STUDIES ON OAT • BREEDING. I9 



There is then apparently, a decided effect of the selection made 

 in 1911 upon the plants grown in 1912. As shown in table 3 

 the apparent effect of the 191 1 selection upon the rows of 1913 

 and 1914 is very much less. The 1914 rows show an excess 

 total deviation in the opposite direction to that of the selection. 

 The average deviation per row shows some differences in the 

 different years. Thus in 1912 the average deviation per row 

 was noticeably greater in the direction of the selection. In both 

 1913 and 1914 (Table 3) the plus rows resulting from the plus 

 selections show a larger average deviation. However, in the. 

 case of the minvis selections the minus deviating rows showed 

 a smaller average deviation than the plus rows. 



Turning to table 4 it is seen that in the 1913 rows there is 

 again a marked effect of the selection. However in the 1914 

 rows there is an excess of the total deviation in the opposite 

 direction to that of the selection in 1912. Here there is, in 

 fact, a decrease in the yield of the plus selections over that of 

 the minus selections. 



In table 5 it is seen that there is an excess of the deviation, 

 in the direction of the selection. 



Considering all three tables it is to be noted that in the yeai^ 

 immediately following the selection there is a more or less 

 pronounced effect of the selection as indicated by the total de- 

 viations. In each case the plus selections have resulted in a 

 larger total plus deviation of the daughter rows and the minus 

 selections in a larger total minus deviation. The effect of the 

 selection upon the rows grown two and three generations after 

 the selection is very much less marked. In two of the three 

 cases the total deviations show an excess in the opposite direc- 

 tion to that of the selection. These rows have of course been 

 subjected to secondary and tertiary plus and minus selections 

 and it is possible that these latter have outweighed the original 

 selection. 



It is also seen that in the year following the selection the 

 average deviation is usually larger in the direction of the selec- 

 tion. The one exception to this, is in table 4 in the minus 

 selections grown in 1913. As measured by the average devia- 

 tion, there does not appear to be a very marked effect of the 

 selection. 



