Sept. 15, 1925 
Plate Counts of Soil Microorganisms 
511 
methods could be carried out on the same square. One even-num¬ 
bered and one odd-numbered square were sampled each day, except 
Sundays and holidays, plates were poured with soil-extract agar and 
other media for comparison, and the colonies were counted after 
seven days at 28° C. 
In the first part of Table IV (July 26 to August 7, squares 1 to 20), 
the counts of the microorganisms are given as found in samples taken 
by the two methods. The average for horizontal sampling is 31.2 
millions, that for the vertical method 28.5 millions. 
If the odd and even squares representing the opposite sides of the 
experimental plot are averaged separately, the same numbers are 
obtained with the horizontal method of sampling (31.3 and 31.2 mil¬ 
lions per gram). But in the case of vertical sampling, the odd squares 
have an average of 29.8 and the even squares 27.1 millions per gram. 
Furthermore, the devi- 
ation of each count ----——---- 
from the average is 
greater with the verti- _ x* ^ ^ 
cal method, the latitude_ ** ^ ^ ^ ^ _ 
being -11.1 to +7.7 Vf]f 43 SO. 3S 47 S9 / 
millions for the vertical, -- 
and -4.6 to +8.1 mil- /<? 3* 46 S3 33 4S S7 3 
lions for the horizontal.-7 r~T 7r"TT~I 
These results indicate ^ _ <g 32 44 S6 3/ _ 6^ _ ^ 
that greater uniformity ^ e 30 *2 S4 £9 4/ S3 7 
as well as slightly higher- 
counts are obtained 4 23 40 S2 27 39 S/ 9 
when samples are taken- 
by the horizontal 2 26 33 S0 2S 37 49 // 
method. It is well __ ?/ /<? /-r 7? 
known that the greatest_ 
activity of the micro¬ 
organisms occurs a few ---------- 
inches below the sur- ^ 
face and it Seems prefer- Fig. L—Experimental plot showing the squares in the order of 
able to take the samples sampling 
from this plane without the admixture of the less densely populated 
soil from higher and deeper layers. In sampling with an auger it is 
very difficult to get equal amounts of soil from the various layers. 
Thoroughly mixing a soil sample is undoubtedly of some value, but 
it does not mix the bacteria sufficiently to assure uniform distribution 
of them. In view of these facts, the horizontal method alone was 
used throughout the remainder of the tests. 
* RESULTS OF FREQUENT PLATING 
From July 26 to September 19, 84 squares (i. e., squares 1 to 60 
and 1 to 24 inclusive) were sampled and plated on soil-extract agar 
and the two other substrates just discussed. The counts for each 
square are given in Table IV. The average for the entire series 
was 28 millions per gram of soil (counted on soil-extract agar). If 
this is taken as 100, the counts for the odd and even squares sampled 
on the same day exhibit certain plus or minus percentage fluctuations 
from the general average, as shown by the third graph of Figure 2. 
/V 
/4 
/6 
/3 
20 
22 
24 
/2 
36 
*3 
60, 
3S 
47 
S9 
/ 
/0 
34 
*6 
S3 
33 
4S 
S7 
3 
3 
32 
44 
S6 
3/ 
43 
SS 
S 
3 
30 
42 
S4 
29 
4/ 
S3 
7 
4 
23 
40 
S2 
27 
39 
S/ 
9 
2 
26 
33 
S0 
2S 
37 
49 
// 
23 
2/ 
/9 
/7 
/S 
/3 
