New YorK AGRICULTURAL EXPERIMENT STATION. OI 
bers per gram approximating less than the above dilution would 
usually be overlooked. 
The differentiation of those bacteria that develop colonies on the 
gelatin plates is based on their production of acid, of liquifying 
enzym or of some characteristic appearance in the colony peculiar to 
the particular type of organism. 
It is fairly easy to separate the acid-producers from those that 
do not possess that power. This may be done by the addition of 
calcium carbonate or litmus to the gelatin. We used litmus in our 
work. The litmus must be sensitive, otherwise the weak acid- 
producers may not be recognized. The condition of the organisms 
at the time of plating the sample also seems to affect the results. 
An illustration of this is found in the case of old cheese. Many of 
the lactic organisms become weakened after the cheese has been in 
the curing room for three or four months. When a sample is 
plated from such cheese these organisms often fail to develop 
enough acid to turn the litmus red. 
The separation on the basis of the liquefication of gelatin is a 
fairly accurate method but the group classed as slow liquefiers may 
be often overlooked. The ability of the liquefiers to produce pro- 
teolytic enzyms is subject to variations which afte not yet fully 
understood. The presence of sugar impedes the liquefaction and 
slight variations in the composition or reaction of the media may 
entirely inhibit it. 
After all the colonies had been separated into four classes by 
means of the acid production and liquefaction further subdivision 
was necessarily based upon differences in size, shape, structure and 
color. These latter characteristics are subject to greater variation 
than either the acid production or the liquefaction, so that here the 
personal equation was necessarily large. Jn order to reduce the 
danger of missing any species present on the plate our aim was to 
recognize more types than were present, the recognition being based 
on very slight variations such as may be found among colonies on a 
plate prepared from a pure culture. We checked the possibility of 
including different species under a single head by isolating two or 
more pure cultures of each provisional group, and relied upon the 
system of classification, to be later described, to determine the dupli- 
cates. While we recognize the limitations of separating bacteria 
into species by means of the methods mentioned above we feel fairly 
certain that we missed very few important species of bacteria among 
those that developed on the plates. 
