58 BULLETIN 1294, U. S. DEPARTMENT OF AGRICULTURE 
tree-destroying insects. The evidence available at the time this 
theory was most widely announced indicated clearly that light burn- — 
ing was of little value in reducing hazard, and indeed that it tended | 
to increase rather than decrease insect attacks. In spite of these © 
facts, one of the large lumber companies felt justified in initiating 
some large-scale experiments to test anew the cost and value of the © 
practice. 
In 1920 an area was selected in the pure western yellow-pine type © 
of Siskiyou County, located on the level or gently rolling plateau — 
surrounding Mount Shasta. The experiment, judging by the manner — 
of its execution, was based on the followmg assumptions: | 
1. That light surface fires caused negligible loss to merchantable — 
trees; and that, therefore, no protection of individual trees was — 
required. 
2. That reasonably complete covering of the surface by the fire 
was necessary, and therefore the burning should be done in summer, 
when fires were certain to cover most of the ground. 
3. That to control the fire it was necessary to divide the area inte 
small burns, each of which could be completed in one night; and 
that, therefore, the area must be blocked by fire lines. 
4, That by burning at night the fires could be controlled and dam- 
age to merchantable timber avoided. 
5. That the cost of burning plus direct. damage to merchantable 
timber would be less than the cost of intensive systematic fire control 
until the timber was harvested one to five years from date of burning, 
ake the cost of controlling a serious epidemic of insects then preva- 
ent on the area. 
In the actual burning the area was divided by fire lines into blocks 
of 160 and later 640 acres. Snags near the line were felled by the 
saw and fires were set in the evening at the edge of the fire line and 
allowed to burn toward the center. 
The operation was carried out in the summer of 1920, and no addi- 
tional work was done in subsequent years. An examination by forest 
officers of the California district was made in July, 1921, by running 
43 miles of mechanically located cruise lines in cardinal directions 
and sampling the area sufficiently to give a representative basis for 
conclusions, or 76 acres in all. Losses were found to be as follows: 
Trees burned down, 0.237 per acre, or 425 board feet per acre killed. 
Trees heat killed, 0.167 per acre, or 220 board feet per acre killed. 
fai bers insect infested following fire, 0.334 per acre, or 440 board feet per acre 
11leqd. 
Trees affected by fire, in all, 0.738 per acre, or 1,085 board feet per acre killed. 
The most serious losses were in trees attacked and killed by insects. 
following the fire. Subtracting the average rate of annual loss for 
this region, where fires have been excluded, which is estimated at 40 
board feet to the acre, we have as the amount of loss that can safely 
be charged directly to the fire 400 board feet to the acre. 
As on other areas, an appreciable loss of merchantable timber due 
to the burning down of previously scarred trees was noted; and again, 
as previously shown, trees lost by this process were almost invariably 
the largest individuals with the highest quality of timber, so that not 
only was the total stand reduced but the average quality of the stand 
as well. (PJ. XII, fig. 2.) Reproduction was destroyed on about 
