EXPERIMENTS. 



In the following pages I shall describe the separate series of exper- 

 iments, beginning with the control experiments and then taking 

 the Greenland experiments in chronological order. These experi- 

 ments are shown in tabular form and the tables are collected later 

 into a summary-table. 



The tables give the following information. Serial number of 

 the experiment, date, height of the barometer, temperature (gas- 

 meter), duration of the experiment in minutes, number of respira- 

 tions during this time, frequency of the respirations per minute cal- 

 culated from the two previous factors; further, the air expired 

 during the experiment in litres. As mentioned, the gas-meter may 

 show as much as 2 litres too little; but it is extremely improbable 

 that this error in the experiments amounts at any time to 1 litre; 

 the error in reading amounts at most to ^ 05 litres. Further, the 

 total amount of air respired per hour at 0° and 760 mm. (dry). As 

 the last is calculated from the previous factors, the possibility of 

 error mentioned will also be present here; the possible error may 

 indeed be multiplied by 2, but even in the most unfavourable cases 

 these errors cannot amount to 1 °lo of the values noted. The error 

 in the barometer and thermometer readings may have some influence 

 on the figures for the total amount of air respired, but these errors 

 are also of quite subordinate importance; in a chance experiment 

 at approximately average pressure and temperature a change in 

 barometric pressure of 05 mm. gives a correction of 04 litres in 

 the reduced volume of air, whilst О!"" on the thermometer gives 

 0-2 litres. 



In the next column is shown the amount of an expiration at 

 the given temperature and then the volume of an expiration at З?"" 

 saturated, calculated from the preceding. The error of the baro- 

 meter and thermometer readings could here amount to 1 cc. and 

 0*5 cc. respectively. Then follows the amount of carbonic acid in 

 the dry, expired air, calculated from the analyses. The error of 



