Sept. 15, 1925 
Plate Counts of Soil Microorganisms 
505 
All the counts given in Table II were made on plates poured with 
soil extract agar. As Whittles used a “mannite-salts agar” (14, V . 
19), it was necessary to adopt this medium to get fully comparable 
results. At the same time, it seemed desirable to check up the 
results given in Table II as to the efficiency of the vibration com¬ 
pared with hand shaking and also in regard to the influence of the 
diluent. The inoculum in this case was a clay-loam field soil. 
Table III .—Comparison of mannite-salts and soil-extract agars, tap water and 
shake medium, vibration and hand shaking 
Suspension by— 
Dilution 
Diluent 
Mannite 
salts 
agar 
Soil 
extract 
agar 
Average 
Average 
Hand shaking. 
Do.. 
Vibrator. 
Do.. 
Average_ 
1 : 250,000 
1 : 250,000 
1 : 200,000 
1 : 200,000 
Tap water.. 
Shake medium_ 
Tap water__ 
Shake medium.. 
Millions 
per gm . 
8.9 
8.3 
11.5 
7.7 
Millions 
per gm. 
10.8 
10.6 
16.6 
11.8 
9.9 
9.5 
14.0 
9.8 
} 9.7 
} 11.9 
9.0 
12.4 
A comparison of the counts made on mannite-salts agar and on 
soil-extract agar, as recorded in Table III, shows a small but dis¬ 
tinct difference in every case in favor of soil-extract agar. With the 
soil-extract agar, the colonies showed less tendency to spread, and 
on the whole they were easier to count. Tap water used as diluent 
gave consistently higher counts, although the difference is small in 
the hand-shaking series. When vibrator and hand shaking are 
compared, the former gave a somewhat higher average, but the 
counts making up this average are much more erratic than those in 
the hand-shaking series, and consequently less valuable. 
In this connection, the fact should not be overlooked that growth 
might take place in the vibrator or in the suspension. When making 
the experiments, not less than 45 minutes elapsed between the addi¬ 
tion of the soil to the vibrator cup and the plating out of the highest 
dilution. In order to ascertain whether growth actually amounts 
to much under these conditions, the following test was made. A 
three-day-old agar slant culture of Bacterium jluorescens was flooded 
with sterile water, the growth scraped off, and this suspension trans¬ 
ferred to a flask of sterile water. After shaking for three minutes, 
four aliquots were taken out, two of which were diluted by hand 
shaking, one in tap water and the other in shake medium with gelatin; 
and two were vibrated with the same diluents. Plates were poured 
from the 2 and 2,000,000 dilutions, incubated at 28° C. and counted 
after three days with the following results: 
1 . Control, diluted with tap water, 600 millions. 
2 . Control, diluted with shake medium (gelatin), 540 millions. 
3. Vibrated, diluted with tap water, 135 millions. 
4. Vibrated, diluted with shake medium (gelatin), 690 millions. 
The vibrator cup leaked at the beginning of the first run and 
some of the inoculum was lost; therefore, the count of 135 millions 
can be disregarded. In spite of this accident, it can safely be as¬ 
sumed that growth is not responsible for a marked increase, especially 
if only one suspension is made in the vibrator, as in the case just given. 
