122 PROCEEDINGS OF THE AMERICAN ACADEMY. 



which air was drawn. Strips of argentic nitrate paper hung in the 

 bell jar acquired in eight days a red tint, and yielded in the Marsh 

 apparatus a ring of arsenic. 



Selrni considered that the results pointed to the formation of an 

 arsine, and that the formation of hydrogen by moulds was confirmed 

 conclusively. 



In the light of recent work there may have been a volatile arsenical 

 compound formed in Selmi's experiments. Yet we cannot accept 

 them as conclusive because of the neglect to provide against the pos- 

 sible reduction of the silver paper by the dust particles in the bell jar. 

 Had the filtered air given the above results, they would have been of 

 more value. Besides this the freedom of the reagents from arsenic is 

 not shown. No mention is made of any alliaceous odor from the 

 decomposing matter. The odor from the silver paper after adding 

 alkali may be analogous to that noticed later by Hamberg, Gosio, and 

 myself, but it is not sufficiently characterized by Selmi to draw any 

 definite conclusion in regard to it. 



The results of Fleck and Hamberg were accepted by many as 

 conclusive, and more recently the evidence of Selmi has been con- 

 sidered corroborative. It was some time before Selmi's results 

 became generally known. His paper did not obtain wide circulation, 

 and the abstracts quoted above treated his work as bearing only on the 

 general question of the development of hydrogen by moulds, and not 

 on the formation of a volatile arsenical compound. 



Professor Chandler of Columbia, in the course of his testimony 

 before a committee of the Massachusetts Legislature in March, 1886, 

 stated that two of his students, Messrs. Morewood and Drummond 

 had, in 1879, under his directions, passed the air from a vessel con- 

 taining Paris green and paste into an argentic nitrate solution for a 

 number of days, and failed to find any arsenic. The mixture was 

 then allowed to decompose in a warm place and the air was tested for 

 arsenic with the same result. They then covered a square yard of 

 paper with Paris green and paste, placed it in a vessel, and drew air 

 over it into argentic nitrate for many days. The result of this was 

 also negative. As no further details were given, and tlie work has 

 not, to my knowledge, been published, no conclusion can be drawn as 

 to the possibility of finding small amounts of arsenic by the method 

 used, whatever that might have been. 



Bartlett,* in 1880, tried the following experiments. He first 



* The Analyst, 1880, p. 81. 



