22 
Although, the room in which the experiments were made was zinc- 
lined and the doors and windows were fairly close fitting, still it was 
not air-tight, so that the conditions were not of that exact quanti- 
tative character as obtains in an operation like the precipitation of 
sulphuric acid by barium chloride. In view of this condition of the 
room, which, of course, allowed the winds to exert a varying influence 
upon the rate of leakage therefrom, the percentages obtained in 
those experiments performed under approximately the same condi- 
tions as to temperature can be considered as agreeing fairly well. 
The results also indicate that the relation between the formalin and 
permanganate in the ratio taken is fairly quantitative and that the 
percentage yield of formaldehyde gas is approximately the same 
whether 400, 600, or 800 c. c. be taken for a charge. Experiments lb 
and lib table give an idea of the rate of leakage of formaldehyde from 
the room. In III, where the room was made more air-tight by past- 
ing paper over the cracks of the door and covering the window frames, 
the per cent is a little higher. 
The percentage yield of formaldehyde with this method is very 
much affected by certain variations in the temperature. This, how- 
ever, is not the case with experiments done at temperatures above 
60° to 65° F. The results then agree rather closely, and it is prob- 
able that with this condition of temperature on a very quiet day with 
a charge of 600 c. c. formalin (35.66 per cent by volume) the weight 
of the gas in the room would be between 38 and 39 per cent of the 
weight in the 600 c. c. of formalin used. This percentage is only 
slightly less than that obtained under similar conditions with the retort 
and autoclave methods. 
Effects of low temperature upon the state of the formaldehyde in the 
room . — Provided the temperature is below 60° or 65° F., the libera- 
tion of formaldehyde by this method produces a hazy or fogg}” condi- 
tion of the air in the room. Tins is most marked just after the process 
of liberation is complete; but it does not entirely disappear during 
the time required to do an experiment, usually three hours. The 
lower the temperature in the room the more marked is this condition, 
which apparently is due to some change in the formaldehyde, result- 
ing from the effects of cold upon it and to a certain extent to the pres- 
ence of the aqueous vapor liberated. The latter, however, would 
soon become invisible. 
The exact change that takes place in the formaldehyde is not plain, 
but evidently there is much polymerization, as is shown by the para- 
form deposited in the room after the experiment is completed. 
The hazy condition in the room is about the same with the formalin- 
permanganate method as with the retort and autoclave; but the 
deposit of paraform is more noticeable with the former than the latter 
two methods. 
