234 KIYOSHI 5IA,sri. 



somes form a duplex series, even though from the above reason it can not be 

 expected, that the homologous chromosomes -will always be of the same size. 



The relation between size of cells and that of chromosomes : — Moreover 

 in Figs. 9-12 where the chromosomes are arranged in accordance to our present 

 method, each of the homologous pah's is found to be almost similar in shape, 

 so that we can fafrly reach the conclusion that there exists a constant relation 

 between their size and their form. 



It has already been stated above that the size of the chromosomes seems 

 to vary in accordance with those of the nni'lei. This relation will be more 

 readily understood from Text-fig. 1, where cm"ve I indicates the relation of the 

 size of chromosomes which are contained in the large cell, while curve II 

 indicates that of the small ones. It is found that these two curves are almost 

 parallel with one another. Judging from this fact it is evident that the size 

 of cells as well as that of the chromosomes vary considerably according to the 

 generation of the spermatogonia, the chi'omosomes in eai'ly generations being 

 probably larger than those in the final generation. 



The sex chromosomes : — It is remarkable that there are two different 

 chromosomes which have no mate. One of them is larger than the other, and 

 differe considerably in form from the neighbouring chromosomes of the series, 

 while the other is the smallest one of the entire series. 



The difference iu size Ijetweeu this smallest chromosome and the neighbour- 

 ing one is considerably lai'ger in comparison with that occun-ing between our 

 other pahs. The above data shows conclusively that these special chromosome» 

 are XY-chromosomes. 



The fr'agmentatiou of the chromosomes iu rabbit :— As stated above the 

 number of chromosomes iu the germ cell of the rabbit is not constant but 

 vaaies considerably. Tom Rath ('94) long ago found that the somatic chi'omo- 

 somes in the dog vary considerably. Tlie same plienomenon was reixirted by 

 Winiwarter ('00) in the somatic cells of tlie rabbit. Haxce ('17 b) found 

 that the number of chronaosomes of the sonuxtic cells in the pig is not 

 constant but varies from forty to forty seven. He atti-ibuted cause of this 

 variation iu number to the fr'agmentatiou of certaiu chixjmosomes, and says : 

 " Since the fragments are fairly uniform iu length, the chromosomes must be 

 educed by more or less equivalent amounts, and consequently we should not 



