240 ABSORPTION-CELLS ON LEAVES. 
inserted above to the green leaves, are metamorphosed into little saucers. In one 
species of Teasel, Dipsacus laciniatus (see fig. 561), and in the North American 
Silphiwm perfoliatum (fig. 56 *) the two sheathing portions (vagin®) of every pair of 
opposite leaves are connate and form comparatively large and deep funnel-shaped 
basins, from the middle of which rises the next higher internode of the stem. In 
several Meadow-rues (Thalictrwm galioides and T. simplex) the secondary leaflets, 
which are opposite one another and shut close, almost like the valves of a mussel, are 
moulded so as to form cavities for the retention of water, and in many Umbellifere, 
such as Heraclewm and Angelica, the vagina of each individual leaf is ventricose 
or inflated, thus forming a sac enveloping the segment of the stem which stands 
above it. 
These basins, saucers, and dishes are always so placed, relatively to their 
surroundings, that the water derived from rain and dew is directed into them from 
the surfaces of the leaves, or by the segment of the stem which rises from their 
centres, and thus it is that the depressions are filled. Whether in all cases much of 
the water accumulated is absorbed is certainly open to doubt. In the case of the 
leaves of the Alchemilla (fig. 527), which exhibit the phenomenon so conspicuously 
that the plant has received the popular name of Dew-cup; the absorption of water 
is, at anyrate, very inconsiderable, and here the retention of the dew secures 
advantages of a different kind to which we shall presently have occasion to return. 
On the other hand, it is established that in the case of basins belonging to tall 
herbaceous plants, particularly such as grow on steppes and prairies where often 
no rain falls for a long interval, the water collected is absorbed by the glandular 
hairs and thin-walled epidermal cells developed within them. The fact of this 
absorption may be proved by a very simple experiment. Let a stem of the 
Silphium, represented in fig. 56°, be cut off beneath the pair of connate leaves, which 
form a basin by their union, and let the cut surface be closed with sealing-wax, so 
that no water can be taken up by the stem from below. If the water accumulated 
in the basin is now emptied out, the leaves shortly become flaccid and droop; but if 
the basin is left full of water, the leaves preserve their freshness a long while and 
do not begin to wither until all the water has evaporated and disappeared from the 
basin. If oil is poured upon the collection of water in the basin, so that evapora- 
tion from the latter is impeded, a constant diminution of the water in the basin is 
observed notwithstanding; this leads to the conclusion that the water in question is 
really taken up by the absorption-cells at the bottom of the basin and conveyed to 
the tissue of the leaf. 
The first thing that strikes one on surveying once more all the plants possessing 
on their aérial organs special contrivances for water-absorption is that a large 
proportion of them have taken up their abode in swamps and on the banks of 
rivers and streams, or if not there, at all events in situations where no danger 
exists of the ground being thoroughly dried up. No doubt this appears to be 
inconsistent. How are we to explain the fact that Gentianez, ashes, willows, alpine 
roses, bog-mosses, &e., are still in need of water from the atmosphere, when they all 
