240 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



If the weights are assumed to have the specific gravity at 8.3,^ Mati- 

 gnon's value for the density leads to a vacuum correction per gram of 

 substance of +0.000141, while the value 4.13 necessitates a correction 

 of +0.000145. The difference amounts to only four ten thousandths 

 of a percent, which corresponds to a difference of only one thousandth of 

 a unit in the atomic weight of neodymium. Hence it is of no impor- 

 tance which correction is employed. 



The following table indicates the corrections for the buoyant effect 

 of the air applied to the weights of the various substances. 



All weighings were made on a nearly new balance, Troemner No. 10, 

 which is employed exclusively in atomic weight researches. Weighings 

 were made by substitution for counterpoises as nearly as possible like 

 the objects weighed both in weight and in volume. The weights were 

 carefully standardized to hundredths of a milligram by the method de- 

 scribed by Richards. 2 



Results and Discussion. 



In order to show that no appreciable error occurred owing either to 

 occlusion by silver chloride or from loss of silver chloride in solution, 

 the ratio between the silver used and the silver chloride obtained has 

 been calculated in the five pairs of analyses for which the data are 

 available. 



1 Baxter, Proc Amer. Acad.. 42, 209 (1906). 



2 Jour. Amer. Chem. See, 22, 144 (1900). 



