In attempting to explain the cause of this increase in the number of leucocytes in the circulating blood, after the ingestion of proteid food, we are at once confronted with the possibility of this apparent increase being relative rather than absolute; as possibly due to a loss of water from the blood, incidental to the marked outpouring of digestive juices accompanying digestive proteolosis. For this view, however, there is very little support. In the first place, the blood would need to become very much thickened by loss of water to account for the increased number of leucocytes observed in the experiments. Furthermore, we know that the outpouring of watery secretions into the intestine during digestion is accompanied by an absorptive current in the opposite direction, which would tend to counteract any tendency towards concentration of the blood, and, indeed, might lead in many cases to a direct dilution of this fluid. Again, if the increase in the number of white corpuscles is to be explained in this manner, there should be a corresponding increase in the number of red blood corpuscles. As a matter of fact, Pohl’s results show that those conditions of diet tending to increase the number of white blood corpuscles are without any noticeable effect on the red corpuscles. Abstraction of water can not, therefore, be the cause of the large number of leucocytes contained in the blood after a proteid diet. Much more plausible is the view that the increase is due to a more rapid transference of the corpuscles from their point of origin, _viz._: from the intestine and from the lymph glands of the mesentery, to the blood. In other words, it seems probable that digestive proteolysis in the stomach and intestine is followed or accompanied by a rapid production of new cells in the lymph spaces surrounding the intestine. If this view is correct, there should be a much larger number of leucocytes in the blood and lymph flowing from the intestine, in an animal in full digestion, than in the arterial blood coming to the intestinal tract. That such is the case is shown by the following experiment taken from the many reported by Pohl: DOG WEIGHING 3,330 GRAMS, IN GOOD DIGESTION. A.M. 8.50. Blood from skin contained 8,330 white corpuscles in cubic mm. 9.10. Blood from skin contained 9,618 white corpuscles in cubic mm. 120 grams meat and 20 c. c. water fed. 10.40. Blood from the skin contained 15,092 white corpuscles in cubic mm. 10.55. Body opened, several loops made with the small intestine, and blood withdrawn from vein and artery without any great loss of blood. WHITE CORPUSCLES IN CUBIC MM. BLOOD. 1. Intestinal loop. 11.35 A. M., venous blood contained 17,077. 11.39 A. M., arterial blood contained 7,649. 2. Intestinal loop. 11.40 A. M., venous blood contained 15,033. 11.40 A. M., arterial blood contained 7,061.