26 



GEORGE HARRISON SHULL 



grain-row classes was attributed in part to the fact that the 1907 season had 

 seemed less favorable in general than 1906. 



It was also noted, as a possible contributory condition, that this was the 

 third season in which this corn was grown on the same area north of the 

 laboratory building, and that ''the yield may have been lessened by the 

 gradual accumulation of injurious substances in the soil." The fact that the 



Fig. 2.2 — Young corn cultures growing in East Garden of the Station for Experimental 

 Evolution in 1911, illustrating that no two were alike despite their descent from a single ear 

 of 1904 by meticulously controlled pollinations that precluded the introduction of pollen 



from any other strain of corn. 



average grain-row numbers were not significantly different in the two years — 

 15.8 in 1906, 16.0 in 1907 — in fact a trifle higher in what was thought to have 

 been the poorer year, does not seem to support these suggested explanations 

 of the observed differences of distribution in the two years. 



My contemporaneous notes proposed an additional explanation, namely, 

 that "each successive generation of close inbreeding still further reduces the 

 strains to their simple constituent biotypes, and as these are weaker than 

 hybrid combinations, this too would tend to lessen the vigor, and this 

 lessened vigor might readily be evidenced by a decrease in the average num- 

 ber of [grain-] rows and the total number of ears in the crop." 



If we accept this latter suggestion as valid, it is clear that the occurrence 



