24 0.04 1 0.06 2 1.564 2 25 0.04 ” 0.11 ” 1.491 ” 26 0.05 ” ? ” 1.512 ” 27 0.03 ” 0.05 ” 1.532 ” [6]Fifteen minute determinations 28 0.05 1 0.08 2 1.340 2 29 0.04 ” 0.05 ” 1.325 ” 30 0.06 ” 0.05 ” 1.335 ” 31 0.00 ” 0.00 ” 1.347 ” 32 0.03 ” 0.03 ” 1.412 ” 33 0.02 2 0.02 3 1.363 3 34 0.02 ” 0.06 ” 1.377 ” 35 0.00 ” 0.02 ” 1.363 ” [5] This marked the time when the oxides first settled, leaving a colorless supernatant liquid. [6] This marked the time when the solution was first decolorized. Inspection of the table will show that whether the oxygen were determined immediately after the subsidence of the oxides (i.e. after fifty minutes) or after three hours the results obtained are practically the same. The tendency of the potassium permanganate to lose strength made the frequent preparation of new samples advisable. Wherever a new sample has been used in the course of the experiments, the fact has been noted in the table by an asterisk. It will be noticed that where a new solution is used the figures for flask no. III are immediately larger and nearer to one and one half atoms of oxygen from each molecule of potassium permanganate. A phosphorus gas-pipette was used for the absorption of the oxygen in all the experiments the results of which are embodied in the foregoing table. In subsequent experiments an alkaline solution of pyrogallol was employed. It is now known that the variation in the composition of the manganese oxide in use had some influence upon the results.