CLASSIFICATION OF DAIRY BACTERIA. 99 
some of the other morphological characters. We have in our 
classification, therefore, sharply separated the motile from the 
non-motile rods and the peritrichic from the monotrichic bacilli. 
In a few of our early described species in which flagella were 
not made out we have found it possible to group them from 
their other characters with later isolated cultures which are 
more carefully described. ' 
The question of the liquefaction of gelatine has also been an 
open one, for the data at our command seem to suggest that 
here, too, we have a character that is somewhat variable. 
Some types show this property of liquefaction only after two 
or three weeks’ growth, and we have considerable data to indi- 
cate that the power of liquefying may be completely lost; in- 
deed, among the organisms isolated from milk we have occa- 
sionally found two that are identical in every respect except in 
this power of liquefaction, and organisms that have such pecul- 
iar character as, for instance, the power of producing a pink 
fluorescence, as to convince us that we are really dealing with 
the same organism, but one in which the property of liquefying 
gelatine is capable of being totally lost. 
In the grouping of bacteria in our key we have used prom- 
inently the power of fermentation of sugars as a means of diag- 
nosis. This characteristic is one which has become recognized 
recently as quite significant, and for the purpose of the study 
of the groups of the bacteria of milk, it is evidently one of 
exceptional importance. The fermenting power of sugar is 
closely related to the action of the organisms upon milk, and 
clearly from the standpoint of dairy bacteriology a grouping of 
bacteria with this as a basis is one of the most practically 
useful methods of grouping these organisms. But here, too, it 
appears to us that variations are common. Some of our groups 
contain organisms which apparently show considerable differ- 
ence in their power of fermenting different sugars. It is com- 
mon to separate in different groups any organisms which would 
ferment dextrose but not saccharose from one that would fer- 
iment the two sugars. As we have compared together the large 
numbers of cultures from various sources that we have made, 
we have found such wide variations in this power of ferment- 
ing sugar that we have been inclined to believe that here, 
too, we have a variable factor, and that whereas the general 
