112 Mr. W. F. Barrett on a Physical 



ably fine and dry, so that the absorptions are certainly below 

 those that might be expected under other circumstances. 



On different days the absorption of the heat from a carbonic 

 oxide flame by common air was found to be as follows : — 



Table II. — Source : small carbonic oxide flame. 



Common air. 

 On March 5 . 



Absorption per 100. 

 . ' . . 18-7 



On April 13 . 

 On April 28 . 

 On May 3 . 



. . . 150 

 . . . 18-1 

 . . . 19-3 



It is here evident that the variations in the amount of carbonic 

 acid in the atmosphere can be detected by this mode of experi- 

 ment : such a variation was strikingly shown by comparing air 

 obtained from Brighton Downs with air taken from the labora- 

 tory. In both cases the air was dried, and the determinations 

 made without altering the position of the source or experimental 

 tube. The results are : — 



Absorption per 100. 

 Air from Brighton Downs . . 13*5 

 Air from the laboratory . . . 17*8 



The foregoing experiments led me to examine the absorption 

 by the carbonic acid contained in our breath. 



An india-rubber bag fitted with a stopcock was filled with air 

 from the lungs. The experimental tube was exhausted and the 

 needle brought to zero, where it steadily remained. Breath from 

 the bag was now passed over sulphuric acid and into the tube ; 

 the needle moved quickly on the side of cold, and finally showed 

 an absorption of 30 o, 8, or 50 per cent. Half the entire radia- 

 tion from the carbonic oxide flame was thus cut off by the car- 

 bonic acid mixed with the air from the lungs. 



Smaller pressures of breath from the same bag were next 

 allowed to enter the tube, the absorptions found are given in the 

 following Table : — 



Table III. — Source : Carbonic oxide flame. 



Tension, in Deflection. Absorption 

 inches. per 100. 



f 1 7*2 120 



barbonic acid of breath < ^ qn-O 33*3 



[_30 30-8 50-0 



The bag used as a reservoir was put to a direct test, and found 

 to have no influence on the absorption. 



