12 



of hydrogen from dry iron, but furnished a theory by which the disci- 

 ples of phlogiston* might nevertheless maintain their ground both in 

 this and other cases. The explanation which his theory afforded in 

 this instance was, that the inflammable air, due to the unperceived moist- 

 ure in the iron filings, or in the air of the vessels, is evolved by the force 

 of double affinities in the following manner — the water, decomposing 

 the iron, combines in part with its basis, and in part with the phlogiston 

 (or dry hydrogen) which it was supposed by hypothesis to contain ; form- 

 ing by the one combination the calx, or oxide, of iron, and by the other, 

 inflammable gas f. Such a representation was not incompatible with 

 any known facts, and Cavendish had his own reasons for giving it on 

 the whole a preference over that which seems to us so much more plain 

 and reasonable : its fault as a theory was, that it was needlessly hypo- 

 thetical, and that it was part of a system overloaded with a multitude 

 of hypotheses. 



Lavoisier was the first to introduce into chemistry a juster language 

 and a safer manner of stating facts ; he caught sight of a principle 

 which has been since laid down by Davy as a genei'al proposition, and 

 has contributed much to the distinctness of chemical science, — the prin- 

 ciple that every body is to be reasoned about as simple till it has been 

 proved by direct evidence to be compound. To Cavendish, trained in 

 the rules of demonstration, and gifted with a sagacity and clearness of 

 conception beyond his fellows, hypothetical thoughts and expressions 

 were no stumbling-block ; and he seems therefore not to have felt how 

 great an obstacle they present to the general movement of science as it 

 floats upon the tide of a thousand understandings. 



If the question then be, who reformed the expressions and logic of 

 chemistry, or who furnished the simple terms in which we now state the 

 elements of water ? the answer is, Lavoisier ; but if it be, who dis- 

 covered and unfolded the most important facts on which that reforma- 



* That Watt derived from Cavendish his views on this subject, is evident 

 from the parenthetical introduction of his altered opinion that inflammable 

 gas was not pure phlogiston, but a combination of phlogiston and water, in 

 the middle of experiments and arguments to prove the contrary, without assign- 

 ing any reason, and after the publication of Cavendish's theory. See Mr. 

 Watt's Thoughts. Phil. Trans, vol. Ixxiv. p. 330. 



t Cavendish assigns as his principal reason for believing inflammable gas 

 to be a compound of this description, that it does not unite with oxygen at 

 common temperatures ; but it is likely that he was influenced also by the re- 

 sult of his experiments on "a diiferent kind of inflammable air, namely, that 

 from charcoal," for which, see Postscript, p. 38. 



