332 Dr. Thomson on the true Weight of [Nov. 



barytes, sulphuric acid, and muriatic acid, are exactly those 

 which I have assigned. 



If the practical chemist wishes now to ascertain whether the 

 numbers assigned by Berzelius for the weiglit of the atoms of 

 these bodies be correct, he may make the experiment as follows: 



1 atom of sulphuric acid = 5-0116 



1 atom potash = 5-89915 



Atom of sulphate of potash = 10*91075 



1 atom muriatic acid = 3-4-265 



1 atom barytes = 9-5693 



Atom of muriate of barytes = 12-9958 



Let him, therefore, dissolve in distilled v/ater 10-91075 grs. of 

 dry sulphate of potash, and 12-9958 grs. of chloride of barium. 

 If he examine the liquid after the sulphate of barytes has sub- 

 sided, he will find that it still contains sulphuric acid; for it will 

 be precipit-dted by nitrate of barytes. I say then that the weights 

 assigned by Berzelius are inaccurate It is obvious that the 

 atom of sulphuric acid is either too high, or that of barytes too 

 low. By comparing his numbers with mine, the reader will see 

 that the weight of the sulphuric acid is a little too high, and that 

 of the barytes a little too low. 



It is obvious from the preceding experiment that the true com- 

 position of sulph-dte of barytes is 



Sulphuric acid 5-00 or 33-8984 



Barytes 9-75 66-1016 



14-75 100-0000 



And this is the constitution of the salt which must hereafter be 

 employed to determine the proportion of sulphuric acid in any 

 sulphate. 



The atomic weights of Dalton and Wollaston may be tried 

 precisely in the same way. It will be found that the liquid left 

 after both will be precipitated, indicating elearly that their num- 

 bers are not quite accurate ; but the precipitate which appears 

 when we take Dr. Wollaston's weights is much smaller than 

 with either those of Berzelius or Dalton, indicating that his 

 numbers are nearest the truth of the three. 



It is obvious from the preceding experiment that the weight 

 of an atom of potash is exactly 6. In the fifth edition of my 

 System of Chemistry, I showed that the mean of the different 

 expei'iments to determine the composition of potash comes 

 exceedingly near this number — so near indeed that nobod)'^ 

 could hesitate to adopt it as the true number. I have now 

 demonstrated that it is absolutely true ; for the error which can 



