ROLE OF FIRE IN CALIFORNIA PINE FORESTS 15 
different forest types. Only with an enormous mass of data would 
it be possible to prove statistically the relative rate of loss in different 
types, because of the great difficulty in finding comparable stands, 
fires of comparable intensity, and comparable sites. But, given fires 
satisfying here conditions, it would probably be found that the fir 
types show higher rates of loss than the pine types. An example of 
maximum damage from heat killing is shown in Plate IV, Figure 1. 
In a many-aged forest the susceptibility of the individual trees to 
heat killing by the burning of the litter, undergrowth, and reproduc- 
tion would appear to vary inversely as the quality of the site, which 
is to say, as the height of the trees, a good site having a greater 
proportion of 'trees whose crowns are safe from ground fires. With 
fires of equal intensity, other than crown fires, the percentage of the 
stand killed should be greater on poor sites and less on good sites. 
It is impossible to give rigid statistical proof that liability to heat 
killing is in inverse ratio to the excellence of the site; nor is it possible 
to state the exact magnitude of the losses caused by fires of given 
intensity on poor sites and on good ones. With the exception of 
second-growth stands, our forests are so broken, so variable, and so 
patchy, and the technical difficulties of measuring intensity of fire are 
so great, that no exactly comparable data have been obtained. 
he only method so far developed for measuring with reasonable 
precision the effect of a given fire on individual trees was evolved in 
the examination of certain of the large burns of 1917 on the Eldorado 
National Forest. Considering only one species, western yellow pine, 
and fire for one season of the year, namely, fall, it was found that 
the crown of a tree was killed for an average distance of 20 feet above 
the highess point of actual burning by the flames. If, for example, 
a yellow pine had a 40-foot crown and the fire had actually scorched 
the trunk up to the base of the crown, then 20 feet of the crown 
succumbed. (PI. IV, fig.2.) Naturally, the level to which the flames 
tended to reach depended on the amount of fuel on the ground. With 
only bear clover, a low-spreading shrub, for example, it was found 
that the flames reached only 10 to 15 feet, but with manzanita, a 
much taller undercover, they reached 40 to 60 feet. The observa- 
tions indicate that more trees would be killed in low stands than in 
tall ones, other things being equal. 
CROWN INJURIES AND RATE OF GROWTH 
In addition to the direct or primary physical damage to merchant- 
able timber which is evident from even a casual examination of 
burned areas, fire in the virgin forests of California may and often 
does start a less conspicuous train of events which is reflected in the 
condition of the stand for many years afterwards. These influences 
ordinarily show themselves either in a reduction of growth or in 
decay that injures the trees without actually killing them. 
EFFECT OF CROWN INJURIES ON GROWTH 
One of the important secondary injuries resulting from fires is the 
reduction of the rate of growth as a result of crown injuries. On 
the Klamath National Forest a study of the effect of surface fires on 
rate of growth was very instructive. In 1910 a fire ran through a 
