The loop which had thus been emptied and returned to moist warm surroundings was left 10 minutes and the normal secretion allowed to collect, and at the end of that period the loop was again emptied with as little exposure as possible and the fluid measured. The cannula was then clamped off and the secretion allowed to gather for a second 10 minutes. This again was drawn off and measured. The quantity secreted was usually fairly constant and quite small, the manipulation undoubtedly increasing it somewhat. When the normal secretion was obtained the purgative salt in an isotonic solution was administered either subcutaneously, intravenously, or locally. The secretion was again allowed to collect, and was drained off and measured after 10 minutes. This was repeated several times and the amounts compared with what was taken as the normal secretion. When the solution was applied locally, a method which is perhaps the most satisfactory, it was allowed to drop on the loops from a pipette, care being taken to have it at body temperature and as nearly as possible isosmotic with the blood. In each case special care was taken to have no interval between the emptying of the loop and the beginning of the succeeding period of 10 minutes. In other words, the loop was always entirely empty at the beginning of each period. The results of a few of these experiments may be seen in the following reports:-- 1. Rabbit. Loop 30 cm. long, upper part of small intestine. Loop contained in beginning[48] 5.0 c.c. Fluid removed after 1st 10 minutes 0.2 c.c. Fluid removed after 2d 10 minutes 0.5 c.c. 2 c.c. m/8 BaCl_{2}, injected subcutaneously. Fluid removed after 1st 10 minutes following injection 4.0 c.c. Fluid removed after 2d 10 minutes following injection 3.4 c.c. Fluid removed after 3d 10 minutes following injection 3.0 c.c. In this rabbit the increased secretion of fluid was accompanied by extremely active peristaltic movements. The faeces could be seen passing along the loops of the lower part of the intestine with great rapidity. Within 30 minutes after the administration of the salt, the passage of faeces to the outside began. This continued for some time, the faeces becoming constantly softer, until finally they were almost entirely unformed. As shown by this experiment, and also by the following ones, the action of barium chloride persists for a considerable length of time. The action of sodium citrate is more transitory. 2. Rabbit. Loop 25 cm. long. Loop contained in beginning 3.0 c.c. fluid deeply bile stained After 1st 10 minutes 1.0 c.c. fluid deeply bile stained After 2d 10 minutes 0.8 c.c. fluid somewhat lighter in color