The Development of Ischnochiton. gQ3 



series (Fig. 23). In this case there is a hastening of the division in 

 the posterior quadrants, a phenomenon which manifests itself from 

 now on. This division is radial and is generally almost completed 

 before a similar cleavage affects the anterior cells. 



Before this latter division is completed spindles appear in the 

 apical cells. In many cases which I noticed particularly the first 

 spindle arose in C and was almost immediately followed by D but 

 I cannot lay any stress upon this point for many variations existed. 

 This cleavage is dexiotropic and results in a beautiful cross of cells 

 lying symmetrically about the animal pole (Fig. 31). From this time 

 onward the cleavages vary more and more, lacking regularity except 

 that in almost every case they occur in the posterior quadrants before 

 the anterior. 



The next cleavage is in the rosette cells of the posterior quadrants 

 ig^i.i.2.1 and Ic^-i-'^-^. (First stages of division shown in Fig. 31.) 

 I have never seen this cleavage in the anterior quadrants. If it 

 occurs I feel confident that it is at a stage later than the one re- 

 presented in Fig. 43. 



All the cells of the apical series divide in the next stage. The 

 outer ones cleave irregularly in every one of the score of cases that 

 I have studied but I think that it may be safely said that the division 

 is leiotropic in the majority of cases. Fig. 41 gives a very good 

 idea of the modifications that arise in this division. The central cells 

 lai'i-i-i etc. also divide in a leiotropic division, their position being 

 shown in Fig. 41. 



In the same figure the cells ic^-^-'^-ii and id^-i'^-i-^ of the 

 posterior arm of the cross may be seen dividing in a perfectly bilateral 

 fashion. 



Beyond this stage of 28 cells I have been unable to continue 

 the cell lineage owing to great irregularities in the time of cleavage, 

 and this difficulty is augmented by the fact that the basal cells of 

 the cross form several small cells that destroy the distinct outlines 

 of the group. However I have been able to follow the cells in a 

 general way to a stage shown in Fig. 43. Here one is reasonably 

 sure of the general position of the series, especially so regarding the 

 anterior arms which remain undivided and no longer exist in contact 

 with the trochoblasts but are cut off from them by the cells of the 

 Molluscan cross of adjacent quadrants uniting below them. Also the 

 posterior arms consisting of several cells may be made out with con- 

 siderable accuracy, and in a general way the entire series may be 



