422 EDMUND B. WILSON 



explanation of the difference of the sexes in respect to the inter- 

 change of sex-limited factors that is proved to take place by the 

 experimental results. 18 



There is no a priori reason why such a process of free interchange 

 between homologous chromosomes may not take place in the 

 somatic or diploid nuclei. It is, however, far simpler to assume 

 that it occurs during or subsequent to synapsis; for it is only at 

 this period that the homologous chromosomes are intimately and 

 regularly associated (cf. Boveri, Strasburger, '08, '09). It is 

 these facts, taken together with the cytological evidence, that 

 lead me to the conclusion already indicated that the synaptic 

 union results in a reorganization of the chromosomes, and that the 

 two moieties of the final prophase-chromosomes are probably not 

 identical with the original conjugants. Janssens's theory of the 

 chiasmatype ('09) gives the only simple mechanical explanation 

 thus far offered as to how such an orderly exchange of materials 

 may be effected; and his conception has recently been applied in 

 an ingenious manner by Morgan ('11 b) to an explanation of 

 both 'repulsion' and 'coupling' in heredity. The chiasma type- 

 theory has been criticized by Gre*goire and by Bonnevie on the 

 ground that a strepsinema stage occurs in the division of somatic 

 nuclei as well as in the maturation-prophases, and that the orig- 

 inal twisting together of the threads is in some cases followed by 

 untwisting, without such aprocessof partial fusion and subsequent 

 secondary splitting as is postulated by Janssens. There are two 

 replies to this. One is that Janssens's theory is not an a priori 

 construction but a conclusion based on a most accurate and de- 

 tailed study of the actual conditions seen in the prophases of 

 Amphibia which prove that such a process as he postulates must 

 here take place. The other is that there is considerable evidence 

 that a twisting together of the conjugating threads takes place 

 in the process of synapsis, leading to a most intimate union of the 



18 It would, however, be rash to generalize this statement at present, for the 

 observations of Buchner on the female Gryllus ('09) and those of Winiwarter and 

 Saintmont ('09) on the cat, have demonstrated a nucleolus-like body in the synap- 

 tic and later stages of the oocytes which may be the .X"X"-bivalent, though this is 

 unproved, and doubt is thrown upon the second of these cases by the recent work 

 of Gutherz ('12). 



