l62 



Journal of Agricultural Research voi. xviii, no. 



An inspection of Table I shows that there are a number of cases in which 

 there is apparent coincidence between the percentage of yellow-berry 



produced in one season and 

 the percentage produced by 

 the same strain the follow- 

 ing season. However, a 

 correlation table between 

 yellow-berry percentages 

 for the two successive sea- 

 sons plotted for 56 strains 



1 that were planted and har- 



2 vested on the same date 

 I gives a correlation coeffi- 



3 cient of only 0.078 ±0.005. 

 3 This extremely low corre- 

 l lation indicates that the 

 J external conditions are the 

 I determining factors to a 



1 degree which the heredi- 

 !; tary tendencies of the plant 

 ^ have little power to modify. 



2 On the other hand, a graph 



1 showing the percentage of 

 >. yellow-berry in 1907 and 



3 1908 for those strains hav- 



2 ing 10 per cent or more of 

 ^ yellow-berry indicates that 

 ^ there is a hereditary rela- 

 u tion; and were the number 

 >■ of cases larger, distinct in- 



» dications of inheritance 



3 



3 would be seen. 

 I The relation between 

 1; date of harvesting and the 

 5 percentage of yellow-berry 



is indicated in figure 2. It 

 is evident that from June 

 25 to July I there was an 

 increase in the percentage 

 of yellow-berry. The total 

 number of rows harvested 

 July 3 and 10 are not suffi- 

 cient to permit definite 

 conclusions with respect to these dates. It appears, however, that in the 

 season of 1908 there was a close relation between the percentages of 



! S S S ? 



