676 E. PH. ALLIS, 



would then join the main infraorbital in a postorbital and not in a 

 postfrontal bone, which is certainly much more in accord with the 

 position of the hyomandibular canal in selachians, which canal seems 

 so evidently the homologue of the hyomandibular part of the canal 

 in Polyodon. It is also in full accord with the position of the anterior 

 end of the horizontal check line of pit-organs of Amia, which line so 

 probably represents the dorsal part of the hyomandibular line of 

 Polyodon (4). Moreover, the several cranial bones of Irolyodon here 

 under discussion would then conform closely, in their relations to 

 each other, with the corresponding bones in Clarias, as given by 

 Pollard (18), and also with those of Trematosaurus as given by 

 Feitsch (8); excepting only in the apparent absence of an extra- 

 scapular bone in both these latter animals, and in the intercalation 

 of a lachrymal, along the supraorbital canal of Trematosaurus^ between 

 the prefrontal and the nasal. Comparison with Clarias and Tremato- 

 saurus would also indicate that those lateral canal ossicles that span, 

 in Polyodon, the space between the bones B'^ and C^ must together 

 represent the prefrontal bone of the fish, the posterior ones alone re- 

 presenting the frontal; the prefrontal thus lying serially along the 

 supraorbital canal between the frontal and the nasal. In Archego- 

 saurus the supraorbital canal seems (8) even to be restricted to the 

 postfrontal and prefrontal, not traversing the frontal at all. However this 

 may be, it is evident that the so-called prefrontal of many descriptions 

 of teleosts is quite probably not the homologue of the similarly named 

 bone of amphibians, the latter bone being either wholly absent in 

 teleosts, being there represented in the posterior one of the two nasal 

 bones sometimes found on each side of the head, or being completely 

 fused with either the frontal or nasal. But, it is needless to say, 

 much more extended, and especially much more careful, investigation 

 of the lateral system must be made before definite conclusions, based 

 on it, can be arrived at. 



The homologies here last above proposed show, if correct, that 

 while Polyodon more closely approaches selachians in the arrangement 

 of its lateral canals than any known teleost or other ganoid, the bones 

 enclosing those canals more closely approach the arrangement found 

 in Amphil)ians than those of any other known living fish, excepting 

 Clarias, which latter fish, according to Huxley (12), closely ap- 

 proaches Coccosteus in its dermal armament. 



The conditions found in Polyodon, moreover, definitely establish 

 the fact that there is a definite lateral canal component in certain of 



