78 
elapsed before the exposures were made. They were variously 
placed in the car for exposure, the exact location being recorded in 
each case. . 
Different lengths of exposure might have been tried, but in our ex- 
periments 2 hours was chosen. Therefore, the car in all experiments 
was opened after 2 hours from the time of beginning to liberate the 
gas. The organisms were removed from the car as soon thereafter as 
possible, usually about 15 minutes, and taken to the laboratory. 
With the exception of B. tuberculosis, they were planted in nutrient 
bouillon, placed in the incubating room, and the results recorded, as 
has been described (p. 18). 
In case of B. tuberculosis , the pieces of carpet and * slips of filter 
paper (Wilson method) were washed in bouillon and the washings 
injected intraperitoneally into guinea pigs. Control pigs were also 
inoculated at the same time. The pigs were weighed before and at 
definite intervals after inoculation; those that did not die within 
about 3 months were chloroformed. A careful post-mortem exam- 
ination was made of all. The gain or loss of the chloroformed pigs is 
given in each experiment ; the figures refer to the weight of the pig 
when inoculated as compared with its weight when chloroformed. 
Although the cars seemed comparatively tight, the small percent- 
age of formaldehyde gas found in the cars on opening showed that 
much of it had escaped during the 2 hours. The temperature and 
relative humidity were taken just before starting an experiment and 
again as soon after it was over as possible. The second reading, how- 
ever, indicates more the condition of the outside atmosphere than any 
change resulting from the experiment, as the car was then being aired. 
I. FO IlMAJL I X-PEKMA N G AN ATE METHOD. 
Formalin and permanganate were used in these experiments in the 
same proportions as given on page 20, namely, 1 c. c. of the former to 
0.5 gram of the latter. The ordinary commercial crystals of perman- 
ganate were used without further powdering. The formalin and per- 
manganate were mixed in galvanized-iron buckets, one being used for 
every 500 c. c. of formalin, and placed in the car as described in each 
experiment. As the quantities of the formalin and permanganate 
used, the condition of the weather, including temperature, relative 
humidity, winds, etc., are given in each individual experiment, fur- 
ther description here is unnecessary. The experiments are as follows: 
