38 BULLETIN 1121, U. S. DEPARTxMENT OF AGRICULTURE. 



numerous factors are involved. In such cases inbred segregates^ 

 homozygous in all factors conducive to vigor would be rare and the 

 degree of skewness would be imperceptible. 



It may be added that even if only one or two factors are involved, 

 the skewness in F, would be imperceptible if environmental in- 

 fluences play an important part in the variation. In the experi- 

 ments on guinea pigs described in the present paper, over 90 per 

 cent of the variation in such characters as size of litter and weight 

 is demonstrably due to factors which are not genetic. No significant 

 differences in skewness or even in variation can be found between 

 inbreds and second generation crossbreds. But none is to be ex- 

 pected under such conditions. 



Finally, the success which some livestock breeders have had with 

 close inbreeding and which Darwin obtained in at least one case 

 with his morning glories indicates that inbred lines are produced 

 occasionally which are thoroughly satisfactory from the standpoint 

 of 'vigor. This point has been demonstrated most conclusively in 

 the extensive experiments of the Wistar Institute conducted by 

 Dr. Helen D. King (1918). In this experiment a strain of rats (two 

 lines since the seventh generation) has been inbred, brother with 

 sister, for 22 generations. Not only has full vigor been maintained 

 but the inbreds have actually come to surpass the random-bred 

 stock of the Wistar Institute in size and fertility. This result she 

 attributes to careful selection. The strain of albino rats used was 

 doubtless also rather homogeneous to begin with. The fact remains 

 that long-continued, intensive inbreeding is not incompatible with 

 a high degree of vigor. ^ 



It thus turns out that as far as the facts of inbreeding and cross- 

 breeding are concerned the distinction between the hypotheses is 

 largely one of wording. The choice between them depends on which 

 involves the fewest unproved assumptions. So far as the writer 

 knows, it has not been demonstrated in any specific case that a 

 heterozygote may show an increase in vigor while the two homo- 

 zygotes are indifi^erent. On the other hand, it has been noted 

 repeatedly that there is a correlation among known Mendelian 

 characters between dominance and vigor, or, looking at it from 

 the other end, between recessiveness and deleterious efi^ect. Collins 

 (1921) prefers the latter form of statement as suggesting better the 

 probable evolutionary significance of the phenomenon. 



Most of the mutations known in DrosopMIa are less vigorous than 

 the normal strain (Morgan, Sturtevant, Muller, and Bridges, 1915)- 

 They are also mostly recessive, at any rate as regards detrimental 



5 These experiments have also given a remarkable demonstration of the success of inbreeding associated 

 with close selection, as a method of modifying a character so dilflcnlt to deal with as sex ratio. Two Unes, 

 separated in the seventh generation, were selected respectively for high and low ratio of males to females. 

 An average sex ratio of about 122 became fixed in one line, about 82 in the other. 



