STUDIES ON OAT BREEDING. II 



From this table it is to be noted that there is a great differ- 

 ence in the average yield per plant for the same pure line in 

 different years. These dift"erences are due in all probability 

 to differences in external conditions. The highest average 

 yield for all the lines was obtained in 191 1. In this year the 

 conditions for the growth of garden plants was very favorable. 

 The lowest yield was obtained in 19 13. This was not a good oat 

 year in many respects. However, two things contributed to make 

 this year's yield particularly low. In the first place it has already 

 been noted that the garden rows were planted only a foot 

 apart in 191 3, while in the two previous years the average 

 distance between rows had been one and one-half feet. In the 

 second place these garden rows were grown on the same piece 

 of ground for the three years, 191 1, 1912 and 1913. Whether 

 soil toxins produced by the same crop in previous years oper- 

 ated to decrease the yield in 1912 and 1913 we are not able to 

 say. At any rate this point s'hould be mentioned as a possible 

 factor. It should further be pointed out that the field plots 

 •during these years showed an increase in the average yield 

 for 1913 over the two preceding years.'" 



Only a part of the decrease in the 19 13 garden yield can be 

 accounted for on the basis of the different spacing. In 19 14 

 the rows were again planted one foot apart but the garden was 

 moved to another plot of ground. The average yields in 1914 

 compare very favorably with those of 191 1 and 1912. 1914 

 was an exceptionally good oat year and no doubt the average 

 yield is somewhat better than can be expected for a series of 

 years. However, this shows that conditions other than spacing 

 may greatly influence the yield of garden plants. It is probable 

 that the results as a whole would have been but little more 

 uniform had the spacing been the same in each year. 



The num.ber of minus and plus selections has been about the 

 same in each pure line each year. It is true that owing to the 

 skew distributions of the plants about their means the plus 

 selections averaged to deviate somewhat farther from the 

 means than did the minus selections. However, in view of 

 the results obtained in the latter portion of this paper it is not 

 probable that these differences have had any influence upon the 

 mean vields in the later vears. 



'^Surface and Barber. Loc. cit. 



