400 
ature. If the milk turns sour in this time and curdles normally it j 
was taken as an indication that antiseptics had probably not been ij 
added. On the other hand, if it did not curdle in this time, under ? 
these conditions, it was regarded as possibly containing preserva- 
tives and was systematically examined for all substances ordinarily | 
employed as milk preservatives by the methods described by Leach, 1 
Yan Slyke, and other well-known authorities on the subject. ■ 
(2) A considerable number of the samples were tested for preserva- 
tives by Blyth’s (11) method for the detection and estimation of pre- ‘ 
servatives in milk. This method is carried out in the following man- 
ner : Ten cubic centimeters of the milk is put into a clean, wide test 1 
tube, and into another tube for purposes of comparison and control S 
are put 10 cubic centimeters of a sample of milk of known purity. To 
each tube 2 cubic centimeters of a strong aqueous solution of blue ' 
litmus is then added, and after plugging with cotton wool the tubes i 
are sterilized by heating to 80° C. for ten minutes. The tubes are 1 
then removed from the sterilizer and cooled to ordinary temperature. ■ 
Each tube is then inoculated with 0.5 cubic centimeter of a solution ' 
containing 0.5 cubic centimeter of sour milk and 100 cubic centimeters 
of water. After thoroughly mixing, the tubes are kept at 15° to 25° ‘j 
C. for twenty- four hours and are then examined. The tubes contain- *1 
ing samples of milk which contain preservatives will be colored blue 
or pink, whereas the tubes containing milks to which no preservatives . | 
have been added will be of the same color as the control experiment | 
with normal milk, viz, white or nearly so. This test dej^ends upon i 
the fact that in the normal souring of milk the colored substances 1 
present in litmus are reduced by the bacteria to colorless (leuco) com- i 
pounds. ' 
(3) All of the samples of milk without exception were tested for 
formaldehyde and boric acid by the methods described by Leach 
(12). The test for formaldehyde described by this author as the 
hydrochloric acid test is capable of readily detecting 1 part of for- j 
inaldeh}"de in 250,000 parts of sweet milk and 1 part in 50,000 in i 
sour milk. It has been shown by Rideal and Foulerton (13) that at ' 
least 1 part of formaldehyde in 50,000 is required to preserve milk i 
for twent 3 ^-four hours, so that this test is capable of detecting much 
smaller quantities of formaldehyde than is ever emploj^ed in prac- >' 
tice. During the month of September, during which time special 
attention was paid to the subject of preservatives in the milk, 20 
cubic centimeters of each sample of the milk was distilled and a few 
cubic centimeters of the distillate collected in a small amount of dis- 
I 
