CEREAL IMPROVEMENT SELECTION 187 



B. NUMBER OF KERNELS IN EQUAL VOLUMES OF DIFFERENT 

 GRADES OF WHEAT 



Three things are shown by this calculation: (1) the 

 difference between the number of kernels of any grade 

 to the unit of weight and unit of volume, (2) the very 

 great difference in number of kernels of the different 

 grades in the same weight or volume in the same variety, 

 and (3) the varietal variation in number of kernels of any 

 grade to the unit of weight or volume. It is evident 

 from these facts that much variation in results must 

 follow the planting of the same number of kernels to the 

 unit area, even from the same grades of seed, if these 

 grades are made indiscriminately on the basis of size or 

 weight, and that the influence of the variety is very great. 

 Differences in rate, date, and space of planting and cli- 

 matic differences may also so affect the experiments as 

 to be a partial cause of some conflicting results that have 

 been obtained in widely different localities. 1 



186. Mass selection. Aside from the selection of 

 seed in bulk through mechanical grading, for a long time 

 there has existed a practice of selecting the best spikes 

 or the best plants from a plat or field, and sowing the seed 



1 In Proc. Amer. Soc. Agron. II, 59-69, 1910, Montgomery has 

 reviewed the methods employed in various series of tests of the 

 value of heavy and light kernels, and shown the widely different 

 conditions under which the many tests were conducted. 



