AQUATIC PLANTS. 
75 
first, illustrated by the Water-lilies, Limnanthemitm, Ilydrocharis^ 
Fotamocjeton natans, &c., is more or less rounded, leathery, with an 
upper surface like that of land -plants, but having the stomata upon it, 
and with a leaf- stalk capable of intercalary growth at its junction with 
the blade. Goebel points out that Limnanthemum presents an advance 
upon Castalia, in that the leaf in the former springs from the peduncle, 
so that food could travel to the seeds direct instead of going down a long 
petiole to a rhizome and then up a long peduncle. The ribbon type of 
submerged leaf occurs in Callitriche, longer and narrower the deeper 
they are below the surface, in Vallisneria and in some of the leaves of 
Sagittaria, Glyceria, &c. The much-divided leaf occurs with others of 
the floating type in Salvinia, Trapa, Water Crowfoots, and in Cabojuba, 
one of the Water-lily family, and by itself in Myriopliyllum, Hottonia, 
Utricidaria, &c. With this type we should perhaps associate the unique 
leaves of the Lattice-leaf of Madagascar {ApGnogeto?i fenestrale, Hook, fil.)^ 
submerged leaves of oblong form, in which most of the cellular tissue 
between the intersecting veins breaks up as the leaf grows, leaving 
perforations. Goebel points out* that here water with its dissolved gases 
has ready access to all parts of the leaf, so that none of the intercellular 
spaces which occur in other aquatic leaves are requisite. The subulate, 
or awl-shaped, type, generally in radical rosettes ascending from the 
bottom of lakes of varying depth, is approximated in the flat leaves of 
Stratiotes, and more characteristically seen in Isoetes, Subularia, Lobelia 
Dortmanna, and Littorella. The last-named exhibits a marked difference, 
its leaves when submerged being longer, more erect, and more cylindric 
than when not so. It is, I believe, a general characteristic of this type 
of leaf to have an apical growing point of long-continued, though limited^ 
growing power. 
The varying directions taken by adaptation might equally be illus- 
trated by the flowering processes of aquatics, where, besides such sub- 
merged types as Zostera, ^Ye have floating "hydrophilous" forms like 
Vallisneria, wind-pollinated flowers like the Bur-reed (Sparganium), and 
insect-pollinated ones like Utricularia. I have perhaps, however, 
already said enough to illustrate the great variety of plants which have 
been made by the conditions under which they find themselves to 
approach one another in certain characters, especially those of their 
tissues and vegetative structures. 
Time does not allow any adequate treatment of the last great problem 
in connection with these adaptations, viz. the extent to which they are 
hereditary. Kerner, after describing many modifications produced by 
change of soil and climate, says t that they were also manifested by the 
descendants of these plants, but only as long as they greiv in the same 
p)lace as their parents,'' and he thus discriminates between inherited 
characters, which he considers specific, and those liable to disappear by 
reversion, which he terms varietal. It nowhere appears, however, that 
in the course of his experiments he did anything to test the question 
whether in a series of generations grown under identical conditions 
characters originally acquired, and liable to disappear by reversion, do 
* Pflanzenbiologischc Schilderiingcn, ii. 319. 
t Op. cit. ii. 514. 
