154 Edward Phelps AUis jr., 



directly backward, and leaves the bone at its hind end, having tra- 

 versed it its full length, along its mesial edge. As it traverses the 

 nasal, it is not, for much the greater part of its length, entirely 

 enclosed in bone. The mesial edge of the nasal bone is simply 

 folded upward and laterally, so to speak, and a gutter, open laterally, 

 is thus formed in which the supraorbital canal lies. At three dif- 

 ferent points this guttei' is bridged over by a naiTow bridge of bone, 

 and at each of these three points there is a sense organ in the canal. 

 Between each two of these three organs there is a primary tube, two of 

 these tubes thus leaving the canal as it traverses the nasal. The 

 anterior tube arises immediately posterior and partly opposite to the 

 first supraorbital sense organ, and is tube No. 2 of the line. It runs 

 laterally and forward, as a simple tube, and opens, by a single pore, 

 on the outer end of a short tubular process, the pore lying dorso-postero- 

 mesial to pore No. 1 infraorbital and related to that pore in the 

 manner already fully described in describing the infraorbital canal. 



Tube No. 3 supraorbital arises from the canal but slightly poste- 

 rior to tube No. 2, and immediately posterior to and partly opposite 

 organ No. 2. It is a long pouch extending backward and laterally 

 nearly parallel to the main canal, and in all the specimens examined, 

 excepting one, it was a blind pouch, without external opening, and so 

 flat and delicate that it might easily be overlooked in dissection. In 

 the one specimen in which it opened on the outer surface of the head 

 by a pore, the pore was small and lay about half-way between the 

 second pore of the line and the posterior nasal aperture. 



Having left the nasal bone at its hind end, the supraorbital canal 

 curves gradually backward and mesially, and is here enclosed in a 

 fibrous tube which has become largely chondrifled. The cartilaginous 

 portion of the tube is attached posteriorly to that part of the skull 

 to which I have already referred as being very probably the frontal 

 part of a fronto-sphenoid bone, and it may be found either as two 

 processes directed forward and slightly laterally, or as a nearly closed 

 cylindrical tube. The ventro-lateral one of the two processes, where 

 there are two, or the ventro-lateral portion of the single process, 

 where there is but one, arises from the side wall of the skull im- 



