formation may be seen to
a minor extent. Even the common _Scolopendrium vulgare_ occasionally
produces small pitchers of this character, as in the varieties named
_perafero-corautum_, Moore, and _peraferum_, Woll.[350]

In carnations leaves may sometimes be seen from both surfaces, from
which project long, sharp-pointed tubular spurs at irregular intervals.
A very singular illustration of this is figured by Trattinick,[351], in
which the leaves, epicalyx, sepals, and petals, were all provided with
tubular spurs.

In _Cephalotus follicularis_ rudimentary or imperfect pitchers may be
frequently met with, in which the stalk of the leaf is tubular and bears
at its extremity a very small rudimentary leaf-blade. It is not in all
cases easy to trace the origin and true nature of the ascidium, as the
venation is sometimes obscure. If there be a single well-marked midrib
the probability is that the case is one of cohesion of the margins of
the leaf; but if the veins are all of about equal size, and radiate from
a common stalk, the pouch-like formation is probably due to dilatation
and hollowing of the petiole. Again, when the result of a union of the
margins of the leaf, the pitcher is generally less regular than when
formed from the hollowed end of a leaf-stalk. Further information is
especially needed as to the mode of development and formation of these
tubular organs, so as to ascertain clearly when they are the result of a
true cupping process, and when of cohesion of the margins of one or more
leaves. (See Cohesion, p. 31. For bibliographical references consult
also A. Braun, 'Flora v. Bot. Zeit.,' 1835, t. xviii, p. 41,
_Aristolochia_.)

=Tubular formations in the flower.=--A similar formation of tubes
happens in some double flowers; for instance, it is not infrequent in
double flowers of _Primula sinensis_, in which tubular petal-like
structures are attached to the inner surface of the corolla; sometimes
these petaloid tubes replace the stamens, while at other times they
appear to have no relation to those organs. In the particular flowers
now alluded to the tubular form seems due to a dilatation, and not to a
cohesion of the margins. (See Cohesion, p. 23.) These tubular petals
resemble in form and colour almost precisely the normal corolla in
miniature, but are not surrounded by a calyx, nor do they contain
stamens, while the less perfect forms show clearly their origin from a
single tube-like organ.

[Illustration: