benefi- 

 cial. Harrowing after oats are up has increased the yield of 

 grain in Nebraska,^ but decreased it in Kansas.^ Oats may be 

 cultivated by sowing two drill rows and missing two, but using 

 the same amount of seed per acre. Sometimes increased yields 

 have been obtained, but usually a decreased yield results from 

 more or less accidental injury to plants. It appears that on 

 gravelly soil especially liable to suffer from drouth, cultivation 

 may increase the yield, but ordinarily it cannot be considered good 

 practice. The Iowa Station found that cutting back oats when 

 showing five leaves decreased the amount of lodging, increased 

 the yield twelve per cent and somewhat delayed ripening.^ 



400. Influence of Size of Seed. — The Ohio Station obtained 

 heavy and light seed by use of a fanning mill and during seven 

 years averaged forty-six bushels with heavy seed, forty-five 

 bushels with common seed and forty-three bushels per acre with 

 light seed. No difference was found in weight per bushel.^ 

 By the same method Kansas Station during eight years obtained 

 thirty-one, thirty and twenty-eight bushels respectively.^ Minne- 

 sota Station sowed two bushels of oats weighing thirty-seven 

 pounds per bushel and two bushels weighing twenty-one pounds 

 per bushel, and obtained sixty-four and fifty-five bushels of grain 

 per acre respectively.^ 



401. Influence of Seed Selection. — In the above tests the 

 selection was by specific gravity and not by weight of individual 



1 Wis. Rpt. 1891, p. 91. 



2 Neb. Rpt. 1899, p. 15. 



8 Kan. Bui. 13 (1890), p. 62 

 < low?. Bui. 45 (1900), p. 2^0. 



5 Ohio Bui. 38, p. 48. 



6 Kan. Bui. 74, p. 199. 



' Minn. Bui. 31, p. 201. 



