'442 Edward Phelps AUis jr., 



The lateral canal of the body begins immediately posterior to 

 the supratemporal commissure, and each organ of the line, so far back 

 as was examined, is enclosed in a short cylindrical tube or scale of 

 bone. The anterior one of these lateral -line scales or ossicles has, 

 as stated above, the position of a lateral extrascapular. 



The preopercular canal joins the main infraorbital directly 

 opposite the lateral end of the supratemporal canal, having, in this, 

 an unusual relation both to the latter canal and to the main 

 infraorbital. 



Anterior and middle head lines of pit organs are found, and also 

 an ethmoidal sensory line, represented in part by a line of pit organs 

 and in part by a short ethmoidal canal. This short ethmoidal canal 

 touches and coalesces, in the middle part of its length, with the 

 corresponding canal of the opposite side of the head, thus forming 

 a short median canal, but not an ethmoidal cross -commissure. 

 This is a distinctly plagiostomian feature, and has already been once 

 referred to. 



In Ophichthys, Myrus, and Muraena the arrangement of the 

 lateral canals does not differ in principle from that found in Conger. 

 In each of these fishes, however, the frontal section of the supra- 

 orbital canal instead of being enclosed in a flbro-cartilaginous tube, 

 is enclosed in what must be the frontal component of the completely 

 fused bones of this part of the skull. 



Gymnarchus niloticus. In this fish (fig. 29) the cranial latero- 

 sensory canals are relatively large and are provided with a limited 

 number, only, of relatively long, unbranching, and closed primary 

 tubes. No one of the canals or tubes of the head, so far as could be 

 determined either by surface observation or by injection, has any 

 communication with the exterior, and the lateral canal of the body 

 also seemed to have no such communication. As the canals of this 

 fish must certainly have been developed exactly as they are in Amia, 

 it follows that the primary tubes must once have opened on the outer 

 surface of the head, and then, for some reason, have all been second- 

 arily closed. The only recorded case that I know of that in any 

 way approaches this condition, is that described in Lota, in which 



