The apparatus as finally developed is based upon the following principle: First, the air leaving the chamber is delivered into a copper box or wind chest, provided with three circular openings. Two of these openings have diameters exactly alike. The other may be adjusted to size, either by an iris diaphragm or still more accurately with a series of metal disks with circular openings of definite, known size. The air leaving this wind chest escapes either into the room or into a vessel in which it may be collected. Since the free discharge from this wind chest is a prerequisite, some device must be attached to insure that the air passing through the two small openings will be discharged against atmospheric pressure. Furthermore, as the air leaving these small openings is to be used for sampling, the sampling cans into which the air is delivered should be provided with some means for removal of the air as rapidly as it is delivered, so as to insure atmospheric pressure in these cans.

Extended experience has shown that, as used with the universal respiration apparatus, the standard size soda-lime and Williams bottles are extraordinarily efficient in removing from an air-current both water vapor and carbon dioxide for long periods of time. These may be used for a ventilation as high as 100 liters per minute, as is done in the universal respiration apparatus when applied to muscular work experiments, but they are best used with a ventilation of approximately 45 to 50 liters per minute. Hence the orifice leading into the sampling cans was made of such size (10 mm. in diameter) as to allow the escape of approximately 45 to 50 liters of air per minute through it. Thus, by connecting to the sampling can a suction pipe leading to a positive blower run by an electric motor, air at the rate of 45 liters per minute could be withdrawn, passed through suitable purifiers, and the carbon dioxide in it collected in a soda-lime bottle, as is done with the universal respiration apparatus. The amount of carbon dioxide thus collected obviously represents the total amount in the sample.

1 Pettenkofer and Voit, Ann. d. Chem. u. Pharm., 1862, Supp. Bd. II, pp. 1 and 52.

Another factor in determining the size of the orifice leading into the sampling can and the total amount of air to be taken care of by the purifying device was that with these bottles, although the combined weight of one soda-lime and one sulphuric-acid bottle is approximately 5,500 grams, it is necessary, to minimize errors in weighing, that not less than 2.5 to 3 grams of carbon dioxide should be absorbed. With a ventilation of 45 liters per minute and the residual carbon dioxide 0.5 per cent, the amount absorbed would correspond to approximately 9 grams of carbon dioxide in a 20-minute period. Even if the ventilation were so adjusted as to have the carbon-dioxide content but 0.25 per cent, the amount absorbed would be approximately 4.5 grams for a period as short as 20 minutes. It is thus possible to use short periods with the respiration apparatus, provided the production of carbon dioxide is sufficient for the carbon-dioxide content of the residual auto be at least 0.25 per cent. Obviously, if the total ventilation of the group chamber could be reduced to but 45 liters per minute, the universal respiration apparatus itself would suffice for the absorption and measurement of the carbon dioxide in the air, and we would have here nothing but a magnified form of the clinical respiration chamber developed and in use in this laboratory.1

But since the use of our large respiration chamber, with a volume of 44,000 liters, involves a much greater production of carbon dioxide than the clinical respiration chamber, with a volume of but 600 liters, a ventilation of 45 liters per minute would be inadequate for the removal of the larger amount of carbon dioxide, because the ventilation of a respiration chamber is dependent upon two factors (1) the amount of carbon dioxide produced and (2) the percentage of residual carbon dioxide desired.

The main problems, therefore, in the aliquoting and analysis of this air-current, are, first, to secure perfect aliquoting, and, second, to insure constancy in both the amount and pressure conditions of the air discharged; finally, since the outdoor air entering the respiration chamber contains a definite percentage of carbon dioxide, the calculation of the total amount involves a knowledge of the volume of air entering the chamber.

1 Benedict and Tompkins, Boston Med. Surg. Journ., 1016, 174, pp. 857, 898, and 939.