542 
Journal oj Agricultural Research 
Vol. XXX, No. 6 
fillment. During the summer of 1923 
data on a series of plots in various 
forest types were taken near Bathurst, 
New Brunswick, and Metis Lake, 
Quebec. 
The studies were concentrated on 
second-growth forests. Mature forests, 
where the mortality was much higher, 
were avoided, owing to the complication 
of other factors, such as periodic cut¬ 
tings. It was also deemed advisable 
to put more time on the second-growth 
stands, since such results would be more 
directly applicable to forest conditions 
of the future. 
EXPLANATION OF DATA 
In the softwood forests one-tenth acre 
plots were used. It was found that 
this size suited the purpose better, 
since the variation in injury was quite 
local and spotted. The percentages of 
balsam, red spruce, and white spruce, 
and the basal area and average diam¬ 
eter are based on the total of softwoods 
only; the percentage of hardwoods is 
that of the total number of trees on the 
plot. The average diameter was com¬ 
puted from the basal area. All trees 
from 3-inch diameter breasthigh and 
up were tallied and all data computed 
on this basis. This naturally gives a 
low expression for average diameter. 
The percentage of mortality was 
computed on the basis of number of 
trees rather than on volume or basal 
area. No volume tables were avail¬ 
able for these small diameters, and it 
was thought that percentages by trees 
would give a better expression of the 
effects of defoliation on the plots, 
especially the suppressed ones. In 
cases where the injury is low, it would 
often be quite negligible if computed 
in volume. 
All increments are expressed in milli¬ 
meters, showing the radius for a 10- 
year period unless otherwise stated. 
In each plot a varying number of cores 
were taken at breastheight (by means 
of an accretion borer), depending on 
the percentage of the species and num¬ 
ber of each dead. At least 10 cores 
from dead and living trees of each spe¬ 
cies were secured, or, if the number of 
trees in such classes was small, at least 
half were taken. No definite method 
of selecting trees was used; the plot 
was gone through and all trees taken 
as encountered selected from all di¬ 
ameter classes. In computing the av¬ 
erage rate of growth for each species 
on the plot, averages were first ob¬ 
tained from the selected cores for both 
dead and living trees. These averages 
were multiplied by the total number of 
dead and living trees, respectively, 
added and divided by the total num¬ 
ber of that species on the 'plot. On 
the first few plots the average rate of 
growth for each species was determined 
from dominant trees only, average- 
diameter trees, all trees or the method 
described. The variation between dif¬ 
ferent plots as expressed by dominant 
trees and average-diameter trees was 
considerable, but that between the last- 
two methods was so slight that the 
one involving less time was adopted. 
The 10-year period previous to bud- 
worm attack was counted back from 
the enlarged ring produced by the de¬ 
foliation. This is very marked on bal¬ 
sam. In spruce, where the enlarge¬ 
ment is less pronounced, the year of 
budworm attack (previously deter¬ 
mined for the region) can be counted 
back from the last ring formed, but this 
is not reliable for balsam, owing to the 
fact that two or three rings may fail to 
form on parts of the circumference and 
not show on the increment core. On 
dead or dying trees, both spruce and 
balsam, the enlarged ring must be util¬ 
ized for orientation, since from two to 
four rings may fail to form previous to 
death. 
CHARACTER OF BUDWORM DE¬ 
FOLIATION 6 
A brief description of the character 
of the budworm feeding will help to 
explain certain results of the tabulated 
data. Previous investigations have 
shown that the severity of defoliation 
is the primary cause of death of the 
trees through inhibiting normal physi¬ 
ological functions. 
The degree of defoliation of the 
spruce and balsam is chiefly due to the 
variation in development of the new 
growth, to the migratory habits of the 
larvae, and to the fact that the old foliage 
of spruce is not consumed by the larvae. 
For normal development, the young 
caterpillars require new foliage as food. 
They begin feeding at the time the 
balsam and white-spruce buds open. 
During epidemics their abundance is 
such that the new growth of balsam is 
consumed by the time the caterpillars 
are half grown; in the later instars they 
are able to subsist on the old needles 
and consume up to 100 per cent of 
these. White spruce furnishes a greater 
abundance of new growth, so that rarely 
do the larvae consume all of it before 
the needles harden. This hardening of 
the needles takes place about the time 
the caterpillars are half grown, causing 
6 For more detail see reference, footnote 3. 
