260 COMPARISONS OF RESPIRATORY EXCHANGE. 



similar, if not greater, errors also occur in their analyses of expired air. 

 It is to be recommended that investigators publish their analyses of 

 atmospheric air and thus indicate the general accuracy of their gas 

 analyses. 



In choosing a respiration apparatus, an investigator must consider 

 whether or not he wishes to use gas-analysis apparatus. Those who do 

 not should select some respiration apparatus which is constructed on 

 the Regnault-Reiset principle, since, if properly manipulated, no gas 

 analyses are necessary, the respiratory exchange being determined 

 directly by either weight or volume. On the contrary, the acquirement 

 of the technique of gas analysis is of great service, even in using an 

 apparatus of the Regnault-Reiset type, as it may be desirable to deter- 

 mine the composition of various portions of the expired air, the residual 

 air, or alveolar air in studies of this character. Furthermore, it is 

 possible at the same time to study the ventilation and the effect upon 

 the respiratory exchange of breathing atmospheres of varying composi- 

 tion. If, then, one has not acquired skill in gas analysis, the field of 

 investigation is very much limited. 



To sum up, therefore, gas analysis requires a great deal of time to 

 carry out and is very tedious; an apparatus for determining the respira- 

 tory exchange which does not require such analysis is accordingly to 

 be preferred. Furthermore, with a method in which the respiratory 

 exchange may be determined directly, the results may be obtained 

 more quickly than with a method involving gas analysis, for it is rarely 

 possible to make such analyses as rapidly as the weighings and the 

 computations can be made by the direct method, and at the same time 

 obtain the necessary records of the pulse, respiration, and other factors 

 included in a complete respiration experiment. The ability to use 

 gas-analysis apparatus, however, extends widely the field of an investi- 

 gator in respiration and respiratory exchange. 



ACCURACY AND INTERPRETATION OF RESULTS. 



In studying the respiratory exchange of man, some standard of 

 accuracy is necessary in order that one may interpret the results and 

 draw inferences from variations which may be found. If an experi- 

 ment with three experimental periods is made with a man in a resting 

 condition and without food for 12 hours or more, a certain constancy of 

 results may be expected. The variations from this constancy are due 

 to three things: Errors in the actual manipulation and the limits of 

 accuracy, due to the apparatus itself; the accidental variations in 

 the metabolism of man; and abnormalities in the respiration, such as 

 dyspnoea, apnoea, and hyperpncea. 



The first source of variation must be eliminated so far as possible by 

 the experimenter. To this end he must observe all the precautions 



