VARIATIONS IN BACTERIA CouUNTS 239 
The results of the various compositions of media compared at all tem- 
peratures showed that 29 per cent of the highest counts were obtained on 
plain agar, 18 per cent on lactose agar, and 53 per cent on dextrose agar. 
Of the lowest counts recorded, 33.7 per cent were obtained on plain 
agar, 46.2 per cent on lactose agar, and 20.1 per cent on dextrose agar. 
On the basis of the 37° plain-agar count, the counts for all samples, as 
represented by the average percentage figures, show variations ranging 
from less than a twofold increase, for lactose agar at 37°, to more than a 
sevenfold increase, for dextrose agar at 30°. 
From these data it appears that dextrose agar at 30° or 20° has distinct 
advantages over any of the other combinations used, for obtaining higher 
counts. There is comparatively little choice between plain agar at 20° 
and 30°, and lactose agar at the same temperatures, altho plain agar at 
30° seems to have a slight advantage in the number of highest counts 
and the average percentage increase. The 37° counts on all of the media 
are decidedly unfavorable in comparison with those of the other incubation 
temperatures. Dextrose agar, however, has evident advantages over 
plain and lactose agar at this temperature. The slight difference which 
is found between the two latter media at 37° is in favor of the lactose agar. 
Variation in counts at 37° C. 
While lower bacteria counts were found at 37° than at the lower incu- 
bation temperatures, it is recognized that longer incubation periods are 
necessary in order to develop higher counts at the lower temperatures. 
At the end of forty-eight hours, the 37° counts have frequently proved 
higher than those resulting from similar incubation periods at the lower 
temperatures. An additional incubation period of two or three days, 
however, has always been necessary in order to develop the higher counts 
at these lower temperatures. On the other hand, there is usually very 
little increase in the 37° count after the first forty-eight-hour period. 
These facts might easily be interpreted to mean that forty-eight hours 
at 37° represents a time-temperature relationship which cannot be reduced 
if the greatest growth in the shortest period of time is desired. This 
particular temperature and incubation period may therefore be looked 
upon as the minimum and cannot be changed without materially affecting 
the usefulness of the results. The exacting demands for inspection work, 
which have determined the use of the 37° forty-eight-hour count, are not 
