360 Milk and Its Products 
sample, and a special moisture scale. The scale is specially adapted 
for moisture work, but may be used as a cream scale in operating the 
Babcock test. 
The scale has a tare weight for balancing ae cup and a large and 
small weight for weighing the sample and obtaining the percentage 
of. moisture. The beam has two rows of figures which give readings 
with the larger weight. The lower row gives readings in grams and 
the upper row in percentages. The smaller weight gives readings in 
grams when the weight is moved from 1 forward. Each notch repre- 
sents .02 grams, the total value of the small scale being .2 grams. When 
the small weight is moved from 0 backward, each notch represents a 
loss of .1 per cent of moisture when 20.2 grams of butter are used. The 
small weight is intended to be used only in moisture work. In using 
the scale for Babcock work, the small weight is not used but is left at 
rest on the figure 1. Then, when the scales are balanced, the small 
weight is negligible. Care must be taken not to let any draft of air, 
as from an open window, strike the scales when in use, as they are so 
sensitive that a very slight current of air would throw them out of 
balance. The scales will give readings in percentages only when 20.2 
grams of butter have been weighed or, in other words, when the large 
weight is on 20 (of the gram scale) and the small weight is on zero. 
The cup used is of cast aluminum and is durable and perfectly 
smooth. The absence of creases or crevices allows it to be cleaned and 
dried thoroughly. 
Operation of the test.—It is necessary that a representative sample 
be taken for a moisture-test. If the butter is sold in tubs, the sample 
should be taken from the tub with a butter trier, after the butter has 
been‘packed. It is best to take three drawings—one from near the edge, 
one from the middle, and one half way between the edge and the mid- 
dle. Some butter-makers test the butter as soon as it is worked. This 
is a mistake, since considerable moisture is lost in the process of print- 
ing and packing. 
Place the sample to be tested in a glass container which has a fairly 
wide mouth, so that the sample can be stirred. A quart fruit jar is 
useful for this purpose. Then hold the container in warm water until 
the butter begins to melt. Remove the container from the warm bath 
and thoroughly mix the melted with the unmelted butter. In the lab- 
oratory a long-bladed cheese knife was found very useful for mixing 
the butter. A wooden stirrer should not be used, as it is likely to take 
up moisture from the sample. The process of melting the butter and 
mixing it with the unmelted butter is repeated until the sample con- 
