294 • MB. W. E. C0LLIN6E ON THE [Apr. 2, 



3. The Occipital Commisstteb passes off as an inward branch 

 immediately in front of the lateral canal, and almost opposite to 

 the operculo-mandibular branch (PI. XXI. figs. 16, 17). Like all 

 the other branches, it is enclosed in a series of more or less 

 isolated cartilages. It traverses the occipital region of the head 

 and forms a connection between the main canals of either side of 

 the head. 



The distribution and branching of the sensory canal system in 

 Conger conger resembles in a many ways that figured in Clarias. 

 The large saccular dilatations are perhaps the most peculiar feature 

 in connection with the form of the canal, etc., but branches occur 

 in Clarias in these several positions. 



This form of canal has not previously been described in the 

 Physostomi,Ibeheve,and it cannot be regarded as at all typical of 

 this suborder. Ley dig (6) was the first to draw attention to these 

 pecuhar cartilages enclosing the canals of this form, and in 1850 

 he figured and described them in Lota vulgaris. 



I shall have to refer to this account later when treating of the 

 suborder Anacanthini and to compare it with Hyrtl's account and 

 some investigations of my own. 



On PI. XXI. figs. 18, 19, I have figured the cartilage which 

 connects the occipital commissm-e with the large saccular dilatation, 

 from which the main canal of the head and the operculo-mandibular 

 branch pass off. Each cartilage forms a drainpipe-like structure, 

 averaging from 20 to 44 millim. in length, the walls of which 

 consist of a dense connective tissue, with numerous cartilage cells 

 scattered throughout. Towards one end and on one side of the 

 cartilage a series of white patches are present, which remind one 

 of the terminal cartilages on the bronchial tube of many vertebrates ; 

 these are denser patches of cartilage. On the dorsal surface and 

 towards the end not possessing these cartilaginous patches is a 

 small ossification. This was not found to be present on all the 

 cartilages. Each cartilage is connected with its fellow by a pro- 

 longation of the connective tissue, as shown in figure 18, these 

 overlapping the sides of the cartilage at the end, a, fig. 19. 



The presence here of a dense connective tissue, isolated chon- 

 drifications, and an ossification afford an interesting example of the 

 manner in which canal-bones may have originated. 



Innervation. 



The specialization in the form of the sensory canals, as has been 

 described in Conger, is accompanied by a greater development of 

 the cranial nerves and an increase in the branching. 



The nerves of the trigeminal group are all large branches, and, 

 excepting the ramus oticus, pass forwards. The branches which 

 innervate the canal system are : — 



1. The ramus ophthalmicus superficialis. 



2. The ramus buccalis. 



3. The ramus maxiUaris. 



4. The ramus oticus. 



