CLASSIFICATION OF DAIRY BACTERIA. 107 
method of replating, and, after purification, they are inoculated, 
upon agar streaks. After about two days’ growth on agar they 
are inoculated into the various other culture media. 
The morphology of the various organisms was determined from 
fresh agar streak cultures although, to determine the formation 
of spores, it was sometimes necessary to use older cultures. To 
determine the motility we have usually used an ordinary bouillon 
culture of 12-24 hours’ growth. We have found that a careful 
study of a hanging drop of such a bouillon culture is most satis- 
factory for this purpose. To determine the presence of flagella 
we proceeded as follows: from an actively growing bouillon cul- 
ture a drop was removed with a platinum loop and spread thinly 
over the surface of solidified agar. This was then incubated at 
37° for 12 hours. A-small quantity was then removed, diluted 
in three successive drops of sterilize water and stained by the 
well known Loeffler method. 
‘The culture media we have used have been those ordinarily 
employed. Our bouillon, gelatine and agar have been made with 
Liebig’s beef extract instead of chopped beef, because of greater 
convenience and uniformity. The fermentation tube test was 
made with dextrose, lactose and saccharose, 1 per cent. of these 
sugars being added to ordinary bouillon. In all the cultures 
tested by us in the last five years we have used all three sugars. 
The mz/k which we have used has in all cases been skimmed 
milk, sterilized by boiling for 10-15 minutes on three successive 
days. Potato cultures have been made by cutting plugs from 
large clean potatoes, slicing them once obliquely, and then soak- 
ing them in running water over night. After this they were 
placed in tubes and sterilized in an autoclave. 
In general the terms which we have used in our descriptions 
have been those suggested by Chester and adopted by The 
American Society of Bacteriologists. In stating whether or not 
an organism produces acidity we refer to the production of acidity 
in dextrose bouillon. In some cases, as will be seen, organisms 
produce acidity in lactose but not in dextrose; but these are rare. 
The detection of acidity has been made by meaus of carefully 
prepared litmus paper. This paper we have prepared ourselves 
from neutral litmus. The test for acidity has been made in 
about two days, and also later. This method has been adopted, 
both for the fermentation tubes and for milk. In determining 
