296 MR, W. E. COLLINGE OK THE f^pr. 2, 



Vn. SUMMAET. 



The more important features set forth in the foregoiDg pages 

 may be summarized as follows : — 



(a) The sensory canal system in the Physostomous Teleostei 

 is widely removed from that present in the Elasmobranchi, but 

 shows many affinities to that in the Ganoidei, e.g. the manner in 

 which the canals and branches are protected and the dermal 

 character of the lateral canal, in many forms. 



(b) The species of the different families differ largely from one 

 another, and in some particulars either agree ^^•ith more specialized 

 or generalized fishes ; e. g. in Sahno the main canal of the head and 

 the supra- and sub-orbital branches &c. are protected by a series 

 of small drainpipe-like canal-bones not unlike those described and 

 figured in Polyodon and Acipenser (3. p. 524), while in Coiw/er and 

 partly in the Siluridce the branching is not unlike that which 

 obtains in many Anacanthinous forms. 



(c) The cluster-pores and primitive pores are in the majority of 

 cases few in number, as are also the sensory organs in the canals 

 and branches. 



(d) In the Siluroids the results obtained support the view that 

 the sensory canal system in this family is one which once was 

 much more complicated in its nature and has since degenerated. 



The facts which favour such a view are : — 



a. The fewness and smallness of the cluster-pores. 



b. The exceedingly minute primitive pores. 



c. The dimensions of the canals and branches. 



d. The many indications, in various bones, of pre-existing 



canals. 



(e) Excepting in Callichihys and Esox all the branches on the 

 head pass off from the main canal. 



(/) The presence of grooves in various stages of development 

 in Pimelodus affords an instance of the manner in which the canals 

 are formed and further how in their earlier condition they resemble 

 that common to the adults of more generalized fishes. 



(f/) The succession of canals in Salmo is exceedingly interesting, 

 showing as it does a series of functionless canals passing through 

 the substance of the bone and which once probably communicated 

 with the surface by a series of pores. By a later anastomosis of 

 these pores and a growth of small drainpipe-Uke ossifications a 

 second and more superficial series of canals arose, which gradually 

 replaced altogether those traversing the deeply-seated cranial 

 elements. 



(h) Wherever the branching of the main canal and branches is 

 great, and the sensory organs in the same more numerous, there is 

 a corresponding increase in the branching of the trigeminal group 

 of nerves. 



(i) The nerve-supply is fairly constant throughout the suborder, 

 but there is some slight interchangeability, which is set forth in the 

 following table : — 



