The apparatus is divided into two functional parts, one for measuring the gaseous exchange, the other for measuring the heat production of the subject. A schematic presentation is here given (Fig. 1).

Fig. 1. - Schematic diagram of the Atwater-Rosa-Benedict respiration calorimeter.

The Gas Analysis

The inner lining of the apparatus presents an air-tight copper box having a capacity of 1123 liters. One end of the box, through which the patient lying on the bed is admitted, may be closed with a glass plate by means of wax. The air within the box is purified by drawing it out of an opening in the box through a rubber tube and forcing it by means of a rotary blower through a system of absorbers, whence it returns again to the box by another rubber tube. It passes (see diagram) first through sulphuric acid (1), which removes the water, then through moist soda lime (2), which removes the carbon dioxid, and next through sulphuric acid (3), which absorbs the moisture taken from the soda lime. If the bottles be previously weighed, the gain in weight of 1 represents water absorbed, and the gain in weight of 2 plus 3 equals the carbon dioxid absorbed. By this method the water and carbon dioxid produced by a man are taken from the air, while oxygen within the chamber is being absorbed by the man himself. This causes a diminution in the volume of the contents of the box. In order to replace the oxygen used, oxygen is automatically fed into the system from an oxygen cylinder which may be weighed before and after the period. The automatic feeding of oxygen into the box is accomplished by means of a spirometer whose interior is connected with the interior of the calorimeter chamber. As the volume of the air in the box decreases, the spirometer falls until a certain point is reached, at which an electric contact releases a clamp, which allows oxygen from the oxygen cylinder to enter the box, causing the spirometer to rise, break its electric contact, and clamp off the oxygen supply. So sensitive is the spirometer to the movement of the patient that a device called a "work adder" has been attached to it, which records the subject's movements.

1 Murlin and Lusk: "Jour, of Biological Chemistry," 1915, xxii, 17.

2 Benedict and Carpenter: Carnegie Institution of Washington, 1910, Publication 123.

3 Williams: "Jour, of Biological Chemistry," 1912, xii, 317.

4 Riche and Soderstrom: "Archives of Internal Medicine," 1915, xv, 805. 5 Lusk: "Archives of Internal Medicine," 1915, xv, 793.

Ventilating System:

Of, Oxygen introduced as consumed by subject.

3, H2SO4 to catch moisture given off by soda lime.

2, Soda lime to remove CO2.

1, H2 SO4 to remove moisture given off by patient.

Bl, Blower to keep air in circulation. Indirect Calorimetry:

Increase in weight of H2SO4 (1) = water elimination of subject.

Increase in weight of soda lime (2) + increase in weight of H2SO4 (3) = CO2 elimination.

Decrease in weight of oxygen tank = oxygen consumption of subject. Heat-absorbing System:

A, Thermometer to record temperature of ingoing water.

B, Thermometer to record temperature of outgoing water.

V, Vacuum jacket.

C, Tank for weighing water which has passed through calorimeter each hour.

W, Thermometer for measuring temperature of wall.

A1, Thermometer for measuring temperature of the air.

R, Rectal thermometer for measuring temperature of subject. Direct Calorimetry:

Average difference of A and B X liters of water + (gm. water eliminated X 0.586) = (change in temperature of wall X hydro-thermal equivalent of box) ± (change of temperature of body X hydrothermal equivalent of body) = total calories produced. Th, thermocouple; Cu, inner copper wall; Cu2, outer copper wall; E, F, dead air-spaces.

At the beginning of an hourly period of experimentation an observer at the table calls "time." At this instant the rotary blower is stopped, the air current switched so as to pass through a new set of weighed absorbers, and then the rotary blower is started again. At the word "time" an operator also turns a pet-cock which cuts off the respiratory chamber from the spirometer cylinder, which is then filled, always to a given point, with oxygen from the oxygen cylinder. The pet-cock is now opened and a freshly weighed oxygen cylinder is placed in the position of the other, which is removed. Repeating these procedures an hour later, one may determine by difference in weight the gain of water and carbon dioxid by the absorbers and the loss of oxygen by the cylinder. The figures are subject to corrections due to (1) gain or loss of water or carbon dioxid content in the box itself during the period, which gain or loss must be added to or subtracted from the increase in weight of the absorber system. This gain or loss of water and carbon dioxid in the box also affects the volume of the air in the box and, therefore, the quantity of oxygen admitted, as do, in addition (2), a change in temperature within the box and (3) a change in barometric pressure. These corrections must be made in order to determine whether oxygen is to be added or subtracted from the quantity which has been furnished from the oxygen cylinder. The result gives the quantity of oxygen which the man has absorbed. It is apparent that all the errors of determination fall on the oxygen, and yet the exactness of the method is witnessed by the close approximation in alcohol check experiments of the theoretic and actual values for oxygen consumed.

If a person in the calorimeter moves even the arm during the critical moments just before "time" is called, the increased local heating of the air may cause the spirometer to rise to a considerable height, of which the air thermometers inside the box fail to make compensatory record, and the oxygen determination will be too low in that hour and too high in the next.

Analysis of the air in the interior of the chamber is made just before the beginning of each hour by passing 10 liters of air from the box through three U tubes containing, respectively, sulphuric acid, soda lime, and sulphuric acid, then through a Bohr gas-meter, and back into the box again. This is called the "residual analysis".

Under the conditions present in the respiration apparatus carbon dioxid is measured with the greatest ease and accuracy. Oxygen is also measured with accuracy if the person within the box lies perfectly quiet for ten minutes before the end of the period, whereas water production is the least accurate of all the determinations on account of the varying hygroscopic condition of the walls, bedding, and other surfaces within the closed spaces of the apparatus.