VARIATIONS IN BACTERIA COUNTS 245 
TABLE 6. Temperatures Recorpep Eacu Hour at DirreRENT PLACES IN A Buiock 
oF TigHTty PACKED PETRI PLATES 
(In degrees centigrade) 
tenn Temperatures recorded after 
Deter- DLA CUTE) | paella e Lees omer Rees Aiea beets eee SD 
mination Bt : 1 9 3 4 5 | 6 7 8 9 
Brar hour | hours | hours | hours] hours | hours | hours! hours} hours - 
a a a ee) ee (ere | OTe (ORI) (DESO | 
il , Cgeeeceetereeests ROO? | S20" P BUCOF 16 be ells Sete Waters Cokes eae arstchahcote| eaters Gig igeee eee 
2. G See eee PO ied | RS ORO cal CONas Opi Meeaeti a | kseate ces [ica neeesueel lem eeicerl | meets |Buetot ay, Teed tee 
Bas Sei bes 0) BF? || SUC? Seb ay On| Ses Gc oe alleonsacleacancllooocadlaacdas 
4. 3 Sea Bil BE) SY ZEN CRS polseh [alan Laxey Io dee alles os celles sure e| |S Seams 
sc es 21 70°) 27 .0° | 31.62 | 33.8° | 35..0° | 35-52 | 85.52 | 36.02 | 36.62] 37.0° 
GMs heya 2NOM 25,0 (28102 | 31202 | 8275 34.42 1.8552 1.35..5- 136.02 || 37 02 
The temperature records clearly indicate the relative length of time 
required by plates in different positions to reach the normal temperature. 
If records from inside the block of plates could have been obtained 
readily, it is quite probable that the length of time necessary to reach 
37° would have proved even greater than that shown by determination 6. 
The rapidity of the heat diffusion apparently did not suffice to heat the 
entire block of plates soon enough to prevent marked irregularities ia 
the counts made from the inner piles of the block. This fact emphasizes 
the greater likelihood of discrepant results from overcrowded incubators, 
or from other causes which in any way reduce the ventilation around the 
interior plates. In the positions represented by determinations 5 and 6, 
the temperatures were below normal for nine hours. This period, during 
which the plates in these positions were below normal, amounts to approxi- 
mately 19 per cent of the entire incubation period. Plates on the outside 
of the pile, however, remained below normal temperature for a much shorter 
period, and consequently their counts were much higher and more nearly 
uniform than those obtained from plates that had not been maintained 
at the normal temperature for the entire period. 
Discussion 
The wide range of variations in the counts obtained by different incu- 
bation temperatures and media emphasizes the inadequacies of any single 
combination of temperature and media for determining the maximum 
