386 J. M. D. OLMSTED 
found extending up from the base of the barbel through a dis- 
tance of not more than 2mm. ‘These taste buds were evidently 
recently regenerated, for their appearance was different from that 
of old buds. They were smaller, the nuclei of the sense cells 
stained more heavily, and the sense cells themselves were shorter 
and more crowded together—characteristics which belong to 
newly formed buds. ‘The nerve at the base of the barbel in this 
region was decidedly more fibrillar and less granular than in the 
portions above, so that regeneration of the nerve had commenced. 
A twenty-day preparation also showed well-formed taste buds 
at the base of the barbel. Here, too, the buds were small, com- 
pact, and heavily staining. They were found at a distance of 
3 mm. from the cut end of the barbel, which in this case was not 
cut so close to the head as in the eighteen-day preparation just 
described. Regeneration had, therefore, in all probability, pro- 
ceeded to a distance of 5 mm. from the true base of the barbel. 
At‘any rate, well-developed taste buds were found at a higher 
level on the twenty-day barbel than on the eighteen-day one. 
Hence regeneration is a progressive process, and probably follows 
(or may be simultaneous with) the regeneration of the nerve. 
The boundary between the region containing perfect taste buds 
and that with only epidermal cells could be readily made out in 
tangential sections. In the part containing taste buds the circles 
of epithelial cells enclosed numerous fine dots, the cross-sections 
of the ‘sense hairs’ (fig. 17), while in the more distal regions the 
circles enclosed only three to five large epidermal cells. The 
change between the two regions was quite abrupt. In a zone 
extending for a short distance above the most distal taste buds, 
there was active mitosis in the germinative layer at the tips of 
the dermal papillae, the first stage in the regeneration of taste 
buds at these points. 
To test out the question of the return of taste buds after 
regeneration of the nerve, five fish were chosen at random from 
some sixty, each of which had had the nerves to the dorsal and 
lateral barbels cut. These were separated from the others, fed, 
and kept in good condition for two months. The second day after 
the operation any one of these four barbels on any fish could be 
