612 
Journal of Agricultural Research 
Vol. XXX, No. 7 
abscissa, and the diameter at other 
points the ordinate. As a rule, values 
are read at 10-foot intervals up the 
tree from the curves of series A, and 
a separate curve is drawn showing the 
relation of d. b. h. to diameter, 10, 20, 
30, and 40 feet above ground level. 
This series of curves so rounds out the 
values in Figure 2, A, that, in effect, 
it takes the lines as shown in Figure 
2 , A, and turning them upon the point 
of convergence as a pivot spaces them 
systematically, placing them upon the 
correct d. b. h. If the lines in Figure 
The spacing between the lines in 
Figure 2, B, depends upon the slope of 
the lines in Figure 2, A (taper of the 
trees). In conical trees they are natur¬ 
ally evenly spaced, as shown in the 
figure, but in full-boled trees they get 
farther and farther apart toward the 
top of the tree, as in such trees rate of 
taper increases upward. In practical 
work, the points determining these 
lines usually fail considerably of falling 
into good lines, so that in drawing an 
average curve errors are very likely to 
be made in giving end values with^a 
50 60 70 8 0 90 100 0 2 4 6 8 10 12 14 
Total Height of Tree (Feet) Diamete r 
Fig. 2.—The main steps in graphic harmonization of taper curves 
2, B, are straight, the corrected spacing 
of the lines in Figure 2, A, will be even; 
if they are curves, the spacing will be 
uneven, but the changes will be regular 
and even (as will occur in cases where 
form changes with diameter). In prac¬ 
tice, lines in Figure 2, B, are nearly 
straight. Their slope depends upon 
the distance between the lines in Figure 
2, A, at various levels, so there is the 
greatest slant at the base of the tree 
where the lines are farthest apart and 
the least toward the tip. 
poor basis more weight than they de¬ 
serve, consequently tilting the lines. 
Marked tilting in any single line will 
be noticed on inspection of the finished 
graph, however, unless all the lines are 
tilted in the same direction (which un¬ 
fortunately is quite possible). The 
effect is equivalent to making the lines 
in Figure 2, A, evenly spaced, but 
either too far apart or not far enough. 
The spacing may be different in differ¬ 
ent height classes, as these are curved 
entirely independently. 
