VI IKTEODFCTIOjN".
wliile slioots of Bamhiisa Tulda, according to Roxbargli, rise to their full size of from
20 "to 70 feet in hciglit in about 30 days. The CULM-SEEATHS, to winch we have
referred as surrounding the young shoots, arc veiy interesting, for they are almost
always of shapes which arc characteristic of the species to which they belong. In
regard to this, Munro says: —" The spathes or large sheaths whicli cover the nodes
"or lower j^ortions of all bamboos vary much in size and appearance, and will, I
"think, afford good characters when they are more studied and better known.
"Dr. Brandis has paid considerable attention to this subject; but these sheaths do
"not appear in general to have attracted the notice of collectors." Kurz, too, held
strongly the opinion that these culm-sheaths were very important in classification,
and his collection of drawings of culm-sheaths, deposited in the Herbarium of
the Eoyal Botanic Garden, Calcutta, has proved of great service to me. Thanks
to the exertions of many friends and to the facilities for collection I have myself
enjo3'ed, there are not many species herein described of "which the sheaths are
unknown; and a glance at the drawings of them, and, still better, at the Herbarium
sheets, will show that almost all have some definite characteristic which is sufiicient
in the absence of the flowers to identify the species. Culm-sheaths have three
princii)al parts. The fiisé part, the sheath proper, corresponding to the petiole
of ordinary leaves, appears in bamboos as a broad expansion with its base attached
at the node of the cubi. Sometmies the sheath is very thin and papery, as in
most Jrtindinariae ¡ sOmetunes it is thick and smooth, as in most Dendrocalami
and Bambnsce; and sometimes it is coriaceous in texture, as in Dinocliloa and
Oxytemnthera Bourdilloni. In regard to clothing, some species {e.g., Dendrocalamm
silckimensis) have a dense felted mat of brittle stiff hairs all over the outer surface,
and from this down to the nearly glabrous sheath of Dendrocalamm giganteas
there is almost every gradation. The second part is the " imperfect blade," corresponding
to the blade of a leaf, and is inserted on the top of the sheath, where
it takes many forms and shapes^ and frequently is decurrent into " auricles," which
often are fringed in various ways with stiff bristles. In most species of Anmdinaria,
Phyllostachjs, Thyrsostachjs, Oxgtenanthera, Dendr o calamus, Melocanna, and
Teinostacliyum the imperfect blade is narrow, frequently recui-ved and long; while in
Bavibusa, Gigantocliloa, Dinocliloa, and some species of Cephalostachyum it is broad,
triangular and much decurrent. The blade of Bamhisa hhasiana is swollen out and
inflated, while in Ochlandra setigera it is scarcely more than a fine needle-like point.
The third part is the ligule, inserted, as in the leaves of all grasses, on the inner
surface at the junction of the sheath and blade. There is, of course, in all parts,
as Kurz has pointed out, a good deal of difference in size and shape, according
as the sheath is taken from the base, the middle, or the top of a cuhn, or from
a side branch; but a little study and experience soon teaches us to recognize the
general characters. Almost the only cases I know of in which the culm-sheath
fails to }4eld a distinguishing character are Bamhusa Tiilda, H. mttans, B. terns
and Gigantochloa macrostachja, in which four species the culm-sheaths arc very
similar in appearance.
INTRODUCTION. vu
The L'EXTES of all bamboos are very similar in general appearance, for, although
some species have usually large leaves and others quite small leaves, the size
depends much on the part of the plant from which they are taken. Thus, in
Dendrocalamus HamiUonii, the leaves of young shoots and the end leaves of strong
branches are usually very large, while those of medium branches arc moderate in
size and those of thin shoots from lower nodes are quite small. In respect to determination
therefore, as ICurz says, "little value can be attached to the size, shape
and "ncrvature of bamboo leaves." Bamboo leaves are usually linear, lanceolate or
oblong-lanceolate in shape; they have usually a short petiole into which the base,
which is frequently unequally cut, extends; the point is usually long acuminate, often
scabrous, sometimes shaggy [Bamhusa hhasiana and B. marginoid); the edges
are often scabrous; the sides glabrous or softly hairy and tlie veins parallel and
prominent. I have described these veins as they are usually seen: first a midril)
or main vein usually thick; secondly a number or pairs of secondary veins easily
seen with the naked eye; and thirdly a number of intermediate veins, usually 5 to 7,
for seeing which a lens is required. Then there are fourthly the transverse veinlets,
and here I wish to note that, so far as my own observation goes, true transverse
veinlets occur only in the genera Artindinaria and Phyllostachys ; in others, -what
appear to be transverse veinlets are not really such, but are caused by glands
which in fresh specimens are seen thi-ough the leaf as pellucid dots and in dry
specimens as raised lines, giving the appearance of cross bars between two neighbouring
intermediate veins. At the base of each leaf, below the petiole, come the " leafsheaths"
and "ligules"; both often giving good characters for the identification of
species; for the leaf-sheaths are often furnished with cilias or Inistles and small am-icles
of various shapes, and the ligules may be of different degrees of prominence, those
of Ochlandra Brandisii and Oigantochloa ligiilata being especially long.
In regard to INFLOUESCENCE, there is great variation among bamboos; sometimes
the spikelets appear on leafy branches, sometimes in gigantic panicles covering a
whole culm; sometimes the spikelets are very few and scanty, sometimes they are
extremely numerous; sometimes they are distant on the branches of the inflorescence,
sometimes congested into large rounded heads. The inflorescence is made up of
spikelets with or -^^ithout bracts. These spikelets vary much in the number and
arrangement of the flowers, but they all contain (I) empty glumes, usually two,
sometimes more, sometimes fewer; then a variable number of flowers, of which the
lowest and the last may often be empty, but consisting of (2) a flowering glume,
generally similar to the empty glumes; (3) a palea which is usually keeled or
convolute and embraces (4) the lodicules which vary in number, being sometimes
absent, as in most Dendrocalami, sometimes very many, as in Ochlandra, but usually
three in number as in most genera: (5) the stamens which are three in number
(as in most Arundinariece), six (as in Biibamlusem, Dendrocalamea. and most
genera of Melocannem), or many (as in most species of Ochlandra, one of wdiich
may have as many as 120); and (6) the ovary., surmounted by the style and
stigmas. In all these parts there is considerable variation, and indeed in the