Olive, Mitotic divisiou of tlie uuciei o£ the Cyaiiophyceae. 21 



splits tlirougliout its entire lengtli into two exactly equivalent 

 lialves'' (AVilson, p. 70). A tliird attribute of a typical mitosis 

 is tlie presence of a fibrous, acliromatic mass known as a 

 spindle. 



It remains, therefore, to be proved, first, that, in the di- 

 vision of tlie central body of tlie Cyaiiophyceae^ a spindle is 

 present; second, tliat tlie clironiatin is arranged, at some time 

 dnring tlie process, in tlie form of a spireme tliread; and, lastly 

 and most important of all, tliat the chromatin granules are 

 halved and tliat an equal number is distributed to each 

 daughter cell. 



It is perhaps advisable, at the very outset, to discuss brieily 

 the staining reactions of the chroniatic and achroniatic elements 

 of the nucleiis. Wliile the writer realizes fully that staining 

 reactions should by iio meaiis constitute the principal argument 

 in Support of the mitotic division of the central body, yet there 

 can be no doubt that a comparison witli the well known results 

 already obtained witli nuclear stains will be of value. The 

 writer is well aware, further, that staining reactions are fre- 

 quently misleading; and that the most credible data concerning 

 complex nuclear phenomena are furnished by observing 

 simply the changes wliich take place. Those most valuable 

 stains, Heidenhain's iron haematoxylin and Flemming's 

 triple stain, wlien used to check and to Supplement each 

 other, in niy opinion, assist as perhaps no other stains can, 

 in the Interpretation of the comjDlex structures witli wliich we 

 have to deal in the central body. Iron haematoxylin gives 

 generally mucli the shaiqier differentiation and is the easier of 

 the two to use; but the objection has been rightly raised that 

 other tliings tlian nuclear structures may be stained by it and 

 that, consequently, great care must be employed in drawing 

 conclusions. I have found, moreover. particularly in the case of 

 tliese algae, that Flemming's triple stain is an exceedingly 

 difiicult combination to handle so as to obtain the best results; 

 but, on the other liand, it fui-nishes us with staining reactions 

 wliich can scarcely be doulited, in the Interpretation of chromatic 

 and achi'omatic elements of the nucleus. 



It is sufficient to say here that the most satisfactory details 

 were obtained with well differentiated iron haematoxylin. The 

 minute black chromosomes stood out, sharply deiined, in a bluisli, 

 or dark, or at times almost invisible acliromatic substance (iigs. 

 8, 18, 26, 27, 32, 34 etc.;. Flemming's triple stain gave dark 

 reddish, or purplish chromatic structures wliich were often poorly 

 differentiated, thus giving an appearance which would lead one 

 to the conclusion that chromatin granules and the acliromatic 

 substance were fused together (figs. 7, 10, 13, 17, 25 etc.). The 

 dense, acliromatic portion of the central body stained with the 

 triple stain dark bluisli, or purplish, or eveii reddish; whereas 

 the hbers which lead from the chromosomes to the cross waUs 

 d<> not readily take any stain sufficiently to bring tlieiii out 



