6o8 THE SCIENTIFIC MO.\TIILY
upon comparative anatomy and embryology as opposed to the triple evidence afforded by these wiences when reenforced by paleontology.
It is through the discovery of primitive types of the fringe-fiDned ganoids, to which Huxley gave the appropriate name CroBsopterygia, that the true ancestn- of the Amphibia and ot the amphibian limb ha*
��Ih^h traced, as due to a partial change of function whereby the pro- pelling fin was pra<lually tranjiforroed into the propelling limb. This implies a long tcrrestrio-aquatie phase in which the fin was partly used for propulsion on muddy surfaces (Fig. IT). In the parallel retrogres- sive evolution of the lung fishes {Leindosiren, Gymnolus) the fringe- finned fishes (Calamoichthi/n) and the bony fishes (Anguilla) the final eel-shaped, finless stage is either approached or actually passed.
The bony fishes (Teloosts), which first emerge as a distinct group in Jurassic time, radiate adaptively int« all the great body tjpes at- tained by the older group.s, more or less closely imitating each in turn, so that it is not easy to distinguish superficially between the armored (atfishes {Loririiria) of the existing South American waters and their prototypes [Cfiihalaspis) of the early Paleozoic. The most extreme specialization in this great group is to be found in the radiations of abyssal fishes into slow- and swift-moving forms inhabiting the great depths of the ocean, adapted to tons of water pressure, to temperatures
�� �
