In the upper part, the radius rotates on itself; while below, it rotates around the ulna. It follows that the forearm, which in all quadrupeds is in a state of permanent pronation, can, in carnivora, take the position of supination, or rather, of demi-supination. In fact, whatever be the mobility of the two bones of the forearm, the movement is not able to bring the palmar surface to the front, but only to direct it towards the median line. =The Articulation of the Wrist.=--Here are found, as in man, three superimposed articulations: the radio-carpal, intercarpal, and carpo-metacarpal. If we remember the movements which take place at the plane of these articulations in man, and take account of the fact that the mobility of the limbs is reduced just in proportion as they are simplified in structure so as to become organs of support only, we can easily comprehend that, in the horse and the ox, and, in a word, animals that have a canon bone, the movements of the wrist are little varied in character, while in carnivoræ, on the other hand, they are relatively more so. We will remember that in the ox and the horse the region of the wrist is called the _knee_. In flexion, the hand is bent backwards; in extension it is carried forwards. These two movements take place especially in the radiocarpal and intercarpal articulations. In the first of these articulations, it is the superior row of the carpus which glides backwards and forwards on the corresponding articular surface of the forearm. In the second articulation, it is the second row which moves; gliding on the inferior articular surfaces of the row above it. This inferior row carries the metacarpus with it; for the carpo-metacarpal articulation is much less mobile than either of the other two. In flexion, the articular surfaces are separated from one another in front; and the changes of form which result from this are noticeable on the anterior surface of the 'knee.' Moreover, at that moment this region contrasts markedly in its outlines with the parts above it and below it--that is to say, with the corresponding surfaces of the forearm and of the canon bone. As for the lateral movements, by which the hand is inclined outwards and inwards in its movements at the wrist, they exist to an appreciable extent in the cat and the dog only; in order to understand this, it is enough to compare the shape of the articular surfaces of this region in carnivora and the horse, for example. In the latter, those surfaces are almost plane; in the cat, on the contrary, they are curved (inferior surface of the forearm, concave; superior border of the carpus, convex). These latter, then, are, in form, similar to those which exist at the same level in the human being; this explains the possibility of analogous movements of the whole hand--that is to say, of the movements of abduction and adduction.