1911.] PUBLIC DOCUMENT — No. 31. 125 



and ioJiiK! in the presence of liydriodie acid or solu1)le iodide is 

 one of the most sensitive in analytical chemistry, forming the 

 characteristic bine iodide of starch. A more delicate indicator 

 conkl not be desired. Since the method was first api)lied to the 

 analysis of arsenicals nnnierons modifications have been devised 

 to insnre complete solution of the arsenic, to prevent oxidation, 

 to eliminate or render innocuous substances that might effect the 

 titration, and to enlarge its field of application so as to readily 

 include the arsenates. The Association of Official Agricultural 

 Chemists began work on insecticides in 1899 and has rendered 

 valuable service. 



New Processes. 



The introduction of the Thorn Smith process ^ marked a turn- 

 ing point m the analysis of arsenicals. It was intended particu- 

 larly for Paris green, and is the official method for that sub- 

 stance. Solution of the arsenic is effected by boiling the sample 

 with a slight excess of sodium hydrate, which readily unites 

 with the free arsenous acid, and also with the combined after 

 displacing the copper. In presence of a reducing substance like 

 sodium arsenite, the copper is precipitated as cuprous oxide and 

 a portion of the arsenous acid oxidized to arsenic. This oxida- 

 tion necessitates a subsequent reduction of the filtrate with 

 hydrochloric acid and potassium iodide (hydriodic acid), and the 

 removal of the excess iodine with thiosulfate. The solution is 

 neutralized with dry sodium carbonate, an excess of sodium bi- 

 carbonate added, and titrated with iodine. The process is 

 accurate, though the double titration is objectionable. 



Avery and Beans devised a very ingenious process - noted 

 for its simplicity. The Paris green is pulverized, solution 

 effected with concentrated hydrochloric acid in the cold, neu- 

 tralized Avith sodium carbonate, the precipitated copper redis- 

 solved with sodium potassium tartrate and titrated as usual. 

 The copper held by the alkaline tartrate colors the solution but 

 does not effect the titration. Plydrochloric acid, however, is a 

 poor solvent for free arsenic, and unreliable, which constitutes a 

 very serious objection to the process. Avery noted this error 

 and advised ^ that samples showing a tendency to separate white 



» Jour. Amer. Chem. Soc, 21, 769 (1899). 3 Jour. Amcr. Chem. Soc, 25, 1096 (1903). 



2 Jour. Amer. Chem. Soc, 23, 485 (1901). 



