APPARATUS AND METHODS USED IN THIS RESEARCH. 45 



variations in temperature, pressure, or actual volume of air inside the 

 system. By reading the millimeter scale over which the pointer from 

 the counterweight of the spirometer bell passes, it can easily be seen 

 whether or not the apparent volume of air in the chamber is altered 

 during a test. 



To make such a test, all of the various parts of the apparatus are 

 connected as in an experiment with an infant, and the ventilating air- 

 current started. After the first moment or two, during which the air 

 throughout the whole system will be attaining equilibrium, the bell on 

 the spirometer should reach a constant level, and thereafter the air 

 should remain absolutely constant unless affected by changes in tem- 

 perature or atmospheric pressure, these being indicated by the readings 

 of the barometer and the two air thermometers. If the changes in the 

 position of the spirometer bell can not be accounted for by temperature 

 or barometer changes, there is obviously a leakage of air into or out of 

 the system, usually the latter. 



To test the efficiency of the apparatus and the absence of a defect in 

 any individual part, especially when assembling the parts or when 

 trying to locate a leak, a water manometer, consisting of two glass tubes 

 connected at the bottom by a short bit of rubber tubing and attached 

 to a suitable standard, is found advantageous, inasmuch as the slightest 

 leak in any individual portion of the apparatus can readily be detected 

 by applying pressure with a bicycle pump. When the apparatus has 

 been properly installed, with accurately fitting rubber gaskets and 

 connections, and suitable inspection given from time to time, there is 

 no occasion for leakage, so that such an occurrence can invariably be 

 ascribed to faulty technique. Since the experiments with infants 

 instantly follow the test of the apparatus, the only disturbance there- 

 after being the removal of the cover which fits into the water seal, it 

 will be seen that these tests should prove an admirable index of the 

 condition of the apparatus during the experimental period. 



Tests for the ejficiencij of the absorbing vessels.— The amount of carbon 

 dioxide given out by the infant is determined by noting the increase 

 in weight of the soda-lime vessel (D, or D 2 ) with its attendant Williams 

 bottle (Ex or E 2 ) ) the degree of absolute moisture in the air when it 

 enters the soda-lime bottle and leaves the Williams bottle should be 

 identical. If, however, the sulphuric acid in the Williams bottle, E x or 

 E 2 , following the soda-lime container, is allowed to accumulate water 

 to such an extent that its efficiency as a water-absorber is somewhat 

 less than that of the Williams bottle, C, preceding the soda-lime con- 

 tainer, it is obvious that there would be a loss of water from the system 

 as a whole and the amount of carbon dioxide thus measured would 

 actually be too small. Conversely, if the air is not as dry before it 

 enters the soda-lime bottle as when it leaves the Williams bottle fol- 

 lowing, there will be an undue increase in the weight of the carbon- 



