FOSSIL FISHES. 19 



lents ; but the Chelonian plastron consists of nine bones, while there are 

 only hve in that of DiniohtJiys. X)f these live the anterior three corres- 

 pond more closely in position with the anterior three of the turtle's ven- 

 tral shield, and have better claims to be considered their equivalents than 

 have the second pair to be regarded as the homologues of the second pair, 

 in the shield of the turtle. The hinder pair of plates in the plastron of 

 Dinichthys are much more free and independent than the second pair of 

 the turtle's shield, and have much less the character of dermal scutes, and 

 more that of internal bones. Still farther the posterior pair of the tur- 

 tle's plates (Xiphoplastrons) are, so far as we know, entirely absent from. 

 the plastron of Dinichthys. It is perhaps possible that the as yet unlo- 

 cated plates of the under side of the body of Dinichthys, described in the 

 notes on D. Terrelli, may have been so associated with those of the plas- 

 tron as to give the ventral armor more similarity to that of the turtles 

 than it now seems to have, but in the present state of our knowledge the 

 differences seem to be not only great but radical. 



3d. The characteristic gular plates of some Amphibians {Archmgosaurus, 

 etc.) have been referred to as offering some similarity to the anterior 

 three bones of the plastron of Dinichthys. 



These Amphibian throat-plates consist of a rhomboidal median one, 

 with a pair somewhat triangular in outline, converging forward and 

 united with the median plate by its antero-lateral margins. Von Meyer 

 considers these plates as homologous with the anterior plates of the plas- 

 tron of the turtles, while Owen compares them with the jugular plates of 

 Megalichthys and Sudis. Ey Prof. Huxley they are regarded as clavi- 

 cles and an interclavicle. Taken by themselves, the anterior three bones 

 of the plastron of Dinichthys are not very unlike, in form and position, the 

 gular plates of Amphibians, but we must know more of the plates which 

 protected the throat of Dinichthys, before we can make the compari- 

 son satisfactorily. Possibly the homologues of the Amphibian gular 

 plates, if any existed in Dinichthys, v^ere placed quite anterior to the 

 plastron. However that maybe, the posterior pair of plates of Z>?"w«^cA- 

 thys are without any representatives in the shield of Archmgosaunos ; a 

 difference so important as to throw doubt over any suggestion of hom- 

 ology. 



4th. Prof. Huxley, in his admirable memoir, " On the Classification of 

 Devonian Fishes," * compares the sternal shield of Goccosteus with those 

 of some existing Siluroid fi.shes — Clarias, Loricaria, etc. — and finds such 

 correspondence in these parts, as well as in the cranial plates and dorsal 

 armor, that'he suggests a-genetic relationship between the ancient Placo- 



* Memoir of the G-eological Survey of the United Kingdom, Decade X. 



