6 
CIRCULAR 816, U. S. DEPARTMENT OF AGRICULTURE 
might be expected, there was a significant difference between field popu¬ 
lations. In all but 2 counties (Fergus and Judith Basin) the differences 
were highly significant. The differences between populations in the 5 
different within-field locations, however, showed a slight tendency toward 
significance in only 4 of the 10 counties, as seems apparent from a com¬ 
parison of the averages in table 2. 
It should make no material difference, therefore, where the units are 
taken within fields if they are fairly well distributed. Some data of this 
type might be considered to need transformation before analysis, but 
population levels were rather uniform, and the total number of egg pods 
found in each designation was high enough so that little or no gain could 
be expected from the transformation. 
The results of an analysis of the data for the field margins are given 
in table 4. These results are similar to those for the fields. Population 
Table 4. —Variance in numbers of egg pods per square foot between field 
margins, between locations within margins, and between units within the 
locations, for 10 counties in north-central Montana in 1940; }/%-square- 
foot sample units (degrees of freedom given in parentheses in column 
headings) 
County 
Between 
field 
margins 1 
(69) 
Between 
locations 
within 
margins (1) 
Interaction, 
field and 
location 
(69) 
Between units 
within 
locations 
(560) 
Blaine. 
1.12 
0.82 
20.54 
0.39 
Cascade. 
.81 
1.04 
.42 
.38 
Chouteau. 
4.23 
1.90 
.87 
1.02 
Fergus. 
2.99 
2 5.32 
1.22 
1.33 
Hill. 
8.62 
2 29.61 
L5.07 
3.29 
Judith Basin. 
.13 
.07 
.04 
.05 
Liberty. 
1.19 
22.76 
.56 
.48 
Pondera. 
6.16 
7.20 
3.98 
4.13 
Teton. 
11.33 
.08 
5.20 
6.49 
Toole. 
.86 
.97 
.46 
.37 
‘Significant at 1-percent level of probability. 
“Significant at 5-percent level of probability. 
variations between field margins were highly significant for all 10 coun¬ 
ties. Populations in the two halves of the margins were not significantly 
different in 7 of the 10 counties. 
In general, the analyses show highly significant variations in the 
numbers of egg pods present in different fields and field margins. On the 
other hand, the variations between locations within fields or field margins 
had very little, if any, significance. 
In order to make a more detailed comparison of the numbers of egg 
pods per square foot, 10 wheat-stubble fields in a single South Dakota 
locality were extensively sampled in 1942. Fifty pairs of }/% -square-foot 
units from the same locations were taken in each. The pairs were taken 
from similar uniformly distributed locations in each field. The popula¬ 
tions in the different fields ranged from 0.14 to 0.93 egg pod per square 
foot. Analysis of variance of the pairs is shown in table 5. 
