Improving Jancso's vital demonstration of Salvarsan, normal neo-tanvarsan (a Japanese neosalvarsan) was injected into a rabbit at the rate of 0.1 g per kilo of its weight, by the four following methods, and the viscera-deposit and the hours of the excretion of salvarsan were investigated. 1. Method with concentrated solution. Normal neo-tanvarsan was injected into the vein of a rabbit; the quantity used being 0.1 g dissolved in 5 cc Aq. dist. per kilo of the weight of the animal. 2. Method with dilute solution. The same quantity of normal neo-tanvarsan dissolved in 50 cc of 0.4% NaCl solution was injected into a vein in the same conditions as in method 1. 3. Method with active serum. The same dose of normal neo tanvarsan dissolved in 5 cc of the serum of a rabbit was injected into a vein in conditions otherwise identical with those of Methods 1. & 2. 4. Method with inactive serum. The serum of a rabbit was rendered inactive by subjection to a constant temperature of 60°C during the period of an hour. In this, normal neo-tanvarsan was disselved at the rate of 0.1g to 5 cc of serum per kilo of the weight of the animal, and injected into a vein. Air embolism reselted in the deaths of the rabbits treated in the abeve ways after periods of, 1, 10, 20, 30, 40, and 70 hours respectively. All the organs were then extracted, and sections made on the freezing microtome. Examination for salvarsan in the cells of the organs was made by the following method:- (A) Fixation in 10% Formalin for three days. (B) Making sections on the freezing microtome. (C) Bathing for 1,5 hours in a mixture of 1 part, 3% silver nitrate, to which had been added sufficient ammonia to destroy the brown turbidity of the liquid; and 1 part pure glycerine. In this developer, the salvarsan and derivates present in the tissues became brown-black in colour. (D) Cleansing for one minute in H(2)O. (E) Bathing in 1% thiosulphate of Soda (Na(2)S(2)O(3)) for 10 minutes. (F) Cleansing in H(2)O. By the above process, the following results were demonstrated. (1) In the case of the dilute injection, there is an even dispersion of salvarsan throughout the organs. In the casc of the injection of the concentrated solution salvarsan was found to be remarkably evident after 10 to 20 hours, as compared with the other methods. In the case of the inactive serum injection, the precipitation of salvarsan was least, viz. the affinity of the organs was least. (2) Thus, in these injections, salvarsan was taken in the Kupffercells of the liver, sinus cells and endothelial cells of the spleen, endothelial cells of the suprarenal gland, endothelial cells of the lymphatic gland, endothelial cells and reticular cells of bone marrow and histiocytare cells of interstitiar intermediate tissues of other organs (lungs, heart, testiculae, salivary glands, small intestine, thymus gland) and histioegtere cells of milk spots of the omentum and some parenchym cells (liver cells of liver and the epithelial cells of the kidney). In short, salvarsan is taken in endoreticular cells and some parenchym cells. (3) Finally, histological observation showed little difference in the injury to tissues, resulting from any these four methods of salvarsan injection. All the injections produced oft degeneration to the liver cells of the liver and epithelial cells of the kidney and congestion of each organ.