118 PROCEEDINGS OF THE AMERICAN ACADEMY. 



ively, putrid cheese, putrid blood, and yeast, and each flask was con- 

 nected with an argentic nitrate solution as before. All four flasks 

 were kept at 32° for six weeks, and air from them was passed through 

 the silver solutions. A slight black precipitate was formed in each, but 

 the liquid on being tested iu the Marsh apparatus showed no arsenic. 

 No odor was observed, and Schmidt and Bretschneider considered the 

 fermentation to have been stopped. They decide that poisoning can- 

 not take place in damp rooms from a volatile arsenical compound, and 

 that it is more likely that the trouble comes from the detachment of 

 dust iu dry rooms. These experiments are more in line with recent 

 work than those preceding, and they are the first that are worthy of 

 much consideration. Certain conditions, not then understood, might 

 have operated against the formation or detection of a gaseous com- 

 pound. Although the Berzelius-Marsh test may have been used, we 

 do not know the limit in Schmidt and Bretschneider's hands, and the 

 treatment of the silver solution is not given. 



Wittstein,* in 1860, advances the idea that the arsenites iu the 

 coloring matter are changed to arseniates at the cost of part of 

 their oxygen (5 As 2 3 = 3 As.,0 3 + As 4 ), and that metallic arsenic is 

 volatilized. 



Sonnenschein.f in 1869, made the following experiment in a damp 

 room on the ground floor in which the paper was very arsenical, the 

 occupant of which was affected by symptoms attributable to arsenical 

 poisoning. The air of the room after passing through a wash-bottle 

 was led through a hard glass tube heated to redness. After many 

 hours' heating there appeared a perceptible mirror, which Sonnen- 

 scheiu concluded to be arsenic, but he was unable to determine the 

 nature of the volatile compound. The deposit, however, received no 

 confirmatory test. 



Fleck, t in 1872, conducted the following series of experiments. 



1. A five-litre tubulated bell jar resting on a ground-glass plate 

 was covered on the inside with paper smeared with Schweinfurth green 

 (about 15 mgr. arsenious oxide per sq. cm.). The paper was fastened 

 by a paste of potato starch. Through the tubulus passed a cork 

 carrying one right-angled tube reaching to the bottom of the bell jar, 

 and another to just below the cork. The jar was closed and allowed 

 to stand. 



* Quoted by Eulenberg, he. cit., p. 414 ; also Schmidt's Jahrb., CX. 88. 



t Handbuch d. gericht. Chemie, I860, p. 153. 



t Zeitschr. f. Biologie, VIII. 444; also, Dingl. Polyt. Jour., CCVII. 146. 



