ox THE ANATOMY OE PISHES. 
261 
any vertebral centrum. On this point, Ramsay Wright (44) remarks, “ the 
Siluroids do not afford the necessary material for determining the homology of the 
incus* on account of its reduced condition” (p. 109). The condition of the ossicle in 
at least some Characinidse is essentially similar, and this led Sagemehl (33) to regfard 
it as a free metamorphosed rib belonging to the second vertebra. We are able to 
show, however, that the Siluridae do afford the necessary evidence for determining the 
nature of the ossicle. Our discovery of an iniercalariuin in the Siluroids Macroncs, 
Liocassis, Pseudohagrus, &c,, which has an ascending process forming part of the wall 
of the neural canal, and a horizontal process which terminates distally in a nodule 
imbedded in the interossicular ligament, at once brings the Siluridm into conformity 
with the Cyprinidse, and establishes the identity of the ossicle in the former family. 
The correctness of this conclusion is confirmed by the fact that the ascending process 
of the ossicle in these Siluroids is situated between the paired foramina for the exit of 
the roots of the second and third spinal nerves from the neural canal. The condition 
of the ossicle, in the majority of Siluroids, is clearly due to degeneration from its more 
primitive condition in Macrones ; and from this point of view such genera as Ccdli- 
chrous and Cryptopterus, in which the ascending process has vanished, and the 
horizontal process and its distal expansion alone remain, must be regarded as 
supplying an intermediate stage between the two extremes. A further question 
arises as to whether, in addition to a modified neural arch, the intercalarium did not 
originally include an element comparable to a transverse process. We are inclined to 
think that it did, and that the horizontal process of the ossicle, when present, repre- 
sents the modified transverse process of the second vertebra. In its origin from the 
neural arch or ascending process, the hoi’izontal process conforms precisely to the 
contiguous transverse processes of the fourth and fifth vertebrm, which spring in 
exactly the same way from the neural arches, and not from the centra of their 
respective vertebrce. 
V. — The Physiology of the Air-Bladder and Weberian Mechanism 
IN THE SiLURIDyE. 
A. Silurid(E normcdes. 
From the morphology of the air-bladder and Weberian ossicles, we may next proceed 
to discuss how far the facts elucidated in the preceding sections of this paj)er throw 
any light on the vexed question of their ph 3 ’siology. 
There is a strong d priori probability that the Weberian mechanism is physio- 
logically related to one of the several functions that have been ascribed to the 
auditory organ or to the air-bladder, but to which of them is a (piestion by no means 
(= iutercaluriuui.) 
