BIGELOW : NUCLEAR CYCLE OF GONIONEMUS MURBACIIII. 309 



metaphase. Since, however, in all the earliest cases studied, I have 

 found from 25 to 28 or 29, and a few of these invariably lie at a differ- 

 ent level from the others, I believe there is no reasonable doubt that the 

 number is actually the same as in somatic cells, in all probability twenty- 

 four (see page 299, and Figs. 9, 10). As the splitting progresses (Figs. 

 11-13), all the daughter chromosomes are usually visible in pole views, 

 so that with the advance of the process the apparent number constantly 

 increases, a condition which at first, when I endeavored to determine the 

 relation of the spermatogonial to the somatic mitoses, proved very em- 

 barrassing. The process of splitting is parallel to the future plane of 

 cell division and thus longitudinal; it commences first at one end of 

 each rod-like chromosome, so that there constantly occur V-shaped fig- 

 ures, many of which are seen in the cells represented in Figures 29 and 

 30. From what we have learned of the structure of the chromosomes 

 of this animal, it is evident that the split must divide each chrornomere 

 into two, and that it is therefore truly ecpiational, not only for the chro- 

 mosome but for its component parts as well. 



It is in the early metaphase that the achromatic figure can first be 

 clearly distinguished (Fig. 31). This, as described by both Aders (:03) 

 and Downing (:05) for Hydra, is exceedingly simple, its most important 

 feature being, as in the case of the somatic cells, the entire absence of 

 any trace of astral radiations. 



The centrosomes, which, as already noted, cannot be demonstrated 

 previous to the metaphase, are exceedingly minute granules (Fig. 31), 

 lying directly at the focus of the spindle rays, without any trace of 

 centrosphere, and situated either at the margin of the cell (Fig. 31) or 

 some distance within its substance (Plate 3, Fig. 32). Preparations fixed 

 with Flemming's mixture show characteristic differences in the appearance 

 of the spindle from those fixed in vom Eath's fluid. In the former case 

 its ground substance appears homogeneous, staining more darkly than 

 ordinary cytoplasm ; the fibres are exceedingly delicate and can be traced 

 only with difficulty. In the latter it presents no difference from cyto- 

 plasm ; the fibres, however, are much thicker, stain darkly, and are 

 plainly composed of series of apposed granules; the centrosomes are 

 larger and more prominent. These differences, moreover, are character- 

 istic of the achromatic figures of all the mitoses studied. 



Anaphase. — The daughter chromosomes now commence their migra- 

 tion toward the poles ; this, however, is not, as a rule accomplished by 

 all the chromosomes simultaneously, for one or two usually outstrip the 

 others (Plate 2, Fig. 31). They may still retain a position parallel to 



