CHEMISTRY — BAXTER. 



211 



the salt before weighing was not considered advisable on account of the high 

 fusing point of neodymium chloride and the consequent danger of contami- 

 nation through attacking of the glass tube by the hydrochloric acid gas. In 

 each weighed portion of chloride the chlorine was determined either by titra- 

 tion against weighed equivalent amounts of silver or by weighing the silver 

 chloride produced. The results of the analyses follow, calculated on the 

 basis of silver 107.88. 



These results are subject to two uncertainties: First, the neodymium chlo- 

 ride, since it was not fused, may have contained moisture. Since, however, 

 the salt was dried wholly by efflorescence, without melting, it is probable — 

 from experience with other salts — that only traces of moisture were retained. 

 Preliminary experiments to confirm this conclusion by determining the loss 

 in weight during fusion of the neodymium chloride were on the whole 

 satisfactory. The second uncertainty is in the vacuum correction of the 

 neodymium chloride. Time has been lacking to redetermine the specific 

 gravity of the salt. In the calculations for the table of results Matignon's 

 value of 4.20 is used. 



The close agreement of the results obtained from so wide a range of mate- 

 rial indicates that the dififerent samples were essentially identical and reason- 

 ably pure. The final result agrees very well with von Welsbach's most recent 

 determination, 144.5, ^^^ with that chosen by the International Committee 

 on Atomic Weights, 144.3. 



The atomic weight of iron was investigated by Mr. Victor Cobb. Very 

 pure metallic iron was converted to ferrous bromide by solution in hydro- 

 bromic acid, and the salt was several times recrystallized from hydrobromic- 

 acid solution. The crystals were dehydrated and fused in a weighed quartz 

 boat in a current of nitrogen and hydrobromic-acid gases. It was found im- 

 possible to use a platinum boat for the purpose, since, owing to partial dis- 

 sociation of the ferrous bromide vapor, an iron-platinum alloy was formed 

 in considerable quantity. Finally, the salt was dissolved in slightly acidu- 

 lated water, and after careful oxidation with a slight deficiency of potassium 



