stock started from some point in the ungulates. I have manifested my repugnance to supposing such an origin, which appears monstrous. But while this repugnance is legitimate when we have regard to recent species—the extremities evolved from the branch—it is less so when we go back to the trunk before the specialization of the ungulates had become as pronounced as it is now. It must be said that nothing is impossible in nature; that things the least probable when we look at the outcome are realized by the most unforeseen processes, the most tortuous roads. What selection by the hand of man has done for pigeons is found done in nature by means the laws and mechanism of which are invisible to us, and which we can only denominate chance.
There is one objection to the descent of man from the ape which I have entertained, and which comes to the support of M. Vogt's thesis. As I have previously said, the primordial type of the mammalia has four limbs, the destination of which is already written out as far back as we can go; all four fitted to walking, but the fore-limbs adapted besides to serve as organs of prehension, while the hinder ones are essentially organs of support and walking. This double specialization goes back to the reptiles, not to speak of the dinosaurians, with which it is very marked. Some amphibians present traces of it. With the most ancient mammalia which are known in all their parts, like the Phenacodus primævus of the Lower Eocene of Wyoming Territory, the fore-limb is well characterized as an organ of prehension and the hinder one as for walking. The humerus is articulated with a narrow glenoid cavity at the upper outer angle of the omoplate, so as to permit the freest motions in different directions; the radius is mobile on the cubitus, around which it performs the turning movement required by the function of the hand; the five fingers are spread out, the thumb is turned more on its axis as if to permit opposition, and the hand continues in a straight line with the forearm. On the other hand, the femur is united, as with us, to a massive pelvis; the articular surfaces of the knee, the knee-pan, and the two immovable bones of the leg, are just what the exclusive function of locomotion requires; the foot is plantigrade, with a prominent heel and close toes, and is articulated perpendicularly by its arch with the leg, as in man. With another contemporary animal of the same bed, the Coryphodon, of which I have only representations of the foot and hand, but those whole, to judge by, these two organs present more resemblance, the foot being a little like a hand, but the differentiation is nevertheless made.
This specialization or differentiation has reached its maximum in man, no other animal showing it in equal degree. In the bird the upper limb has become a wing, that is, a function of locomo-
