70 Nebraska Agricultural Exp. Station, Research Bui. 11 
increased yield for the large or heavy seed. Two per cent could 
easily be interpreted as experimental error if the tests were not 
so consistent in their indications. It is not likely that the yields 
are absolutely accurate. The conclusion would seem justified 
that the practical use of the fanning mill in seed preparation is to 
remove weed seeds and trash, when present. If the seed is well 
cleaned at the threshing machine, little further is to be gained 
by fanning mill grading. 
(e) When large and small seeds are alternated in the row at 
the normal planting rate and grown thus in competition, plants 
from the small seeds are reduced in relative yield as a result of 
the competition. As an average for the two tests bearing on thia 
point, this competition in favor of the large seed amounts to 13 
per cent. This suggests a natural elimination (within a mass 
variety) of poorly adapted types, which produce unduly small or 
light-weight seed. 
