R12 
By means of an eye piece micrometer the internal diameter of the 
tube is carefully measured and then turning the eye piece so that the 
scale is lengthwise of the tube all of the globules included in a cer- 
tain number of divisions are counted. ‘The slide is then moved 
along until a new field appears and another count is made. This is 
repeated till five or six observations have been made with each tube, 
this being usually sufficient. If the tube has been carefully selected 
there is no need of making more than one measurement of its diam- 
eter, as it will be sufficiently uniform throughout the length used 
to make no appreciable difference. In a tube .1 millimeter in diam- 
eter the number of globules which will be found in a length of .1 
millimeter will range, with milk diluted to 50 times its volume, from 
50 to 100. 
This operation is usually repeated with two or three tubes con- 
taining the same milk and the results calculated for a tube of uni- 
form diameter and length, so that they may be directly compared. 
To facilitate calculations the standard diameter of the tube is taken 
as 100 divisions of the micrometer and the length as 50 divisions, this 
being the length of the field that has been included in one observa- 
tion. The number of globules in a tube of this length and diame- 
ter of 100 divisions, is found by the formula ae in which » 
is the number of globules found in the standard length of tube, and 
d is the measured diameter of the tube. The numbers thus found 
are sufficient for the comparison of any two milks, but in order to 
make them of general application the numbers should represent the 
globules in some definite volume of undiluted milk. They can be 
reduced to this standard by multiplying them by a constant number 
which combines the ratio between the volume of the above tube and 
the volume desired, and also the dilution of the milk. ‘The figures 
given in this report are calculated for .0001 cubic millimeter of milk, 
this being a convenient unit and one which does not give numbers 
too large to be comprehended. 

It may be interesting to the curious to know that ten ciphers add- _ 
ed to the number of globulesin .0001 cubic millimeter, will give the 
number in one litre or approximately one quart of milk. 
The average volume of the globules may be obtained by dividing 
the volume of fat, as derived from analysis, in .0001 cubic millime- 
ter of milk by the number of globules which it contains. But as rela- 
tive values only are of use, it is sufficient to divide the per cent. of 
fat by the number of globules. The relative volumes in this re- 
port have deen derived in this way, but, to avoid a long decimal, 
the numbers given have been multiplied by 10,000. To be strictly 
accurate the specific gravity of the milk and the butter fats should be 
considered, but as this would scarcely change the relation of the fig- 
ures, it is omitted. With this modification the numbers given rep- 
resent in absolute volume ten billionths of a cubic millimeter. 
An example that has occurred in actual practice is presented in or- 
der to illustrate the methods used and to indicate how great varia- 
tion may be expected in duplicate determinations. It has been se- 
lected on account of the wide difference in the diameter of the tubes 
used, and is in no respect better than may be expected if due care 

