270 
when applied to methods will still further improve the quality of 
the butter made. 
It may not be out of place here to state in a general way some of 
the properties of butter which distinguish it from cream. Butter 
consists essentially of the fatty portions of milk which have coalesced 
during the process of churning, and have entangled with it some of 
the casein and other matters which the cream contains, so that the 
chemical composition of the two varies only in the relative amount 
of the different ingredients. In extreme cases they may be almost 
identical so far as their chemical composition is concerned, and yet 
no one would be likely to mistake the one for the other, their physi- 
cal composition being very different. In cream the fat globules are 
all separate, a thin film of serum preventing them coming in actual 
contact; in butter these films are broken, the fat of the adjoining 
globules being in contact and adhering. Cream is an emulsion, but- 
ter a mass of fat retaining in its structure a portion of serum or other 
emulsifying liquid. Cream, with care, may be diluted indefinitely 
with water, the globules of fat becoming uniformly distributed 
throughout its volume. Butter on the other hand when placed in 
water will retain its form and in no way be affected unless the agita- 
tion is violent and in no case will it have the character of an emulsion. 
There have within the last few years been quite a number of meas- 
urements made of the size of the fat globules in milk from different 
animals and from different breeds. While there is no reason for 
doubting the conclusions derived from these measurements, viz.: that 
the size of globules varies largely in the milk from different breeds, 
the largest globules being associated with those noted for butter; and 
further that large globules are, in general, associated with facility 
of separation of cream and ease and economy of churning, still the 
methods, by which the results have been interpreted might lead to 
serious error, in so far as the size of the globules in the different 
cases have been taken proportional to thier average diameter as ob- 
tained from a number of micrometer measurements. This would 
be sufficiently accurate for most purposes if the globules were nearly 
uniform in size. But as the volumes of spheres vary not directly as 
their diameters, but as the cubes of their diameters, and as the aver- 
age of a series of numbers does not determine what the average of 
their cubes will be, it follows with bodies varying to the extent often 
occurring in the fat globules of milk that their average diameter 
conveys no definite idea of their volume, which is really in most 
cases what is sought. An example willmake thisclear. If we have 
three spheres whose average diameter is 2, their individual diame- 
ters may be 2, 2,2 or 1,2, 3 or 1, 1, 4 in which cases their average 
volumes would be as 8 : 12 : 22 respectively. 
The average volume of milk globules, owing to their variable size, 
is always greater than the volume indicated by their average diam- 
eter, the difference between two volumes being greatest when the 
globules vary most in size. It is therefore evident that if we wish 
to study the influence of size of globule upon the character of milk, 
their volume rather than their average diameter should be considered. 
The volume of milk globules may be dedetermined directly from 
their individual diameters or their average volume may be found 

