Problem of Xeromorphy in Marsh Plants . 853 
4. The leaf surface is much reduced. It is of course a very common 
thing to find the upper leaves of the shoots of herbaceous plants considerably 
smaller than those lower on the stem. But undoubtedly the effect of this 
diminution in size of the upper, more exposed leaves will be to reduce con- 
siderably the output of water from the terminal portions of the shoot. It has 
indeed been suggested above that this reduction in size of the leaf (or at 
least of its component cells) is in part due to development under conditions 
which strictly limit the available supply of water. In Spiraea , the leaves 
which present the greatest transpiring surface are placed in relatively sheltered, 
humid strata of the vegetation. 1 Though all shoots exhibit periodicity in 
the size of leaves, the actual sizes of the latter vary greatly according to the 
conditions under which they are grown (see Text-fig. 5). 
5, When Spiraea leaves are exposed to strong insolation, they tend to 
assume the profile position. This is taken up, not by the leaf or leaflet as 
a whole, but by the two halves of a leaflet separately. A marked ventral 
curvature of the leaflet is thus produced. This will probably tend to 
a further reduction of transpiration. The phenomenon is a frequent one in 
marsh plants as well as others, but need not be fully discussed here, as it 
will form the subject of a forthcoming paper. 
At first sight the more numerous stomata of the upper leaves suggest 
that they may counterbalance the characters which would otherwise reduce 
transpiration. But it has been shown (p. 838) that in all probability 
the possible transpiration per unit area of leaf surface, if the stomata alone 
are considered, would be little if any more from the upper than from the 
lower leaves. The balance of characters, then, shows that the upper parts 
of the shoots of .S'. Ulmaria are distinctly more xerophytic than the lower. 
But in any case the degree of xeromorphy is slight. The question now arises, 
is this degree of xeromorphy necessary or even advantageous to the plant ? 2 
It may be regarded as an axiom that where functions such as transpira- 
tion are concerned, the need for special regulatory devices in a given species 
will be determined, not by the average conditions, but by what may be 
termed the normal extremes which the species is called upon to face 
in nature. In other words, to be successful, a plant must always be prepared 
for such emergencies as it may occasionally have to meet during the 
vegetative season. 3 
1 Yapp (’ 09 ), p. 281. 
2 And if a dense pubescence is necessary to the type-form, why is it not equally so to the less 
hairy variety denudata ? On this point all that can be said (though it is not advanced as being 
a complete answer) is that var. denudata appears to be a less successful form. It is rare in nature, 
while the type is common. So far as my experience goes, it usually occurs only in sheltered spots. 
Culture experiments on the roof of the College at Aberystwyth show that, especially if the soil is wet, 
it is much less able to endure the effect of wind than the hairy form. Moreover, after all, the 
hairiness of the two forms is merely a matter of degree. 
3 A. R. Wallace (’ 10 ), p. 258, goes even further than this. He says : 1 A dominant species has 
