Timherlake — Swarm-Spores of Hydrodidyon. 501 



mediate contact with the cytoplasm, which has not, however, 

 penetrated to the interior of the nuclear cavity. The nucleole 

 disappears so far as I have been able to observe before the equa- 

 torial plate stage is reached. 



The history of the spindle in these early stages I have not 

 been able to make out at all. When the equatorial plate stage 

 is reached the spindle is distinct. It usually ends in two 

 sharply defined poles, at either of which there is a small spher- 

 ical, densely stained body ; but there are no indications of polar 

 radiations to form an aster (Fig. 9). The spindle fibres gener- 

 ally come to a sharp point apparently just at the surface of the 

 body so as to give the appearance of a distinct body lying rather 

 in contact with the end of the spindle than forming a part of it. 



Owing to the impossibility of observing the early stages of 

 spindle formation the origin of the above described bodies, as 

 well as their relation to spindle development, could not be made 

 out. During the period of the reconstruction of the daughter 

 nuclei when the spindle disappears, the bodies also become in- 

 distinguishable. Whether they are the homologues of centro- 

 somes is, in view of the scarcity of data in connection with their 

 history, of course, not evident, still the constancy with which 

 they appear in the equatorial plate stages and early metaphases 

 indicates that they bear the same relation to the process of di- 

 vision as the centrosomes in other cells. I shall apply the name 

 centrosomes to them in the subsequent discussion. In many 

 cases the whole spindle seems to lie in a clear cavity, as if the 

 nuclear membrane persisted throughout the greater part of the 

 division process, but the boundary of this cavity is always quite 

 iiTcgular and I am inclined to think that it is the product of 

 the fusion of the vacuoles of the cytoplasm surrounding the 

 nucleus. (Figs. 9, 11, 12.) In other cases the cytoplasm is 

 apparently in immediate contact with the spindle. (Fig. 10.) 

 The chromatin material in the equatorial plate forms so com- 

 pact a mass as to render it almost impossible to distinguish the 

 individual chromosomes, and consequently to make out the 

 method of the separation of the daughter chromosomes. 



In the metaphases the daughter chromosomes go back, to their 

 respective poles in dense groups, all the individuals of each 



