It is necessary, then, that the oxygen cylinder be weighed. This is done after first closing the pet-cock on the end of the pipe conducting the gas beneath the floor of the calorimeter room, slipping the glass joint in the rubber pipe leading from the reduction valve to the pet-cock, and breaking the connections between the two rubber pipes, the one from the pet-cock and the other to the reduction valve, also breaking the electrical connection leading to the magnet on the cylinder. The cylinder is then ready to swing freely without any connections to either oxygen pipe or electrical wires. It is then weighed, the loss in weight being noted by removing the brass weights on the shelf attached to the counterpoise. It is important to see that there is a sufficient number of brass weights always on the shelf to allow for a maximum loss of weight of oxygen from the cylinder during a given period. Since the cylinders contain not far from 4 to 5 kilograms of oxygen, in balancing the cylinders at the start it is customary to place at least 4 kilograms of brass weights on the shelf and then adjust the counterpoise so as to allow for the gradual removal of these weights as the oxygen is withdrawn. As soon after the beginning of the period as possible, the U-tubes are weighed on the analytical balance, and if they have not gained too much they are connected ready for the next analysis. If they have already absorbed too much water or carbon dioxide, they are replaced by freshly filled tubes. Immediately at the end of the experimental period the barometer is carefully set and read, and the reading is verified by another assistant. Throughout the whole experiment an assistant counts the pulse of the subject frequently, by means of the stethoscope, and records the respiration rate by noting the lesser fluctuations of the tambour pointer on the smoked paper. These observations are recorded every few minutes in a book kept especially for this purpose. A most excellent preservation of the record of the minor muscular movements is obtained by dipping the smoked paper on the kymograph drum in a solution of resin and alcohol. The lesser movements on the paper indicate the respiration rate, but every minor muscular movement, such as moving the arm or shifting the body in any way, is shown by a large deflection of the pointer out of the regular zone of vibration. These records of the minor muscular activity are of great importance in interpreting the results of the chemical and physical determinations. FOOTNOTES: [5] W. O. Atwater and F. G. Benedict: A respiration calorimeter with appliances for the direct determination of oxygen. Carnegie Institution of Washington Publication No. 42, p. 91. (1905.) Francis G. Benedict: The influence of inanition on metabolism. Carnegie Institution of Washington Publication No. 77, p. 451. (1907.)