SANGER. — VOLATILE COMPOUNDS OF ARSENIC. 147 



In considering the nature of the compound, the only definite facts 

 obtained are that it is formed in presence of oxygen, that the develop- 

 ment is best from carbohydrates, and that arsenious or arsenic acid 

 and their salts are best suited to the development. The only conclu- 

 sion that may be drawn is that the volatile compound is an organic 

 derivative of arsenious or arsenic acids. Yet the properties are such 

 that it has little resemblance to any of these derivatives thus far 

 isolated. Much work remains to be done before the composition can 

 be determined, and Gosio is continuing his research to that end. 



In this connection the researches of Selmi * are interesting. From 

 a corpse, in which the presence of arsenic was established, he isolated 

 a ptomaine in small quantity, but the test for arsenic in it was nega- 

 tive. In the stomach of a hog,f saturated with an arsenic solution 

 and left to decompose, he was able to isolate two bodies containing 

 arsenic. The first was obtained by distillation with steam, and had 

 an intensely poisonous action like strychnine. The second was found 

 in the residue from distillation, and the action of this resembled that 

 of the ptomaines. In the urine of a dog t poisoned by arsenic Selmi 

 finds a volatile arsenic compound with tetanizing action, and considers 

 it identical with the first of the two previously described. In no case 

 were these compounds found in quantity sufficient to make an attempt 

 at determining their composition. Husemann,§ referring to Selmi's 

 work, thinks that an arsenical ptomaine could be generated from 

 arsenical paper and paste, yet the properties of the volatile compound 

 and the fact that it is best formed in non-nitrogenous ground seem 

 to be against the formation of a ptomaine from these materials. 



Washington University Chemical Laboratory, 

 Saint Louis, September, 1893. 



* Atti della R. Accad. dei Lincei, [3], II., June 2, 1878. 

 t Mem. d. Accad. d. Scienze, Bologna, [4], I. 299. 

 t Ibid., [4], II. 3. 

 § Arch. d. Pliarm., CCXIX. 415. 



