126 EXPERIMENT STATION. [Jan. 



arsenic slionld be treated with N/2 hydrochloric acid, 5 to 10 

 cubic centimeters for each ,1 of a gram, and boiled gently. In 

 case arsenic remains undissolved, a cold saturated solution of 

 sodium acetate, 3 grams salt for each .1 of a gram of substance, 

 is added, and boiling continued until solution is effected. By 

 another modification ^ suggested by Avery, and reported by 

 Thatcher,- 1 gram sami:)le is boiled five minutes with 25 cubic 

 centimeters of sodium acetate solution (1-2), dissolving the free 

 arsenic which is removed by filtration. The residue is dissolved 

 in dilute hydrochloric acid and both solutions titrated. 



Haywood attempted, in several ways,^ to modify the original 

 A very-Beans process so as to insure solution of the free arsenic. 

 After treating the sample with a slight excess of hydrochloric 

 acid at laboratory temperature, sodium carbonate was added 

 and the solution boiled. In another case sodium bicarbonate 

 was employed, but the results were unsatisfactory in both in- 

 stances, due to more or less reduction of copper and accompany- 

 ing oxidation of arsenous acid. Accurate results were secured, 

 however, by filtering off the hydrochloric acid solution and boil- 

 ing the residue with 5 grams of sodium bicarbonate, titrating 

 both solutions. 



Haywood proposed still another modification ^ which might 

 be considered a simplified Avery-Thatcher process ; ^ .4 of a 

 gram sample is boiled ten minutes with 25 cubic centimeters 

 sodium acetate solution (1-2) to dissolve free arsenic, and con- 

 centrated hydrochloric acid carefully added UJitil solution is 

 effected. After neutralizing with a solution of sodium carbonate, 

 avoiding an excess, alkaline tartrate and sodium bicarbonate are 

 added and titrated as usual. 



The Avery, Avery-Thatcher and Avery-Haywood processes 

 employ the same reagents, dift'cring only in their application. 

 The co-operative investigation ^ of the association in 1904 showed 

 that the three above modifications, together with the Haywood, 

 gave closely agreeing results, with little, if any, advantage in the 



* Optional official method, Assoc. Off. Agr. Chem. 



2 Proc. Assoc. Off. Agr. Chem., 20, 196 (1903). 



3 Jour. Amer. Chem. Soc, 25, 963 (1903). 



* Proc. Assoc. Off. Agr. Chem., 20, 197 (1903). Optional official method of the association. 



* Loco citato. . 



« Proc. Assoc. Off. Agr. Chem., 21, 98 (1904). 



