2 BULLETIN 361, U. S. DEPARTMENT OF AGRICULTURE. 
of the sediment test they never had a means of quickly determining 
the exact amount. It has also been a fact long and fairly well estab- 
lished that milk containing sediment or visible dirt, such as manure, 
hair, etc., was produced under insanitary conditions, but when these 
ingredients were not present in the milk no field inspector could 
determine its purity. 
Upon the adoption of the sediment test as a means of detecting 
insanitary milk at the milk-receiving stations, the producers un- 
doubtedly began to use methods calculated to remove the visible 
dirt. Such methods have been resorted to as straining the milk 
through cotton, cheesecloth, and Canton flannel to prevent the 
detection of visible dirt at the station by the field inspector. These 
methods have so changed the value of the sediment test as a means 
of judging pure milk that when no sediment or visible dirt can be 
detected it is often almost impossible to state whether the milk is 
produced under sanitary conditions or not. In order to determine 
whether the sediment test could be wholly relied upon as a means 
of detecting insanitary milk at milk-receiving stations, an experiment 
was conducted with this purpose in view. 
OBJECT OF THE WORK. 
The object of this experiment was to prove whether milk contain- 
ing little or no visible dirt, as often occurs when filtered through 
certain substances by gravity, was free from a large number of bac- 
teria. It was decided that by comparing the bacterial count with 
the sediment test (also when milk was filtered through various 
utensils) certain information could be obtained regarding this points 
OUTLINE OF EXPERIMENT. 
Briefly, the experiment was conducted as follows : 
Three of what we considered the most practical sediment-test 
apparatuses were used, namely, the Gerber, the Wizzard, and the 
Lorenz. The Gerber apparatus was selected because it represents 
a gravity method. The average length of time required for one 
pint of milk to pass through the disk by this method was 15 minutes. 
The Wizzard was selected as a pressure type which could be easily car- 
ried for field work and attached to the milk bottle without removing 
the milk. By this method the time required for the milk to pass 
through the disk was about two minutes; its disadvantage was that 
when the pressure was applied there was no means of holding the 
apparatus securely to the bottle. The Lorenz apparatus was se- 
lected as a pressure type in which the milk is placed in the metal 
container and the pressure applied. The time required by this 
