has, indeed, been shown that, when the nerve going to a salivary gland is stimulated, the gland shrinks, notwithstanding the great dilation of its bloodvessels. Under the influence of the stimulation the granules in the gland-cells imbibe water, swell up, and escape from the cells. The cells discharge their accumulated stores, in the first instance, more rapidly than they take up materials (even fluid) from the blood. For its knowledge (if the term may pass) of what is wanted the gland-cell is dependent upon messages which reach it through the nervous system. These messages take origin in the endings of the sensory nerves of the mouth, pass up to the brain, and are reflected down the nerves to the gland. So accurate is the information conveyed to the glands, that when a horse transfers the work of mastication from one side of its mouth to the other, as it is in the habit of doing about every quarter of an hour, the flow of saliva from the parotid gland on the masticating side is increased; on the other side it is diminished. Two or three times as much saliva is poured out on the one side as on the other. Not only is the amount of saliva poured out in response to stimulation proportional to the needs of mastication, but the kind of saliva is adapted to the nature of the food. In a dog—and this is an observation which can be made only on an animal which lives on a mixed diet—it is possible to determine the amount of the two kinds of saliva secreted and the relation of flow to food. When meat is given to the animal, the submaxillary gland yields its secretion; when it is fed on biscuit, abundance of the watery parotid saliva is poured forth. A mouthful of sand also causes the parotid saliva to flow, in order that the sand may be washed out of the mouth.