METHODS OF MEASURING RESPIRATORY EXCHANGE 19 



work and rest respectively, the quantities of carbon dioxide given in 

 columns (a) and (ft) of Table III were eliminated. These figures show 

 rather wide variations and the question to be decided is : can it be 

 concluded from them that the elimination of carbon dioxide is in- 

 creased during mental work. The simplest way to treat the figures is 

 to form the ratio between the CO 2 elimination during work and dur- 

 ing rest. This has been given in column (c) and shows that on an 

 average the excretion is 2*4 per cent, higher during mental work than 

 during rest. The deviations of the individual results from this average 

 are given in column (d) and squared in column (<?). The sum of the 

 squares is 1832, the standard deviation therefore p = 9-3 per cent., and 

 the mean error of the series 2'O per cent., which shows that it can 

 not be concluded that there is any increase in carbon dioxide elimina- 

 tion during mental work though there may possibly be one amounting 

 to anything from o to 6 per cent. Double the mean error is usually 

 taken as the limits between which the true result must fall, provided 

 always that the deviations are of a purely " accidental " character. 



The units employed for expressing the respiratory exchange of 

 animals are rather various and no definite usage has so far become 

 established. The most rational plan, and that which is best adapted 

 for comparisons with other quantitative work on metabolism, probably 

 is to give the weights of gases absorbed or eliminated in a given time. 

 This possesses the advantage, moreover, that the figures require no 

 further qualification and that no doubt can arise about their meaning. 1 

 Usually, however, the quantities of gases are expressed by their 

 volumes. The volume of a gas being absolutely indefinite, the expres- 

 sion by volume requires a further statement of conditions regarding 

 temperature and pressure. In gas exchange work it is customary to 

 give the volume measured dry at 760 mm. mercury pressure (tempera- 

 ture of mercury o) and o temperature, and when no other conditions 

 have been definitely mentioned quantities given by volume must be 

 presumed to have been reduced to 760 mm. and o. The actual measure- 

 ments are practically always made at higher temperatures, at varying 

 barometric pressures, and with gases either completely or partially satu- 

 rated with water vapour. A very convenient table for reducing gas 

 volumes measured moist (that is saturated) at pressures from 740 to 

 775 mm. and temperatures from 10 to 25 has been given by Hal- 



1 When gas quantities are expressed by weight the definition of the respiratory quotient 

 must be slightly modified. The respiratory quotient is then the ratio of the weight of O ? 

 in the CO 2 produced to the weight of O a absorbed. 



2 * 



