28 



BRITISH FOSSILS. 



belonging to either of these orders is at present known in the fossil 

 state ; that they are represented by a very small number of genera 

 and species in our existing Fauna ; finally, that the Pharyngo- 

 branchii, from their very nature, could hardly be preserved in a 

 recognizable state, even in such fine mud as that of the Oxford olsij, 

 or the Solenhofen slates ; and that of the Marsipobranchii nothing 

 but the horny teeth could be expected to escape destruction. 

 Lepidosiren, on the other hand, might have left as definite traces of 

 its existence as Diioterus, and hence its entire absence in the fossil 

 state is a negative fact of greater value. 



3. The Elasmobranchii abounded, teeth and spines testifying to 

 the numerous and diverse genera which haunted the Devonian seas. 

 It is more difficult to say to what sections of the order these genera 

 belonged, as the only Devonian Elasmobranch whose whole structure 

 can be restored with any certainty is Fleuracanthus, a fish which 

 belongs to a family distinct from any now living. 



4. The Ganoidei, as I have endeavoured to show above, are 

 largely represented by a suborder, the Crossopterygidge, which drops 

 into comparative insignificance in later ages. Of the existence of 

 Amiadse there is no evidence, and even if we include Tharsis, 

 Thvissops, and Leptolepis under this suborder, they are scanty in all 

 later formations ; but what is much more remarkable is the appa- 

 rent, entire, or almost entire, absence of the Lepidosteidse, a suborder 

 which obtains such a prodigious development in the Mesozoic epoch. 

 The nature of the Acanthodidse, and the question whether there is 

 any reason to suspect the existence of Chondrostei during the 

 Devonian epoch will be considered by-and-bye.* 



* The determinatiorx of the characters of the families of Lepidosteidsc and of the limits 

 of the suborder is a difficult problem, of which I hope to treat more fully hereafter. 

 One interesting fact results from my investigations, so far as they have hitherto gone, 

 viz., that Lepidosteus belongs to a totally distinct family from its Mesozoic allies, 

 whether "Sauroids" or "Lepidoids." The Pyenodonts and HoplopleuridfE do not appear to 

 me to belong to the Lepidosteidffi, and I doubt their being true Ganoids. Tor the present 

 I propose the following as a sketch of an arrangement of the Lepidosteidse. 



Lepidosteid^, 



Heterocercal Ganoids with rhomboidal scales ; branchiostegal rays ; non-lobate 

 paired fins ; a preoperculum and an interoperculum. 

 Fam. 1. Lepidosteini. 

 Maxilla divided into many pieces ; branchiostegal rays few and not enamelled. 

 Lepidosteus. 

 Fam. 2. Lepidotini. 



Maxilla in one piece ; branchiostegal rays many and enamelled ; the anterior ones 

 taking the form of broad plates. 



(a) CEchmodus, Tetragonolepis, Dapedius, Lepidotus, S^c. 



(b) Eugnathus, Pachycornms, Oxygnathus, Sfc. 



(c) Aspidorhynchus. 



Perhaps the genera marked a, b, c, should torm distinct sub-families. 



