54 BULLETIN 1294, U. S, DEPARTMENT OF AGRICULTURE 
private lands in the Mother Lode region. It is known that certain 
of these stands in Nevada County have been light burned more or 
less regularly for many years. Some of the results have already been 
noted in considering the influence of fire on second-growth forests. 
From these results it is clear that, as with uncontroiled fires in the vir- 
gin forest, repeated light surface burns not only do not permanently 
reduce the amount of inflammable material but actually tend to 
increaseit. (PI. XII fig. 1.) Asplendid illustration of the dubious 
value of light burning as a reducer of hazard is afforded by a study 
of the Cement Hill fire of November 26,1919. This fire, which had 
been set on one of the periodically light-burned areas, became sud- 
denly accelerated by a heavy wind and covered 600 acres, developing 
into localized crown fires on several exposed slopes and wiping out 
every tree on areas of 2 to 4 acres. 
It is noteworthy that although the fire occurred late in the fall, the 
damage was as heavy on the areas where light burning had been prac- 
ticed regularly as it was in the stands where fire exclusion has been 
maintaimed. Nor is this an isolated example of the harmful results 
from light burning, for it is a fact that summer fires in these light- 
burned forests are of not infrequent occurrence and are gradually wip- 
ing out the stand. As far as second-growth forests are concerned, the 
data prove that even repeated surface fires under the most favorable 
conditions do not reduce hazard sufficiently to prevent crown fires. 
In all the virgin forests in which light burning had been started 
from 1910 to 1917 the practice was deeb ed after the initial burn. 
This cessation made it impossible to study the effects of repeated burns 
in reducing hazard except by depending on chance fires or on an 
experimental burn. Therefore an area on the Plumas National Forest 
was selected and an experimental light burn stated by the Forest 
Service in the spring of 1919. 
Some 200 acres in the mixed conifer type with wide variations of 
slope, aspect, forest types and site quality were selected. Before the 
burning was begun a careful cruise of the original stand was made. 
In addition, a number of sample plots were established to determine 
depth of litter, density of reproduction, number of fire-scarred trees, 
size of scars, and character and distribution of underbrush. A fire 
line was then constructed around the area and burning was done at 
See from this line toward the center. Fires were set also along the 
ridges in an effort to force downhill burning rather than uphill. 
re area has been burned twice—once in May, 1919, and again in 
ay, 1920. 
ires burned freely but lightly on the warmer and drier aspects. 
Because of the abundance of squaw carpet, they did not cover all the 
ground, although a reasonably complete burn was obtained. Simul- 
taneously, fires were set on the cooler aspects; but in spite of repeated 
efforts, only small patches could be burned. This difficulty of burn- 
ing has been experienced both times the burning was done, and of 
the 200 acres within the area only 126 have been burned over to date. 
As on the Sierra County area, the largest amount of inflammable 
material exists on the coolest slopes, where the amount of under- 
growth and reproduction is materially greater than on the dry south- 
erly slopes. Even at the same elevation, a period of at least two 
weeks must elapse after south slopes burn freely before fire will 
