Still we have not exhausted all the fossil remains of the Burdiehouse rocks. Mingled among the stems of ferns and lepidodendra, and the scattered valves of the cyprides, lie the scales, teeth, and bones, of several large ganoidal fishes, along with entire specimens of the smaller genera. The scales of holoptychius are especially abundant, often crowded together by dozens, and probably not far out of the arrangement they had when grouped on the body of the living animal. Detached teeth of the same fish also frequently occur along with disjointed internal bones. The remains of the contemporary megalichthys likewise abound, more particularly the scales, which have a fine nut-brown colour, and dot the surface of the rock with their bright glittering enamel. Several other smaller ganoids may be met with, especially a small and elegant species of Palæoniscus (_P. Robisoni_), and one of Eurynotus, a fish remarkable for the great size of its dorsal fin. Not uncommon, too, are the ichthyodoralites of a gigantic placoid--the _Gyracanthus formosus_--with all their delicately-fretted ornament and a peculiar crystalline glistening surface when broken across, whereby the smallest fragment can be easily distinguished from any other bone in the limestone. Such are the ichthyic remains of the Burdiehouse beds; what deductions can be legitimately drawn from them? As before, we must have recourse to the analogy of living nature. The existing ganoidal fishes chiefly inhabit lakes and rivers, especially near the confluence of the latter with the ocean. They feed on the decaying matter brought down from the land, or on the minute Crustacea that swarm upon the river-bottom. If, as seems probable, the ancient ganoids had habits similar to those of their present representatives, then the rocks wherein their remains occur abundantly may have originated on river-bottoms, and such may have been the case at Burdiehouse. So that here again we have corroborative evidence of the fluviatile origin of the limestone in question. But besides the remains of ganoidal fishes there occur the defensive spines of placoids. Now, the placoids are emphatically marine fishes, and the sole living representative of the most ancient genera of this order is the Port-Jackson shark, that haunts the seas round Australia. The ichthyodorulites of Burdiehouse, therefore, if we would apply analogy consistently, must be regarded as the relics of marine species. And this conclusion, too, will be found in entire harmony with those already obtained, for if we are right in assuming the Burdiehouse strata to have originated at a river-bottom, particularly near the sea, we may expect to find the remains of marine predaceous fishes imbedded in the sediment that gathered there, just as the teeth of the shark may be preserved among the mud forming in the upper reaches of many British estuaries, seeing that not a few instances are known where that fish