LAVOISIER. 253 



rests ; but it is equally undeniable, that from less elabo- 

 rate experiments Mr. Watt had before him drawn the 

 inference then so startling, that it required all the 

 boldness of the philosophic character to venture upon 

 it the inference that water was not a simple element, 

 but a combination of oxygen with inflammable air, 

 thence called hydrogen gas. That Mr. Watt first gene- 

 ralized the facts so as to arrive at this great truth, I 

 think, has been proved as clearly as any position in the 

 history of physical science. (' Life of Watt/ Historical 

 note in Appendix. Eloge of Watt by Arago.) It is 

 equally certain from the examination of Mr. Cavendish's 

 papers, and from the publication lately made of his 

 journals, first, that he never so clearly as Mr. Watt drew 

 the inference from his experiments; and, secondly, that 

 though those experiments were made before Mr. Watt's 

 inferences, yet Mr. Cavendish's conclusion was not drawn 

 even privately by himself, till after Mr. Watt's inference 

 had been made known to many others. * l 



In 1783, after Mr. Cavendish's experiment had been 

 made, and after Mr. Watt's theory had been formed upon 

 the experiments of Warltire and Priestley, and of Mr. 

 Watt himself, Sir Charles Blagden went to Paris. The 

 experiments of Mr. Cavendish were made in 1781, the 

 theory of Mr. Watt was contained in a letter which was 



* Mr. Harcourt's publication, contrary indeed to his design, has 

 greatly strengthened the evidence in Mr. Watt's favour. ('Life 

 of Watt,' in vol. i., p. 201.) Professor Robison's article in the 

 'Encyclopaedia Britannica' gives an opinion coinciding with mine; 

 and it was published thirteen years before Mr. Cavendish's death. 

 I first stated that opinion in a published form in 1803-4. ('Edin- 

 burgh Review,' vol. iii.) See the Appendix to this Life, in which 

 some account is given of the extraordinary errors and carelessness 

 about facts, which distinguish M. Cuvier's Eloge of Mr. Cavendish. 



