Sorting and blending of unit characters. 227 



blend hybrid (nut ell a), occurs in the zj^gote or fertilized egg. Therefore 

 it is of a very different type from that which takes place in Mendelian 

 segregation, due according to general belief to a qualitative, or differ- 

 ential division or reaction, in the gonotokonts. The qualitative division, 

 or segregation of "unit" characters in Mendelian segregation, is by 

 many believed to find its explanation in the peculiar cytological pro- 

 cesses occurring in the meiotic divisions (heterot3T)ic and homöotypic — 

 together called "allotyiDic" divisions by Strasburger, 3, 1904) giving 

 rise to the pollen grains and embryo sac. No such qualitative division 

 is known to take place regularly in the fertilized egg, and cannot be 

 invoked to explain zygotic segregation. In the angiosperms the first 

 divisions of the fertilized egg do not result in free cell or free nuclear- 

 formation, any one or several of which may develop embryos, but some 

 of the products of the first divisions are definitely left behind in the 

 suspensor and other cells. In certain of the gymnosperms there is a 

 regular process in the division of the fertilized egg nucleus into four 

 or more free nuclei, several of which are capable of forming independent 

 embryos. But here, as in the angiosperms, no regular cytological pro- 

 cesses occur which are at all comparable to the meiotic divisions in the 

 gonotokonts, upon which a theory of segregation of unit characters could 

 be buut up. 



Very little is known of the cytological process in the fertilized 

 egg of plants so far as it relates to the more critical stages in the 

 organization of the nuclear figure for the first division and the behavior 

 of the paternal and maternal chromosomes in the organizatiqn of a work- 

 ing relation in the new diploid nucleus. Guignard (1890) describes 

 and figures the spindle for the first division of the fertilized egg in 

 Lilium Martagon, which is preceded by a double gnarl stage represent- 

 ing the paternal and maternal nuclear chromatin skeins, but the separ- 

 ate chromatin skeins were not observed in the successive divisions. 

 The formation of the paternal and maternal chromosomes and their 

 mutual behavior in the organization of the diploid nuclear complex for 

 reciprocal action in the new organism also was not observed. In Pimis 

 strobus Margaret C. Ferguson (1901, 1904) has shown that while the 

 sperm and egg nuclei are in close contact the spindle is organized be- 

 tween them, while the paternal and maternal chromosomes form from 

 the chromatin network of the sperm and egg respectively and move to 

 the nuclear plate of the spindle. In two successive divisions the paternal 

 and maternal chromosomes are formed in separate groups. But their 



15* 



