404 Edward Phelps AUis jr., 



know of, being- the tubular ossicle that encloses the post-squamosal 

 section of the main infraorbital canal in the Muraenidae; which 

 ossicle is apparently a latero-sensory, unit without related sense organ. 

 But, however this may be, it is the unit of the latero-sensory system, 

 as above defined, that alone can be used in any attempt to establish, 

 by means of this system, the homologies of the cranial bones; and so 

 used, the course and position of the canals, and above all the number, 

 position and innervation of the sense organs in those canals, is cer- 

 tainly of great value. 



That the latero-sensory canals, in addition to their value for the 

 determination of the homologies of the bones they traverse, are so 

 constant in position as to even be of classificatory value, was Garman's 

 [39] opinion, and R. G. Harrison's recent work [45] would certainly 

 tend to confirm this. 



Ganoidei Holostei. 



Amia calva. The latero-sensory canals of Amia may be said to 

 present a complete and typical teleostean arrangement. 



The supraorbital canal of this fish develops, as I have shown 

 [9], in two somewhat separate portions; an anterior portion, containing 

 three organs that later become enclosed in the nasal bone, and a 

 posterior portion, containing four organs that later become enclosed 

 in the frontal. Posterior to the posterior one of these seven organs 

 there is a posterior continuation of the supraorbital sensory line that 

 never becomes enclosed in bone, remaining always as a line of sur- 

 face organs to which I gave the name of the anterior head line of 

 pit organs. The supraorbital sensory line was, in the youngest spe- 

 cimens that I examined, wholly separate from and independent of the 

 main infraorbital line, and the supraorbital canal, when first enclosed, 

 was also wholly independent of the main infraorbital canal. Later 

 the supraorbital canal anastomoses, by its penultimate primary tube, 

 with that primary tube of the main infraorbital canal that leaves 

 the latter canal between its postfrontal and squamosal sections. This 

 maimer or place of anastomosis, it may here be stated, is not con- 



