144 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Experiment 18. — Three 200 c.c. Erlenmeyer flasks were fitted as 

 in Exp. 17, with rubber stoppers and tubes, the latter plugged with 

 cotton wool. A wall paper was selected having 115 mgr. arsenious 

 oxide per square meter, the color, a dark red, suggesting an aniline com- 

 pound mordanted with an arseniate. 3 square decimeters of this were 

 cut into strips and placed between several slices of potato, which were 

 put into each flask. The total amount of arsenic used was 3.45 mgr., 

 reckoned as As 2 3 . The flasks, after addition of a little water, were 

 sterilized for two and a half hours at 100-105°, and showed no evi- 

 dence of mould for six days afterward. The potato, which was still 

 moist and impregnated with the red coloring matter, was then inocu- 

 lated with the culture in the tube marked a patata contenente una 

 striscia di tappezzeria arsenicale," and the flasks were connected with 

 a series of absorbents similar to that of Exp. 17. The current of air 

 was not drawn through the system until the flasks had stood for a 

 week, the temperature being from 20 to 25°. No mould appeared for 

 four days, and the growth was then very slow, was confined to the first 

 two flasks, and the amount was very small. Indeed, after 17 days 

 had passed without much apparent action, the silver solution being 

 but slightly affected, I disconnected the flasks, not intending to pursue 

 the experiment further. Yet on opening the flasks the alliaceous 

 odor though faint, was perfectly plain, and I proceeded to test the 

 first silver solution. This was poured out, and the very slight deposit 

 removed with a little nitric acid was added to the solution. The latter 

 was evaporated with nitric acid nearly to fusion, taken up with dilute 

 nitric acid, desilverized, and prepared as usual for the Marsh flask. 

 This had been in action 40 minutes, and was free from arsenic. At 

 the end of an hour after introducing the solution there was a small but 

 perfectly plain mirror, which I estimated at 0.005 mgr. The second 

 silver solution was not examined. 



Dr. Farlow kindly examined the first flask of the series, and found 

 "P. brevicaule in good condition but not pure, for there was another 

 species of Penicillium present not in very good fruit. The second 

 species may have been a small form of P. glaucum, but I do not 

 think I can say certainly from the material examined, which was too 

 young." 



The amount of the volatile compound determined in this experi- 

 ment is 0.14% of the total present. This estimate is more accurate 

 than in Exp. 17, and the chief error is in the absorbent, though the 

 amount is only a part of what might have been formed in the course 

 of time. 



