LINKAGE 391 



some reason is limited in these cases and some others to the sex which is 

 heterozygous for the sex-factors. On the contrary, in the grasshopper, 

 Apotettix, Nabours (1919) has shown that some crossing over occurs in 

 both sexes, and the same appears to be true in Primula (Gregory; Alten- 

 burg 1916), the rat (Castle and Wright 1916), and Zea (Emerson). In 

 Paratettix, in which no crossing over has been demonstrated, Miss 

 Harman (1920) reports that the homologous chromosomes do not 

 conjugate until the end of the prophase, and suggests that their indepen- 

 dence during the early stages may account for the absence of crossing 

 over. 



General Discussion. In the foregoing pages a brief account has been 

 given of the main points in the theory developed by those who have made 

 the most thoroughgoing attempt to relate the phenomena of heredity to a 

 visible cell mechanism. To follow out the details of its application does 

 not lie within the scope of this chapter : it is here intended only to furnish 

 a starting point for cytological studies in this field by indicating the 

 common ground upon which cytology and genetics meet. It is import- 

 ant, however, to differentiate between evidence which is genetical and 

 that which is cytological in nature; and further to remind ourselves to 

 what extent observed fact and hypothesis respectively have been woven 

 into the theory. Caution is particularly necessary in this latter regard, 

 since the general nature of many of our ideas of inheritance is traceable 

 in part to the speculative theories of Weismann. Weismann's theories 

 of heredity and development, which are summarized in the next chapter, 

 were primarily "corpuscular" or " particulate " theories: the phenomena 

 of heredity and development were referred to distinct material units 

 which in some way were able to bring about the development of the 

 heritable characters in the individual and their transmission from one 

 generation to the next. Bearing in mind the phenomena of inheritance 

 reviewed in the preceding chapters, especially the behavior of the Men- 

 delian characters, it is difficult to escape the conclusion that differential 

 factors of some sort, which in an unknown manner initiate the series of 

 reactions resulting in the several characters, are carried in the nucleus. 

 To determine the nature of these factors and to discover the real relation 

 existing between them and the developed characters are among our 

 greatest problems. That the factors or genes are discrete units is a 

 hypothesis which is not only plausible, but has also proved itself to be 

 most useful. If such factors exist, the chromosomes afford a means of 

 precisely the kind required to account for the observed distribution of 

 characters throughout a series of generations. Hence from Roux and 

 Weismann onward the factors have been lodged in the chromosomes. 



But it is when these factors are directly sought with the aid of the 

 microscope that disappointment is met. The frequently observed 

 granules or chromomeres in the chromatin thread or chromosome are 



