this respect the cultural characteristics of
_Lactobacillus keffir_ would seem to give promise of success in the
synthesis of the keffir grain. Fig. 11 is a photo-micrograph of an
extremely thin section through a keffir grain, after a preceding
treatment with saffranin. The matrix is composed entirely of long thin
bacilli (_Bacillus caucasicus_), while the peripheral portions, which
are more deeply stained, consist to a large extent of dense masses of
yeast cells with occasional streptococci. In a normal grain the latter
organisms are present on the surface or in the cavities and grooves of
the grain, and only to a less extent in the matrix. Nikolaiewa[57]
claimed to have isolated a hitherto unknown bacillus capable of
coagulating milk by acid production, _Bacterium caucasicum_, not
identical with, but related to Freudenreich's _Bacillus caucasicus_, and
also a torula. Although no experiments were carried out, Nikolaiewa
asserts that this organism forms the matrix of the grains. He was able
to produce a beverage resembling keffir, just as Freudenreich and
Essaulow did with entirely different organisms, but his product would
appear to have been slightly too acid and to have lacked the
characteristic aroma of the normal product. In the course of an
extensive series of experiments Kuntze[58] found the following
organisms:

(_a_) True lactic acid forming bacteria, _Streptococcus
acidi lactici_ (Grotenfeldt).

(_b_) Bacteria of the group _Bacterium acidi lactici_
(Hueppe) and _Bacterium lactis aerogenes_.

(_c_) Various torula and yeast species.

(_d_) Two species of butyric acid bacteria, _Bacillus
esterificans_ and _Bacillus keffir_ (Kuntze).

His conclusions are: 1. In any case the presence of a yeast capable of
directly fermenting milk sugar is not essential. 2. The significance of
the presence of yeast lies in the fact that stimulation of the lactic
bacteria occurs; further, the yeast exerts a regulating influence upon
the rapidity of the fermentation proper. The variety is of minor
importance, provided always that the yeast does not produce an
unpleasant flavour. By the use of mixed cultures of _Bacillus
esterificans_, _Bacillus keffir_, and _Streptococcus acidi lactici_, and
a keffir yeast, Kuntze obtained a product that possessed to the fullest
degree all the characteristic properties of a normal keffir. In such
cultures he was successful in obtaining the formation of keffir-like
grains. Keffir f