( 583 ) 
XXVIII.— 0?i a New Method of Testing 3Iilk. By Dr. 
Augustus Voelckke, F.R.S. 
The milk of healthy cows, it is well known, varies in compo- 
sition and quality to some extent, according to a number of 
conditions, such as the age and breed of the cows, the time of 
the year when they are milked, distance from the time of calving, 
the liberal or scanty supply of food, and its quality and 
suitability for producing either rich or watery milk, and other 
circumstances to which it is not necessary to refer in particular 
in this place. 
Whilst, however, under ordinary conditions of proper feeding, 
the proportions of curd, milk-sugar, mineral matters, and water 
do not vary to any large extent in cows' milk, the percentage of 
butter-fat, its most valuable constituent, is subject to con- 
siderable fluctuations, which proportionately affect the quality 
and economic value of milk. 
The estimation of pure butter-fat in milk by a ready method, 
requiring on the part of the operator no special training in 
chemical manipulation, has long been felt a great desideratum ; 
for there can be no doubt that such a method would be a boon 
to milk-contractors, to the makers of butter, and to all persons to 
whom it is matter of importance to be able to ascertain without 
delay whether milk is rich or poor in butter-fat. Butter-fat, it 
need hardly be mentioned, occurs in milk in the form of cream- 
globules, which rise to the surface as cream on allowing milk 
to stand at rest for a period of from six to twelve hours. 
Instruments for measuring the percentage of cream in milk 
are called " creamometers." They are made either in the form 
of a cylindrical measuring-glass, on a glass foot, and divided 
into 100 equal measures, or in the form of graduated wide 
glass tubes. The gradations proceed downwards, from a point 
near the open end, marked zero, and each division indicates 
1 per cent, of cream. Such creamometers have long been used, 
and are still employed, for testing the quality of milk. I 
showed, as far back as 1863, that the indications given by 
creamometers are unreliable in forming a correct estimate of 
the richness of milk in butter-fat. 
There are two circumstances which seriously interfere with 
the practical use of the creamometer. The first is, that the 
cream which rises from different kinds of milk often varies 
greatly in composition. Cream may contain not more than 
18 per cent, of pure butter, or it may contain from 24 per cent, to 
28 per cent. The cream of small breeds of cattle is generally 
richer than that of large breeds ; and milk, although it throws 
