Oct. i, 1917 Reproduction from Seed Stored in the Forest Floor 
11 
Crass — No reproduction 
,Area of dense//\ 
j reproduction o/a 
y noi 
most often occurred, not in a solid unbroken cover, but in various¬ 
sized patches with very irregular and ramifying boundaries. Where the 
reproduction was lacking, the ground was covered with grasses, herbaceous 
plants, and shrubs, evidencing an uninterrupted growth since the burn 
was formed. The occurrence of these two types of cover made an inter¬ 
laced pattern resembling mosaic over the entire bum, although each type 
often expanded solidly over a slope or basin many acres in extent. 
Everywhere the feature that was most striking was the sharp line of 
demarcation between the reproduction and the grass areas. For all its 
tortuous windings the boundary was always distinct. Obviously such a 
condition could not 
have resulted from any 
process of overhead 
seeding, but must rath¬ 
er have been produced 
by some action on the 
surface of the ground 
itself. The idea of 
ground fire suggested 
itself. One who has 
seen ground fire burn¬ 
ing in forest duff will 
remember that it bums 
very irregularly, here 
leaving an island and 
there forming a deep 
bay between two points 
of unburned ground. 
When at length the 
smoldering fire is 
stopped, the result is 
just such a mosaic of 
burned and unbumed 
territory as has been 
described for the re¬ 
production and grass territory. A representative spot in the Columbia 
bum is shown on Plate 2, A, and the sketch in figure 4 is a graphic 
portrayal of the same spot. 
The likelihood that any part of the forest floor will burn depends on 
a number of varying factors, such as the quantity and kind of humus 
soil and its moisture content. Favorable conditions for a ground fire 
are shown in Plate 2, C, and the results of ground fire in Plate 2, D. 
Most severe ground fires occur on dry sites, provided those sites have a 
sufficient quantity of duff to carry fire at all. Accordingly the repro¬ 
duction occurs most densely in the moist sites and is open or lacking on 
exposed dry sites, although this may be partly due to the fact that these 
Or css - No reproduction 
Result of ground fire 
Fig. 4.—Ground plan of the spot shown in Plate 2, A. The conform¬ 
ance of this irregular boundary line with the erratic margin of 
ordinary ground fire is clearly evident. Constant repetition of 
this feature over the entire Columbia bum could lead to but one 
conclusion, that where ground fire burned there resulted no re¬ 
production, excepting in proximity to living seed trees. 
