300 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



cells are represented in Figures 9 and 10, both of which afforded a com- 

 paratively accurate count; they also show the dumb-bell form of many 

 of the chromosomes, which is usually if not always lost in the later 

 metaphase. In the equatorial plate the chromosomes lie with their long 

 axes parallel to the future plane of cell division, and therefore when seen 

 endwise present the aspect of minute circular bodies (Fig. 11). Longi- 

 tudinal splitting now takes place, and the daughter chromosomes begin 

 individually their migration toward the poles of the spindle ; frequently 

 they retain their position transverse to the axis of the spindle (Fig. 11) ; 

 but usually they revolve, so as to lie parallel to the axis of the spindle. 

 As I have already stated, in the description of the resting stage, no ar- 

 choplasmic structures or centrosomes are then to be seen, nor do any 

 such appear until the formation of the equatorial plate. The spindle 

 figure can then be traced, and at each of its poles is visible a minute 

 granule, the centrosome, without any trace of a surrounding sphere. As 

 to the origin of this body I am wholly in the dark, nor can I offer any 

 evidence as to its fate, except that in the early anaphase it can no longer 

 be demonstrated. The entire achromatic figure is extremely simple, 

 consisting in the metaphase merely of the centrosomes and spindle fibres, 

 without any trace whatever of astral radiations. The spindle fibres are 

 very delicate, even in vom Rath preparations, and do not appear to be 

 of granular nature ; but with the separation of the chromosomes there 

 are formed interzonal filaments, which are very much thicker than the 

 spindle fibres and are clearly formed of rows of distinct granules (Plate 

 2, Fig. 12). During the migration of the daughter chromosomes, which 

 usually takes place in an irregular manner, one or two outstripping the 

 others, the polar portions of the spindle fibres grow less and less appar- 

 ent, while the interzonal filaments become even thicker, presenting, par- 

 ticularly in cells fixed in the vom Rath fluid, the appearance of very 

 definite rods, often bent outward somewhat at the centre, as is shown in 

 Figure 12. Finally, as the chromosomes closely approach the poles (the 

 centrosomes are now no longer to be seen), the polar parts of the fibres 

 disappear altogether, leaving in their place merely a small homogeneous 

 mass of archoplasmic substance (Fig. 12) ; the interzonal filaments, on 

 the contrary, persist until after the constriction of the cell is completed 

 (Fig. 13). Certainly, then, these cells afford no evidence whatever in 

 support of the well-known theory of cell division through fibrillar con- 

 traction ; and we must look to some agency other than that of a contrac- 

 tion of the spindle fibres to effect the migration of the chromosomes. 

 Furthermore, there can, I think, be no doubt that spindle fibres and in- 

 terzonal filaments are in this case structures of very diverse nature. 



