Class
mid-points
in 10 cm.
Actual
6242
Fitted
2471
Actual
Fitted
3
45. 56
39. 58
41. 64
37. 64
4
23.86
23.99
23. 67
23. 62
5
13.23
14. 53
13.80
14. 76
6
6. 73
8. 80
7.37
9.21
7
4.38
5. 33
5. 34
5. 76
8
2.42
3.23
2. 67
3. 60
9
1. 53
1. 96
2.39
2.25
10
1. 00
1. 19
1.46
1.40
11
0. 52
0. 72
0. 81
0. 87
12
0.42
0. 44
0.49
0. 55
13 and
0. 35
0.26
0. 36
0. 34
over
Total
100
100
100
100
Table 5.
Percentages per diameter class for 6242 and 2471 trees
In table 6 the frequencies are given for both distributions. The logarithmic
transformation required for fitting caused an underestimation of the total number
of trees. This may partly be remedied by extending the upper limit of the range.
The extension was accomplished in this case by re-grouping the trees in
diameter classes of 30 cm. Both re-grouped distributions were fitted to a
negative exponential and again highly significant fits were obtained as may be
seen in table 7.
N
Equation
D. F
Residual
Variance
t
6242
log N
= 4.560182 -
0.224739 D
2
0. 004590
22.25139
2471
log N
= 4.133425 -
0.214037 D
2
0. 004 6 5 6
20. 98745
Table 7.
Regression equations for diameter distributions in 30 cm classes