SANGER. — VOLATILE COMPOUNDS OF ARSENIC. 143 



appeared in 30 minutes which was at its maximum in 60. The 

 amount formed was 0.025 mgr., making 0.095, or probably 0.1 mgr. 

 in all, from this solution. 



It was possible that the first silver solution had not absorbed all 

 the volatile compound, as was conjectured in Hamberg's case. Bear- 

 ing in mind the probable volatilization of arsenic in the method 

 employed in the first solution, the second, which contained little or no 

 deposit, was evaporated with considerable nitric acid to incipient 

 fusion, during which no alliaceous odor was noticed. The residue 

 was taken up with dilute nitric acid, precipitated with hydrochloric 

 acid, and the filtrate, after evaporation with sulphuric acid and dilu- 

 tion, added to a Marsh flask in which the absence of arsenic had 

 been shown by a 40-minute run. After an hour, the mirror, which 

 appeared slowly, was at its maximum, and was estimated at 0.015 

 mgr. This shows conclusively that argentic nitrate in such dilution 

 does not absorb the volatile compound with any degree of completeness. 



To prove that no arsenic could have entered from the air of the 

 room, the rear solution of silver was treated in the same manner, and 

 gave no trace of arsenic. 



The total amount of arsenic (as As 2 3 ) obtained from this series 

 of flasks was 0.12 mgr. No further satisfactory quantitative result 

 can be adduced for many reasons. During the interval between Exps. 

 17a and 17 b, much of the compound may have escaped; the silver 

 solutions did not absorb all of the compound, and the method of 

 treating the first solution at least was incorrect, being based on the 

 old assumption that arseniuretted hydrogen was the product. Hence 

 0.12 mgr. does not represent by any means what may have been 

 formed. Yet if one reckons 300 mgr. of arsenic (as Ae 2 3 ) to the 

 flasks, the amount recovered is 0.04% of this, a proportion which for 

 all we know may be 100 times too small. 



The flasks were now uncorked. The odor, though fainter, was 

 distinctly perceptible and was confirmed by others in the laboratory. 

 Dr. W. G. Farlow has had the kindness to examine the growth for 

 me, and reports that the first two flasks contained only P. brevicaule 

 " fruiting and in good condition," while the third, which I suspected 

 was not homogeneous, contained also P. glaucam, though the amount 

 " as compared with the P. brevicaule is less than one would suppose 

 on looking at the flask without examining microscopically." 



Though the above experiment confirmed the results of Gosio to my 

 complete satisfaction, the following trial is of great interest on account 

 of the small amount of arsenic and its source. 



