192 A. C. WALTON 



An examination of 762 nuclei in the late prophase stage 

 showed two in which there were nine tetrad chromosomes of 

 like size and one small diad chromosome, the latter being 

 identical with the chromosome formed from the smaller chro- 

 matin mass of the early spermatocytes (fig. 4a). The rest of 

 the nuclei examined showed eight tetrad chromosomes of like 

 size and one that was about one-third larger (fig. 4b). This 

 chromosome was apparently a tetrad with one end longer than 

 the other. Among these 760 nuclei exliibiting this condition, 

 there were 78 with a more or less distinct constriction midway of 

 the large chromosome, varying from a very slight superficial 

 indentation to an almost complete separation of the elements. 



From this evidence it seems probable that, as in the case of 

 A. megalocephala (Boveri, '09; Edwards, '10), there is a con- 

 sistent attachment — generally an indistinguishable union — of the 

 small chromosome to the end of an autosome. Such a composite 

 chromosome has a hexad form in A. felis, the small chromosome 

 being a diad and the autosome a tetrad in form. The method 

 of development of this diad chromosome from a small separate 

 chromatin mass is so completely analogous to the formation 

 of the X-type idiosome complex of A. canis from a similar 

 separate chromatic mass in the early spermatocytic nuclei, that 

 the interpretation of this chromosome as an idiosome of the 

 X-type seems consistent. Thus we find that early in the 

 development of the spermatocytes the haploid number of 

 chromosomes, nine, clearly defines itself; eight of these nor- 

 mally being tetrads and one a hexad (fig. 4b). The hexad is 

 formed by the attachment of the idiosome to the end of one of 

 the autosomes. 



In the metaphase of the first spermatocytic division the 

 chromosomes are arranged in the equatorial plate, with the 

 hexad chromosome apparently more centrally located than any 

 of the autosome tetrads (figs. 5a and 5b). The great variation 

 in the position is probably due to the antagonistic action of the 

 idiosome and autosome components of the hexad which pre- 

 vented either a distinctly peripheral or a distinctly central loca- 

 tion. In three cases in which the idiosome was found unattached 

 to an autosome, it occupied a definitely central position (fig. 5c). 



