408 W. Batesou. 



are sometimes confounded. The latter difficulty, for instance makes 

 it impossible to unravel the copious information which exists respecting 

 the heredity of deaf-mutism. As we now know, the first preliminary 

 in attempting- research in genetics is a strict analj^sis of the data. 



Cytological appearances: sterility. 



It is beyond the purpose of tliis article, which deals primarily 

 with the results of breeding, to discuss the relation of Mendelian 

 observations to the facts of cell- division. Nevertheless it should be 

 lemarked that all that has been observed by cytologists is con- 

 sistent with the results of experimental genetics. The recognition 

 of a definite dilferentiation among the chromosomes (see especially 

 Sutton (134)) is probably an important advance, though until we 

 can positively recognize characters in the zygote as associated with 

 some visible cytological elements we must be cautious in forming 

 positive conclusions as to the relation of cytological appearances to 

 the phenomena of heredity. Such an observation has however been 

 lately made by E. B. Wilson (138) who has found evidence of a 

 Adsible dimor])hism among the sperm cells of certain insects, and has 

 shown that this dimorphism corresponds with difi'erences existing 

 between the somatic cells of the males and females of these species. 

 This discovery, which is I believe at present unique, raises high 

 anticipations for the future of cytological research. 



To extend this line of inquiry cases are needed in which types 

 having recognizably dissimilar cytological constituents give a fertile Fj . 

 Such cases seem to be extremely rare, but in Drosera Rosenberg 

 (130, 131) has found an instance which goes some way towards 

 satisfying the required conditions and in that remarkable example he 

 has actually seen the segregation of chromosomes retaining their 

 parental characteristics. 



There is still no positive proof that segregation occurs at the 

 reduction division, but all the facts known point to that conclusion. 

 Indeed we can scarcely doubt that this is the critical moment. In- 

 vestigations of sterile forms (Guy er (124). Gregory (123)) show 

 that it is often at this division that abnormality begins. Nevertheless 

 I am not yet convinced that segregation is exclusively confined to 

 that division. Knowing, as we do, that the "sterility" of hybrids may 

 show itself in every degree ranging from occasional imperfection in 

 the pollen-grains to inability to flower and sometimes to almost com- 



