312 A FURTHER CONTRIBUTION TO THE 
I am indebted to a suggestion of Mr. Godlee, of University College, London, 
has three advantages over boiling by direct flame; it avoids frothing, which 
in the case of milk is extremely troublesome, and also the sputtering to which 
Dr. Roberts, of Manchester, has drawn attention ;1 it prevents any loss of 
water by evaporation, and so disposes of the vexatious question of whether the 
specific gravity of the liquid has not been so raised as to render it unfavourable 
for organic development ; and lastly, it avoids any ‘ burning’ of the milk with 
its accompanying chemical changes. 
The milk having been allowed to cool completely, a portion of it is decanted 
off into experimental glasses. These are plain ‘ liqueur-glasses ’, each provided 
with a glass cap shaped like a small evaporating dish (made to order at any 
glass work) and covered with a small glass shade standing on a square piece of 
plate-glass. The glass plate has the double advantage of allowing the glass 
to be removed without disturbing the glass shade, and also of preventing the 
air beneath the shade from acquiring an accidental odour such as is derived 
from wood or other porous substances, and interferes with judging of chemical 
changes by the sense of smell. 
The glass shade and glass cap have in combination the effect of perfectly 
excluding all living dust, although, as neither cap nor shade is made to fit 
closely, a constant free interchange by diffusion between the air in the liqueur- 
glass and the gases of the atmosphere is permitted. Hence, provided always 
that the liqueur-glass and its cap are free from living organisms to begin with, 
and that the contained liquid is similarly circumstanced, the latter will remain 
for an indefinite period unchanged except by gradual loss from evaporation, 
till at length in the course of months it dries up into a solid mass. 
Further I have found as a matter of experience that if the glass shade and 
cap are raised, in a part of the room free from draughts, for the purpose of 
inoculation of the liquid or withdrawal of a small quantity for examination, 
there is practically no risk of the accidental introduction of organisms, provided 
of course that the operations be nimbly executed and that any piece of apparatus 
introduced into the glass have been suitably purified. For it appears that 
organic germs are not nearly so abundant in the atmosphere as is sometimes 
assumed, and only a very small fraction of the portions of dust with which the 
air of an occupied room is loaded have such germs adhering to them. Thus, 
in one instance the sole result of exposure of a glass of uncontaminated urine 
for half an hour in my study was three plants of three different kinds of fila- 
mentous fungi, each growing from one point and enlarging thence in all directions, 
while the liquid remained otherwise unchanged in aspect, a fact which may 
* See Nature, February 20, 1873. 
