106 VITALITY AND EFFICIENCY WITH RESTRICTED DIET. 



a slight increase in the air tension inside the wind chest and the escape 

 of air through the three orifices will be determined in large part by this 

 pressure, which is tested by a manometer attached to the petcock 

 (p), fig. 9, and is more or less roughly proportional to it. Tests have 

 shown that with orifices of the same standard size, the amount of air 

 delivered will vary du-ectly in all cases with the pressure, although it 

 may not necessarily be proportional to the pressure. It is of import- 

 ance, however, to secure a condition whereby a sample of air delivered 

 through a small orifice can be collected, withdrawn, and the carbon 

 dioxide absorbed, and yet have the discharge through this small orifice 

 under the same physical conditions, so far as pressure and tension are 

 concerned, as the air passing through the large opening, B. In other 

 words, the slight pressure inside the wind chest, due to the rotary 

 air-impeller, renders it particularly necessary to make sure that on the 

 discharge side of the three openings the pressure is always the same, 

 i. e., absolutely atmospheric. 



To build a different type of wind chest for each aliquot was im- 

 practicable. It was accordingly arranged to make the large opening 

 adjustable, using caps of various sizes to fit over the aperture. By 

 this means the total amount of air leaving the wind chest is reduced 

 and, incidentally, the amount of air discharged into the sampling cans 

 through the 10-mm. openings is slightly modified by the slight increase 

 in pressure inside the wind chest due to the fact that the free discharge 

 is somewhat hindered by the reduction in size of the large aperture. 



The maximum pressure inside the wind chest, even when all openings 

 are closed, is, however, so small that relatively large variations in the 

 amount of air leaving the wind chest influence but slightly the actual 

 amount of air discharged into the separate cans. With the rotary air- 

 impeller used and the 10-nam. openings, the amount of air discharged 

 is reasonably constant at about 45 liters per minute, practically inde- 

 pendent of the size of the main opening. 



DUPLICATION OF SAMPLES. 



With nearly all types of respiration apparatus, even those con- 

 nected with chambers, the kind of experiment ordinarily employed is 

 such that a repetition is usually easily made and with relatively slight 

 expense, particularly with animals. On the other hand, when one pro- 

 poses working with a large group of human individuals, to duplicate an 

 experimental session is an expensive procedure; hence it was necessary 

 at the outset to provide for duplicate analyses and dupUcate samples. 

 For this reason, two sampling cans were provided. The second can 

 (C2, fig. 8) is connected with an exact duplicate of the absorbing system 

 in series with the can Ci. This second absorbing system has a 

 separate Crowell blower (shown in the upper part of fig. 11 as F2), its 

 duplicate large Williams bottles for absorbing water vapor, and its 



