176 WILLIAM WALLACE. 



the effect of formic-form ol-alcoliol, the second illustrates 

 the action of Perenyi's fluid. Such fluids as Perenyi's, 

 Gilson's mercuro-nitric; and bichromate-osmic-nitric (Allen's) 

 are therefore not suited to the study of the yolk nucleus. 



The body in the ova of fishes which, following Cunning- 

 ham (1897) and others, I take to be the centrosphere, is also 

 indicated in the figs. 7, 8, 12, 21. It is a dense, round body, 

 and, at a certain stage in the growth of the egg, is found in 

 contact with the germinal vesicle (fig. 7), and then looks and 

 stains very like a nucleolus. By this time the true yolk 

 nucleus has attained some development. Puring the growth 

 of the egg the " centrosphere," as we may provisionally call 

 it, moves away from the germinal vesicle towards the surface 

 of the egg. In Zoarces it degenerates in the zone of oil 

 globules long before true yolk spheres appear. In the plaice 

 it is last seen as a body, " recalling the form of an octopus" 

 (Cunningham), with amoeboid margin directed outwardly, in 

 the zone of developing yolk spheres. In Cymatogaster it 

 is said to remain over during the segmentation of the egg 

 (Hubbard, Eigenmann). Stained with Mann's methyl blue- 

 eosin the centrosphere is seen to be composed of two 

 substances — a sponge-work staining purple, and contained in 

 the meshes of this numerous granules, which stain red with 

 eosin like the nucleoli. 



As to the occurrence of the true yolk nucleus in teleostean 

 fishes, I have observed it in the eggs of Zoarces, Sygna- 

 thus, Sal mo, and Zeus, but I have not found it in the 

 eggs of Pleuronectes platessa and P. limanda, 

 although the latter were fixed in Bles's fluid and picro- 

 formalin, just those reagents which were used to demonstrate 

 the existence of this body in the other species. The "cen- 

 trosphere," on the other hand, occurs in the ova of all 

 species examined. It is most easily demonstrated in Pleu- 

 ronectes, in which it is not obscured by a yolk nucleus. 



Since the breaking up of the yolk nucleus is closely asso- 

 ciated in time and space (see figs. 12, 37) with the first 

 appearance of oil in the cytoplasm, and since the yolk 



