4) Through the second hole of the rubber stopper closing the flask was passed a short piece of glass tubing bent obliquely just above the stopper and tightly connected with a rubber tube about 50 cm. long, DD. This tube was clamped near its lower end with a Mohr’s pinch-cock, while the upper was connected with the stem of a small funnel, F. When an experiment was to be conducted, the azotometer C, was connected to the tube, B, and tied tightly, a piece of good rubber tubing being used for the connection. The reservoir of the azotometer, M, was then raised until the mercury had driven out all the air in C, and the stop-cock closed. Next the funnel F was filled with water, the pinch-cock closing D opened and the water allowed to flow down and drive out all the air in the tube. The latter was then clamped and thus kept full of water. The apparatus being now ready, the flask A, containing the substances in the desired quantity, was made fast to B and D by means of its stopper. The stop-cock of C opened. The flask A was heated in a water-bath for any desired time, the oxygen being collected in C as fast as it was given off. At the end of the experiment, all the gases which remained in A and B were driven over into C by lowering the reservoir M and opening the pinch-cock at D. When this was accomplished the stop-cock of C was closed and the gases brought under atmospheric pressure by bringing the mercury in C and M to the same level. It will be seen that in this gas volume was included the air which A and B contained at the beginning of the experiment. The oxygen evolved from the contents of the flask was determined thus: After reading off the total volume of air and oxygen and reducing to normal conditions, the total volume of oxygen was obtained by absorption with phosphorus or pyrogallol, and from the residual nitrogen could be calculated from the volume of air to be deducted the total volume of gases. The remainder is the volume of oxygen sought.” Action of manganese dioxide on potassium permanganate in acid solution. Three sets of apparatus like the one described were used. The flask of apparatus No. I contained manganese dioxide and dilute nitric acid. The flask of apparatus No. II contained potassium permanganate and dilute nitric acid. The flask of apparatus No. III contained potassium permanganate and dilute nitric acid and manganese dioxide. The manganese dioxide used in these experiments was precipitated from a solution of potassium permanganate by a dilute solution of manganous sulfate in the manner described on page 34 and then washed and dried at 100°C. An illustration of the proportions in which the different substances were brought together in each of the three flasks is to be found in the following statement of the quantities used in the first experiment.