MUSEUM OF COMPARATIVE ZOOLOGY. 99 



fish possessing a villous layer, than to ignore the evident constancy be- 

 tween villi and granulosa cells, and to assume an extensive multipli- 

 cation of the latter in eggs during the period of their growth from a 

 diameter of .7 mm. to .75 mm. 1 



The latest paper dealing with the capsular membrane of Perca is that 

 of Owsjannikow ('85, pp. 3-8), who reaffirms Miiller's claim that the 

 " tubules " are hollow structures, and corroborates Ransom's discovery 

 that their inner ends are divided into branches which penetrate the 

 pore-canals of the zona radiata. 



On gold or silver preparations of ovarian eggs, one finds the granu- 

 losa cells l bounded by broad lines of a precipitate, so that there must 

 be present a large amount of intercellular substance. The cells them- 

 selves lie, as previous authors have shown, over the beginnings of the 

 corkscrew-shaped canals. These " beginnings," in the fully developed 

 egg, are not at all cells, and have no nuclei ; they are little funnels, with 

 the narrow end sunk as it were into the jelly. The finely granular 

 substance lies more compactly in the bottoms of the funnels; it is 

 scantier on the margin, and in many places extends beyond the rim, 

 of the funnel. This tissue (granulosa cells) often presents the appear- 

 ance of stellate cells joined together by numerous processes, and sepa- 

 rated from one another by abundant intercellular substance. The 

 more closely packed molecules at the bottom of the funnel had given 

 occasion, he says, to the assumption that there was a nucleus there. 



1 In his account of the Graafian follicle, not always readily harmonized with his 

 figures, and sometimes obscure, he claims the presence of a greater number of epi- 

 thelial (or endothelial, for he recognizes no difference between the two) cell layers 

 than have usually been admitted. Thus, if I rightly understand the explanation 

 of Fig. 4, Taf. I., there are in Perca, e. g., two epithelial cell layers between the 

 vascular layer and the capsular membrane, — an outer layer of flat endothelium and 

 a deeper layer of cylindrical granulosa cells; but in the text (pp. 4-8) he speaks of 

 only two cell layers, an inner and an outer granulosa (!), which are separated by 

 the vascular layer. From his description of the latter as the source of the new 

 eggs, there can be no doubt that it is the " germinal epithelium " of authors. I 

 can reconcile this apparent contradiction between text and figures only on the 

 assumption that Figure 4 and its explanation belong to a period in Owsjannikow's 

 studies when he was not as yet convinced of the error of taking " the granular accu- 

 mulation in the bottom of the funnels of the 'tortuous canals ' for a cell nucleus." 

 That he ultimately supposed that to be an error appears from his description, and 

 the statement (p. 5), "Die am Grunde des Kelches dichter an einander gelagerten 

 Moleceln gaben die Veranlassung dort einen Kern anzunehmen." 



In the case of Osmerus, moreover, he recognizes the existence of two layers of 

 large flat endothelial cells in addition to the granulosa. (Compare his explanation 

 of Fig. 8, Taf. I.) 



