BEGINNINGS OF THE HETEROSIS CONCEPT 



27 



of essentially the same average numbers of grain-rows in the two years gives 

 only a specious indication of the relative climatic and soil effectiveness in 

 these two seasons. It must mean simjjly that the diminution of grain-row 

 numbers produced by increasing homozygosity happened to be balanced by 

 the increased frequencies in the higher classes, produced by the gradual ac- 

 cumulation by selection of more potent hybrid combinations. 



TABLE 2.3 



GRAIN-ROW COUNTS AND HEIGHTS OF PLANTS IN 

 THE CULTURES OF 1907 



A truer measure of the relative favorableness of the two seasons for growth 

 and productiveness of these cultures can be derived from a study of the 

 middle classes with 12, 14, 16, and 18 grain-rows. These grain-row classes 

 making up 80 per cent of the 1906 crop and 61.5 per cent of the 1907 crop, 

 must be relatively free from most of the distortion assumed to be produced 

 either by increasing homozygosity or by the accumulation of the more po- 

 tent hybrid combinations. If we average these four grain-row classes by them- 

 selves for the two years, we find that in 1906 their average was 15.5 grain- 

 rows, and for 1907 only 15.0, thus agreeing with my general impression 

 that 1907 was the less favorable year. 



With the fundamental change in my understanding of the nature of my 

 corn population came a reorientation of the experiment. I found myself at 



