4o8 



Journal of Agricultural Research 



Vol. XIX, No. 9 



two years, however, is less than 0.5 mm., so that the data coincide far 

 beyond any reasonable expectation. 



The reason that errors in this connection are suggested is that it is not 

 plausible that the differences in soil or water would affect the kernels by 





/4? 



Fig. 4. — Graph showing lateral and dorsoventral diameters of the ovary tip as compared with length, 

 lateral diameter, and dorseventral diameter of the kernel for the 26 days following flowering in 1916. 



this time, since there obviously is no insufficiency of nutrients for this 

 primary growth. These data have an incidental bearing on the value of 

 observations on the length and diameter of kernels. Such observations 

 are used frequently to identify varieties. In 1914, the author (2) stated 



