tiles, lost some of their special characters, becoming more remote from the mammalian type. The pelvis became weaker and more open. The shoulder-girdle lost parts in some orders, and gained some expansion in others. The vertebræ became more perfectly articulated by the bodies, and in the case of the snakes by the arches also. Finally, the ribs lost one of their points of articulation with the vertebræ, and the jaws became looser and more open, and especially adapted to swallowing large bodies whole. The history consists of a successive departure from the mammalian type, and a running into a specialization, which, in some cases, means degeneration.
A curious specialization which supervened on the reptilan type is that of the birds. Various saurians exhibit unmistakable approximations to the birds. The land-saurians include types that walked on the hind legs and had many bird-like characters of the hinder feet and of the pelvis. The flying saurians present affinities in the same direction. The class of birds presents many perfections both general and special. Their brains are larger than those of reptiles, and they acquired warm blood. In their own specialty of flight they display wonderful power, while the highest orders add that vocal skill which makes them so pleasing to man. Here is a good example of advance in evolution. Cases of degenerate evolution are to be found in birds, but they are few.
The next lowest class, that of the Batrachians, prevailed during the coal-measure period. They expanded enormously during the Permian, and were worthy contemporaries in size and numbers of the theromorph reptiles. Their numbers diminished subsequently, as the record now stands, though some of the species maintained their bulk during the Triassic period. In modern times they are comparatively insignificant; frogs, toads, salamanders, sirens, and cœcilias not playing an important part in the existing fauna. In tracing the successive changes of structure of these creatures, one is forced to believe that degeneracy has played an important part. The bones of the skull have so diminished in number as to leave it in some cases in a condition comparable to that of the primitive fishes. In not a few modern types the metamorphosis is never completed, the animals remaining permanently breathers in the water. Whatever we may call such changes, they are plainly a specialization which has carried them further and further away from their starting-point; and, as in the case of the reptiles, this starting-point has been near to orders higher than itself. The Permian Batrachia are nearer in structure to the Permian reptiles (Theromorpha) than any subsequent form of Batrachia has been (Fig. 3).
Professor Agassiz pointed out that the early fishes presented relations to other vertebrata, as I have since shown to be true of the Batrachia and Reptilia. Some of the primitive fishes he called "sauroid" or "reptilian" fishes. Batrachian fishes would have been a more accu-
