212 W. G. VanNarne — Embryology of Eustylochus. 



that time they lie in the cytoplasm near the nuclear membrane, which 

 is usually more or less indented, and at a little later stage begins to 

 disappear at these points. They lie widely separated but not necessa- 

 rily at opposite poles of the germinal vesicle (Figs. 2, 3a and 3J), and 

 are not connected by a central spindle. The similarity of this mode 

 of appearance to that found by Coe (1) in the nemertean Cerehratulus 

 and by Griffin (12) in Thalassema is very great. The observations 

 of Coe are of especial importance, for the preparations on which his 

 work is based have surely never been excelled in beauty and clear- 

 ness of detail, if indeed ihey have been equalled. 



When the centrosomes, which stain black with the iron-hsema- 

 toxylin, first become distinguishable, they are surrounded by a more 

 or less distinct centrosphere, which stains with cytoplasmic stains 

 and runs out into the few short aster rays developed at this early 

 stage. It may be of some significance that there is no distinct 

 line of demarcation between the rays and centrosphere, although in 

 later stages this ceases to be the case. Moreover in this eai'ly stage, 

 the rays seem to have but little connection with the cytoplasmic 

 reticulum, but as they become longer they branch and grade imper- 

 ceptibly into the reticulum with which they are continuous (Figs. 4 

 and 5), as Wilson (31) has so clearly shown in Toxopneustes. 



These phenomena suggest that the rays are at first outgrowths of 

 the centrosphere, but grow by the continuous addition and rearrange- 

 ment of the microsomes of the reticulum. The branching which 

 later becomes noticeable is largely determined by the position of the 

 yolk globules. This would indicate a difference of origin of the 

 basal and peripheral portions of the aster rays. The indications 

 are, however, that the centrosphere itself is differentiated from the 

 cytoplasm surrounding the centrosome by the influence of the latter, 

 making the ultimate origin of both portions of the aster rays the 

 same. 



The term centrosphere, as I use it, is equivalent to the " centrosom '' 

 of Boveri and to the " zone medullaire de la sphere attractive " of 

 Van Beneden and to the " couche medullaire (de la sphere attrac- 

 tive) " of Van der Stricht. Such a structure as the " couche " or 

 " zone corticale de la sphere attractive " of Van Beneden, which is 

 also described by Van der Stricht in Thysanozoon and by Fran- 

 cotte, I do not find in well preserved eggs, and I find no evidence of 

 any differentiated body external to and surrounding the centrosphere 

 (Van der Stricht's "couche medullaire"), except the aster rays 

 extending out into the cytoplasm. In poorly preserved eggs I have 



