40 ■ MAINE AGRICULTURAL EXPERIMENT STATION. I915. 



13. On the whole the results obtained in this study give no 

 evidence that selection for three years lias modified any of the 

 characters studied. The one or two apparent exceptions dis- 

 cussed in the paper might very easily arise in chance distribu- 

 tions where so small a number of years are considered. The 

 weight of the evidence against an effect of selection far out- 

 weighs the evidence for such an effect. It must be concluded 

 that in the present material and for the characters studied, selec- 

 tion for threv years has produced no effect zuhich can be detected 

 by the methods used. 



14. It is proposed to continue a portion of these selections 

 and later to attempt an individual analysis rather than the statis- 

 tical one presented in this paper. 



