DIFFERENTIATION, HEREDITY, SEX 305 



cross. And in the second place, the phenomenon of " domi- 

 nance" appears even in these early stages of development, and 

 a hybrid may show certain clearly maternal characters and yet 

 in other respects closely resemble the paternal type (Stein- 

 bruck, Driesch, Boveri, Loeb and Moore). Great variability is 

 often the rule and frequently it is impossible to say whether 

 either parental trait really appears purely. It should be pointed 

 out, first, that it frequently happens in Mendelian inheritance 

 that true hybrids are either purely maternal or purely paternal 

 with respect to single traits, and second, that only after synap- 

 sis, which occurs in the germ cells of the mature hybrid organ- 

 ism, are the paternal and maternal chromosomes really brought 

 into complete relation. 



On the whole, then, while there are some few results difficult 

 of favorable interpretation, we may say that the evidence from 

 hybridization, though at first distinctly opposed to the hypothe- 

 sis of nuclear determination, at present affords the strongest 

 support of this hypothesis, and indicates that normally the 

 characters of a hybrid are determined by both of the germ 

 nuclei, and that when nuclear material from only one parent is 

 functional the characters of the so-called hybrid are deter- 

 mined thereby. 



We might mention one further possible interpretation of 

 some of the results opposed to this conclusion, and emphasized 

 by Conklin and others, namely, that in some cases, at least, the 

 fundamental or general traits of an organism may be deter- 

 mined immediately by the cytoplasmic structure of the ovum 

 alone or chiefly, while the nuclei are equally concerned in the 

 determination of the more particular specific or individual 

 traits, often appearing relatively late in development. Such 

 a possibility seems to be indicated by many of the facts of 

 germinal localization already described, and it may be that some 

 of the results indicated above, non-conformable with the 

 hypothesis of nuclear determination, point in the same direc- 

 tion. Conklin writes (Science, XXVII, 89-99) : " At the time of 

 fertilization the hereditary potencies of the two germ cells are 

 not equal, all the early development, including the polarity, 



