268 bulletin: museum of comparative zoology. 



by focusing, there being no sharp or definite limit to the area. The 

 outline of the centrosome does not become regular until the chromosomes 

 reach the metaphase. The general appearance of this early centrosome 

 in Haminea is similar to that shown by Conklin (:01, Diagram A, B) 

 for Crepidula ; he, however, shows only four bodies surrounding a clear 

 area. But the further history of these bodies in Haminea does not 

 agree with the changes that take place in the corresponding stages in 

 Crepidula. By following the successive stages shown in Figures 4-6 

 and 9-13 (Plates 1-3), it will be seen that the spherical masses com- 

 posing the centrosome become smaller and more regular in outline ; the 

 clear area diminishes in size until it disappears, and the five bodies 

 meantime become a small, deeply staining, homogeneous centrosome. 

 These changes result in the formation of a body which in its reaction 

 to stains, its density, and its further development shows that it is prob- 

 ably homogeneous. While it is possible that this body has been formed 

 by the differentiation of one of these five bodies from the others, it is 

 more probable that it has resulted from a fusion and condensation of all 

 of the bodies into one. 



A polar view of the centrosome (Plate 2, Fig. 5) shows these bodies 

 at their maximum size ; whereas an oblique section (Fig. 6) usually shows 

 only part of the bodies. Their reaction to stain is another noticeable 

 characteristic. In Figure 4, the earliest stage observed, they take a very 

 dark stain in iron-hsematoxylin. In the next two stages (Figs. 9, 10) 

 the stain is not as dark ; this may be due in part to the smaller size of 

 the bodies, but it seems as if the bodies were changing chemically and 

 therefore reacted differently as they passed through these changes. 



It is difficult to decide with certainty how these bodies become con- 

 verted into a small, homogeneous granule, like a typical centrosome. In 

 sections of the ovotestis of animals that have just begun to lay, one often 

 finds in a single section from fifty to one hundred eggs in approximately 

 the same stage of development ; so the difficulty has not been so much 

 want of material as inability to interpret satisfactorily the conditions pre- 

 sented. An attempt was made to ascertain whether any of these bodies 

 pass off in toto into the cytoplasm, but there is no satisfactory evidence of 

 this. Owing to the large number of stages studied, it was possible to draw 

 an inference as to what was taking place. It seemed evident that these 

 bodies were becoming less clearly defined, as well as smaller in size. 

 There is no evidence, however, that they break up and then pass into 

 the cytoplasm ; nor is there anything to show that in Haminea (as in 

 Unio, Lillie, :0l) one of these bodies becomes centrally located and 



