20 SOUTHERN FIELD CROPS 



be reduced by about half a bushel per acre, and the open- 

 furrow method makes possible an even greater reduction. 

 The earlier the date of sowing and the more complete the 

 preparation of the land, the smaller may be the quantity of 

 seed employed. 



24. Size of seed. — Scores of experiments have been 

 made to determine the size of seed or grain to sow for the 

 best agricultural results, most of which show a distinct 

 advantage from sowing large or hea\'y^ seed. 



Zavitz secured the following results (American Breeders' As- 

 sociation, Vol. II, p. 121) : — 



After selecting seed for thirteen j-ears, the large seed being 

 taken each year from the plot sown with large grains, the small 

 grains continuously from the plots sown with smaU seed, the crop 

 from the large seed yielded 65.5 bushels per acre as compared 

 with 44.7 bushels from the small seed ; the crop from the large 

 seed weighed .35.5 pounds per bushel as compared wiWl 24.3 

 pounds per bushel from the continuous sowing of light seed. The 

 deterioration due to sowing poor seed is still better shown by the 

 fact that the crop from the large seed required only 1149 grains 

 to weigh an ounce while that from the light seed required 2066 

 grains. 



Among the puljlications summarizing the experiments on this 

 point are the foUo^^Tug : — 



Nel)raska Expr. Station Bull. 104. 



Ohio Expr. Station Bull. 3S. 



Kansas Expr. Station Bull. 74. 



Canada Expr. Farms, Rpt. 1901. 



Ontario Agr. College and Expr. Farms. Rpt. 1903. 



25. Separation of grains by fanning. — It should be 

 borne in mind that there is a tendency for any one oat 

 plant to bear as many heavy seeds as light seeds. This is 

 because each spikelet usually matures one larsie and one 



