DESTRUCTION OF BACTERIA BY CHEMICALS. 153 
few drops of the culture from which all lumps have 
been filtered are added. At intervals of one, five, ten, 
fifteen, and thirty minutes, one hour, and so on, a small 
platinum loopful of the mixture is taken from each tube 
and inoculated into 10 c.c. of lukewarm gelatin, from 
which plate cultures are made. The results obtained are 
signified as follows: x per cent. of the disinfectant in 
watery solution and at 2 temperature kills the organism 
in twenty minutes, y per cent. kills in one minute, and 
soon. If there be any doubt whether the trace of the 
disinfectant carried over with the platinum loops may 
have rendered the gelatin unsuitable for growth, thus 
falsifying results, control cultures should be made with 
vigorous bacteria in gelatin to which a similar trace of 
the disinfectant has been added. 
The disinfectant to be examined should always be 
dissolved in an inert fluid, such as water; if on account 
of its being difficultly soluble in water, it is necessary 
to use alcohol for its solution, control experiments may 
be required to determine the action of the alcohol on the 
organism. Sometimes, as in the case of corrosive sub- 
limate, the chemical unites with the cell substance to 
form an unstable compound, which inhibits the growth 
of the organism only while the union exists. In-some 
tests it is necessary to break up this union and note 
then whether the organism is alive or dead. 
In the above determinations the absolute strength of 
the disinfectant required is considerably less when cul- 
ture media rich in albumin are employed than when the 
opposite is the case. Thus creolin (Pearsons), when 
bouillon is used as a culture medium, stops develop- 
ment in the proportion of 1 to 15,000 or 1 to 5000; 
when ox-serum is used only in the proportion of 1 to 
