10 Dr. Thomson on the true Atomic Weight of [Jan. 



racy of my own experiments, and not suspecting any inaccuracy 

 in Mr. Dalton's, I naturally concluded that phosphuretted 

 hydrogen gas was capable of uniting with three proportions of 

 oxygen gas ; namely, 



1 volume phosphuretted hydrogen with 1*0 volume oxygen 



1 1-5 



1 ■ 2-0 



The first two of these proportions were my own ; the last wa» 

 Dalton's. Now phosphuretted hydrogen gas is composed of 



1 volume hydrogen gas "I j i • . i 



T 1 u 1^ ° y condensed nito one volume. 



1 volume phosphorous vapour/ 



The volume of hydrogen gas requires half a volume of oxygen 

 gas to convert it mto water. Therefore, we have 



1 volume phosphorus uniting with 0*5 volume oxygen 



1 1-0 



1 1-5 



This is the same thing as saying that an atom of phosphorus 

 unites with one atom, two atoms, and three atoms of oxygen. I 

 had demonstrated that an atom of phosphorus weighs 1*5. 

 Hence it was obvious that the weights of these three compounds 

 of oxygen and phosphorus must be as follows : 



First compound 2*5 



Second ditto 3*5 



Third ditto 4*5 



I concluded that these three compounds were hypophospho- 

 rous acid, phosphorous acid, and phosphoric acid. This recon- 

 ciled my own experiments with those of Dalton and Berzelius. 

 I was induced by this reasoning, which will be admitted to be 

 suflSciently plausible, to alter the proof sheet, and to bring it to 

 the state in which it appeared in the fifth edition of my System 

 of Chemistry. 



Soon after this I went to Glasgow, and nearly a year elapsed 

 before I was in possession of a laboratory, or had it in my power 

 to return to the subject. As soon as I had the means I tried 

 Mr. Dalton's experiment, and was a good deal surprised to find 

 it inaccurate, 1 was unable to obtain a complete combustion of 

 a mixture of one volume phosphuretted hydrogen gas and two- 

 volumes of oxygen. I was now satisfied that my original views 

 on the subject were correct; and Davy^s paper on the subject, 

 which appeared about this time, and which agreed exactly with 

 my original views, served to confirm my opinion. Berzelius's 

 experiments still remained to be accounted for. I was satisfied 

 that they must be inaccurate; but the difficulty was to hit 

 upon a method of demonstrating them to be so. The experi- 



