214 



KIYOSHI IIASUI. 



the late prophase many long and variously carved cLromosomes appear 

 wliicli are entangled wUb each other (Fig. 86). In the metaphase it is 

 difScult to malie out every chromosome, for the long, curved chromosomes 

 overlap each othar, uotwithstauding tha perfect separation of the individual 

 ones (Fig. 87,!. 



It has been found that the number of chromosomes is not constant but is 

 considerably variable (Figs. 87-91). For coimting the chi'omosomes only thoss 

 cells were selected which were clearly foimd to be imcu'; and ia which the 

 chi'omosomes were well separated (Figs. 87-91). It is therefore clear that 

 the variation of the uumljer of chromosomes is not caused by the use of 

 poorly prepared materials. The seventeen cells studied in this connection 

 fall into tlie groups : — 



Table III, Showing variati'm in the number of chromosomes. 



As in the mouse, the chromosomes vary considerably in size and form 

 (Figs. 87, 88). The Lirger chromosomes are considerably long and variously 

 curved, while the smaller ones appear as round bodies (.Figs. 87, 89). Some 

 larger ones are sti'ongly curved and sometimes the transverse constrictions 

 can be seen at their ends (Figs. 87-91)- It is remarkable that in those 

 cells which coutain a large number of the chi'omosomes, the smaller ones 

 increase in number, while in those haviug forty eight chromosomes only 

 fom- smaller ones can be seen (Figs. 89-91). From the above facts it is 

 conceivable that this difference iu the number of chi-omosomes is probably 

 due to the influence of the fixatives. 



The arrangement and the dividing method uf the chromosomes are equal 

 to those (jf the mouse (Figs. 92-95). As all the chtomosomes assume simply 

 the straight rod-shape iu the anaphase, it seems most probable that the 



