12 BULLETIN 642, U. S. DEPARTMENT OF AGRICULTURE. 
a week. It will be noticed that the walls and ceiling were extremely 
dirty. Hay protruded through the open ceiling, and the gutter was 
completely filled with manure which was scattered over the floor 
toward the stanchions. 
The condition of the cows is well illustrated by figure 7. Their 
flanks were caked with manure, but the visible dirt was removed from 
the udders, which were washed before each milking. After having 
been thoroughly washed with clean water and a cloth they were wiped 
with another clean, damp cloth rinsed in clean water. The flanks of 
the cows, which were liable to be rubbed by the milker’s arm, also 
were kept clean. : : 
During the period from November 10 to 24, 1915, 18 samples of 
milk were examined from the open and 23 samples from the small-top 
pail. The results in Table 3 show that the average number of bac- 
teria in milk from the open pail was 6,166 bacteria per cubic centi- 
meter, compared with 2,886 from the small-top pail. 
In the experiment 12 samples of milk were drawn directly from 
the udder into sterile tubes at about the middle of the milking. The 
average count of the milk from the four cows was 987 bacteria per 
cubic centimeter. The difference between that average and the aver- 
age in the milk from the open pail is 5,179, which represents the 
number of bacteria per cubic centimeter introduced through external 
contamination. ‘The average number of bacteria per cubic centimeter 
of 23 samples from the small-top pail was 2,886; subtracting 987 from 
the count, the remainder, 1,889 bacteria per cubic centimeter, repre- 
sents the number added through external contamination. 
TABLE 3.—Bacteria per cubic centimeter in dual samples of fresh milk produced 
under conditions described in Experiment No. 3. 
| i 
|Sample| Open Small- | Sample) Open Small- 
Date. | No. pail. top pail. | Date. No. pail. | top pail 
1915 1915 
Nov.0; pps 1 3,500 2,400 | Noy 19° p. miS-- 2. 16) [esos 4,500 
Nove tian 2| 38,000] 3,100 || Nov. 20,a.m........- 17|  3,600| 4,300 
Nov. ll, p. M........- 3 1, 700 2,400 |] Nov. 20, p. m-....-.-- 15 [eee 1, 900 
Noyei2.a. mee 4 2,100 45300 || Nova22 armas anes 19 5, 600 2,600 
Noy. 12-1 5 | 11,200 1,500 || Nov. 22, p. m......... BS). ienereaee 5, 000 
Noy. d3)a2: nie eS 6 4,300 1: S00 NOVA 23 9a. In ae see 21 8,500 4,10) 
Noy. 43, ). m=: 7 2,800 1,000 || Nov. 23, p. m........- 22 | ae 2,709 
Novei5> pom es 8 5, 100 2,900 |} Nov. 24,a. m....._-.: 23 2, 800 4, (09 
Woys16,/a: m2 "c.052.- 9 2,600 2,700 | 
Noy-46spiemce— AO eee 1.200 Average of milk 
Noy. 17am 11 3, 200 2,700 Samples! 2-2 |pssseee= 6, 166 2, 886 
Novisl(, pe mls eee 12 i, 700 2,800 | Average of 12 
Novis asa 13 4,600 2,100 | udder samples.|.-....-- 987 987 
Nove 184pimee eee 14 3, 590 2,000 a SSS SS 
Noyes 19panm a5. 15 6, 200 3, 800 Difference: : = - 2 4aaseso-- 5,179 1,809 
The results show that when the udders were washed clean and the 
small-top, sterilized pail is used it is possible to produce milk of a 
bacterial count closely corresponding to the number found in milk 
