The maturation aud fertilization of the egg of Cerebratulus. 459 



US to the interesting inquiry as to whether the centrosphere pushes 

 the aster- fibres before it or whether it absorbs their inner portions 

 into its own substance. From the fact that we can easily follow the 

 fibre into the body of the centrosphere, and that here it becomes 

 gradually thinner and fainter until it disappears at an indefinite point, 

 and as the centrosphere increases in size this point becomes just so 

 much farther removed from the centrosome, as well as for reasons to 

 be noted below, it seems extremely probable that the substance of 

 the inner ends of the aster-fibres is actually used in the formation of 

 the centrosphere. 



While the chromosomes are splitting each of the minute centro- 

 somes, situated in the centres of the comparatively enormous centro- 

 spheres, divide into two in preparation, even at this early period, for 

 the division into the four-cell stage. As the daughter-chromosomes 

 move apart they gradually become much shorter and thicker (Fig. 36). 

 The equatorial portions of the spindle-fibres become more irregular 

 and larger. In this region they begin to stain much more deeply 

 than in their other portions. Numbers of tine granules also appear 

 in this region, and these likewise stain very deeply. It is possible that 

 these granules are in part, actually small particles of chromatin which, 

 as authors have believed, have been cast off from the dividing chromo- 

 somes, but it is certain that they mostly originate from the breaking 

 up of the spindle-fibres. 



During the anaphase the centrospheres continue to increase in 

 size, and their vesicular structure becomes more and more apparent. 

 The minute centrosomes which, as we have seen, divided at a very 

 early stage, begin to separate somewhat at right angles to the axis 

 of the spindle. As they move apart, it is seen that each is provided 

 with, or imbedded in, a minute, homogeneous sphere which stains red 

 with Bordeaux. Connecting the two divided centrosomes is sometimes 

 seen a very delicate fibrous structure which bears a certain resemblance 

 to a central spindle, although it is too minute and delicate to be 

 determined with accuracy (Fig. 36). 



As the chromosomes swell out into an equal number of small, 

 «val vesicles (Fig. 38) the daughter centrosomes separate more widely, 

 and all traces of connecting fibres are lost. They move outwards to 

 that side of the centrosphere which lies farthest from the chromo- 

 somes, and about each may be seen the faintest indication of radiations 

 — only four to six rays in a section. The radiations are exceedingly 

 delicate, and are directly continuous with the little sphere of proto- 



