174 THE DECOMPOSITION OF CARBONIC-ACID GAS. [MEMOIR X. 



mand attentive consideration. Moreover, in other plants, 

 as in experiments 4 and 5, the amount of oxygen is rela- 

 tively greater, and between it and the nitrogen there 

 does not appear any exact proportion. 



In order to ascertain whether decompositions taking 

 place under absorbent media, as bichromate of potash, 

 yield the same results as indicated in the foregoing 

 table, I made several analyses of gas collected under 

 those circumstances. The presence of the absorbent me- 

 dium did not seem to exert any influence whatever, the 

 general results coming out as though it had not been 

 employed. 



It was also found that the alkaline carbonates and 

 bicarbonates could be decomposed by leaves in yellow 

 light. The alkaline bicarbouates, as is well known, un- 

 dergo decomposition by a slight elevation of temperature. 

 When boiled in water they gradually give off their sec- 

 ond atom of carbonic acid, and slowly pass into the con- 

 dition of neutral carbonate. In the experiments I made 

 with them the boiling was not continued long enough to 

 affect to any extent the constitution of the salt, and in 

 each case any portion of carbonic acid extricated during 

 cooling was removed by the air-pump. A few leaves 

 placed in this solution showed no effect if kept in the 

 dark, but if brought into the sunshine there was a copi- 

 ous evolution of gas-bubbles, which, on detonation with 

 hydrogen, proved to be rich in oxygen gas. One experi- 

 ment gave 88 of oxygen, 12 of nitrogen. 



In a subsequent communication to the Philosophical 

 Magazine (Sept., 1844), I gave additional analyses of 

 the gas emitted in yellow light as follows : 



Five tubes, each three eighths of an inch in diameter 

 and six inches long, were inverted in a small trough of 

 water containing carbonic acid, with which the tubes 

 were also filled. Some blades of grass nearly of the 



