Problem of Xeromorphy in Marsh Plants. 819 
the shade is not too dense. 1 It readily colonizes moist ground which has 
been recently disturbed ; and hence frequently establishes itself at the foot 
of railway embankments, &c. 2 
§ 5. The Distribution of Pubescence on the Leaves. 
The various British Floras state that the type form of Spiraea Ulmaria 
has pinnate leaves covered with white down beneath. Consequently, in 
September of 1906 I was surprised to find that the Spiraeas at Wicken Fen 
varied markedly in this respect. Some leaves were completely hairy on 
the lower surface ; others were glabrous, while all possible transitions were to 
be found (see PI. LXXXII). The reasons for this extreme variability were 
by no means obvious, so a large number of specimens were grown for some 
years. It was hoped that if any explanation for such a marked phenomenon 
were forthcoming, it might throw light on the general problem of xero- 
morphy in marsh plants. 
The experiments (confirmed by many field observations) revealed the 
existence of a regular periodicity in the production of hairs. The rules 
governing hair formation may now be stated. It should be borne in mind 
that all the following descriptions refer to the hairy type form of X. Ulmaria , 
and not to its glabrous variety, denudata. 
(a) Seedlings. In all cases the seedlings are glabrous. This applies to 
every leaf formed during the first year. The plant, if small and weak, may 
remain in this juvenile glabrous condition for two or more years ; though 
if under favourable conditions, it may form normal hairy shoots in the 
second year. 
(b) Erect flowering shoots of adult plants. When a winter bud opens 
early in the year (about March), the first leaves unfolded are juvenile in 
type, i. e. they are small and glabrous. After a variable number (usually 
from 3 to 5 3 ) of these glabrous radical leaves have unfolded, they are 
followed, about the end of April or the beginning of May, by one or two 
leaves which are partly hairy. The latter are in turn succeeded, as the 
flowering shoot elongates and becomes erect, by leaves which are completely 
covered with hairs on the lower surface. 4 The pubescence becomes more 
dense as the erect stem is ascended. (PI. LXXXI shows all the successive 
leaves formed on one flowering shoot. Leaves 1-7 are radical, and leaves 
8-16 cauline. Hairy parts show white, and glabrous dark.) The plant 
1 Adamson (’12), pp. 358, &c., gives a good account of various S. Ulmaria societies in a 
Cambridgeshire woodland. 
2 At least some of the other species of Spiraea mentioned in § 3 are also swamp plants : e. g. 
S. salicifolia and S. tomentosa. See Britton and Brown (’97), p. 196. 
3 I have seen as many as twelve, but this is quite exceptional. 
4 The flowering shoot usually begins to elongate about the middle of May, by which time most 
plants already show some hairy leaves. 
3 H 2- 
