418 
This method of killing and preserving was adhered to for all the 
material employed in this investigation and proved entirely satisfactory. 
Apart from these two cases all the experiments went without a hitch. 
All of the remaining nine specimens revived very quickly and it was 
upon these that the observations here recorded were made. Later, 
when sections of the material were examined, it was ascertained that 
in no case had the cut failed to break the fibre. 
Other specimens were kept under observation under precisely 
similar conditions to serve for control. Some of these had not been 
in any way disturbed, other were anzesthetized exactly like those 
actually used for the experiments but were then returned uninjured 
to the tank excepting one in which a slight incision was made which 
penetrated the skin but was not allowed to damage the filum terminale. 
This specimen did not show the characteristic reaction of those spe- 
cimens in which the terminal filament and ReIssNErR’s fibre were in- 
jured and, indeed, behaved in no way differently from the other control 
specimens. 
In this way, it was hoped that it would be possible to eliminate 
any peculiarities in behaviour, which might be observed, which were 
not directly due to the operation but which might conceivably have 
resulted from change in the environment or other causes. 
VI. Observations upon the living Animals. 
Since, for the reasons already given, most of the experiments were 
performed upon skates, and also because in these animals the cha- 
racteristic reaction appeared more quickly and was more marked, I 
propose to deal first with the observations made upon these fish. 
The specimens were all apparently healthy and, so far as could 
be discovered, in no way abnormal, the selection being governed simply 
by considerations of size, or rather of length, which varied from eight 
to fifteen inches. 
When returned to the tank, immediately after the operation, they 
were almost always quite inert and sank gently to the bottom but, in 
most cases, they revived very quickly (within a few minutes). Their 
recovery was occasionally followed by a series of frantic dashes, during 
which they would blunder heavily into the sides and glass front of 
the tank, against which latter they would often remain swimming con- 
tinuously at the surface, vertically poised, for a relatively considerable 
period. The control specimens behaved, after speedy recovery from 
the anesthetic, in very similar fashion. This activity was usually 
followed in the subjects of the experiments by marked sluggishness. 
