[ 168 ] 



Finding that candles burn very well in air in 

 which plants had grown a long time, and having 

 had fome reafon to think, that there was fomething 

 attending vegetation, which reftored air that had 

 been injured by refpiration, I thought it was pof- 

 fible that the fame procefs might alfo reftore the air 

 that had been injured by the burning of candles. 



Accordingly, on the 17th of Auguft, 1771, I 

 put a fprig of mint into a quantity of air, in which 

 a wax candle had burned out, and found that, on 

 the 27th of the fame month, another candle burned 

 perfectly well in it. This experiment I repeated, with- 

 out the leaft variation in the event, not lefs than 

 eight or ten times in the remainder of the fummer. 

 Several times I divided the quantity of air in which 

 the candle had burned out, into two parts, and 

 putting the plant into one of them, left the other 

 in the fame expofure, contained, alfo, in a glafs 

 veffel immerfed in water, but without any plant ; 

 and never failed to find, that a candle would burn 

 in the former, but not in the latter. I generally 

 found that five or fix days were fufficient to reftore 

 this air, when the plant was in its vigour ; whereas 

 I have kept this kind of air in glafs vefTels, immerfed in 

 water many months, without being able to perceive 

 that the leaft alteration had been made in it. I have 

 alfo tried a great variety of experiments upon it, as 

 by condenfing, rarefying, expoling to the light and 

 heat, &c. and throwing into it the effluvia of many 

 different fubftances, but without any effect. 



Experiments made in the year 1772, abundantly 

 confirmed my conclulion concerning the reftoration 

 of air, in which candles had burned out by plants 



growing 



