While the calculation of the several areas of the three discharge openings from the wind chest gives an approximate estimate of the relative amount of air that will be discharged, on the fundamental assumption that the air thus expelled will be directly proportional to the area of a cross-section of the orifice, nevertheless in any accurate use of the apparatus, such a gross assumption can not go unchallenged. It accordingly became necessary from the earliest development of this apparatus to check quantitatively the relationship between the two 10-mm. openings and the large opening with its various reductions. For this purpose a small auxiliary chamber (approximately 1 cubic meter in size) was connected with the pipe leading to the air-impeller (a, fig. 8) on the sampling apparatus. Into this chamber a known amount of carbon dioxide was admitted from a steel cylinder of liquefied carbon dioxide. These cylinders held approximately 2.5 kilograms each. The gas was found to be of an extraordinarily high degree of purity, with practically no appreciable correction for the slight amount of dissolved air. By weighing this cylinder of carbon dioxide on a balance capable of recording 1 centigram, 100 to 400 grams could be admitted as rapidly or as slowly as desired. Repeated tests with the subsidiary chamber showed that the aliquoting device functioned perfectly, regardless of whether the carbon dioxide was rushing into the chamber as rapidly as possible or entering very slowly over a long period.

By knowing (1) the amount of carbon dioxide admitted to the chamber as shown by the difference in weight of the cylinder at the beginning and end of the test, (2) the amount of carbon dioxide absorbed in the soda-lime bottle, and (3) the volume of air passing through the gas meter, the ratio could be computed between the 10-mm. openings and the particular opening used for the large discharge from the wind chest. For experimental purposes several disks with orifices of varying size were prepared. These disks were made of brass, were 1.5 mm. thick, and 110 mm. in diameter. They had a rim of sheet brass 40 mm. wide, soldered into the edge, thus forming a cup which could be dropped into the annular space filled with water surrounding the large opening of the wind chest. Disks were made with orifices of approximately 60, 40, 29, 22, and 16 mm., respectively. The true diameters of these disks, as found from a series of exact caliperings, are as follows: 16-mm. disk, diameter 15.92 mm.; 22-mm. disk, diameter 22.03 mm.; 29-mm. disk, diameter 28.75 mm.; 40-mm. disk, diameter 40.00 mm.; 60-mm. disk, diameter 59.52 mm. A careful calipering of the main discharge opening showed an average internal diameter of 97.00 mm. The true diameter of the 10-mm. openings was 10.11 mm. In all subsequent references to these various openings, they are given the number representing the nearest millimeter. It so happens that in the work with the diet squads we had occasion to use only the 60 and 40 mm. disks.

The tests with the subsidiary chamber showed that a definite proportion of the carbon dioxide admitted was invariably recovered in the soda-lime bottle, this amount being constant for the individual disk used in the main discharge opening, but varying with the size of the disk. Furthermore, it was found that the amount of carbon dioxide collected in both sets of absorbers was invariably the same, proving that the discharge of air through the two 10-mm. openings was identical.

With the subsidiary chamber various other tests were made to study the influence of both slightly decreased and slightly increased pressure on the intake side of the main blower. Furthermore, since variations in line voltage are to be expected with consequent change in shaft speed, it was important to note whether such changes affected the aliquot or not. Changes in line voltage and shaft speed of the blower were therefore studied, but the aliquoting device of the wind chest with its three openings showed invariably the same proportion of air discharged through the individual openings.

All of the preliminary tests were carried out with the subsidiary chamber. After the tests were completed and the sampling apparatus attached to the group respiration chamber, quantitative tests were made by introducing liquefied carbon dioxide into the larger chamber. With the subsidiary chamber the residual air could be considered as atmospheric, inasmuch as but a few seconds were required to ventilate the chamber completely and sweep out any carbon dioxide admitted. But in the tests with the group chamber it was necessary to make residual analyses of the air inside the chamber at the beginning and end of each test, as with the disks 60 mm. or less in diameter, a long time would be required to sweep out completely the carbon dioxide admitted and reduce the percentage of carbon dioxide in the air to that of outside air.

These tests have been repeated many times throughout the two years that the apparatus was in the process of development and a standard factor found for each size of disk to indicate the percentage of the total carbon dioxide withdrawn from the chamber which was collected in the soda-lime bottles, due correction being made for the carbon dioxide in the ingoing air. The ratio for the individual disks was invariably constant, irrespective of the chamber used, barometric pressure, speed of admitting carbon dioxide, and shaft speed of the rotary blower and consequent pressure inside the wind chest; at least, the ratios were constant within the ranges of pressure change possible with the rotary air-impeller, for we deal here with actual pressures corresponding to no more than those equivalent to a few millimeters in water pressure.

The ratios found for the several disks are as follows: 16-mm. disk, 21.21 per cent of air passing through absorbers; 22-mm. disk, 14.37 per cent; 29-mm. disk, 10.00 per cent; 40-mm. disk, 5.60 per cent; 60-mm. disk, 2.54 per cent. The ratio for the main discharge opening (97 mm.) was 1.14 per cent.

The carbon-dioxide tests with the subsidiary chamber were primarily made to establish the several factors. They were then in a sense reversed in that the factors thus established were employed in determining the known amounts of carbon - dioxide admitted into the large respiration chamber; the agreement of these checks is all that could be desired.