40 BULLETIN 642, U. S. DEPARTMENT OF AGRICULTURE. 
sterilizer was demonstrated to the farmers, who were then left 
to operate it without further assistance. As the work was carried on 
during the summer and the cows were in pasture most of the time, 
Fic. 19.—Interior of barn at farm No. 27. 
their udders were 
always practically 
free from visible 
dirt; therefore that 
factor received no 
further considera- 
tion. 
At each milking 
the milk was stirred 
by means of a long, 
sterile pipette, after 
which samples were 
taken from each can 
and a composite 
sample made, which 
was immediately placed in ice water and plated within an hour and 
a half. For a period of three weeks samples were taken at each 
farm under the usual conditions (open pails, not sterilized) and for 
the three weeks following, when small-top pails, sterilized, were in 
operation. 
Table 18 shows the bacterial count of each sample obtained from 
the night’s milk, fresh, on each farm during the entire six weeks, also 
