6 Thomson, AtmospJicric Pressure on Animal Metabolism. 



Tubes were connected to the right and to the left tube of 

 the T-piecc respectively ; these were attached to the longer 

 of two straight glass tubes passing through rubber stop- 

 pers, one going to the bottom of the bottle, the other short 

 glass tube only passing through the rubber cork, and 

 these were connected with a corresponding but reversed 

 T-piece and two way glass tai:», and the stem of this 

 T-piece was connected to the large bottle communicating 

 with the meter and the water vacuum pump : thus, by 

 turning on the water to the vacuum pump, air was drawn 

 through the bell jar, and, through one or the other of the 

 litre bottles standing side by side according to the direction 

 in which both of the two-way taps were turned, and by 

 reversing both of these taps the stream of air could be 

 complete)}' shut off from one bottle and passed through 

 the other. This enabled me to take samples of the air 

 without interfering with the continuity of the experiment. 



The carbon dioxide in the air in the litre bottles was 

 determined by running in through the straight smaller 

 glass tube which penetrated the stopper, 15 c.c. of yV 

 N. baryta water solution, shaking vigorously and then 

 titrating with y^y N. hydrochloric acid solution. 



The current of air passing through the bell jar was 

 increased till the air in the bottles contained about 20 

 parts by volume of CO2 per 10,000 of air. It was found 

 necessary to pass air at the rate of about 80 to lOO litres 

 equal to 2-8 to 3-2 cubic feet per half hour to give this 

 result, the current being passed alternately through the 

 two bottles till the COo found was about constant. 



The total weight of CO., in the air used per hour was 

 then calculated, and this was further calculated into 

 grammes of CO._. per 1,000 grammes weight of animal, 

 which was always weighed immediately before putting it 

 under the bell jar. 



