64 Journal of Comparative Neurology. 



purpose than chromic acid. Subhmate might be even better, 

 but has not been employed. (Plate IX, Figs. 9-1 1.) 



In the buffalo-fish and black-horse (American Cyprinoids) 

 the nidulus is very small and elongated rather than circular, but 

 the essential structure is similar, though the fibre-connections are 

 much more obvious by reason of the fact that there is less suf- 

 fusion and less tendency to aggregation of the cells. 



In the channel cat fish the characteristic appearance is lost 

 and the elements appear in a transparent mass with all distinct- 

 ness. There is much less definite aggregation of the small cells, 

 while the larger ones are among the most distinctive switch cells 

 yet seen. In Lucioperca the sections made by me in a Berlin 

 laboratory, were less perfectly fixed because of imperfect appli- 

 ances and here, for a time, I was tempted to believe in an actual 

 gelatinous glomerule, such as has been described. But sufficient 

 care demonstrates in these sections the same fusion of elements 

 exactly like those surrounding the spots. These cells are larger 

 and more uniform in size than in the drum and for this reason 

 fewer enter any single glomerule. All transitions were observed. 



In Blicca (a European C'yprinoid), on the other hand, the 

 suffused splotches are absent, although the nidulus is enormously 

 developed. Instead of these glomerules we have beautiful illus- 

 trations of the rosette clusters of fusiform cells which the writer 

 has so frequently called attention to as indicating regions of cell 

 multiplication. The entire body retains an embryonic simplicity 

 and is bounded on most sides by a dense zone of radially placed 

 cells of the same sort as those within. The arrangement is pre- 

 cisely such as we have observed in the embryo of Belonc. That 

 there is no mistake in the identification of the body is proven by 

 the relations of the commissura transversa, which passes through 

 the midst without obvious cellular connection, as in all other 

 cases, the same course being observed in the case of Lucioperca, 

 where the glomerulary structure is so evident. In the black-horse 

 [Cyclepfus) there are the most beautiful examples of this clustered 

 arrangement. In this case also the organ is very large but re- 

 tains an embryonic character. But instead of a spherical mass 

 with a clothing of densely arranged radial cells, the coating is 

 convoluted, forming a body very like the olives. The position 

 of the nidulus is indicated in Fig. 9, Plate XX, of Vol. I, by 



