g
water their roots, stems, and sometimes their leaves, become excessively
elongated, as in _Ranunculus fluitans_, the flower-stalks of _Valisneria
spiralis_, &c. These are cases of variation rather than of malformation,
but are none the less curious, or sometimes perplexing; thus, Lapeyrouse
described, in his 'Supplement a la flore des Pyrenees,' p. 27, under
the name _Potamogeton bifolium_, a plant which Mr. Bentham subsequently
discovered to be nothing but a flowerless variety of _Vicia Faba_
distorted by its growth in water.[510]

=Elongation of the root.=--This, as already remarked, is more often a
variation than a malformation, and is usually due to the presence of
water at a distance necessitating growth at the extremities of the root,
or to the presence of some obstacle, such as a stone, to avoid which the
root elongates till it has passed the obstruction. Occasionally in
Crocus corms some of the fibrils may be met with much lengthened and
thickened, and invested with a fleshy sheath. It is not certain,
however, that these structures are roots; possibly, nay probably, they
may be processes from the stem thrust downwards into the soil, similar
to the formations already described in the tulip (p. 85, fig. 39).

=Elongation of the inflorescence.=--Under this heading it is necessary
to consider lengthening of the common rachis in the case of an aggregate
inflorescence, and lengthening of the individual flower-stalks, whether
they be solitary or portions of a multiple inflorescence. The two
phenomena may occur together, but they are quite as often independent
one of the other. Thus, among _Umbelliferae_ the umbels are occasionally
met with supported on unusually long stalks, while the pedicels of the
individual flowers may or not be increased in length; so also with some
of the Composites, or the heads of flowers of some _Leguminosae_,
_Trifolium repens_, &c. &c.

Another illustration of the sort is that recorded by M. Fournier,
wherein the usually umbellate inflorescence of _Pelargonium_ was,
through the lengthening of the main stalk, transformed into a raceme.
Among Composites a similar change may sometimes be met with.

MM. Clos and De Schoenefeld have recorded the existence of a variety of
the sweet chestnut (_Castanea_)in which the female catkins were as long,
and bore nearly as many flowers, as the male spikes. This is stated to
be of constant occurrence in some localities, and to be accompanied by a
dimini