12 
solution to which had been added a trace of formaldehyde; the con- 
tents of the tube soon acquired a pink color. It seemed evident that 
formaldehyde is wholly absorbed from air passing through a series of 
tubes containing cyanide solution. This conclusion was further 
strengthened by experiments on the absorption of formaldehyde from 
air by water alone, which will be mentioned presently. 
It is a well-known fact that a solution of cyanide undergoes slow 
oxidation and other changes when exposed to the atmosphere, and it 
was suspected that in the experiments the long-continued bubbling 
of air through the solution would have the effect of diminishing its 
titre. It was found in several trials that when air was passed through 
10 c. c. of a nearly decinormal solution of potassium cyanide at the 
same rate and for the same length of time as in the experiments its 
equivalent in decinormal silver nitrate was diminished by nearly 0.1 
c. c. This correction was applied in the calculation of the result of 
each determination. Although it may not be absolutely correct, } 7 et 
it can not be far from correct. As described above, the cyanide tube 
was followed by a tube containing silver nitrate, which latter was of 
course used in the subsequent titration. 
After nearly all the results given further on had been obtained by 
the cyanide method, there came to our notice an article by Trillat , a 
in which the author states that formaldehyde can be completely 
absorbed from air by passing it through a sufficient number of absorp- 
tion apparatus containing water alone. This suggestion was put to 
the test, and it was found that three water tubes absorbed all of the 
formaldelrpde, the last tube containing only a trace. The reason for 
not having chosen water at the start as an absorbing medium was that 
it was supposed a solution containing a substance, as cyanide, which 
could combine chemically with formaldehyde would be better suited 
as an absorbing agent than pure water. Owing to circumstances it 
was not possible to repeat all of the experiments and use water tubes 
to absorb the formaldeh 3 T de, but in several instances the absorption 
from the same charged air of the room was carried out in cyanide tubes 
and in water tubes, side by side. Although three tubes are sufficient, 
five tubes of water were used for greater precaution. In these par- 
allel experiments the percentage results in the case of absorption by 
water were from 1.5 to 2 per cent higher than those in the case of ab- 
sorption by cyanide. 
We can not account for this difference, unless it is involved in the 
uncertainty of the correction made on account of the change of titre 
of the cyanide solution, due to air bubbling through it, which is referred 
to on page 12. Without making the correction, the two results were 
identical in some cases. It is possible also that the greater simplicity 
a Trillat, A. : Presence normal de la formaldehyde dans les produits de la combustion 
incomplete. Rev. d’hyg., vol. 27, no. 2, Feb. 20, 1905. 
