STUDIES ON OAT BREEDING. 37 



SUMAIARY. 



The chief points brought out in this paper may be summarized 

 as follows : 



1. The present study attempts to analyze the results of 

 three successive years of selection within pure lines of oats. 

 Twenty-eight pure lines representing 13 varieties have been 

 used in this work. In the four years 621 garden rows have been 

 grown involving over 12,500 plants. The characters studied 

 were weight of grain, weight of plant and of straw, height of 

 plant and the number of culms. Only two characters, yield of 

 grain and height of plant, are analyzed in detail in this paper. 

 The remaining characters show essentially the same things. 



2. It is pointed out that the oat flower is practically always 

 self-fertilized. It is shown that if this is true every oat plant 

 must be regarded as homozygous for all of its characters. Con- 

 sequently the oat plant fulfills all the requirements in the original 

 definition of a pure line. 



3. The characters studied are subject to rather wide fluctua- 

 tions due to environment. For this reason it is not possible to 

 compare the absolute values of these characters froin year to 

 year. Instead the deviations of the plants and rows from the 

 mean of their pure line in the given year have been used. 



4. As a first approach to the problem in hand we have de- 

 termined the number of rows, grown from plus selections, 

 which deviated in the plus direction and likewise the number 

 deviating in the minus direction. The same thing has been done 

 for rows grown from minus selected plants. Thus for each 

 selection there are four classes of rows. (Cf. Tables 3, 4 and 

 5). Also the sum of the deviations of the rows in each class 

 has been determined and likewise the average deviation in each 

 class. 



5. From these data for the yield of grain, it is pointed out 

 that usually in the year next following a given selection there is 

 an excess deviation in the direction of the selection. This appar- 



