ts
at the poles of the bacilli. It is extremely difficult to cultivate, and
forms flat, small greyish colonies of irregular outline. The bright
refractive spots above referred to are, however, granules taking the
usual stains quite readily, and not spores as supposed by Kern.

Freudenreich also found three other organisms--a yeast and two
streptococci. The yeast, to which he gave the name _Saccharomyces_
(Torula) _keffir_, forms small oval or roundish cells 2-3 mu wide and 3-5
mu long. The optimum temperature would seem to be about 72 deg. F.; the
maximum 82 deg. F. This organism is unable to ferment milk directly, but is
able to decompose maltose and glucose with gas production. It does not
coagulate milk, but imparts to it a characteristic taste and is unable
to withstand desiccation for more than a few days.

Of the two streptococci isolated, _Streptococcus a_ resembles organisms
of the group _Streptococcus lacticus_ in appearance, but is able to
ferment milk, with weak acid and gas production, and is capable of
inducing coagulation.

[Illustration: FIG. 11.--Section through a Kephir
Grain--highly magnified.]

Contrary to what one would expect in an organism existing in keffir
grains, this streptococcus is as little able to withstand desiccation as
the above-mentioned yeast. _Streptococcus b_ forms smaller cells as well
as smaller colonies than _Streptococcus a_, but produces more lactic
acid and more gas, and retains its vitality after desiccation. The
relation of these four organisms is, according to E. von Freudenreich,
as follows: _Sacch. keffir_ is unable to ferment directly milk or
lactose, so that its growth must be preceded by that of _Streptococcus
b_. _Streptococcus a_ does not seem to play this part, but, unlike
_Streptococcus b_, is able to coagulate milk on its own account. By the
combined action of the yeast and the two streptococci, then, milk can be
coagulated, milk sugar inverted, acid and gas produced by the
streptococci, while gas and alcohol are formed by the activity of the
yeast. The role of _Bacillus caucasicus_ is unknown, but it would seem
to play a part in the formation of the keffir grain itself. By means of
mixed cultures of the above organisms Freudenreich was successful in
obtaining a fermented product possessing in all respects the
characteristic properties of normal keffir. On the other hand,
experiments to induce the formation of keffir grains gave negative
results, but in