SANGER. — VOLATILE COMPOUNDS OP ARSENIC. 135 



theory. He does not mention Hamberg's first paper, and has appar- 

 ently seen only the wretched abstract of the second quoted above, for 

 lie is unaware of some important points in that paper which were 

 similar to his own experience. He quotes the work of Fleck and of 

 Giglioli, and that of Johannson * and Binz f on the tolerance of arsenic 

 by saprophytes. But the most important result he considers to have 

 been obtained by BischofF, though his judgment is influenced undoubt- 

 edly by the fact that his own results have shown the odor to be a 

 positive indication of the presence of the volatile compound, whereas, 

 in Bischotf's experiment it was only an assumption, and not backed 

 by experimental proof. Other investigators than the above are not 

 mentioned. 



Gosio at the inception of his work considered that a volatile com- 

 pound could be formed by the action of mould on arsenical organic 

 matter, though positive proof was wanting, and the mechanism of the 

 reaction was entirely inexplicable. He set himself the following plan 

 of work : — 



1. To show whether from arsenical culture ground, exposed to 

 spontaneous inoculation from various surroundings, there could be 

 developed a volatile arsenical compound. 



2. If so, to isolate the germs which could effect this transformation 

 and to characterize them. 



3. To discover in what arsenical compounds the activity of the 

 organisms manifests itself most markedly ; whether this is to be 

 extended to products used in the arts ; and to find out what condi- 

 tions favor and what retard the action. 



4. To study the volatile compound. 



5. To describe the mechanism by which this gas is produced 

 through the biological energy of micro-organisms. 



The first question was answered by an experiment similar to Bis- 

 choff's. Potato pulp containing from 0.5 to 1% of arsenious oxide was 

 exposed to the air of a cellar. Abundance of mould appeared in a 

 short time, and at the end of a week an intense alliaceous odor was 

 developed. The pulp was then placed in a glass vessel, through 

 which air was drawn for two weeks into a hot solution of argentic 

 nitrate, the air being filtered by cotton wool. The usual blackening 

 took place, on which Gosio very properly lays little stress, and the 

 solution, freed from silver by hydrochloric acid and filtration, gave 

 " characteristic reactions for arsenic." 



* Arcliiv f. cxper. Pathol., II. 503. t Ibid., XI. 200 ; XIV. 345. 



