CLASSIFICATION OF DAIRY BACTERIA. IOI 
tolerably constant, but paler yellows are of far less significance. 
In our classification the color of yellow or yellowish cannot, 
therefore, be .regarded as particularly important, and we are 
convinced that many of the varieties described as yellow are 
identical with others described as white. Where we have good 
reasons for believing this is the case, we have mentioned the 
fact in the following pages by cross references. 
In our opinion the other characteristics commonly used for 
describing bacteria are less significant. ‘The appearance of the 
colony in gelatine is subject to wide variations, and while some- 
times it is useful, it cannot be relied upon as a rule as of very 
great importance. ‘The growth on potato varies more or less 
widely with the nature of the potato, and so decided are these 
variations that we have ceased to place much confidence upon 
the characteristics here described. The growthin bouillon has 
appeared to us to be of less importance than some have thought, 
even the formation of a scum being somewhat variable in differ- 
ent cultures of the same organism. 
The question of nomenclature is, of course, a puzzling one in 
dealing with a new group of organisms, whose inter-relations 
are as problematical as that among bacteria. Nevertheless, it 
is desirable that certain names should be given in our classifica- 
tion. In the following pages, we have adopted the following 
plan. We have named each of the general groups of organisms 
which we have recognized as more or less clearly marked types. 
‘The names that we have given to these are in all cases where 
possible names already applied to the chief members of the group. 
In many cases we cannot find among described bacteria any which 
agree closely enough to admit of identification, and in regard to 
many others, while there are among described types some which 
agree with ours as far as they go, the published descriptions are 
too incomplete to admit identification. In both of these cases 
we have given new names. Inour names we have in practically 
all cases used the word /actis in thename. Weare fully aware 
of the objections to using three names for a species, but in these 
cases the manifest advantage of indicating by the name the fact 
that the organisms are associated with and found in milk prod- 
ucts is so great that we regard it as sufficient to overcome the 
disadvantage. We have used this method in all cases except a 
few well known organisms where it was manifestly unwise. 
