INFLUENCE OF ILLUMINATING GAS ON BACTERIA AND FUNGI 25 
was stored over water until desired for use. At the last, one experi- 
ment was run with gas which had been stored in a mercury seal 
gasometer and so had not been subjected to the alteration of composi- 
tion due to absorption by the water and the giving up to it of gases in 
the water. The results of the last test were corroborative of the 
former ones. The analysis of the gas used in this last test is here given. 
COo 1.0% O2.... 1.2% N2 (difference) .... 5.7% 
C„H,n.... 0.4% CO... 91.7% 
In view of the high percentage of CO in the gas used it has not seemed 
desirable to correct the values by calculation. The gas used in the 
earlier tests was probably lower in CO than that used in the last one. 
For this reason the results obtained in the last test have been used 
instead of the others where a variation occurred which seemed to be 
due to a higher CO content in the last quantity of gas. 
Effect on the Different Organisms 
Bacillus suhtilis. — No effect of carbon monoxide on B. suhtilis 
could be observed at lo percent or below, but at 25 percent and 
above the same sort of very thin colony was produced as was produced 
in illuminating gas at the higher concentrations. As in the case of 
illuminating gas, also, the colony did not usually undergo change to 
the normal air type when returned to the air, but often a normal 
colony would start up at some point and invade the slant from that 
center. Using this organism as a measure of toxicity, therefore, 
carbon monoxide would appear to be something like twice as toxic 
as illuminating gas. 
Bacillus pyocyaneus. — This organism grew in all the concentra- 
tions of carbon monoxide used although its development was slowed 
down to some extent in the higher percentages. It was difficult to 
determine any definite place at which the inhibition set in, owing 
to the indefinite tint of the colonies, although it seems likely that it 
should be placed at about 25 percent. 
Bacillus Kieliensis. — B. Kieliensis was found capable of growing 
in all the test conditions with carbon monoxide. The first pronounced 
retarding effect occurred at 25 percent, although it was small and 
not altogether uniform. The color was sometimes paler at 10 percent 
and usually so at 25 percent, while at 75 percent and higher the 
