132 PROCEEDINGS OF THE AMERICAN ACADEMY. 



solution. Air was passed for eight days, and the reduction was some- 

 what less than before. The test of the solution was entirely negative. 



Experiment 8. — In order to try the effect of alkaline fermentation 

 about 0.5 sq. m. of brown wrapping paper was smeared with flour 

 paste containing 8.5 grams arsenious oxide and a quantity of lime, 

 and placed in the same bell jar without removing the preparation used 

 in Exps. 5-7. The total amount of arsenious oxide was therefore 

 about 15.5 grams. Air was passed for 17 days. There was much 

 mould in the bell jar, and the silver solution was slightly reduced. 

 No arsenic, however, was found in it. 



Experiment 9. — In this and in the following experiment the main 

 deviation from the previous trials was in the amount and character of 

 the arsenic compound, and the duration of fermentation. In a large 

 bell jar, fitted with a similar system of absorbents to that of Exp. 1, 

 was placed about 1 sq. m. of a wall paper containing only a small 

 amount of arsenic, and on this was smeared a paste of flour and syrup, 

 to which had been added 10 grams arsenious oxide and part of a yeast 

 cake. After action for 36 days, during which time air was drawn 

 through slowly, the silver solution was examined but gave no arsenic. 

 Very slight reduction had taken place, the light not being as strong as 

 in the previous experiments. 



Experiment 10. — The conditions of this experiment were exactly 

 the same as in the preceding, with the substitution of 10 grams of 

 Paris green for the arsenious oxide. The result was also the same. 



Direct Experiments on the Air of Rooms. — Experiment 11. — The 

 room in which this experiment was performed was a large attic 

 chamber. The surface of paper exposed was about 60-70 sq. m., each 

 containing 280 mgr. arsenious oxide. The frieze and ceiling con- 

 tained traces only, 0.8 mgr. and 1.0 mgr. respectively, so that the total 

 amount of arsenic may be estimated at about 18 grams. The occupant 

 of the room spent much of his time in it, and showed symptoms attrib- 

 utable to chronic arsenical poisoning. The room was at about 25°. 



In order to absorb the volatile compound, if present, a washing 

 bottle similar to those used in the previous experiments was partly 

 filled with a 2°f solution of argentic nitrate and connected with an 

 aspirator. In order to free the air from arsenical dust, it passed, 

 before reaching the silver solution, through two tubes filled with 

 cotton wool. In four days, the room being closed, about 750 litres of 

 air were drawn throuo-h the bottle. There were then a few black 

 specks in the solution. This, after filtration, was analyzed as before, 



