660 MR. L. A. BOODLE ON THE STRUCTURE OF LEAVES 
naturally to a harder and dwarfer habit. Luerssen also (in 
Rabenhorst’s Kryptogamenflora, vol. iil. p. 108) states that sunny 
and dry localities favour the production of very hairy forms, 
while damp and especially shady places tend to show plants 
which are little or scarcely hairy. 
Among other distinctive characters, one may mention that in 
exposed plants the leaves are darker green than those of shaded 
plants, the veins are more sunk and the sub-division of the 
secondary pinne is usually not carried so far; while in quite 
sheltered plants the proportionate number of sterile leaves is 
very much greater than in exposed plants, and the sori in the 
former are commonly small. 
During the years 1902 and 1903 I made observations on the 
structure of the leaf of the bracken in different natural habitats, 
and I also carried out a cultural experiment, the results of which 
bear on the same subject. 
In describing these observations, it will be most convenient to 
begin with a comparison of the structure in plants from two 
kinds of habitat *, which rank as opposite extremes. 
If we compare two leaves t, one from a dry situation fully 
exposed to sun and wind, and the other from the deep shade and 
shelter of a dense copse, we find that the structure is rather 
strikingly different. In the exposed leaf, as compared with the 
sheltered one, the outer wall of the upper epidermis is consider- 
ably thicker, there is a well differentiated hypoderm (either con- 
tinuous or nearly so), the thickness of the leaf is considerably 
greater, the palisade-tissue usually occupies a distinctly greater 
proportion of the mesophyll and its cells are more elongated, 
while the spongy tissue usually appears less lacunar. In the 
sheltered leaf there is practically no hypoderm, e.g. a few cells 
accompanying the midrib and none elsewhere, or sometimes two 
or three cells accompanying some of the veins in addition, while 
a certain number of epidermal cells may contain chlorophyll. 
There may sometimes also be no differentiated palisade. The 
difference between these two types of leaf may be illustrated by 
* Observations were made on bracken-leaves in the following places:—Sheen 
Common, Richmond Park, and localities near Windsor, near Kemsing in Kent, 
and near Woodmancote, Telscombe, and Crowborough in Sussex. 
t For the structural comparison of two leaves the parts usually chosen were 
the lowest segments on the lowest-but-one secondary pinna of the lowest primary 
pinna, 
