By E. MOLISABI.


[Illustration]

The author, after criticising the various methods for estimating fat
in milk which have been proposed from time to time, agrees with Stokes
(_Analyst_, 1885, p. 48), Eustace Hill (_Analyst_, 1891, p. 67), and
Bondzynsky (_Landwirth Jahrb. der Schweiz_, 1889), that the method of
Werner Schmid is the simplest, most rapid, and convenient hitherto
introduced. The conditions tending to inaccuracy are: The employment
of ether containing alcohol; boiling the mixture of milk and acid too
long, when a caramel-like body is formed, soluble in ether; the
difficulty of reading off the volume of ether left in the tube, owing
to the gradations of the instrument being obscured by the flocculent
layer of casein; when only a portion of the ether is used, fat may be
left behind in the acid mixture, as shown by Allen (_Chem. Zeit._,
1891, p. 331). The author believes that by the invention of the simple
apparatus represented in the accompanying figure, he has rendered the
process both accurate and convenient. This consists of a flask B of
about 75 c.c. capacity, which has a glass tap fused on, with two
capillary tubes attached, the one passing upward, the other downward.
The neck of flask B is ground into the neck of flask A, which holds
about 90 c.c. Either of the flasks can be placed in communication with
the external air by the opening _a_. The ether must be previously
washed with one or two tenths of its volume of water, to remove traces
of alcohol. The operation is performed as follows: 10 c.c. of well
mixed milk are weighed in (or measured into) flask A, 10 c.c. of
hydrochloric acid added, and the mixture heated to boiling on an