276 



from some common early ancestor. It might then be expected that 

 they would be found in Amia and Polypterus. 



In Amia, there is, in fact, a maxillary valve, in its proper teleostean 

 place , between the premaxillary and vomerine teeth, but there is no 

 slightest external indication of a mandibular valve. In Polypterus neither 

 valve is found as such. There are, however, in this latter fish, two flat, 

 toothed, crescentic surfaces that lie in a horizontal position between 

 the anterior portions of the rami of the upper and lower jaws, and 

 that are directly opposed to each other when the mouth is shut, thus 

 having exactly the positions and relations of the maxillary and mandi- 

 bular breathing valves of Teleosts. One of these surfaces, in Poly- 

 pterus, is formed by the so-called vomer bones of the fish, and the 

 other by the presplenial ossicles of Huxley's and van Wijhe's (No. 47, 

 p. 254) descriptions. Of these presplenial bones van Wijhe says that 

 they have the same loose spongy composition as the vomers, and that, 

 like the latter bones, they are furnished with small sharp teeth. Of 

 the so-called vomer bones Traquair says, that they are loosely bound 

 to the overlying cartilage of the floor of the nasal chamber, and so 

 tightly bound to the maxillary and ectopterygoid, that each bone "shares, 

 with its fellow, all the movements of the facial apparatus on the cra- 

 nium". The vomer, in fact, rests upon and is rigidly bound to the 

 ventral surface of the palatal plate of the maxillary, extending nearly 

 to the hind edge of the plate, where it abuts against, and is firmly 

 but not rigidly bound to, the anterior end of the ectopterygoid, slightly 

 overlapping ventrally that end of the latter bone. The anterior half of 

 the vomer lies superficial, that is ventral, to the palatine plate of the 

 premaxillary, but it is only loosely connected with that bone, and so 

 moves independently of it. The curved external edge of the bone lies 

 concentric with and slightly internal to the inner surfaces of the tooth- 

 bearing parts of the maxillary and premaxillary, and is loosely con- 

 nected with them by the mucous tissues of the roof of the mouth. 

 Between this edge of the vomer and the mesial surface of the maxillary, 

 the ventral surface of the palatal process of the latter bone is slightly 

 grooved. Along the full length of the internal edge of the vomer a 

 strong fold of mucous tissue turns outward, that is laterally or forward, 

 as the case may be, dorsal to the bone, between it and the overlying 

 cartilage of the chondrocranium. The furrow thus formed, although 

 shallow, is relatively important, and, in my opinion, there is hardly 

 a doubt that the mucous fold it separates from the rest of the dorsal 

 surface of the mouth is a part of the maxillary breathing valve of the fish. 

 The vomer bones are then simply ossifications of this valve. A similar 



