41 



spring, what the results would be in the fall, because we found it to 

 hold almost invariably true that the ten percent best lot of corn in 

 the community would produce from ten to twelve bushels per acre more 

 than the average corn in that community. 



Another line of work which was carried on was what was known as 

 the imported seed test. In Iowa for the last nineteen years there has been 

 held an annual corn show. It is one of the big events in the year. The 

 state is divided into several districts and for seven years of the time that 

 these county tests were going on we would get the prize winning samples 

 at the State show. That is, we would get corn from the men who had the 

 prize winning samples and plant those side by side with this corn grown 

 in the community, and we found this significant condition, that in an average 

 year the home grown seed, not the best but the average home grown seed, 

 would produce from two to two and a half bushels per acre more than the 

 supposedly best seed of the state, and among the home grown corn there 

 would be that ten percent which would yield about twelve and a half to 

 thirteen bushels per acre more than this supposedly best corn of the state. 

 Now, we attributed that in those days to the so-called fact that it does 

 not pay to take corn away from the community in which it has been devel- 

 oped, and we thought that that was the principal reason why the corn 

 which was winning the prizes at the corn show was a lower yielding corn 



Two of The Low Yielding 1 Samples 



No. 99 was by far the lowest yielding sample in 1919 and 1920. It was not 

 in the 1921 test, but could not have come out of last place for the three year 

 average, even though it had gone to first place in 1921. No, 35 and No. 99 are 

 both typical of many low yielding lots. 



