CELL DIVISION IN EGGS OF CREPIDULA. 543 



In the withdrawal of the archiplasm of the astral radiations into the spindle, 

 certain portions may be left isolated along the rays, especially if these rays are 

 very long, as is the case in the 1-cell and 2-cell stages. These isolated portions 

 of archiplasm become the centers of new radiations, thus forming cytasters. 

 These cytasters sometimes contain deeply staining granules, which have the 

 the appearance of centrioles, and which may be derived from the axial filament of 

 the radiation, but so far as my observations go, they never become the poles of 

 spindles nor undergo division. On the other hand cytasters frequently fuse 

 together, as Mead (1898) and Morgan (1899) have determined, whereas nuclear 

 asters often divide, but rarely fuse. If very many cytasters are present in a cell 

 the astral systems at the poles of the spindles are smaller than usual, because 

 the archiplasm which largely composed these systems is distributed to many cen- 

 ters; this fact has been observed and commented upon by both Morgan and Wil- 

 son. Cytasters appear best developed during periods of mitosis and if a long 

 resting period follows, the place of the cytaster is taken by a faintly staining 

 vesicle of archiplasm or achromatin (figs. 186-188, 191, 192). The substance thus 

 isolated forms a delicate membrane and gives rise to a small "nucleus without 

 chromatin" (R. Hertwig). I have not determined positively that such an achro- 

 matic vesicle can give rise to another astral system at the next mitosis but 

 such seems to be the case. Where cytasters occur during interkinesis it is 

 invariably found that a large amount of archiplasm is distributed through the 

 cell, as for example after the maturation divisions and before the first cleavage. 

 It is well known that a great amount of achromatin escapes from the germinal 

 vesicle at the time of the first maturation division, and this achromatin is widely 

 distributed through the egg along the astral radiations. When the germ nuclei 

 have grown large and have therefore absorbed a great deal of achromatin no 

 cytasters are formed, especially in the vicinity of the nuclei, but when the nuclei 

 are small and much achromatin is still scattered through the cell, cytasters may 

 appear. The cytasters are therefore, in my opinion, isolated portions of archiplasm, 

 derived in large part from escaped achromatin, which take the aster form during 

 mitosis and the vesicular form during resting periods (figs. 183-193). 



In addition to the presence of cytasters in eggs exposed to hypertonic sea- 

 water, several nuclear asters may be present and since all asters within the same 

 cell are in divisional activity at the same time, multipolar mitoses thus arise 

 (figs. 203, 219, et al). These polyasters are found only in cells in which cleavage 

 has been temporarily suppressed while centrosomal division proceeded, or in 

 cells in which cleavage has been permanently suppressed, while the centrosomes 

 have continued to divide after the eggs were returned to normal sea water; in 

 the latter case their number is generally proportional to the length of time the 

 eggs have been in normal sea water. These polyasters pass through periods 

 of division, alternating with periods of rest, and there is reason to believe that 

 they have arisen by regular division from an original single centrosome and 

 aster in each cell; such division would give rise to, first amphiasters, then tetras- 



