278 bulletin: museum of comparative zoology. 



which becomes the centre of one of the newly formed asters. This body 

 increases in size and passes through a metamorphosis in the following 

 cleavage generation similar to that just described. In his description, 

 Boveri does not use the term "centriole," but the function of these gran- 

 ules is to give origin to a new centrosome ; so that when we consider 

 them in the light of their fate, it is permissible to apply to them the term 

 " centriole." The interpretation of a body, it seems to me, should depend 

 quite as much on its function as on its form. For this reason I think 

 that we may use the term "centriole" in describing these changes in 

 Haminea. The granules in the centroplasm in Echinus and in Haminea 

 are similar, except that in Echinus they are more numerous. 



The history of the centriole in Haminea up to this stage indicates 

 that it appears under a variety of forms ; and the succeeding changes, 

 to be described in the next two sections, reveal still greater variations. 

 Conklin (:0l) finds that in Crepidula the centriole may be formed anew 

 in the centroplasm (his medullary zone) in both maturation and cleav- 

 age stages. 'His observations on Crepidula and mine on Haminea 

 suggest that Boveri's contention in regard to the independence and 

 individuality of the centriole is open to serious question. 



In Figure 26 the centriole is well formed and of considerable size; in 

 the next stage (Fig. 27) it is larger, and the centroplasm, the margin of 

 which has become more crenate, is more extensive. A section through 

 the pole at right angles to the axis of a spindle in which the chromo- 

 somes are in a condition similar to those of Figure 27, shows the cen- 

 triole in the form of a dumb-bell, the two bodies being not only close 

 together, but still nnited to each other (Fig. 27 a). 



If an attempt is made to correlate the changes in the centrosome with 

 those in the chromosomes, it becomes evident that the relations of the 

 two are not constant. The centrosome in Figure 28 (Plate 5), for 

 example, seems to have failed to develop as rapidly as usual. For while 

 the centriole usually divides in the early anaphase of chromosome meta- 

 morphosis, in Figure 28 the centriole granules have not yet fused into 

 a single body, — a change which must take place before the centriole 

 divides. A number of similar cases have been observed, from which it 

 may be concluded that the progressive changes in the centrosome and in 

 the chromosomes take place more or less independently. However, we 

 are able to find a connected series of changes in the centrosome which 

 may be correlated in general with the changes in the chromosomes, and 

 for this reason the series illustrated in Figures 19-29 (Plates 4, 5) is 

 arranged in accordance with the known changes in the chromatin. 



