STUDIES ON OAT BREEDING. 1 5 



From this table it is seen, for example, that the first row 

 (No. 163) resulted from a plus selection in 191 1 which was 

 II. 6 grams above the mean of that year, and from a plus selec- 

 tion of 15.66 gms. in 19 12. The average yield of this row 

 itself, however, showed a minus deviation of 1.21 gms. Thus 

 this row deviated in the opposite direction from that for which 

 it had been selected. The next row No. 162 with practically the 

 same kind of selection gave a plus deviation of 2.66 grams. 



The first question to arise is whether plus selections have, 

 on the average, resulted in rows with a greater plus deviation 

 and likewise whether the minus selections have given an excess 

 of negative deviations. In order to make clear the method of 

 procedure we will first examine the results of the single small 

 line given in Table 2. 



In this table is will first be noted that the total plus deviation 

 of the selected plants (1911 and 1912) is in each year consider- 

 ably in excess of the total minus deviation. As noted above 

 this arises from the skew distribution of the individual plants 

 about their means. The range extends much farther in the plus 

 direction than in the minus and hence allows the selection of 

 larger plus deviations. 



From the sums of the columns in the second part of the 

 table (1913 rows) the effect of either the 1911 or the 1912 

 selection can be determined and likewise the combined effect 

 of two plus or two minus selections. Thus the 1913 rows 

 resulting from plus selections in 191 1 show a total plus devia- 

 tion of 3.04 gms. and a total minus deviation of 2.43 grams. 

 The 191 1 minus selections show zero plus deviations and 0.60 

 minus deviations. Thus the deviations of the 1913 rows show 

 a tendency to follow the line of the 191 1 selections. The 1912 

 plus selections give a total plus deviation of 2.66 gms. and a 

 total minus deviation of 1.42 gms. The 19 12 minus selec- 

 tions give 0.38 gms. plus and 1.61 gms. minus, again show- 

 ing a marked tendency to follow the direction of the selection 

 in 1912. We further note that where the selection has been 

 plus in both years there is an excess of plus deviations and 

 with two minus selections an excess of minus deviations. 



Of course the number of rows in this one line is too small 

 to have any significance in itself but it serves to show the 

 method used. It has been pointed out above that by using the 



