296 JAMES F. CROW 



great length of time, but in any particular population such factors might be 

 important. 



IS INCREASED SIZE ADAPTIVE? 



The foregoing arguments are based on the assumption that heterosis is 

 measurable in terms of increased selective advantage. The selection may be 

 natural or man-imposed. This assumption would appear to be reasonable for 

 such factors as fertility and resistance to disease. It also would apply to in- 

 crease in size or yield, if the direction of selection in the past were in this di- 

 rection, as in corn. However, it is questionable whether the increase in size 

 that is sometimes observed in variety hybrids is really adaptive. 



Mather (1943) and especially Dobzhansky (1950) have emphasized that 

 increased size does not necessarily result in increased fitness in natural popu- 

 lations. Dobzhansky proposed the words euheterosis and luxuriance, re- 

 spectively, for increased selective advantage and for mere non-adaptive in- 

 crease in size. In these terms this discussion has dealt entirely with eu- 

 heterosis. 



If euheterosis occurs in species or variety crosses, it is very difficult to 

 explain. It raises the troublesome question: How can the hybrid between 

 two well adapted strains be better adapted than its parents when there has 

 been no selection in the past for its adaptation? It may be that euheterosis 

 is developed only under some form of selection, as in the inversion heterozy- 

 gotes studied by Dobzhansky, or in the series of hybrids between inbred lines 

 of corn selected for combining ability. 



If large size is not advantageous, luxuriance may be due to the covering 

 of recessive factors which were acting as size bottlenecks and had been 

 selected into the population because of this. Each of the parents might have 

 its growth limited by or held in check by a series of factors, and if some of 

 these were recessive, increased size would be found in the hybrids. 



SUMMARY 



Since the earliest attempts to explain hybrid vigor in Mendelian terms 

 there have been two principal hypotheses. The first of these is the domi- 

 nance hypothesis. This notes the observed correlation between recessiveness 

 and detrimental effect and attributes the increased vigor of heterozygosity 

 to the covering of deleterious recessive factors by their dominant alleles. 

 The alternative hypothesis, the overdominance hypothesis, assumes that 

 heterozygosity per se is important — that there exist loci at which the hetero- 

 zygote is superior to either homozygote. 



It is clear that the dominance hypothesis is adequate to explain the de- 

 terioration that results from inbreeding and the recovery of vigor on out- 

 crossing, but it is difficult to explain how the hybrids could greatly exceed in 

 fitness the equilibrium populations from which their parents were derived. 

 The overdominance hypothesis demands the assumption of a kind of gene 



