420 EDMUND B. WILSON 



received fresh support in recent years from several sources. Direct 

 experimental evidence that is nearly if not quite demonstrative 

 has been produced through the work of Boveri ('07) on multipolar 

 mitosis, of Baltzer ('09, '10) on reciprocal crosses in sea-urchins, 

 and of Herbst ('09) and Godlewski ('11) on the combined effects of 

 artificial parthenogenesis and fertilization in hybridization. In- 

 direct but very strong evidence has been given by the demonstra- 

 tion of a constant relation between particular chromosomes and 

 sex and especially sex-limited characters (cf. Wilson, '11 a, b, 

 Morgan, '11, Gulick, '11). This evidence in no manner precludes 

 the view that the protoplasmic substances are also concerned in 

 determination — indeed experimental embryology and cytology 

 have produced very clear evidence that such is the case. The 

 study of the nucleus, and especially of the chromosomes, offers 

 however one of the most available paths of approach to a study of 

 the activity of the germ-cells in determination, and for a detailed 

 analysis of genetic problems in their cytological aspects. 



It has been widely assumed that the Mendelian segregation 

 depends upon a disjunction of chromatin-elements in the reduction- 

 division, as was originally suggested by Guyer, Sutton, Boveri 

 and Cannon. It has, however, becomes obvious from the experi- 

 mental data that if this be so, these elements can not be individual 

 chromosomes of fixed composition. This was first seen to follow 

 from the fact, now apparently well determined in certain cases, 

 that the number of independent allelomorph-pairs may be greater 

 than the number of chromosome-pairs. More recently the same 

 result is demonstrated by the work of Bateson and Punnett ('11) 

 on 'coupling' and 'repulsion' in certain plants, and by that of 

 Morgan ('11 a) on sex-limited characters in Drosophila, which 

 proves that an interchange of unit-factors must to some extent 

 take place between homologous bearers of these factors in the 

 germ-cells. The results of Morgan in particular nevertheless 

 bring strong support to the view that the chromosomes are such 

 bearers of unit-factors ; for the whole series of phenomena deter- 

 mined in Drosophila, complicated as they seem, become at once 

 intelligible under the assumption that certain factors necessary 



