The many new discoveries of extinct forms so often intermediate, not only between the larger groups, but between many of the lesser ones as well, are making the classification of the vertebrates increasingly difficult. At one time it was sufficient to define a reptile as a cold-blooded animal with a single occipital condyle, that is, with a single articular surface between the skull and the first vertebra of the neck; a mammal as a warm-blooded animal with two articular surfaces; but these definitions are no longer strictly correct. Connecting links do not break down classification, as one might think, but they do often spoil our fine systems and compel our classifiers to take a wider view of nature than their own narrow province affords. We can never hope that most, or even the greater part, of all the animals which have lived in the past will ever become known to us, even imperfectly. Doubtless the species of the past geological ages outnumbered many times, perhaps hundreds of times, all those now living, since many of these latter are merely the remnants of far more varied and extensive faunas. At times the conditions for the preservation of the remains of animal life have been more favorable than at others, and, under such favorable conditions, a fairly good glimpse is sometimes given us of the fauna of some isolated epoch and locality in the earth’s history. Those animals which lived in and about the water have been preserved in greater numbers and more perfectly than the strictly land animals, since fossils are due to the preserving action of water, with few exceptions. Of those animals which lived upon the land or in the air only the rarest of accidents carried the skeletons into the lakes, seas, and oceans. And, even when they had been covered by sediments at the bottoms of lakes and seas and hidden away from adverse agencies, it has often happened that the great erosions of later ages have carried away and destroyed the rocks in which they were inclosed. The records of long intervals of time have thus been lost in all parts of the world. That we are able to obtain even an imperfectly continuous history is due to the fact that the intervals thus lost are not everywhere contemporaneous, that the missing records of one place may be filled out in part elsewhere. But this substitution of records from a distance can never make the history complete. If, in human history, we had only the records for one century in China, for another in England, and for yet another in South America, how imperfect indeed would be our knowledge of human progress. Animals and plants are never quite alike in remote regions, and they never have been. The living reptiles of North and South America are today almost entirely different, and, were their fossil remains to be discovered a million years hence, it would be very difficult to decide that they had once lived contemporaneously; difficult, though perhaps not impossible,