82 A RESPIRATION CALORIMETER. 



The air entering the respiration chamber is free from carbon dioxide 

 and water, and contains a larger percentage of oxygen than that in the 

 chamber itself. Theoretically, therefore, there will be a space about 

 the tube conducting air into the chamber that will have varying pro- 

 portions of the different constituents ; practically, however, the slight 

 differences at this point are neglected. 



The distribution of the water vapor in the chamber is very uneven, 

 for we have the moist surface of the skin of the subject and the very 

 moist absorbing system in the upper part of the chamber, and conse- 

 quently, with the tendency of moist air to rise, probably the upper 

 half of the air in the chamber contains a greater amount of water than 

 the lower half. That this difference in composition is sufficient to 

 affect the results of experiments with man is very much to be doubted, 

 and in alcohol check experiments the rate of evolution of water is so 

 slow that in all probability natural diffusion produces a nearly uniform 

 moisture content throughout the whole chamber. In all these calcu- 

 lations, therefore, it is assumed that each portion of air is of uniform 

 composition. 



DA.TA USED IN CALCULATING RELATION OF WEIGHTS AND VOLUMES OF GASES. 



The atomic weights of hydrogen and oxygen employed in these 

 computations are those derived by Morley, 1 as follows : Oxygen = 16; 

 hydrogen = 1.00762. According to the same authority, the weight of 

 i liter of oxygen at 45 latitude is 1.42900 grams. Corrected for 

 gravity, this becomes, at Middletown, 1.42853 grams. From these 

 data the weight of i liter of hydrogen is readily computed as 0.089964 

 gram, and water vapor as 0.80423 gram. 



The atomic weight used for carbon is that given by Clarke 2 as 12.001. 

 The weight of i liter of carbon dioxide is accordingly 1.96427 grams. 



For the weights of a liter of nitrogen and air, data given by three 

 observers, Von Jolly, Leduc, and Rayleigh, 3 for the weight of a liter of 

 oxygen and nitrogen have been averaged. In this way the weight of i 

 liter of nitrogen at Middletown has been computed as 1.25668 grams. 

 Similarly the weight of a liter of dry air, which, according to Rayleigh 

 and Ramsey, 3 contains 20.91 per cent by volume of oxygen and 79.09 

 per cent by volume of nitrogen, is taken as 1.29264 grams. 



The figures obtained for the weight of i liter of water vapor are on 

 the assumption that it is a perfect gas down too and obeys the law of 

 expansion and contraction due to pressure and temperature. Accord- 

 Smithsonian Contributions to Knowledge (1895), 980, p. 109. 

 2 Smithsonian Misc. Coll. (1882), 27, p. 56. 

 1 Smithsonian Contributions to Knowledge (1896), 1033, p. 14. 



