342 
Notes. 
It is used as follows. A table-spoonful of the Kephir-like grains 
is put into a litre-bottle — a champagne-bottle does well — with three 
table-spoonfuls of moist sugar ( cassonade ), and the bottle quite filled 
with water, securely corked and tied. 
It is then left for three days, when a violent fermentation is found 
to be in play, so much C0 2 escaping that the pressure may burst the 
flask or blow the cork out violently if care be not taken. 
The liquor is now strained into a second strong flask and securely 
corked down, the sago-like grains being returned with more sugar 
and water to the first flask to repeat the process. In four days it is 
ready for drinking, and is a slightly turbid, violently effervescing, 
lemonade-like drink, to which all kinds of curative and medicinal 
properties were ascribed by the missionary. 
A microscopic examination of the grains show that they consist 
almost entirely of the sheathed form of a Bacterium so like B. vermi- 
forme that I have little doubt it will prove to be the same. In much 
smaller quantity I found a yeast with long sausage-shaped pyriform 
and oval cells not very like S. pyriformis , and probably different 
from it. 
On following the directions, I found the fermentation to occur 
exactly as described. 
On carrying the experiments further, I was impressed by the 
particular stress laid on the direction to fill up the bottles and cork 
thoroughly, since this seemed to imply the necessity for keeping the 
organism out of contact with air. I therefore decided to try the 
following method. A bottle of Schweppe's soda-water was carefully 
opened, and a supply of ordinary sugar together with some of the 
‘ Paris Kephir ’ (as I term it in my notes) added, and the flask, quite 
full, rapidly corked and tied. To my surprise the fermentation at 
once began, and all the sugar disappeared in a few days, the gas- 
pressure being tremendous. 
The fermentation was almost entirely due to the Bacterium , very 
little yeast being present and apparently not increasing, and the 
conditions show that no oxygen is necessary to start the action. 
Here we have clearly a case of an aerobic Bacterium , capable 
of fermenting sugar to carbonic acid and some other organic acid 
— the liquid has a pleasant acid flavour at the end of the fermentation 
— -and requiring merely such traces of nitrogen as would be present 
in moist sugar. 
