Sept. 15, 1925 
Plate Counts oj Soil Microorganisms 
513 
Table VII .—Average counts of combination of squares demonstrating the uniformity 
of the soil 
Squares 
Average 
Squares 
Average 
1-47-60-36.____ 
27.8 
28.6 
30.7 
30.0 
30.6 
28.4 
59-35-48-12. 
27.4 
27.1 
27.2 
30.2 
30.0 
28.4 
3-45-58-34... 
57-33-46-10.. 
5-43-56-32.... 
55-31-44-8. 
7-41-54-30.. 
53-29-42-6.—. 
9-39-52-28... 
51-27-40-4. 
11-37-50-26... 
49-25-38-2... 
Average...__ 
A verage 
29.3 
28.4 
The curves for air temperature and soil moisture are included in 
Figure 2 for comparison with the plate counts. The averages again 
are calculated as 100 and the daily fluctuations computed accordingly. 
The moisture curve shows three large fluctuations which can be like¬ 
wise recognized in the plate-count curve whose gradual falling off 
is in general agreement with the analogous tendency of the temper¬ 
ature curve. In addition, slight soil differences have undoubtedly 
contributed to the special deviations of the plate-count curves. 
As counterpart to the results obtained on soil-extract agar, curves 
representing the plate counts of the same samples on mannite-salts 
agar are shown in Figure 3. The fluctuations as well as the sample 
deviations are much greater on this medium, the former varying 
from 133 per cent above to 66 per cent below the average. Two 
factors may have been mainly responsible for this result—the selective 
character of the medium, and the smaller counts obtained, as shown 
in Table IV. 
For comparison with these writers’ results, a graph of results 
obtained at the Rothamsted Experiment Station has been included 
in Figure 3 showing the percentage fluctuations in the counts made 
by Cutler and his collaborators (3) on corresponding days of the year 
and using the same mannite-salts agar. It is interesting to note that 
the majority of the counts fall within the range of +20 per cent to 
— 40 per cent and compare well with these present findings with the 
same medium, but they, too, are not as uniform as when soil-extract 
agar was used. 
After the completion of these present experiments of the writers, 
Matthews (7), working at Rothamsted on the partial sterilization of 
soil by antiseptics, has reported that “ the large number of counts 
made as controls on untreated soils failed to show, under the con¬ 
ditions of the experiment, any great saw-edged curve such as Mr. 
Cutler has proved for soils under field conditions.” The counts 
were made in this case on gelatin plates. 
PLATE COUNTS FROM COMPOSITE SAMPLES 
In making plate counts of soil organisms it is customary to make 
one series of dilutions, usually starting with 10 gms. of soil in 100 or 
several hundred cubic centimeters of water. Several plates aro 
then poured from the highest dilution, three being probably the 
most common number; Thiele ( 9 ) recommended 20, and Waksman 
{12) used 10 plates. Other workers have used from 2 to 10 plates 
to get’ a fair average, and have usually started from one composite 
sample in one series of dilutions. 
