10 INTLLODDCIORY ESSAY.
ID a few species, sucli as C. ormtus and 0. conirostris the leaflets gradually
dimiiiisli ill siie towards, and become abortive at the apex of the leaf, while the
rachis becomes moro "clawed" there than it is lower down, without, however, the
development of a terminal portion devoid of leaflets.
I f in the higher part of the stem of an adult Calamus the leaves are destitute
•of a cirrus, then it is certain that the species is never cirriferous except as an
abnormality, such as 1 have observed in C. heteroideusy in its VAR. pallens, \ and in
C. exilis, where occasionally a very slender cirrus appears at the apex of their
otherwise constantly paripiniiate leaves.
Wliole groups of species are characterised by the absence of cirii from their
leaves, and it never happens that a species with nou-cirriferous leaves in the upper
part of its stem is cirriferous lower down: the opposite condition is, however, of
common occurrence. For example, in a terminal portioa of C. asperrimus, one
metre in length, with five fully expanded leaves, the lowest of these leaves ended
in two leaflets with no rudiment of a cirrus between them, the next above had a
ecarcely aculeolate cirrus 7 cm. in length, while the sixth had the cirrus 55 cm.
long and densely armed with 3-nate claws. In the intermediate leaves the cirrus
exhibited a gradual transition between these two extremes.
When there is no cirrus the leaves end in two equal or sub-ecjual leaflets
which are usually the smallest on the whole leaf, and may be perfectly free at the
base or may be, to a greater or less extent, connate. Sometimes a very short
rudimentary cirrus appears normally between the two terminal leaflets, as in 0.
Blumei and C. rhomboideus ; still more rarely the leaf terminates in a small solitary
•undivided leaflet as in G. ramostssimus.
In most Calami the leaves are pinnate and the leaflets very regularly disposed
on both sides of the racliis, but in this case I have frequently observed all the
leaflets on one side slightly smaller than the corresponding leaflets of the opposite
side, and at times even more remotely inserted on the rachis. Very frequently the
inequidistant, geminate or attached in couples, as in C. latifoliua, or
in three, four, or even larger numbers along the rachis, as in C.
gracilis, C. fasciculaltis, etc.
In Calami the leaflets are always symmetric at the base except in C. Blumei
and to a less degree in C. tomenlosus; they never are sigmoid or falcate.
In shape the leaflets vary a good deal. Most frequently they are broadly linear^
more or less ensiform, or narrowly lanceolate and acuminate, where the proportion
of breadth to length varies from 1 : 10 to 1 : 20. There is, however, no lack of other
forms, such as lanceolate, oblanceolate, ovate, obovate, elliptic, oblong, spathulate,
etc. When the leaflets are comparatively broad they are often convex on one side
and concave on the other, or are spoon-shaped, specially towards their apices : In a
small group, that to which C. rhomboidetts belongs, they are more or less lozenge
shaped.
The leaflets almost always terminate in a symmetric, more or less acuminate and
bristly-brushed tip; very often, however, we may observe here a small notch or
indentation on the lower margin not very far from the apex ; this notch in many
THE LEAVES. 11
cases IS inconspicuous or even obliterated, while in others it is rather conspicuous
as in C. ausiralis, C. pseudo-tenuis and especially in C. Lobhiams. I suppose that this
notch may possibly be the last trace of the derivation of an acuminate
leaflet from a truncate one. Among Calami leaflets with a truncate transversely
cut or premorse end, analogous to those of a typical Ptyclmpei-ma, are only found
in C. caryotoides.
The leaflets in Calami, without a single exception, are attached to the rachis by
a very narrow point, tapering always more or less gradually or even abruptly
towards the base where the lamina, commonly doubled downwards, forms just at
the junction with the rachis a small cavity on the lower surface, within which in
many cases nestles a small swelling which resembles an extra-nuptial nectary. In
Calamus the leaflets are never decurrent along the rachis.
Very important charactcrs are derived from the main nerves or " c o s t » " of
the leaflets, as their varying distribution appears to correspond to a varying grouping
of the fibro-vascular bundles in the axial parts of the plant.
Most of the Calami have narrow leaflets with one or three primary nerves or
costie, running throughout their entire length, sharply prominent on the upper surface
and usually less distinct on the lower. In not a few species there are five main or
primary nerves, less often seven to niiie, and even, in 0. rhomboideus, twelve; but,
whatever the number of the costse may be, the central or midrib is almost always
the sti-ongest. All the primary nerves in the leaflets of Calamus traverse the entire
length of the blade and meet at its apex, except in the species of the C. rhomloideus
group, where the nerves nearer to the margins usually disappear at different
levels and do not reach the apex.
Between the primary nerves we find interposed others more numerous and more
slender; very frequently one of these secondary nerves runs along or very near the
lower margin so as usually to slightly thicken it; in the leaflets of Calamits a
primary nerve is never found running along the margin.
When the leaflets are rather broad, as is the case in C. 'pahstris, C. latifolius,
C. javensis, etc., their lower margin is often bordered on the upper surface by a
polished shining band a few millimetres wide; sometimes a few more similar bands
are found also in different places along the lamina; these lines correspond with
the portions of the leaflet that are outermost in the leaf-bud. As the leaflets before
the leaves expand are longitudinally plicate and closely packed together, these
polished bands look as if they had been produced by friction as they were being
forced out from the central terminal shoot.
The primary and secondary nerves are connected by means of transverse
veinlets which are moro or less interrupted and anastomosed. In some species this
secondary venation forms a very elegant network and is very conspicuous, the
component veinlets being close together, shghtly sinuous, thin but very well
defined and continuous, that is to say crossing transversely the whole width of
the lamina; good instances of this style of venation are seen in C. insiguis,
C. heteracanthus, etc. Usually these transverse veinlets are more distinct on the upper
than on the lower surface.
AKN. ROT. BOT. SARD. CALCuriA VOL. XI.