120 F. E. Fritsch and E. J. Salisbury. 
penetrating to an average depth of over nine inches, whereas in the 
case of Erica tetralix, which as already stated almost invariably reju¬ 
venates, the average depth is under six inches ; and the same is true 
for Molinia and Carex pilulifera, whose average rooting depth is 
about three inches. It is, nevertheless, a fact that Ulex nanus , which 
so commonly survives, has deeper roots than any of the Ericaceous 
forms, whilst Erica cinerea, which is frequently destroyed, has rela¬ 
tively shallow roots (see Table II). 
TABLE II. 
Rooting depths (in inches) of plants on the Hindhead heath. 
(Heights of aerial organs in brackets). 
Calluna 18 (35); 10 (32); 8 (32); 6-5 (26); 6 (24); 6 (18). 
Erica cinerea :— 9'5 (8); 6 (30); 3 (20*5); 3 (16). 
Erica tetralix :— 6'5 (11) ; 4'5 (9). 
Ulex nanus :— 24 (24); 12 (32). 
A careful examination of dead and rejuvenating stools discloses 
the fact that the latter fall into two categories, viz., (a) those in 
which the branching of the crown takes place beneath, or in close 
contact with, the general surface of the soil, and (b) those in which 
branching, although occurring above the surface, is so profuse as 
rapidly to accumulate a mound-like protective covering of humus. 
In the case of Ulex nanus the former condition generally obtains, 
and it appears that such few specimens as are killed have a 
relatively exposed crown. This point seems to be the determining 
factor as to whether a Calluna sprouts or not; thus all sprouting 
specimens of this plant examined proved to have a crown that was 
more or less completely buried, whilst the numerous dead specimens 
were found to have branched above the surface of the soil (Plate I, 
Phots. 3,4). The same features were observed in Erica cinerea, but, 
since a large percentage of plants of this species have buried crowns, 
the proportion of survivors is greater. 
Erica tetralix affords an example of the second category. Here 
numerous branches arise just above the surface of the ground 
forming a trap for humus, so that after a few years they are 
embedded. The efficiency of this covering is demonstrated by the 
apparent absence of non-rejuvenating E. tetralix. Tussock-forms, 
like Molinia coerulea and Carex pilulifera, probably owe their immun¬ 
ity to the same feature. 
The cause of survival in the case of Vaccinium and Pteris is 
slightly different, for here the perennating rhizomes are situated 
