112 Dr. E. Turner 07i some Aiomic Weights. 



nitrate of silver. Calculating from these elements, and with 

 108 as the equivalent of silver, we shall find li'OS as the 

 equivalent of nitrogen. It will be l^'OiG if the equivalent of 

 silver be 108-1. 



2. As a mean of three closely corresponding analyses, made 

 by converting the nitrate into sulphate of lead, I find that 

 100 parts of sulphate of lead correspond to 109*307 of nitric 

 of lead. Calculating the equivalent of nitrogen, on the pre- 

 sumption that 103*5 and 40 are the respective equivalents of 

 lead and sulphuric acid, we shall find it to be 14*1 01. 



Berzelius calculates his equivalent of nitrogen from an 

 analysis of nitrate of lead, and estimates it at 14*18. The 

 difference between us principally depends on a different esti- 

 mate of the composition of the oxide of lead ; and until this 

 point shall be settled with more precision than at present, no 

 certain inference can be deduced from the analysis of the 

 nitrate. I have more confidence in the estimate from nitrate 

 of silver, and feel little doubt that 14 is a very close approxi- 

 mation. Some analyses of nitrate of baryta, but which are 

 not fully in a state for publication, induce me to believe that 

 the real equivalent of nitrogen is nearer 14 than 14*1. 



Barium. — From the analysis of chloride of barium, published 

 in my Essay on that compound, no inference could at first be 

 drawn in consequence of the uncertainty respecting the equi- 

 valent of chlorine. Now, however, that we have reason to 

 take 35*45 as the equivalent of chlorine, it follows from my 

 analysis that the equivalent of barium is 68*76 ; and according 

 to the analysis of chloride of barium by Berzelius, it is 68*588. 

 I believe the equivalent of barium is intermediate between 

 68*6 and 68*8, and in the absence of more exact knowledge 

 68*7 may be taken as a very good approximation. 



The o-eneral conclusions which I deduce from the preceding 

 account are the following : 



1. The atomic weights commonly used by British chemists 

 have been adopted without due inquiry, and several of the 

 most important ones are erroneous. 



2. The hypothesis, that all equivalents are multiples by a 

 whole number of the equivalent of hydrogen, is inconsistent 

 with the present state of chemical knowledge, being at vari- 

 ance with experiment. 



3 The subjoined equivalents are very nearly correct :■ — 



Lead .^.....,,.,,.f. 103*5 



Silver .^..^—..-M-i-.... 108 



Barium -. ^.i.'.*..,-- .68*7 



Chlorine ../;«»;'.. 35*45 



Nitrogen ;.....;.. 14 



J 



