elsewhere.[5] Consequently, as a result of this experience, in planning the calorimeters for the Nutrition Laboratory it was decided to design them for special types of experiments. The first calorimeter to be constructed was one which would have general use in experiments during rest and, indeed, during experiments with the subject sitting quietly in the chair. [Illustration: FIG. 9.--Vertical cross-section of chair calorimeter from front to back, showing structural steel supporting the calorimeter and the large balance above for weighing the subject inside the calorimeter. The chair, method of suspension, and apparatus for raising and lowering are shown. Part of the heat-absorbers is shown, and their general direction. The ingoing and outgoing air-pipes and direction of ventilation are also indicated. The positions of the food-aperture and wire mat and asbestos support are seen. Surrounding the calorimeter are the asbestos outside and hair-felt lining.] It may well be asked why the first calorimeter was not constructed of such a type as to permit the subject assuming a position on a couch or sofa, such as is used by Zuntz and his collaborators in their research on the respiratory exchange, or the position of complete muscular rest introduced by Johansson and his associates. While the body positions maintained by Zuntz and Johansson may be the best positions for experiments of short duration, it was found, as a result of a large number of experiments, that subjects could be more comfortable and quiet for periods of from 6 to 8 hours by sitting, somewhat inclined, in a comfortable arm-chair, provided with a foot-rest. With this in mind the first calorimeter was constructed so as to hold an arm-chair with a foot-rest so adjusted that the air-space between the body of the subject and the walls of the chamber could be cut down to the minimum and thus increase the accuracy of the determination of oxygen. That the volume has been very materially reduced may be seen from the fact that the total volume of the first calorimeter to be described is less than 1,400 liters, or about one-third that of the Middletown apparatus. GENERAL CONSTRUCTION. A horizontal cross-section of the apparatus is shown in fig. 7, and a vertical cross-section facing the front is given in fig. 8. Other details of structural steel are seen in fig. 9. In constructing the new chambers, the earlier wood construction, with its tendency to warp and its general non-rigidity, was avoided by the use of structural steel, and hence in this calorimeter no use whatever is made of wood other than the wood of the chair.