_Experiment No. 3._--On February 13, 1906, a rabbit weighing 992 grams was fed with a concentrated aqueous extract of 500 grams of the fresh _Astragalus mollissimus_, collected in September and preserved in chloroform water. Before feeding, the rabbit's ears were warm and the rabbit struggled when any attempt was made to turn him on his back. The temperature at 10.50 a.m., the time of feeding, was 103.5°F.; at 11.15 a.m., 102.5°F. At 11.30 a.m. the rabbit was breathing very rapidly and would stay on his back for some time if placed so. The temperature at this time was 102.6°F. Both pupils, the one exposed to the light and the one protected, were contracted. At 12.02 p.m. convulsive movements of the legs appeared. The rabbit made one leap, the temperature rose to 103.6°F., and after a few convulsive movements of the limbs the anus relaxed and a small stool appeared, the pupils dilated, and the animal died at 12.06 p.m. _Experiment No. 4._--The feeding of the extract of 464 grams induced a fall in temperature of 2.4°F. in three hours, and the rabbit died several hours later (at night). _Experiment No. 5._--March 2, 1906, a rabbit weighing 928 grams was fed with a concentrated extract of 500 grams of the fresh seeds and pods of _Astragalus mollissimus_, made in September, 1905, and preserved with chloroform water. This animal died in one hour and seven minutes. The animal showed the usual post-mortem conditions. It was thus found that the aqueous extract of 500 grams of the fresh _Astragalus mollissimus_ would cause death in about one hour in rabbits weighing about 2 pounds (907 grams), these rabbits showing constant clinical symptoms--urination, paralysis, more or less convulsive muscular twitchings, often terminating in general convulsions, drowsiness, and stupor, with more or less anesthesia. The pupils at the time of death were often unequal. At first there was usually a slight rise in temperature, but this was soon succeeded by a fall. Often there were soft stools. The post-mortem lesions in these cases were marked congestion, with hemorrhages in the stomach walls and a secretion of thick mucus. The portions of the stomach walls most affected were the dependent portions near the cardiac end. The intestines showed dilatation of the blood vessels. The mesenteric vessels and also the vessels in the cerebral portions of the dura were markedly dilated; in some cases there were clots, especially at the posterior portion of the brain, between the cerebrum and the cerebellum. At times there were clots over the dorsal portion of the cord. On cutting into the brain the brain substance itself did not appear to be congested. The cord seemed about normal, but the vessels of its membranes were well marked. The other organs showed nothing characteristic macroscopically. These experiments were repeated many times and found to be constant.