GERM THEORY OF FERMENTATIVE CHANGES 315 
the caoutchouc screen, while the end of the syphon extends into the glass. 
The assistant is now directed to relax his hold upon the caoutchouc junction 
above, when the milk at once flows into the glass, and when this is about two- 
thirds filled, the flow is again arrested by the assistant, the glass removed, the 
cap, held in the other hand of the operator, is reapplied, and the glass placed 
again under cover of the shade. 
All the glasses having been charged, the caoutchouc screen is slipped off 
and a piece of carbolized cotton tied over the end of the syphon, which being 
raised to a higher level than the fluid in the flask the assistant finally relaxes 
his hold and the syphon empties itself into the flask, becoming occupied by air 
filtered by the cotton tied over the extremity. When at any future time another 
set of glasses are to be charged, all that is needful is to remove the cotton-wool 
from the end of the syphon, re-adapt the syringe by means of a caoutchouc 
adapter, steeped for a short time in carbolic lotion, and then proceed as before. 
In this way we avoid the great loss of time involved in providing a fresh syphon 
for every fresh decanting, as I did in the earlier experiments. 
The other experimental fluids employed in the observations about to be 
related were Pasteur’s solution, turnip infusion, an ‘artificial milk’, consisting 
of a solution of sugar of milk and white of egg in water, and urine. 
In preparing the Pasteur’s solution for this set of experiments I deviated 
from Pasteur’s formula in two respects; viz. the proportion of the water and 
the source of the mineral salts. I doubled the quantity of water, so as to make 
the liquid, as I hoped, more favourable for the growth of some organisms, more 
especially after loss by protracted evaporation as occurs in my experiments, 
and tap water was employed instead of distilled, so as to afford greater variety 
of saline material. For the yeast ash, which every one who has tried must 
have found extremely troublesome to prepare, I substituted the same weight 
of ashes left after burning a large amount of loppings from various kinds of 
trees and shrubs ; the liquid obtained by lixiviation being filtered, and a quantity 
used in proportion to the estimated weight of dissolved solids. It seemed to me 
that the salts obtained in this way would be more likely to afford suitable 
pabulum for the growth of different organisms than those derived from one 
particular species of fungus. Thus, my Pasteur’s solution had the following 
composition :— 
Water from the tap. ; . 5000 grains 
Lump sugar ae ; : r w250-* 4. 
Tartrate of ammonia . BOs! 
Salts from wood ashes . . , Da 
It happened that the alkalinity of the ashes exactly counterbalanced the acidity 
