4^^ SEDGWICK AND WINSLOW. — BACILLUS OF TYPHOID FEVER. 
In a lialf or three-quarters of an hour more. The time occupied by the whole 
process of freezing is recorded in the tabulation of each experiment. The tem- 
perature, in the first set of experiments with " Race A," was observed by means 
of three mercury thermometers inserted in different parts of the liquid, and at the 
time when the tubes froze the thermometers registered 6°— 7° below zero, C. In 
later experiments the temperature was observed by means of a minimum regis- 
tering spirit thermometer fastened to the inside of the cover of the inner cylinder, 
which recorded the temperature of the air just above the liquid in which the tubes 
were immersed. Partly on this account and probably partly because of its greater 
quickness of response, this thermometer gave lower records than did the mercury 
inslruinents in the first experiments. The readings of the spirit thermometer are 
given in the tables for each set of tubes. 
As soon as the tubes froze, they were removed from the freezer and either thawed 
at once or kept frozen in an ice-chest for a few hours, or placed in a cold-storage ware- 
house where they were kept for the longer periods at a temperature one or two 
degrees below zero, C. After the frozen condition had been maintained for the de- 
sired length of time, the contents of the tubes were thawed, shaken up, and sampled, 
again in duplicate. As a rule the samples taken from the thawed tubes were planted 
directly, while those made before freezing Avere diluted, one to ten, with sterilized 
water. All plates, for these quantitative determinations, were planted with common 
nutrient agar-agar, containing 1.25 per cent agar, LOO per cent Witte's peptone, and 
.25 percent salt, and having an acidity equal to 1.50 per cent. As the counts to be 
made were chiefly comparative, agar was preferred to any other medium, on account 
of its freedom from liquefaction. The plates were allowed to develop at the room 
temperature except in certain special cases to be noted later. Those made from the 
unfrozen water showed their maximum growth in three days and were counted after 
that Interval. Those made from the thawed ice, however, w^ere found to develop 
more slowly ; for them five days was generally found sufficient, although after the 
longer periods of freezing as much as ten days was sometimes allowed. The plates 
wure finally counted with the aid of a hand lens. 
In many of the sets of experiments a control tube was included, which was treated 
just like the others except that it was not inoculated. Each series of tubes Includes 
two lots of eight or ten each, frozen on two different days. 
The cultures were grown in bouillon (containing 1.00 per cent peptone, .25 per 
cent salt, and 1.00 pur cent acid), and were changed twice or three times a week. In 
the carlior experiments the tubes were inoculated from a culture grown at the room 
