564 
SANITARY NOTES ON POTABLE WATER. 
latter in contact with the filtered water until all materials 
might be thoroughly disinfected, the meat would remain 
fresh. This led to the following modification. 
Two vessels were charged with the three materials—spongy 
iron, pyrolusite, and sand—as before, and water was passed 
through them for ten days. Now it would have been im¬ 
possible to introduce the meat at its former place without 
disturbing the filtering media or exposing them to reinfec¬ 
tion. Each filter was, therefore, after the washing had been 
completed, connected with a glass flask, provided with an 
inlet tube passing to the bottom, and an exit tube from the 
top of the flasks. Otherwise the flasks were sealed air-tight 
after the meat had been placed in them. The inlet tues were 
connected by india-rubber tubing with the exit tubes from 
the filtering arrangement. None of the usual precautions as 
to disinfection of either flasks or tubing were employed, as 
I concluded that if these minutiae were of consequence the 
putrefactive matter adhering to the meat would certainly 
do its work. 
After passing water through the vessels and flasks between 
three and four weeks, the meat in one of the latter was found 
to be as fresh as the first day, but I feel hesitation in posi¬ 
tively asserting whether or not the meat in the other flask 
exhibited a very faint putrid smell. If so, the explanation 
may perhaps be that some of the india-rubber tubing I used 
had certainly before been employed in connection with ex¬ 
periments on hay infusion. It might be argued that any 
preservation had been due to inactivity of the bacteria owing 
to the cold, the experiment being made in December and 
January last. I do not believe this, for the vessels stood 
in my laboratory only two yards from the fireplace; and in 
the course of the experiments immediately preceding these, 
to which I referred before, the meat became putrid, notwith¬ 
standing that the temperature was then about the same as 
during the succeeding experiments. 
Nevertheless, another set of experiments was commenced 
towards the end of January last, certain precautions being 
taken to guard against any possible error. 
One filtering vessel was, as before, filled with spongy iron, 
pyrolusite, and sand, and another, for the sake of comparison, 
under otherwise like conditions, with animal charcoal, as I 
know' from my former experience that, unless the cold pre¬ 
vented it, the charcoal would cause the meat to become 
putrid. A third vessel w 7 as lastly charged with spongy iron 
in a modified manner,, namely, so that the filtered water 
when leaving the vessel still contained some little oxygen in 
