220 Dr. Thomson's reply to Berzelius's 



Had the number for barytes been 9*5 instead of 9"75, there 

 would have existed a very obvious analogy among them all. 

 They would all have terminated in 0"5, or they would all have 

 been multiples of 4 hydrogen. This analogy struck me at 

 an early period of my investigations, and I was anxious to find 

 the weight of barytes only 9*5. The experiments of Berze- 

 lius rather favoured the idea ; according to his analysis the 

 constituents of sulphate of barytes are 



Sulphuric acid 5 



Barytes 9'55 



But Klaproth's analysis, made with great care, gave 

 Sulphuric acid .... 5 



Barytes lO'Ol 



But when I mixed together sulphate of potash and chloride 

 of barium, 1 found in many trials, that 11 of the former and 

 13"25 of the latter were the weights which decomposed each 

 other completely. When I employed only 13 of chloride of 

 barium (the weight, if barytes be only 9"5), there was always a 

 residue of sulphuric acid in the solution. Even IS'l'iiS chlo- 

 ride left a residue of sulphuric acid ; showing clearly that 

 the weight of barytes is more than 9*625. 



Berzelius, in the French edition of his tables published in 

 1819, gives for the weight of barium 17"1386; and under the 

 name oxidum baryticum, we have the four following numbers, 

 which are all obviously multiples of the first. 

 •19-1386 

 38-2772 

 57-4.158 

 76-5544 

 In his new table, published since he had an opportunity of 

 seeing my First Principles, I observe a vast number of changes. 

 He has aliandoned a great deal for which he had formerly 

 stickled ; and though he has not had the candour to acknow- 

 ledge as much, I see the great impression which my views have 

 made upon him. I am uncharitable enough to believe, that 

 it was in order to prevent his countrymen and the Germans 

 from being aware of the benefit which he derived from my 

 labours, that his attack upon me was made. I had touched 

 his selfish feelings, and disturbed those dreams of chemical 

 sovereignty in which he has been evidently indulging. In his 

 new table he gives the atom of barytes 



9-5688 



This is about jTfth part less than my determination. It was 



impossible that my error could have amounted to 2 per cent. 



It could not have been greater than y^^^^dth part at the utmost. 



But there is a circumstance of which I was not aware when 



Ide- 



