386 
Eeport of the Chemist of the 
nature and are not indicated by chemical analysis. A low viscosity of 
the milk serum, accompanied by large fat globules, supplies favorable 
conditions for the economical production of butter. As the coefficient 
of viscosity for the fat of milk increases with the size of the globules, 
it follows that the greater the difference between the viscosity of the 
whole milk and that of the skim-milk the more valuable will the milk 
be for butter, provided, of course, that the viscosity of the skim-milk 
is not very high. 
When warm milk is run through a centrifugal, the skim-milk and 
cream being caught in the same vessel and thoroughly mixed, the 
product is less viscous than the original milk. This appears to be 
chiefly due to the breaking up of the fat globules. The extent to which 
this division takes place is shown by the following determination of 
the number of globules before and after separation: 
Viscosity. 
Number of 
globules in 
.oooi c. m. m. 
Per cent of 
globules with 
diameter less 
than one divi- 
sion of mi- 
crometer. 
Milk 
267 
149 
52.7 
After separation 
248 
201 
60.3 
After second separation 
249 
174 
70.0 
The second separation was effected at too low a temperature and 
resulted in partially churning the cream. Under the microscope 
many of the fat globules appeared distorted and adhered to each 
other; this partially accounts for the smaller number of globules and 
increased viscosity found after this separation. Although the per cent 
of small globules after the second separation is considerably increased, 
their number remains almost exactly the same as before, showing that 
in the process of churning the large globules are the ones which first 
unite to form granules of butter. 
The effect which this division of the fat globules has upon the 
quantity and quality of butter made from centrifugal cream is at 
present uncertain, no direct tests having been made. This pecu- 
liarity, however, may explain why the same manipulation which 
produces high-grade butter from ordinary cream often fails when 
applied to centrifugal cream. 
The hypothesis that the butter fats are derived directly from the 
food, and are transferred without change to the milk, as advanced by 
