Calibrating the Glassware 99 
the fat is measured. With this last it is possible 
to read easily to .01 of 1 per cent. 
The fat in the solid milk products, as butter and 
cheese, may also be conveniently determined by the 
Babcock test. Since butter or cheese cannot be 
measured, it is necessary that the sample to be 
tested be weighed. Balances sensitive to .1 of a 
gram are sufficiently delicate. Hither 18 grams of 
the substance may be weighed, in which case the 
percentage of fat. is read directly from the bottle; 
or, what is more convenient, any amount from 4 to 
8 grams may be taken. In the latter case, the 
observed reading of fat bears the same proportion 
to the percentage of fat in the substance taken that 
the weight of the sample taken bears to 18; and the 
percentage is found by multiplying the observed read- 
ing by 18 and dividing the result by the weight of 
the sample taken. 
In testing butter and cheese, it is convenient to 
use the bottles with detachable necks. A _ little 
water should be added to the bottle before the 
acid is put in, to aid in the solution of cheese. 
Calibration of glassware.—The correctness of the 
graduation of the glassware may be tested with more 
or less accuracy according to the means at hand. 
The bottles are all graduated ‘on the assumption 
that the tubes are of uniform caliber. The O and 
10 points are determined experimentally, and the in-| 
tervening space equally divided into 50 divisions 
with a dividing engine. The spaces should, there- 
fore, be of uniform size, and if the eye can detect 
