THE DETERMINATION OF BACTERIA IN ICE CREAM. 3 
melting being hastened by frequent shaking with a circular motion. 
At the end of the 15-minute period each flask containing the melted 
ice cream was again shaken 30 times, with a circular motion, in order 
to mix the sample thoroughly and to shake out as much air as possible. 
One cubic centimeter of melted ice cream was then removed from 
each flask and placed in 99 c. c. of sterile water at a temperature of 
40° C. (104° F.) The water in all the dilution bottles was at a 
temperature of 40° C. (104° F.) in order to keep the fat in a melted 
condition. 
The pipettes were so graduated as to deliver 1 cubic centimeter 
between two marks. This avoids the necessity of blowing out the 
pipette or immersing the end in the dilution water, and therefore 
eliminates the introduction of varying quantities of melted cream 
which adhere to the pipette. 
The dilutions were made in the usual way, using 99 c. c. and 9 c. c, 
respectively, of sterile water. Each dilution bottle or tube was shaken 
25 times, and great care was taken to measure the quantity accu- 
rately in the pipettes. Standard beef-infusion agar was used, and 
sufficient medium was prepared to last through the work; conse- 
quently no factor of variation was introduced by the plating medium. 
The plates were incubated at 30° G. (86° F.) for a period of five days, 
after which the duplicate plates were counted. 
VARIATION IN THE BACTERIAL CONTENT OF COMMERCIAL ICE CREAM. 
VARIATION IN DUPLICATE SAMPLES FROM VARIOUS PARTS OF THE SAME LOT. 
In our first experiment twenty- two 1 -gallon lots of ice cream were 
obtained from seven different manufacturers. This cream, as in- 
tended, was of different flavors, was made in different ways, and 
included products containing different ingredients and varying 
percentages of butterfat. 
The complete results obtained from a study of these samples are 
shown in Table I, in which is recorded the percentage of fat in the 
ice cream from each manufacturer, also the presence or absence of 
gelatin, the flavor of each lot, the dilution used in plating, the num- 
ber of colonies found on each of the duplicate plates, and the calcu- 
lated average number of bacteria in a cubic centimeter of melted 
ice cream. Where there is a blank space in the number, of colonies 
on duplicate plates no count could be made on account of " spreaders," 
which entirely obscured the colonies. Every count that could be 
obtained is included in this table of results, and no count was left 
out as being a "freak" result. 
A study of the table shows that the bacterial counts of the nine 
samples from as many different positions in each gallon lot of ice 
cream check remarkably well with one another. 
