196 EXPERIMENT STATION. l-Ian. 



the reason that ordinary porcelaneous arsenic in pulverulent 

 condition is diftieult to moisten and of low solubility, but 

 more particularly because it would induce a secondary reaction 

 from lack of base to satisfy the acid that was previously com- 

 bined with the lime. Sodium arsenite, !N^aAs02, is readily 

 soluble and proved a satisfactory source of arsenic. A salt 

 of fair quality can be procured on the market, or is easily 

 prepared by adding 1 part of arsenous oxide to a boiling solu- 

 tion of 1.45 parts of sal soda, or an equivalent amount of soda 

 in the form of anhydrous carbonate, bicarbonate or hydroxide. 

 A slight excess of arsenic is required to insure complete volatili- 

 zation of the carbonic acid. 



Na2C03 • 10 H2O + AS2O3 = 2 NaAs02 + CO2 + 10 H2O. 



The resulting arsenite should be free from arsenates, car- 

 bonates, sulfates or other acids forming insoluble compounds 

 with lime. 



Any decision as to concentration of solutions is naturally 

 more or less arbitrary ; dilution tends to make difficult pre- 

 cipitation with considerable loss of salt, and the opposite an 

 unwieldj' precipitate with greater occlusion. As a compromise 

 solutions of l/> molecular strength (M/2) were finally adopted. 

 Another factor studied was the influence of temperature of 

 solutions on the resulting precipitate, ranging from that of the 

 laboratory to nearly boiling point at the moment of precipita- 

 tion. Room temperature with two hours' standing gave a 

 product of practically the same composition, and of probably 

 better physical characteristics, tlian the higher temperatures 

 and was considered more desirable. 



As the alkalinity of the soda in sodium arsenite is not de- 

 stroyed by the arsenous acid, it should be run into the calcium 

 chloride solution slowly with constant agitation in order to 

 prevent any precipitation of calcium hydroxide. An excess 

 (10 per cent.) of sodium arsenite was found desirable to per- 

 fect the salt. After standing several hours the liquor was re- 

 moved by means of a Buchner funnel, and the lime arsenite 

 washed rapidly with cold water until nearly free from chlorides. 

 A centrifuge or filter ])ress might give equally good or better 

 results provided the work was done rapidly. Undue washing 

 was avoided, as it was thought safer to retain a small amount 



