CH. vii THE BURNING OF INFLAMMABLE AIR 113 



The observation which led to the discovery of the com- 

 pound nature of water was made by Priestley, in 1781, in 

 the course of "a mere random experiment, made to 

 entertain a few philosophical friends " (Experiments and 

 Observations, 1781, V. Appendix, p. 398). This consisted 

 in exploding a mixture of common air and inflammable air by 

 means of an electric spark in a closed glass vessel. After 

 the explosion the sides of the vessel were seen to become 

 covered with dew. Further experiments made by Mr. 

 Warltire (ibid., pp. 395-397) with a copper vessel appeared 

 to indicate a slight loss of weight, owing, it was thought, 

 to the heat which had escaped. This experiment attracted 

 the attention of Cavendish, who repeated it in 1781, and 

 published an account of his work in the Philosophical 

 Transactions of the Royal Society for 1784, under the title 

 " Experiments on Air" (A.C.R. III). 



Cavendish (1781) prepares water by burning inflam- 

 mable air with common air. Using a large explosion vessel 

 capable of holding 24,000 grains of water, Priestley's experi- 

 ment " was repeated several times with different proportions 

 of common and inflammable air." The best conditions were 

 reached when 423 measures of inflammable air were exploded 

 with 1000 measures of common air, whereby the volume of 

 the air was reduced to 811 measures, and practically all the 

 oxygen was removed. But, whatever the proportions in 

 which the gases were mixed, Cavendish "could never perceive 

 a loss of weight of more than one-fifth of a grain, and com- 

 monly none at all." Although the supposed loss of weight 

 could not be confirmed, Cavendish found himself in agree- 

 ment with Priestley when he noticed that : 



"In all the experiments, the inside of the glass globe 

 became dewy . . . ; but not the least sooty matter could be 

 perceived" (A. C. R. III. 13). 



" The better to examine the nature of this dew, 500,000 

 grain measures of inflammable air were burnt with about 2\ 



I 



