ON THE ESTIMATION OF SUGAR OF MILK. 371 
caustic potash. We have then to determine with care the 
strength of the alkaline solution, which is effected by ascer- 
taining the quantity of sugar required to decolorise a given 
quantity of the liquor. It is important to observe, that 
in this operation sugar of milk, and not cane sugar, is to be 
used. 
I have made a few experiments with the view of estab- 
lishing a test liquor, the value of which shall be known. 
The following proportions have always afforded a liquor 
twenty cubic centimetres, of which correspond to two 
decigrammes of whey : 
Take of Crystallised sulphate of copper ... 10 parts. 
Ditto bitartrate of potash ... 10 <{ 
Caustic potash ........ 30 " 
Distilled water 200 " 
The liquid, after being filtered, is limpid, and of an in- 
tense blue color. 
Preparation of the Whey. — In estimating the sugar of 
milk it is necessary to separate the fatty matter and the 
caseine by coagulation. This is easily effected by putting 
fifty or sixty grammes of the milk into a flask, adding a few 
drops of acetic acid, and then raising the temperature to 
40° or 50° Cent. A transparent liquid will be obtained by 
filtration. According to my experiments, 1000 parts of 
milk yield 923 parts of whey, 1000 parts of which latter 
contain about 57 parts of sugar of milk. 
The testing of the Whey. — Twenty cubic centimetres of 
the liquid to be tested is transferred by means of pipette to 
a flask, which is preferable to a capsule because the change 
of color can be readily perceived. The liquid is to be 
brought to the boiling point. On the other hand, a burette, 
each division of which is the fifth of a cubic centimetre, is 
to be filled with the whey; and this is to be added, drop 
by drop, to the test liquor, keeping it continually agitated 
and heated from time to time, until the blue color has en- 
tirely disappeared. In the first instance, a yellow precipi- 
tate of hydrated suboxide of copper is formed, which 
