90 STORRS AGRICULTURAL EXPERIMENT STATION 

Fic. 31—Photograph showing filthy condition of cows where experiments 
given in table 21 were made. 
better on the day the last samples were taken than they had 
been on the preceding days yet even here the relative differ- 
ence is very striking. While the samples from the open pail 
averaged considerably over 3,000,000 bacteria per cubic centi- 
meter the samples from the covered pail averaged but a little 
over 100,000 or practically 3 per cent. of the number found 
in the milk from the open pail. This means that over thirty- 
three times the number of bacteria fell into the milk drawn 
into the open pail as fell into the milk drawn into the covered 
pail. The real significance of this is apparent when we re- 
member that the number of bacteria in fresh milk is an in- 
dication of the amount of dirt which has gotten into the milk. 
The diagram given in connection with this table shows in a 
striking way the relative germ content of these two series of 
samples. The length of the lines in this diagram indicate 
the two averages so that the difference in their length repre- 
sents graphically the difference between 103,000 and 3,439,- 
000 bacteria per cubic centimeter of milk. Comparing the 
results obtained in these three different barns it is evident 
that the relative value of the use of the covered pail as a 
means of excluding dirt and bacterial contamination varies 
