Herrick, Gustatory Paths in Fishes. 



429 



very thick, thorny and much branched, interlacing with the root 

 fibers to form a very dense neuropil in the most superficial lay- 

 er of the vagal lobe. 



Fig. ^2. Transverse section through the vagal lobe of Anieiurus nebulosus, 

 illustrating a secondary gustatory neurone similar to that shown in Fig. 30. 

 GoLGi method. X 60. 



As the figures show, they are very unlike the correspond- 

 ing cells of the vagal lobe of the larger cyprinoid fishes, whose 

 minute cell bodies are closely crowded around the periphery 

 with the bushy dendrites thrust directly inward and giving off 

 at the apex the neurite. These cells are exceedingly numerous 

 in the carp and relatively few in the cat fish, and this fact doubt- 

 less explains the difference in form. Probably neither type is 

 very near the primitive form. Johnston figures in his Petro- 

 myzon paper ('02, Figs. 7 and 21) chief cells of the vagal lobe of 

 Lampetra whose cell bodies lie close to the ventricle and give 

 off the neurite directly, while the busy dendrite is directed to- 

 ward the periphery. Houser ('oi) unfortunately secured no 

 impregnations of the chief gustatory cells of the shark, his prep- 



