74 
TABLE 10. 
Experiments in a 24-liter bottle. 
Formalin-permanganate method. 
Formalin 0.5 c. c. and permanganate 0.25 gram. 
Moisture removed by calcium chloride. 
Varying quantities of moisture introduced. 
Ordinary summer temperatures. 
B. coli communis exposed on filter paper. 
[+ means growth; — no growth.] 
Tempera- 
ture at 
time of 
experi- 
ment. 
Quantity of water in- 
jected" as aqueous 
vapor. 
Absolute 
humidity, Relative 
grains per humidity, 
cubic foot. 
Time of exposure in minutes, and results. 
5 
10 
15 
20 
30 
45 
60 
75 
90 
120 
77.0 
6. 17 
62 
+ 
-1- 
• 
77.5 
5. 44 
54 
+ 
+ 
+ 
+ 
75.0 
4. 64 
50 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
TABLE 11. 
Experiments in a 24-liter bottle. 
Formalin-permanganate method. 
Formalin 0.5 c. c. and permanganate 0.25 gram. 
Moisture removed by calcium chloride. 
Varying quantities of moisture introduced. 
Ordinary summer temperatures. 
B. subtilis exposed on filter paper. 
[4- means growth; - - no growth.] 
Tempera- 
ture at 
Quantity of water in- 
jected as aqueous 
vapor. 
Time of exposure in minutes, and results. 
time of 
experi- 
Absolute 
humidity, 
grains per 
cubic foot. 
Relative 
humidity. 
ment. 
O 
10 
15 
20 
CO 
o 
45 
60 
75 
90 
120 
77.0 
6.17 
62 
+ 
4- 
+ 
+ 
+ 
— 
— 



77.5 
5. 44 
54 
+ 
+ 
+ 
+ 
+ 
+ 
— 
— 
— 
— 
75.0 
4.64 
50 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+' 
+ 
+ 
It will be observed from the above experiments that with a relative 
humidity between 50 and 60 per cent both B. coli communis and B. 
subtilis were killed within a reasonably short time, but below 50 per 
cent neither organism was killed within 2 hours. 
Ihc^e results demonstrate the important part that humidity plays 
in formaldehyde disinfection and that the minimum limit lies between 
50 and 60 per cent. This is a slightly smaller percentage than was 
found necessary in the room experiments to accomplish disinfection; 
but of course the conditions in the bottle were unusually favorable 
for germicidal action and slightly better results would therefore be 
expected. 
The different temperatures at which the experiments were per- 
formed may account for some of the little variations in the results of 
the above tables (8-11). - 
