THE RESPIRATION APPARATUS. 43 



mercury is drained out of the valve into the reservoir. Nevertheless, 

 it is conceivable that a few globules might adhere to the metal work 

 and the mercury gradually find its way into the air current. In the 

 first set of valves, however, it is highly probable that any mercury 

 vapor passing through the absorber system would be absorbed, and 

 consequently there remains only the possibility of the vaporization of 

 mercury from the valves beyond the absorbers. 



Since the volume of air confined in the system is used over and over 

 again, it might at first glance be considered an ideal place for the accu- 

 mulation of mercury vapor. Two circumstances militate against this 

 assumption. In the first place, the air is continually being withdrawn 

 from the chamber, and any mercury vapor remaining in it would 

 be absorbed along with the water in the water- absorbers. Secondly, 

 the air traverses a relatively long metal pipe galvanized inside, so that 

 the tendency for amalgamation would be very great. Likewise the 

 copper walls of the chamber would tend to retain the mercury. As a 

 matter of fact, in none of the experiments thus far made, in which 

 different subjects have remained in the chamber for periods varying 

 from i to 13 days, have indications of mercurial poisoning ever been 

 noted, and it seems probable that no appreciable quantity of mercury 

 vapor enters the respiration chamber. 



Proportion of water vapor in the air. Since the ventilating current 

 of air enters the respiration chamber absolutely dry, the possible effect 

 of such dry air on the mucous membranes of the throat and nose is of 

 importance, especially in long-continued experiments. In the imme- 

 diate vicinity of the pipe which conducts the air into the chamber, 

 unquestionably the air is extremely dry. In the course of a very short 

 time, however, diffusion produces a uniformity in the composition of 

 the air in the chamber which is probably pretty evenly distributed 

 throughout the whole volume. 



The total volume of air in the chamber is not far from 5,000 liters, 

 and if saturated with water vapor at 20 C. it would contain about 85 

 grams of water vapor. Generally the amount of water vapor present 

 in the residual air is not far from 40 grams, although at times it may 

 be as low as 25 grams. Obviously, then, there is only about 50 per 

 cent saturation under ordinary conditions, and at times as low as 30 

 per cent. This is not unduly dry air, and yet experience has shown 

 that it is capable of producing certain physical effects that can be 

 attributed only to excessively dry air. The subjects very frequently 

 complain of being rather cooler than when in the air of the labora- 

 tory. This is explained by the fact that there is much more rapid 

 vaporization of water from the lungs and skin and consequently a low- 



