370 MILK AND MILK PRODUCTS 
The shaking of the sample must be done thoroughly and quickly, 
and the test bottle inverted several times, so that the liquid in the neck 
becomes thoroughly mixed. By pressing in the rubber stopper the height 
of the liquid can be brought to about the zero point on the scale. 
If only a few samples have to be analyzed and the room is warm, 
the test bottles can be put into the centrifuge without any preliminary 
heating, otherwise the test bottles must be warmed for a few minutes 
(not longer) in the water bath at a temperature of 60° to 65°C. When 
the temperature rises higher than this, say above 70° C., the rubber stop- 
per is liable to be blown out of the test bottle. After the test bottles 
have been heated they are arranged symmetrically in the centrifuge and 
whirled for three to four minutes at a speed of about 1000 r.p.m. When 
the centrifuge has a heating arrangement attached to it, the preliminary 
warming is not, of course, neccessary. When the test bottles are taken 
out of the centrifuge, they are again placed in the water bath at a tem- 
perature of 60° to 65° C., and left there for several minutes before being 
read; where the centrifuge is heated, the tubes can be read off as taken 
from the centrifuge. 
By carefully screwing in the rubber stopper, or even by pressing it, 
the lower limit of the fat column is brought onto one of the main 
divisions of the scale, and then, by holding the test bottle against the 
light the height of the column of fat can be accurately ascertained. 
The lowest point of the meniscus is taken as the level when reading the 
upper surface of the fat in a sample of whole milk, and the middle of the 
meniscus for separated milk. 
If the column of fat is not clear and sharply defined, the sample 
must be again whirled in the centrifuge. 
Hach division on the scale is equivalent to 0.1 per cent, so it is very ' 
easy to read to 0.05 per cent, or, with a lens, to 0.025 per cent. If the 
number which is read off is multiplied by 0.1, then the percentage quan- 
tity of fat in the milk is obtained; e.g., if the number on the scale was 
36.5, then the percentage of fat is 3.65. (Milk and Dairy Products, 
Barthel; translated by Goodwin, p. 71.) 
Arla-Jensen (1916) advises the use of chemically pure amyl alcohol. 
PROTEINS IN MILK 
Procedure. Measure 5 ¢.c. of the milk into a 500 c.c. Kjeldahl flask 
delivering it to the bottom of the flask. Add 25 c.c. of concentrated 
sulphuric acid and 0.7 gm. mercuric oxide or the equivalent of pure 
mercury. Heat the flask directly over a free flame in a hood until the 
