120 ME. F. J. COLE ON THE STRUCTURE AND MORPHOLOGY OF 



been found necessary to study details, and to include in the investigation whole 

 subjects not contemplated in the beginning." The more important additions are the 

 sections on the lateralis accessorius (which I trust is here correctly understood for the 

 fii'st time), the phytogeny of the lateral canals, the relations of the latter to the auditory 

 organ, the metamerism of the lateral sense organs, and the morphology of the " facial " 

 ganglion and Jacobson's anastomosis. Hence the somewhat comprehensive title of the 

 present communication. 



C. Sensory Canals in General. 



The sensory canals of the Cod differ in no essential respect from those of the typical 

 form. Indeed the uniformity that exists in this connection, even amongst fishes of 

 widelv different classification, is one of the most remarkable incidents in the history of 

 this remarkable system. The same fundamental plan may be seen in all, whether we 

 take an Elasmobranch, a Holocephalan, a Dipnoan, a "Ganoid," or a " Teleostean." 

 Even in the Palaeozoic Coccosteaus and Asterolepids, we see, thanks to the admirable and 

 laborious researches of Traquair, the same generalised type. In Coccosteus, for example, 

 we may consider the horizontal canal traversing the anterior dorso-lateral, ihe external 

 occipital, and the central plates, as the lateral or body canal. This, after a break, gives 

 rise to a supra-orbital canal, traversing the central and pre-orl)ital plates, and an infra- 

 orbital, situated on the central, post-orbital, and maxillary ossicles. Tliere is also a 

 supra-temporal commissure situated on the two centrals, and this probably marked the 

 anterior extremity of the lateral canal. The hyomandibular series is represented by 

 canals on the post-orbital, marginal, and maxillary, and perhaps by the other canals in 

 this region. In PtericJithi/s, again, the sensory canals are very simplified, being reduced 

 to representatives of the lateral and iafra-orbital canals, the former being connected 

 anteriorly on the median-occipital plate by the supra-temj)oral commissure. Here the 

 practical absence of a supra-orbital canal is perhaps correlated with the curious condition 

 of the orbits in the Asterolepid fishes. 



As I have elsewhere maintained (46, p. 181), and as the pioneer researches of Allis 

 must be held to have established, the sensory canal systems of fishes, both recent and 

 fossil, may be reduced to a common type. Pending the production of weighty evidence 

 to the contrary, this must be considered as one of the best-established conclusions 

 attained by the careful study of these organs by many able observers. This common 

 type may be summarised as follows : — 



(1) A supra-orbital canal passing, as its name implies, over the eye. Innervated by 

 the superficial ophthalmic lateral line nerve. Anteriorly may anastomose with the 

 infra-orbital canal or end blindly ; posteriorly anastomoses either Avith the infra-orbital 

 or lateral canal, but more often with the former. 



(2) An infra-orbital canal coursing underneath the eye. Innervated by the buccal 

 -l-otic lateral line nerves. Anteriorly may anastomose with the supra-orbital canal or 

 end blindly ; posteriorly anastomoses either with the lateral canal only or with both 

 the supra-orbital and lateral canals (usually). 



