142 bulletin: museum of comparative zoology. 



section. In cyclostomes there are two grooves symmetrically placed on 

 either side of the median plane. As I have elsewhere shown (Sargent, 

 lOS"), it is a conspicuous structure in cyclostomes, selachians, and rep- 

 tiles, less prominent in teleosts, birds, and amphibia, and inconspicuous 

 in mammals. The specialized ependyma of the groove consists of char- 

 acteristic elongated ependymal cells, and is sharply marked off from the 

 usual ependyma lining other portions of the ventricle. In general it 

 acts as a support for the constituent elements of the fibre of Reissner? 

 which pass through it into the ventricle, and as an ' anchorage ' for the 

 fibre as a whole. This ependymal structure was, I believe, first ob- 

 served by Fulliquet ('86, p. 100, pi. 4, fig. 17). He briefly described 

 it as " a la pai'tie dorsale du ventricule un boquet de cellules beaucoup 

 plus grosses, fusiformes (cde.),'' It has since been noticed by Rabl- 

 Ruckhard ('87), Edinger ('92), Gage ('93), and Dendy (:02). 



4. Sense Organs. 



Groschuff ('97) has described the occurrence on the floor of the cen- 

 tral canal of groups of cells similar to the sensory buds seen in the 

 integument of many animals. In studying the central canal in a large 

 series of vertebrates, I have observed that sometimes the ependymal or 

 epithelial cells forming the floor of the canal are grouped together so as 

 to resemble such sensory buds. Such an apparent grouping of the cells 

 is particularly likely to occur at any angle in the floor or wall of the 

 canal. Tliis grouping I believe to be a purely accidental occurrence of 

 no functional significance. Occasionally in Amia an ependymal cell 

 may have a long, pointed process resembling somewhat a sensory hair, 

 projecting into the canal. This condition may, however, be the result 

 of forcing the cell into the lumen by the artificial pressure applied to 

 the outside of the cord in removing it from the animal. 



//. Contained Material. 



Within the lumen of the central canal and in the bi-ain ventricles I 

 find in the preparations that I have studied at least six distinguishable 

 materials. 



(1) The coagulated and shnmhen mass oniji?iating from, the cerebro- 

 spinal fluid. (See Sargent, :00, Figs. 3, 7, 9.) This fluid, according to 

 the analyses given by Hammarsten ('98), contains about one per cent of 

 solids. The coagulum may assume a great variety of forms and ap- 

 pearances under the action of various fixing and staining agents. It 

 may be diffuse, finely or coarsely granular (o^>. cit., Fig. 3), in clot-like 



