38 bulletin: museum of comparative zoology. 



is not derived from the body of the centrosome by a simple process of 

 transformation, as has been maintained by several authors, but rather, 

 that it is caused by a formative power exerted by the centrosome over 

 the latent archoplasm contained in the cytolymph. That is, its origin 

 is similar to that of the astral rays, toward the formation of which it 

 later often contributes some of its own material. It is to be especially 

 noted that during all the early stages of the development of the cen- 

 trosphere, some of the astral rays may be seen to traverse the centro- 

 sphere and to abut upon the smaller enclosed body, the centrosome. 

 The same phenomenon is to be observed occasionally in the later meta- 

 phase and anaphase. These facts seem to preclude the possibility of 

 the outer zone being considered the true centrosome, and to justify the 

 application of that term to the small dark staining body within. Shortly 

 after the formation of the first maturation spindle the structure at the 

 centre of the aster seems to have reached its highest development (Figs. 

 50 a, 50 b). It consists of an elongated dumbbell-shaped centrosome 

 enclosed in a homogeneous or very finely granular centrosphere of an 

 oval form. 



In the metaphase of the large type of spermatocyte the long diameter 

 of this oval mass is arranged in no fixed relation to the diameter of the 

 spindle, although later, in the telophase, a constant relation is very 

 apparent (Plate 3, Figs. 37-39). It is to be noted, however, that during 

 the rotation of the division figure, which in the large type of spermato- 

 cyte seems invariably to occur during this stage (metaphase), the long 

 diameter of the centrosome is parallel to the direction of its move- 

 ment. This is shown in Figures 33, 34, 35. It is also noteworthy that, 

 in the revolution of the spindle, its poles (centrosomes and asters) 

 always seem to precede, while the other parts — mantle fibres, spindle 

 fibres, and chromosomes — lag behind. This suggests very strongly 

 that it is the centrosome which is the originator and director of this 

 interesting change in the relative positions of the various parts of the 

 cell. In this, as in numerous other instances of cell activity, this small 

 body appears indubitably to act as the directive or dynamic centre of 

 the cell. 



By the time the rotation of the spindle is completed the centrosome 

 and other structures at the centre of the aster have changed considerably 

 in appearance. The centrosome is always at this time of a distinctly 

 dumbbell shape. It consists of two aggregations of centroplasm con- 

 nected by a bar of the same material, as shown in Figure 50 b. Vei*y 

 often this connecting bar or thread is relatively of considerable length, 



