NATURAL REGENERATION OF DOUGLAS FIR. ree 
covered about 60,000 acres, swept over this region. While the fire was 
running up the main Cispus Valley and up the North Fork and Nig- 
gerhead Creeks, it traveled at the rate of 8 to 4 miles an hour. It 
swept through the patches of green timber that had been left by 
the 1902 fire and entirely killed many of them. The dead snags and 
down logs in the old burn caused a very hot fire, and when it reached 
the boundaries of the 1902 fire it usually killed the green timber 
on a strip from a few chains to over half a mile in width. This 
fringe of fire-killed timber and patches of timber killed by the 1918 
fire that had remained after the 1902 burn provided definite checks on 
the source of young growth following a first or second fire. The 
Douglas fir seed crop of 1918, being general and unusually heavy, 
afforded an excellent opportunity for restocking, because the fire con- 
sumed the vegetation on the old burn, and where it killed the green 
timber left mineral soil, which was an exceptionally good seed bed 
for the germination of the 1918 seed crop when it matured in the 
fall. The fire occurred so early in the season, however, that there 
was no possible chance for seed to mature before the trees were killed 
in the burn. of 1918. Any reproduction found on the area after this 
fire must of necessity have been from seed that was in the forest 
floor from previous seed crops, or it must have come from seed blown 
by wind or carried by other agencies from the adjoining green 
timber. 
The reproduction following the burn of 1902 was 16 years old at 
the time it was killed by the 1918 fire. A few Douglas firs were 
found that had been producing cones for two or three years before 
the fire, but the amount of seed produced on these areas was very 
small. The mineral soil was either exposed or had a vegetative cover. 
In the area burned, the loose mineral soil shifted and buried the seed. 
This has been found to be an important factor in reproduction in 
some places, and is no doubt one of the underlying causes of the 
scattered reproduction found in the 1902 burn that was reburned in 
1918. Perhaps the most striking features on this area, from the 
standpoint of reproduction, are the establishment of seedlings where 
the green timber was killed by the 1918 fire and the practically com- 
plete absence of reproduction where the 1902 fire was followed by 
the 1918 fire. 
The studies of 1914 and 1915 provided an accurate history of this 
area before the big burn of 1918. For this reason the effects of the 
fire of 1918 could be analyzed without any speculation as to what 
might have occurred before the fire. The relation of the occurrence 
of seedlings to the areas of timber killed by the 1918 fire was care- 
fully analyzed by using belt transects, which were run from pvints 
out in the 1902 burn to the timber killed in 1918, and through this 
killed timber to the present stand of green timber. 
These transects showed that there was practically no reproduction 
out in the twice-burned area, and, where the lines approached to 
within 1, 2, or 3 chains of the edge of the timber killed in 1918 the 
reproduction appeared. ‘The stands of young growth varied in den- 
sity from a few hundred to nearly 40,000 to the acre without relation 
to the remaining green timber, but with a noticeable relation to soil 
and topographic conditions. 
The principal factor in the variation in the density of reproduction 
apparently was the moisture of the soil. On some of the transect 
