288 bulletin: museum of comparative zoology. 



approaches as near the animal pole as the centrosome, but terminates at 

 one side of that structure in the outer edge of the cortical layer of the 

 sphere ; the inner end bifurcates, one branch being directed toward the 

 centrosome, the other Ij'ing tangent to the cortical layer of the sphere. 

 The three ends of this fibre terminate in the dense cortical layer, so 

 that it is difficult to determine its exact extent. 



The cytoplasm has now encroached on the region formerly occupied 

 by the spindle ; the chromosomes do not lie in the section containing the 

 two centrosouies. It will be noticed that in Figure 45 the spindle fibres 

 have a different origin from the astral rays, since the spindle fibres arise 

 from the centrioles, but the rays from the cytoplasm. In Figure 46 

 the astral rays have not formed on the side toward the spindle, as they 

 usually do. The large fibre seen in Figure 46 breaks up and disappears 

 iu the cytoplasm. With the disappearance of this fibre, the centro- 

 somes, which may be assumed to have been held together by it, quickly 

 separate, the deeper one moving most. The outer aster, in fact, moves 

 only a little nearer the periphery (compare Plate 7, Figs. 46 and 47 ; 

 Plate 11, Figs. 77 and 79), whereas the inner one migrates a consider- 

 able distance into the egg, causing a movement of the deutoplasm which 

 is thei-eby excluded from a long, narrow region that will soon be 

 occupied by the nascent spindle. 



During the separation of the asters, the cortical layer (st. ctx.) of the 

 sphere has increased in extent and defiuiteness. The primary astral 

 rays have increased in length and number, but there is as yet very little 

 evidence of secondary rays. The wall of the old centrosome, as has 

 been said, has gradually broken up into small particles, which can no 

 longer be distinguished from the granules in the cytoplasm (Figs. 

 41-44). During these changes the size of the old centrosome increases 

 greatly, but the staining reaction of the portion not occupied by the 

 spindle remains the same (Fig. 38). As the old wall disappears, the 

 surrounding cytoplasm approaches the maturation figure until it comes 

 into direct contact with it. The behavior of the cortical and medullary 

 layers shows how intimately these are associated with the centrosome. 

 The cortical and medullary layers either increase in size so as com- 

 pletely to surround the new achromatic figure (Plate 7, Figs. 42, 43 ; 

 Plate 9, Fig. 58 ; Plate 11, Fig. 73), or they may be confined to the 

 inner pole (Fig. 45), in which case they are not as large as in Figure 38. 

 On the other hand, both layers may disappear at an earlier stage. 



Figure 41 (Plate 7) shows the fate of the astral rays of the parent 

 centrosome ; for in addition to the fine short rays centring iu the new 



