STUDIES ON OAT BREEDING. 177 



in the remaining years the rate was two bushels by weight. It 

 is possible that this difference in seeding- accounts for the in- 

 creased yield. 



Another thing that may have influenced the yield is, that the 

 seed of all these varieties was obtained in Canada and the 

 Middle West. There is a persistent opinion among many farm- 

 ers that foreign grown seed will yield better than home grown 

 seed of the same variety. Where this question has been tested 

 by experiments little or no grounds have been found for such 

 a belief. For example the Iowa Experiment Station'^ tested 

 twenty varieties by importing seed of each one, every year for 

 three years and comparing with home grown seed. Their results 

 show that on the whole the imported seed was no better than 

 home grown. In fact the varieties usually did better after they 

 had become acclimated by one or two seasons' growth in Iowa 

 In view of these and other results we are not inclined to attach 

 any significance to the suggestion that the better yields of 19 lo 

 were due to imported seed. As a matter of fact the results of 

 the 1912 and 1913 tests contradicted this suggestion. The seven 

 varieties imported from out of the state in 1912 all yielded very 

 poorly in that year. In 1913, however, the majority of these 

 varieties grown from our own seed yielded very well (cf. tables 

 5 and 6). 



INTER- AND INTRA-SEASONAL VARIATION. 



The question of the reliability of the results of a series of 

 variety tests is one which has attracted more or less attention. 

 Recently Harris" has proposed the inter-plot correlation as a 

 measure of the reliability of such tests. However, it appears 

 to us that the variation constants (standard deviation and co- 

 efficient of variation) together with the probable error of the 

 mean give a very desirable measure of the confidence which can 

 be placed in the average of a series of such tests. These con- 

 stants together with the mean production form a very good 

 guide as to the desirability of a given variety. 



"Burnett, L. C Some Data for Oat Growers. Iowa Agr. Exp. Sta. 

 Bui. No. 128, pp. 93-127, 1912. 



"Harris, J. A. On the Significance of Variety Tests. Science, N. S., 

 36, pp. 318-320, 1912. 



