324 Dr. Thomas Thomson on the Atomic Weight 



multiplied by the specific heat amounts to the constant 

 quantity 0*375 to be correct. My atomic weight for nickel 

 deduced from a very simple experiment, which I still con- 

 sider as susceptible of greater accuracy than anyanalysis 

 whatever, was 3-25. Now, 3*25 x 0-1035 = 0-337 ; while 

 3-625 X 0-1035 = 0-375. At the same time I stated, that 

 I had never been able, by the most careful experiments, to 

 obtain any number for nickel ever approaching to 3*625, 

 which would make the atom of nickel higher than that of 

 iron. I made a hasty experiment in the month of January 

 last, which I have stated in a note. — Records of General 

 Science, iii. 255. But it was unsatisfactory; and my 

 numerous occupations, at that busy season of the year, in 

 the college of Glasgow, put it out of my power, at that 

 time, to prosecute the experiment. I, therefore, left the 

 investigation till summer should arrive, and give me leisure 

 to resume the subject. I have just finished the experi- 

 ments which I had projected (July 10th, 1836), and pro- 

 pose in this paper to state the results which I obtained. 



In my First Principles of Chemistry , vol. i. p. 358, I have 

 stated the experiments from which I deduced 3*25 to be 

 the atomic weight of nickel. I found, that when a solution 

 of 17-125 grains of pure crystals of sulphate of nickel is 

 mixed with a solution of 13*25 grains of chloride of barium, 

 a double decomposition takes place, and after the sulphate 

 of barytes has fallen to the bottom, the supernatant 

 liquid, tested by solutions of sulphate of soda and chloride 

 of barium, remains transparent, and, consequently, contains 

 no appreciable quantity either of sulphuric acid or barytes. 



This experiment I still find to be perfectly correct. I 

 employed in repeating it sulphate of nickel, which had been 

 purified by five successive crystallizations. The crystals 

 were small but well formed, and they were rendered as dry 

 as possible without losing any of their water of crystalliza- 

 tion by pressure between folds of bloating paper and ex- 

 posure to the air. The chloride of barium was prepared 

 on purpose for the experiment. It was perfectly pure, and 

 rendered anhydrous by exposure to a red heat in a platinum 

 crucible. 



The conclusion which I drew from this experiment was, 

 that 17*125 grains of crystallized sulphate of nickel con- 



