124 PROCEEDINGS OF THE AMERICAN ACADEMY. 



the assumption that chronic arsenical poisoning was due to dust alone. 

 He used moist bread and saccharine liquids, to which arsenious oxide 

 had been added. The preparations were exposed to the air, and, 

 after the mould had formed, were placed in a vessel through which a 

 current of air was passed into an absorbent. Argentic nitrate and 

 auric chloride were used, and in some experiments the air was led 

 through a hot Berzelius-Marsh tube. No arsenic could be detected 

 in the air above the mould, nor was there any odor. 



Bischoff,* in 1882, during the examination of some fodder with 

 which arsenic had been mixed with intention of poisoning, placed a 

 part of it, while still moist, in a covered glass dish. After some 

 weeks he observed that mould had collected in the mass, and, on 

 opening the dish, noticed an odor which, from its garlic nature, he 

 concluded to be due to arseniuretted hydrogen. Strips of paper 

 moistened with argentic nitrate were at once turned brown. This 

 was considered by Bischoff to confirm the theory of the formation of 

 arseniuretted hydrogen during fermentation. It is greatly to be 

 regretted that this experiment, which has actually furnished a clue to 

 the recent successful investigations, should not have been carried out 

 properly. There was undoubtedly a volatile arsenic compound pres- 

 ent, yet one cannot, in such an important question as this, accept 

 an odor or a blackening of silver paper as indicative of arsenic. 



Hamberg,f in 1886, published the result of an investigation, extend- 

 ing over nearly nine years and a half, on the change produced in ar- 

 senious oxide in contact with decaying animal matter. I give the 

 experiment in some detail, because an incorrect and misleading idea of 

 it seems to have been obtained from the abstract quoted. 



In a twelve-litre flask were placed portions of a body ; lungs, liver, 

 kidneys, and intestines, mixed with broken glass and sand. The mix- 

 ture was moistened with a solution (it is not stated whether it was 

 acid or alkaline) of one gram of arsenious oxide, covered with sand and 

 aluminous earth, and the whole saturated with water. The flask was 

 connected with (1) a tube containing cotton wool, (2) a tube con- 

 taining test paper, (3) an absorption tube with a 4<% solution of 

 argentic nitrate, (4) a U tube to catch any of the silver solution 

 which might be mechanically carried over, and to the last was attached 

 a Finkener aspirator. Air was drawn from outside the laboratory 



* Vierteljalir f. gericlit. Med., N. F., XXXVII. 163. 



f Pharm. Zeitsclir. f. Russland, XXV. 770; Bellamy t. k. svens.k. vetenskabs- 

 akad. Handl.,Bd. 12,11., No. 3; Fres. Zeitsclir. f. analyt. Cliem., XXVI. 788, Ref. 



