

vin THE BURNING OF INFLAMMABLE GASES 145 



spark, and appeared to be almost wholly converted into 

 carbonic acid. (New System of Chemical Philosophy, 1810, 

 II. 382.) 



Berzelius and Dulong- (1820) calculate the composition 

 by weight of carbonic anhydride. The earliest estimations 

 of the proportion by weight of carbon and of oxygen in 

 carbonic anhydride, were (like the early numbers for the 

 composition of water) dependent on a knowledge of the 

 densities of the gases concerned in the combustion. 



As the volume of carbonic anhydride produced was equal to 

 the volume of oxygen used, the composition of the anhydride 

 could be deduced directly from a knowledge of the densities 

 of these two gases. This method was employed by Berzelius 

 and Dulong, working in Berthollet's laboratory at Arcueil 

 (Ann. de Chimie, 1820, 15, 393), in deducing the value 

 which was accepted as authoritative from 1820 to 1840. 

 Taking the density of oxygen as 1*1026 times that of air, 

 and that of carbonic anhydride as i'524, it followed from 

 the equality of the volumes of the two gases that 1*524 

 parts by weight of carbonic anhydride were made up of 

 1*1026 parts of oxygen and 1*524- 1*1026 = 0*42 14 part of 

 carbon. The percentage composition of the anhydride was 

 therefore : 



Oxygen = 72'35% . 



Lavoisier's combustion of charcoal. Berzelius's numbers 

 were in close agreement with those which had been given 

 nearly 40 years earlier by Lavoisier ("Decomposition and 

 Recomposition of Water," 1783; Works, II. 345) as a 

 result of a careful quantitative investigation of the reduction 

 of red lead by charcoal. In this experiment the weight of 

 oxygen was deduced from the loss in weight of the oxide ; 

 the weight of fixed air or carbonic anhydride was calculated 

 from its volume and density. Lavoisier found that : 



