208 INBREEDING AND OUTBREEDING 



improbable. Surgical castrations of insects which have 

 not affected the secondary sexual characters even though 

 made in the early larval stages, are not conclusive because 

 of the possibility of the primary sex organs having a 

 marked influence on development very early in the life 

 cycle ; and parasitic castration brings in another variable 

 through the presence of the alien organism. 



Again, there is a presumable advantage in bisexual 

 reproduction in having sex-linked characters. We say 

 presumable advantage, for all of the relationships be- 

 tween sex and sex-linked characters are not clear. The 

 facts are these: One sex is always heterozygous for the 

 sex determiner and the factors linked with it. Even if 

 there be no actual advantage in the heterozygous condi- 

 tion, if heterosis prove to be only an expression of the 

 meeting of dominant characters, a possible advantage still 

 accrues to this phenomenon because the mechanism con- 

 tributes toward mixing of germ plasms. As an example, 

 let us take the Drosophila type of sex determination. 

 There the sperm is of two kinds ; the one containing the 

 sex chromosome and its sex-linked factors, the other lack- 

 ing it. The eggs are all alike, each bearing the sex 

 chromosome. It follows, then, that the male always re- 

 ceives this chromosome from his mother who may have 

 received it from either her father or mother. Moreover, 

 further variability may be derived from the linkage 

 breaks which occur always in the female. This last phe- 

 nomenon is hardly worthy of special mention, however, 

 until it is shown to be typical of bisexual reproduction. 



This short reconnaissance presents only the facts on 

 the role of reproduction in evolution as they are affected 

 directly or indirectly by inbreeding and outbreeding. A 



