26 STORRS AGRICULTURAL EXPERIMENT STATION. 
Neither Group VIII. nor IX. has seemed to be of any signifi- 
cance in our studies, since both disappear rapidly from the milk. 
In next to the last column of our tables we have placed 
the few miscellaneous species found sporadically in different 
samples. Frequently these have been clearly distinguished 
and separately enumerated in our studies of the plates, but 
since they are present in small numbers only and are likewise 
sporadic in the different samples, we have not regarded it as 
necessary to tabulate them specially. In this column, also, are 
included the undetermined colonies. In our study of the plates 
we first determined the total number of colonies, then the 
number of colonies of each clearly distinguished group. ‘The 
difference between the sum of these distinguishable colonies and 
the total numbers gave the number of indeterminate colonies. 
As will be seen by the tables that follow, it was possible in some 
cases to differentiate practically all of the colonies, so that the 
number of undetermined was zero; while in others the differ- 
entiation was more difficult, and this number was high. ‘The 
accuracy of the differentiation of the group thus varies inversely 
with the percentage of the undetermined colonies. Whenever 
this number is high, it has generally been due to the difficulty 
of distinguishing between the neutral bacteria and the lactic 
bacteria of Groups I. and II. Usually most of the undeter- 
mined should be added to Group I., since they were commonly 
faintly acid colonies, not clearly distinguishable as Bact. lactis 
acidt, A closer approximation to a correct result may thus 
usually be obtained by adding the first and the next to the last 
columns. 
In the tables that follow, both total numbers and percentage 
are given. In giving the total numbers only the significant 
numbers are recorded. If the number is under 1,000, it is 
given to the nearest 100; if over 100,000, it is given to the 
nearest 10,000; 1f over 1,000,000, to the nearest 100,000; and if 
over 10,000,000, to the nearest 1,000,000 only. ‘The percentage 
is printed in black faced type, and is calculated to the nearest 
tenth of one per cent. ‘This close calculation was necessary 
to bring into the tables at all some of the groups present in 
small numbers only, for these were frequently found in numbers 
representing less than one per cent. 
