Osteolepis, Phaneropleuron, &c.  To this group is also to be ascribed the huge Onychodus (fig. 102, d and e), with its large, rounded, overlapping scales, an inch in diameter, and its powerful pointed teeth.  It is to be remembered, however, that some of these “Fringe-finned” Ganoids are probably referable to the small but singular group of the “Mud-fishes” (Dipnoi), represented at the present day by the singular Lepidosiren of South America and Africa, and the Ceratodus of the rivers of Queensland.

[Illustration:  Fig. 105.—­A, Polypterus, a recent Ganoid fish; B, Osteolepis, a Devonian Ganoid; a a, Pectoral fins, showing the fin-rays arranged round a central lobe.]

[Illusration:  Fig. 106.—­Holoptychius nobilissimus, restored.  Old Red Sandstone, Scotland.  A, Scale of the same.]

Leaving the Ganoid fishes, it still remains to be noticed that the Devonian deposits have yielded the remains of a number of fishes more or less closely allied to the existing Sharks, Rays, and Chimoeroe (the Elasmobranchii).  The majority of the forms here alluded to are allied not to the true Sharks and Dog-fishes, but to the more peaceable “Port Jackson Sharks,” with their blunt teeth, adapted for crushing the shells of Molluscs.  The collective name of “Cestracionts” is applied to these; and we have evidence of their past existence in the Devonian seas both by their teeth, and by the defensive spines which were implanted in front of a greater or less number of the fins.  These are bony spines, often variously grooved, serrated, or ornamented, with hollow bases, implanted in the integument, and capable of being erected or depressed at will.  Many of these “fin-spines” have been preserved to us in the fossil condition, and the Devonian rocks have yielded examples belonging to many genera.  As some of the true Sharks and Dog-fishes, some of the Ganoids, and even some Bony Fishes, possess similar defences, it is often a matter of some uncertainty to what group a given spine is to be referred.  One of these spines, belonging to the genus Machoeracanthus, from the Devonian rocks of America, has been figured in a previous illustration (fig. 102, f).

In conclusion, a very few words may be said as to the validity of the Devonian series as an independent system of rocks, preserving in its successive strata the record of an independent system of life.  Some high authorities have been inclined to the view that the Devonian formation has in nature no actual existence, but that it is made up partly of beds which should be referred to the summit of the Upper Silurian, and partly of beds which properly belong to the base of the Carboniferous.  This view seems to have been arrived at in consequence of a too exclusive study of the Devonian series of the British Isles, where the physical succession is not wholly clear, and where there is a striking discrepancy between the organic