240 G. C. SuppLen, W. A. WHITING, AND P. A. Downs - 
operative in research work, in which the longer incubation periods an 
lower temperatures are generally used. These longer incubation period 
at lower temperatures are used for the purpose of obtaining maximur 
counts where ‘immediate results are not essential. Since this is the mai 
object, greater variations in time or temperature may be used withou 
affecting the results to the same degree that they would be affected b 
similar variations in the 37° forty-eight-hour counts. 
Recognizing the importance of maintaining the correct temperatur 
for the 37° count, an attempt was made to find out what variations i 
counts would result from possible differences of temperature due to pilin 
the individual plates in a compact mass, as compared with the result 
obtained by so arranging the plates as to allow free circulation of ai 
around each one. The former condition occasionally exists in any labor- 
atory, particularly when a large number of plates must be incubated at 
the same time. In these experiments, the capacity of the incubator was 
only about half utilized. The sources of heat in the incubator were at 
the bottom, two and one-half inches below the temporary floor, at the top, 
and on two sides. Variations in temperature to which individual plates 
might be exposed, therefore, would be due to the slow diffusion of the 
heat resulting from the diminished ventilation around the piles of plates. 
For these comparisons, two samples of market milk containing approxi- 
mately the same number of bacteria, and with no apparent differences in 
flora, were selected. A sufficient volume of a single dilution was made 
so that about 200 plates could be prepared from the same bottle. The 
samples were diluted with the object of obtaining between 30 -and 400 
colonies to the plate. All plates were poured from the same batch of plain 
agar. They were placed in the incubator so that consecutive numbers 
would lie next to each other. The average of the counts from plates 
bearing two consecutive numbers was considered as the count for a single 
sample. 
In testing the effect of the free circulation of air during incubation, 
200 plates were arranged in layers on wire screens with about one-half 
inch air space between each layer. The bottom layer was one and one- 
half inches from the floor of the incubator. A uniform temperature of 
37° at two inches from the top, and the usual ventilation of the incubator 
were maintained thruout the forty-eight-hour period. The counts obtained 
from these samples are shown in table 38. 
