CAUSE OF HYBEID VIGOE 177 



the first generation, the data given in Table VII are ob- 

 tained. Summing up the results of this tabulation, it will 

 be found that eight individuals are completely homo- 

 zygous and reach the same development as either parent, 

 six units ; eight are heterozygous in all three chromosome 

 pairs and duplicate the twelve-unit growth of the first 

 generation; the remaining forty-eight ' individuals fall 

 into equal-sized groups, developing to eight and ten units, 

 respeotively. In other words, the distribution is sym- 

 metrical, and this sjonmetry remains, however many 

 chromosomes are involved. 



It should also be noted that the mean development 

 of the second generation is nine units, which is an excess 

 of just half of the excess of the first generation over the 

 parent. The extra growth derived by crossing the two 

 different types has diminished 50 per cent. In the third 

 generation, from a representative sample of the second 

 generation, it can be shown that this excess again dimin- 

 ishes 50 per cent., so that the effect on the average is only 

 25 per cent, as great in this generation as in the first, and 

 so on, in subsequent generations, until the effect dimin- 

 ishes to a negligible quantity in about the eighth genera- 

 tion. This is in fair agreement with the actual results 

 obtained by inbreeding maize, as it ought to be, because 

 the development attained by each individual varies di- 

 rectly with the number of heterozygous factors. 



In the preceding illustration of the way heterosis 

 may be brought about, perfect dominance was assumed. 

 Moreover, breaks in linkage with the formation of 

 new linkage groups were not considered. All these 

 things enter as complicating factors. Perfect domi- 

 nance, except in more or less superficial characters, 



12 



