186 JOURNAL OF COMPARATIVE NEUROLOGY. 
those of previous investigators, except that the spinal ganglia 
regenerated only in the most proximal part of the new-formed 
cord. Scosso (29) (1891) obtained positive results on frog 
larvae and tritons although the spinal cords of other vertebrates 
gave no evidence of regeneration. 
The cases of the adult. salamander and triton are very 
similar. Caporaso (6) (1887), working with the latter animal, 
found that after cutting off part of the tail the spinal cord 
degenerated cephalad for a short distance, the cells lining the 
central canal remaining intact. These ependyma cells then 
proliferated, restoring first the degenerated area cephalad to the 
injury, and then extending backward to restore the part lost by 
the amputation of the tail. Within the new formed mass 
ganglion cells were differentiated. Scosso (29) and BARFURTH 
(2) have also reported positive results on the triton. MULLER 
(26) (1864) observed in the regenerating tail of the salamander 
a new cord, which resembled in appearance the lost part, but 
which did not establish functional connection with the new tail. 
_ The results of work on the frog are much less conclusive. 
Mastus and VANLAIR (24) (1869), after removing 2 mm. of the 
spinal cord of a frog, found that the cavity was subsequently 
filled with tissue containing what they took to be nerve cells and 
nerve fibers. Later Masius (23) (1880) reasserted his former 
conclusions, butas neither Scopso (29) (1891) nor MaRINESCO 
(22) (1894) have been able to confirm his observations, it seems 
doubtful whether the spinal cord of the adult frog possesses 
even this degree of regenerative capacity. 
In brief, all amphibian larvae and the adult triton can 
perfectly restore a lost portion of the spinal cord; the sala- 
mander accomplishes this only incompletely; and it is an open 
question whether the frog can do so at all. 
Spinal Cord of Reptiles. 
Lizards have been studied by MU ier (26) (1864) and 
FRalIssE (12) (1885), each of whom found the lost portion of 
the spinal cord partially reformed, but not so completely as to 
be of functional value. On the other hand, Scosso (29) (1891) 
