72 STORRS AGRICULTURAL EXPERIMENT STATION. 
determinations it was impossible to determine the dirt by this 
method where so many samples were handled each day. An 
arbitrary scale for determining the dirt was therefore adopted. 
By this method each sample of milk was marked according to 
the dirt which was visible in the bottom of the sample bottle 
after it had stood a certain length of time. The scale used ran 
from 1 to 7. A sample containing a very slight amount of 
sediment was marked 1; one containing slightly more was 
marked 2; still more 3, and so on up the scale, 7 indicating 
samples containing the greatest amount of dirt. These figures, 
as will be seen, represent the relative amount of visible dirt 
which the different samples contained and are valuable simply 
as a means of comparison. In certain cases centrifugal tests 
were made to determine the amount of dirt, and in certain 
others the weight of the dried dirt was determined by the 
method described in Bulletin No. 25 of this Station, page Io. 
These results will be given later in connection with certain 
other data. It was found that the marks given on the basis of 
the above arbitrary scale compared very closely with the actual 
amounts of dirt determined by weight. It is believed, there- 
fore, that this method gives a fairly accurate basis for compar- 
ison of the dirt which the milk contained. 
After the plates had been incubated in the laboratory for 
six days at 20° C. they were studied and the following deter- 
minations made: 
1. The total number of bacteria present. 
2. The number of acid producing bacteria. 
3. ‘The number of rapid liquefying bacteria. 
4. ‘The number of slow liquefying bacteria.. 
5. The species or types comprising the greater part of the 
organisms present. 
After these facts were ascertained the percentage of acid pro- 
ducing bacteria and the percentage of liquefiers taking the two 
groups together was also determined. | | 
It will be readily understood that in order to give the results 
of such a line of experiments it will be necessary to use a vast 
number of figures. In fact it is practically impossible to pre- 
sent the data thus obtained in any other than in tabular form. 

