326 REV. S. GRAHAM BIRKS 



Osteolepis itself, but also in Thursius, Diplopterus and 

 Glyptopomus, so that Megalichthys is unique among these 

 genera of its family in this respect, but agrees in it with 

 Rhizodopsis and the Holoptychiidae. Strangely enough 

 Gyroptychius, one of the Rhizodontidae, appears to have a 

 pineal foramen. 



In Megalichthys coccolepis the distinction drawn between the 

 mandibles of the Rhizodontidas and Osteolepidae is weakened, 

 if not completely broken down, by the fact that the angular, 

 and at least one infra-dentary, remain separate from the 

 dentary in this species. This is clearly seen in Plate XVI., 

 which illustrates a specimen not described elsewhere. 



In both Megalichthys and Rhizodopsis the first dorsal fin is 

 nearly opposite the pelvic fins, and the second dorsal is almost 

 opposite the anal ; but in Osteolepis the first dorsal fin is in 

 advance of the pelvic pair, and the second dorsal is opposed 

 to the space between the pelvics and the anal. The tail is 

 hetero-diphycercal in Megalichthys, heterocercal in Rhizodopsis, 

 and strongly heterocercal in Osteolepis. Ring vertebras occur 

 in all three genera. A comparison of figures of the skulls of 

 the genera from the dorsal aspect will illustrate very clearly 

 the great resemblance between the skulls of Megalichthys and 

 Rhizodopsis. 



It has been recently shewn that there is a fundamental 

 resemblance among the palates of the early Stegocephalian 

 Amphibia which also applies to that of Megalichthys, and that 

 the vomerine tusks of this genus are parallelled in the same 

 group. It may be added that the condyle at the base of the 

 skull of Megalichthys resembles that of Loxomma in its con- 

 cavity (PI. XV., fig. 2, and text-fig. i). 



When we turn to the dorsal surface of the skull of Megal- 

 ichthys we find certain obvious dissimilarities between it and a 

 generalised Stegocephalian type. The forward position of 

 the orbits of the fish and the absence of the parietal foramen 

 are noteworthy, the latter character indicating that Megalichthys 

 is certainly not the direct ancestor of the Stegocephalia. In the 



