58 ATOMIC WEIGHT DETERMINATIONS. 



pure materials, and passed through a solution of litharge 

 in potash, and ov6r a coarse powder of caustic potash 

 before use. The resultant water was caught in calcic chlo- 

 ride and weighed. The determination was also confirmed 

 by experiments on the specific gravity of oxygen and 

 hydrogen. The minimum result for hydrogen was 0.9934, 

 the maximum 1.0086. {Thomson's Annals of Phil., ^, 1821, 

 48.) 



T. Thomson: 1 {0 = 16); 6.25 (0=100). 



Thomson found the Sp. Gr. of H = 0.0694. Taking that 

 of O as 1.1111 on theoretical grounds (the supposed com- 

 pound nature of air, etc.,) he calculates the above value. 

 [Erdmann's Journ. fiir Prak. Che?n., 8, 1836, 374; Records 

 of Gen. Sci., R. D. Thomson, 1836, 179.) 



J. Dumas : 1.0012 (O = 16) ; 6.2575 (O = 100). 



Determined by nineteen experiments on the reduction of 

 cupric oxide with pure hydrogen. The gas was made from 

 pure materials and was passed through solutions of plumbic 

 nitrate and argentic sulphate, and over potash, and dried 

 with cold sulphuric acid or with phosphoric acid. The 

 weighings of the oxide and of the reduced copper were 

 made in vacuo. [Dumas corrected the results obtained for 

 the air contained in the sulphuric acid, but does not ex- 

 plain how he estimated it, while certain other possible cor- 

 rections are not mentioned.] The mean of the corrected 

 results is 12.515. The extreme difference is 0.09 for = 

 100. Without the correction for absorbed air the mean is 

 12.533, [or 1.00264]; maximum 12.583; minimum 12.481. 

 {Paris Oomptes Rend., U, 1842, 537.) 



Erdmann and Marchand : 1.0016 (0 = 16); 6.26 

 (O = 100). 



Determined by eight experiments on the reduction of 

 cupric oxide with hydrogen, the number is the mean of the 

 results. In four of the experiments the correction for 

 vacuum was calculated. These gave H = 12.548 ; extreme 

 difference, 0.067. In four experiments the weighings were 

 made in vacuo. These gave a mean of 12.492, with an ex- 

 treme difference of 0.015. The oxide employed was either 

 copper scale or was produced from cupric nitrate. The 

 hydrogen was made from pure zinc and sulphuric acid, and 

 was purified with potash in solution and in lumps, mercuric 

 chloride, sulphuric acid, and chloride of calcium. In the 



