110 
The effects of temperature seem to be principally upon the state of 
the formaldehyde after it is liberated: that is. below a certain point it 
polymerizes. This seems to take place below about 65° F. with the 
formalin-permanganate and below about 60° F. with the other 
methods. Although the germicidal power of formaldehyde is not 
entirely lost at these temperatures, it is so diminished as to be of very 
little practical value for disinfection purposes. 
The role played by the relative humidity in formaldehyde disinfec- 
tion is more important than that of any other other influencing fac- 
tors. In the absence of moisture, formaldehyde is practically inert as 
a germicide. The minimum percentage of relative humidity that in- 
variably gives efficient results in disinfecting with formaldehyde can 
not be stated exactlv. since it varies slightly according to other con- 
ditions. such as temperature, wind, the quantity of gas present, the 
space to be disinfected, etc. 
However, it appears from experience that the relative humidity in the 
air before starting an experiment should not be lower than as follows: 
1 For an ordinary well-closed room. 60 per cent: 2 for ordinary 
railway cars. 65 per cent. TFith a high temperature, a large quantity 
of formaldehyde, and a long exposure, efficient results may be ob- 
tained with slightly lower percentages of humidity than just stated. 
There is not much difference in the amount of moisture given off 
by the different methods. This form of moisture, however, does not 
seem to answer for disinfection purposes so well as the natural humid- 
ity in the atmosphere. This conclusion is indicated in some experi- 
ments in which the relative humidity before starting the experiment 
was below the point above stated, and while the moisture given off 
raised the percentage of humidity to a point where efficient results 
would be expected, the organisms usually escaped destruction. Con- 
sequently. the exact value of the moisture given off in the process of 
liberating the formaldehyde can not be positively stated, but it ap- 
pears that too much trust must not be placed in it. 
In the experiments in a bottle very good results were* obtained with 
a relative humidity between 50 and 60 per cent, which is slightly less 
than that required in the room and car experiments. Of course, the 
conditions in the bottle were much more favorable for efficient results 
than usually obtain in ordinary formaldehyde disinfection. 
The results of our chemical determinations must be interpreted as 
representing the quantities of formaldehyde gas that may be expected 
to be present in a fairly air-tight room when charged with 600 c. c. of 
formalin by the various methods described and after a certain interval 
of time which varies with the method used. 
According to the maximum quantity of gas fotmd. on an average 
warm day and regardless of the time required to evolve it. the meth- 
ods may be arranged as follows : 
