12 



RESEARCHES UPON ATOMIC WEIGHTS. 



ever, that the column of bulbs is ample protection against loss by spattering, 

 even when considerable effervescence takes place during the solution. 



After the silver was dissolved, the solution was diluted somewhat and heated 

 until free from nitrous fumes. Then it was further diluted until not more con- 

 centrated than I per cent, and was slowly added to the i per cent solution of 



cadmiiun chloride in the precipitating flask. After sev- 

 eral minutes' shaking it was allowed to stand several days, 

 with occasional shaking, until the solution was perfectly 

 clear. Two 30 c.c. portions of the clear liquid were then 

 pipetted into test tubes of similar size. To one portion 

 was added i mg. of silver nitrate in the form of hundredth 

 normal solution, to the other an equivalent amount of 

 hydrochloric-acid solution, and the tubes were examined 

 at frequent intervals in the nephelometer until the ratio 

 of the opalescence shown by the two tubes became 

 constant. Richards and Wells have shown that when 

 equivalent amounts of silver and chloride have been used 

 in the original precipitation, the nephelometer tubes show 

 equal opalescence. If this was not the case in the first 

 examination, the contents of the tubes were returned to 

 the precipitating flask and either standard silver nitrate 

 solution or standard hydrochloric-acid solution was added 

 and the shaking and testing repeated until the amounts of 

 chloride and silver in the solution were equivalent. 



Finally a considerable excess of silver nitrate was 

 added to the analysis to precipitate dissolved silver 

 chloride, and the silver chloride was determined gravimetrically as previously 

 described. Correction was of course made for chloride introduced in the course 

 of the nephelometric tests. 



A vacuum correction of +0.000152 gm. was applied for every apparent gram 

 of cadmium chloride, of +0.000071 gm. for every apparent gram of silver chlo- 

 ride, and of —0.000031 gm. for every apparent gram of silver.^ 



All weighings were made by substitution, with tare vessels as nearly as pos- 

 sible like those weighed. The gold-plated brass weights were twice carefully 

 standardized to hundredths of a milligram. 



Fig. 2. — Flask for 

 dissolving silver. 



^ The specific gravity of cadmium chloride has been found to be 4.047. Baxter and Hines: 

 Amer. Chem. Jour., 31, 220 (1904). Richards and Stull have determined the specific grav- 

 ity of silver chloride to be 5.56, and Richards and Wells that of silver to be 10.49. P^b. Car, 

 Inst., No. 28, II (1905); Jotir. Amer. Chem. Soc, 27, 466; Zeit. anorg. Chem., 47, 64. 



The specific gravity of the weights is assumed to be 8.3. (See page 40.) The use of this 

 low value for the specific gravity of the weights has led to slight changes in the vacuum cor- 

 rections of cadmium and silver chlorides from the values used in the original publication of 

 this paper. 



