NEW YorRK AGRICULTURAL EXPERIMENT STATION. Tle 
The above table shows that the machine milk contained 
about 43 times as many bacteria as that drawn by hand. 
It is also interesting to observe that under these conditions 
the simple rinsing which the machine received after the even- 
ing milking was practically as efficient as the laborious wash- 
ing and steaming which were given it after the morning 
milking. 
It will be noted that the germ content of the hand-drawn 
milk in the morning is more than three times as great as that 
obtained at night. This is undoubtedly due to the superficial 
manner in which the cows were cleaned and to the dust in the 
air as the result of their being cleaned and fed immediately 
before milking in the morning as was explained under the 
heading of technique. 
There is no known, reason for the extremely high numbers 
found in the hand-drawn milk on the morning of August 25. 
Since it was so out of keeping with the other results it was 
given in the table but not included in the averages. 
The data in the above table were restricted to the period 
including August 20 to 30 since it was during this time that 
samples were taken both morning ‘and evening and the experi- 
mental conditions for accurate comparative work were at their 
best. During the study of this question with the Globe ma- 
chine there were examined 39 samples of machine-drawn milk 
having an average germ content of 81,100 per cc. and 36 sam- 
ples of hand-drawn milk with an average content of 16,800 
per cc. 7 , 
In these studies with the Globe machine the results obtained 
from its use were contrasted with those obtained by milking 
by hand into a 13-inch open pail under comparable barn con- 
ditions. It soon became evident that as a basis of comparison 
the results obtained with an open pail were of little use. This 
was partly because there was little constancy in the results 
obtained with an open pail ‘and largely because the factors 
which govern the results obtained by the two methods of milk- 
