674 E. PH. ALLIS, 



they lie superficial to the mandible, the relation of the entire hyo- 

 mandibular canal being markedly to the hyoid arch and not to the 

 mandibular one. This is exactly what the innervation of the canal 

 would lead one to expect, and it is clearly evident that the constant 

 relation of the canal to the mandible in teleosts must be a secondary 

 one. It is also evident that the dentary of Polyodon, being wholly 

 unrelated to the lateral canal and its enclosing bones, can not be the 

 exact homologue of the dentary of teleosts. The lateral canal com- 

 ponent of this latter bone seems, in fact, to be a hyoid element that 

 becomes secondarily grafted on a membrane bone developed in some 

 relation to Meckel's cartilage. 



The homologies of the several bones oi the dorsal surface of 

 the head of the fish can now be considered. 



Here it is evident, first of all, that Bridge's bone JB^, not being 

 in any way directly related to the supraorbital canal, can not alone 

 be the homologue of the frontal bone of other fishes. It may, however, 

 represent a part of that bone, if it be assumed that the bone is nor- 

 mally formed of two separate components of différent origin, one de- 

 veloped in relation to the supraorbital canal and the other not so 

 related, and that these two components are found separate and distinct 

 in Polyodon. The little spicules of bone that unite bone B^ and the 

 series of ossicles that form bone 3 certainly favour this assumption, 

 but it seems to me more probable that bone JB^ is a parietal, the 

 presence, in all my specimens, of a median opening between the bones 

 of opposite sides of the head, resembling a parietal foramen, being 

 decidedly in favour of this interpretation. The bones B^ then become 

 paired supraoccipitals, instead of parietals, and this is in full accord 

 with their relations to the suprascapulars and extrascapulars, which 

 relations would otherwise be unusual. 



The suprascapular, correctly identified by Bridge, has the normal 

 relations of that bone to the lateral canals, but its long anterior arm, 

 with the related descending lateral plate, are unusual. Bridge con- 

 siders this descending plate to represent the pedicle or leg of the 

 bone of teleosts. It might perhaps also contain the intercalar of those 

 fishes, that bone being the one with which the pedicle of the supra- 

 scapular normally articulates. It is here to be remembered that the 

 intercalar of all Uving fishes, so far as I know, though generally 

 homologized with the opisthotic of higher animals, is primarily a 

 purely dermal bone (2). It can not therefore be represented in the 

 purely primary ossification that Bridge describes as the opisthotic. 



