effects of this or that condition, we have performed many of our experiments on sleeping animals, having beforehand convinced ourselves by frequent repetition that sleep exercises no restraining influence on the working of the gastric glands. When we recall to mind the failure of our attempts to obtain a secretion of gastric juice by any stimulation whatever of the buccal mucous membrane, and at the same time see how constant and intense the action of this psychic impression is, we are forced to the inevitable conclusion that in our sham feeding experiment the whole secretory effect is due to the psychic stimulus, that is to say, to the keen desire on the part of the animal for food and the satisfaction of enjoying it. In view of the importance of the act of eating, which even now is apparent, but which will become still more obvious when the succeeding periods of secretion are investigated, we have spared neither time nor trouble to arrive at a correct explanation of the mechanism of this factor. We have, therefore, taken in hand a number of modifications of the sham feeding experiment, and these investigations have confirmed the opinion at which we had arrived. If, for instance, the dog has been prepared by a long fast of two to three days, a very intense secretion of gastric juice will always be obtained by the sham feeding experiment, no matter what may be given it to eat, whether boiled or raw flesh, bread or coagulated egg-white, etc. The dog, however, which has not fasted, that is to say has been fed fifteen to twenty hours before, will pick and choose amongst the different foods, eating one with great greed, tolerating another, and refusing altogether a third, and, corresponding therewith, the amount and quality of the gastric juice will manifest wide variations. The more eagerly the dog eats the more juice will be secreted and the greater the digestive power which it possesses. The majority of dogs prefer flesh to bread, and correspondingly less juice will be produced by sham feeding with bread than with flesh. Sometimes, however, we find dogs which will devour bread with greater appetite than flesh. In these cases one obtains more and stronger juice in sham feeding with bread than with flesh. Here is a case in point: a dog is given boiled meat which has been cut into pieces of definite size, and the pieces follow each other at regular intervals of time. The animal eats, but soon, from its behaviour, you see that it develops no particular greed for the meal, and this observation is confirmed by the fact that after fifteen to twenty minutes it ceases taking the flesh. The secretion of juice has meanwhile either not begun at all, or only after a longer interval than five minutes, and remains scanty to the end. Now wait till the secretion has stopped and give the same dog raw flesh, either forthwith or next day, in pieces of the same size and at the same rate as before. The raw meat tastes excellently to the dog; it eats for hours