102 CYTOKINESIS. 



have been unable to determine why the centrosomes and spheres in this single case 

 move under other cells and thus give rise to a middle layer. 



(5). Generally these telokinetic movements continue throughout the Avhole 

 of the period which is commonly called the " rest." They grow less and less evi- 

 dent, however, as the prophase of the next division approaches and, for a brief period 

 before the next cleavage begins, cease altogether. This brief period we may call 

 the "pause" (Fol '96). Dui'ing the pause the nuclei frequently lie in that portion 

 of the cytoplasm which will form the larger of the two daughter cells at the next 

 division. Thus in the pause preceding the first subdivision of the first quartette the 

 nuclei lie as close as ^jossible to the animal pole, figs. 76, 91, 92, 93, and these portions 

 of the cells become the large cephaloblasts at the following cleavage ; in the pause 

 preceding the second division of the first quartette, the nuclei lie some distance from 

 the animal pole in those portions of the cells which will become the large basal cells 

 at the next division, fig. 96, text figs. XIV, XV. This signifies more than that the 

 nucleus lies in the center of its working sphere, since the nucleus does not lie in the 

 center of the cytoplasm, but always in a position which has reference to the future 

 division; the equality or inequality of the division is alreadj- predetermined before 

 any trace of that division has appeared. 



(6). Finally, the movements in telokinesis are in some way caused by the 

 polarity of the protoplasm of each cell ; in fact every blastomere behaves much as 

 does the entire egg before cleavage begins, its substance rotating until the cytoplasm, 

 nucleus, centrosome and sphere are carried to that portion of the cell nearest the 

 animal pole. During the cleavage the spindles lie in man}' directions and cells are 

 formed in many positions, but after every division the original jjolarity of each cell 

 is, as far as possible, restored. Further, this rotation may be associated with the 

 movement of the poles of the spindle, through chemotropic influence, to a free 

 surface of the cell. The fiict that the spheres become pressed against the cell mem- 

 brane and that in this position they undergo changes in form and staining reactions, 

 staining more deeply and becoming more coarsely granular, suggests that they here 

 undergo some chemical change, probably an oxydation.' This factor, however, will 

 not account for the fact that the spheres move in a predetermined course as near as 

 possible to the animal ]3ole and that the whole cell contents move with them ; this 

 movement is evidently reducible to that class of movements which brings about the 

 polarity of the egg, but the causes of these movements I am unable at present to 

 analyze further. 



Comparisons. 



(a) Protozoa and Protopliyia. — Lauterborn ('96) has observed that the nuclei 

 and centrosomes rotate through an angle of 180° at the close of division in diatoms. 



^ Attempts to determine experimentally whether the spheres move to a free surface under the 

 influence of oxygen have so far been inconclusive, since all movements, as well as other developmental 

 processes cease in the complete absence of oxygen (e. (/., in an atmosphere of hydrogen). However in 

 sea water which has been boiled in order to drive off contained gases and then cooled in stoppered tubes, 

 eggs develop irregularly, the micromeres no longer being arranged in a one-layered epithelium over the 

 yolk, but forming irregular heaps and masses, sucli as would result from the failure of the spheres to 

 move to a free surface. 



