Further Observations on Heath Association . 119 
old) to the burnt area (about 7 years old). It may be well to mention 
that, since the first account of the Hindhead Heath was published, 
we have collected local information as to the actual occurrence of 
such a fire leaving the “ central ridge ” untouched. The fact that 
later fires ( e.g ., F 3) may sweep across such lines of demarcation of 
former fires would appear to indicate that direction of the wind is a 
prime cause in their formation. 
As a natural consequence of the firing of the vegetation, the 
mutual shelter afforded by Callnna, Erica, etc., is removed, so that 
the new growth suffers from the full rigours of exposure. At the 
same time the inhibiting effect, upon other plants, of the shade cast 
by tall Callnna (6, p. 104) ceases; as a consequence, invasion by 
light-loving species becomes possible. Hence we find that, whilst 
growth of the shrubby members is slow on a recently burnt area, 
the primary flora exhibits much greater diversity than in subsequent 
phases. 
(2) Killing of certain constituents of the vegetation. 
Although the heat generated by a heath-fire is very considerable, 
it is not sufficient to destroy all the members of the heath-flora, and 
indeed its effects in this respect are selective. Thus, Ulex europceus, 
U. nanus, Vaccinium myrtillus, Erica tetralix, Pteris aquilitta, 
Molinia ccerulea, and all the trees, commonly rejuvenate from old 
stools after burning. In the case of Erica cinerea this would appear 
to be more rare, whilst sprouting stools of Callnna are few and far 
between. On a given area it was found that all the Callunas (54) 
were dead, whilst of the 57 Ulex nanus plants over 87% had survived. 
Sprouting Carexpilulifera is but seldom found, which is in conformity 
with its sparse occurrence in the closed heath-association. These 
facts are illustrated by the following table showing the number of 
sprouting stools of four different species on a given area. 
TABLE I. 
Ulex nanus. Erica cinerea. Vaccinium. 1 Calluna. 
Area F 2 ... 974 ... 91 ... 14 ... 3 
Area F 3 ... 160 ... 69 ... — ... — 
It was previously suggested (3, p. 159) that the selective killing 
bX fire might be related to the rooting depths of the forms concerned, 
but this view has not been supported by subsequent observation. 
Thus, a number of tall Calluna specimens were found to have roots 
1 Vaccinium was very sparsely scattered on the first, and absent from the 
second, of these areas. 
