PROTOZOAN NUCLEI. 391 



disintegrate during the earlier stages of mitosis, forming smaller 

 and smaller chromatin bodies, the final result being a great 

 number of minute chromatin granules, which, as in Chilomonas 

 or Euglena, are found distributed throughout the nucleus. 

 The chromatin granules later unite to form distinct chromo- 

 somes. The formation of the chromosomes is entirely different 

 from the account of the process given by Ishikawa ('94). The 

 granules of chromatin unite in lines which are focussed at one 

 side of the nucleus ; these lines are the chromosomes, and they 

 are subsequently divided through the agency of a complicated 

 mitotic process in which centrospheres, central spindles and 

 centrosomes play an important part.^ In the early stage of 

 division, when the chromatin is scattered throughout the cell in 

 the form of minute chromatin granules, the nucleus of Noctiliica 

 is obviously comparable with the nucleus of the intermediate 

 type, while the vegetative condition can be conceived as due to 

 the coalescence of the chromatin granules to form the large 

 reservoirs. An essential difference in the nucleus of Noctiluca^ 

 however, is found in the absence of an intra-nuclear central body. 

 The place of this important mitotic agent is taken by a large 

 cytoplasmic sphere lying just outside the nuclear membrane. 

 This sphere, during mitosis, plays the same part as the intra- 

 nuclear body of the lower flagellates, but in a much more com- 

 plicated way. While the chromatin granules are fusing to 

 form the chromosomes the sphere divides to form a dumb-bell 

 shaped body consisting of two daughter-spheres and connecting 

 fibrous substance forming the ''central spindle" (Fig. 23). 

 The nucleus then bends around in the form of a U until it al- 

 most completely surrounds the central spindle. The chrom- 

 osomes, focussed at the side of the nucleus which was turned 

 towards the cytoplasmic sphere, now form a nearly continuous 

 line or ring — the nuclear plate — around the central spindle (Fig. 

 2'i^). At this period it can be found by sections that the nuclear 



^ For a description of the process of mitosis in Nodiluca see my paper, 

 now in press, which will shortly appear in the Journal of Morphology on " Mitosis 

 in Nodiluca miliaris and its Bearing on the Nuclear Relations of the Metazoa and 

 the Protozoa." 



