360 ON THE LACTIC FERMENTATION 
that such particles, including perhaps some atmospheric organisms, might be 
deposited upon the upper part of the vessel, and so fail to be destroyed when 
the liquid was heated. But why should success be more likely with Pasteur’s 
solution than with milk ? Simply for this reason, that milk is a material which 
serves as a pabulum for almost all organisms. I once met with a bacterium, 
but only once, that would not live 
in milk; for extremely numerous 
as the varieties of bacteria appear 
to be, almost all of them seem to 
thrive in that lquid, whereas it is 
a common thing to find bacteria 
which, if put alive into Pasteur’s 
solution, will not grow in it at all. 
The defect in our method having 
been discovered, there was no great 
difficulty in correcting it. It was 
done by substituting for the funnel 
this syphon (see Fig. 6), consist- 
ing of two glass tubes (S and T) 
connected by a tube of india-rubber 
(U), with a stopcock (V) in the course 
of the caoutchouc tubing. The sy- 
phon is first completely filled with 
water, the temperature of which 
should be higher than that of the 
air, so that there may be no dissolved 
air given off toform bubbles. Then 
suppose this (W) to be the fluid that 
we wish to introduce into the flask 
(X). We pass one leg of the syphon 
FiG._6. into it ; then turn the tap and permit 
a sufficient amount of fluid to flow 
out to ensure that all the water has escaped from the syphon; then turn 
off the stopcock and proceed in the manner already described for the funnel. 
The tube (T) having been washed with strong carbolic lotion, carbolized_rag is 
wrapped round its lower extremity, and this is applied to the mouth of the 
flask as the cotton cap is removed ; the tube (T) is pushed steadily down to 
the bottom of the flask through the carbolized rag (Y), the stopcock is turned 
on and the liquid passes into the flask without the smallest air-bubble accom- 
