2 2 INTRODFCTOEY ESSAY.
In the male spifcelets the partial iiifloroaoenccs are usually branched two or
three times and tlie division or branchlets which remain ia their basal portions may
be considered to be ooinpound or branched spikes, while the ultiuiato divisions of
these are the .true or simple spikelets.
As regards size, the different divisions of the inflorescence and the spikelets follow
tlie general rule; the upper branches and spikelets are gradually Buialler than the
lower ones and every partial iaflorescence assumes a more or less pyramidal form,
The terminal spikolet of each division is usually considerably larger than the others.
The male spikelets are almost always shorter and more densely furnished with
flowers than the femalo ones.
The axis oE the male spikelets is clotbod with spathels which are commonly
broadly and asymmetrically infundibaliform (PLATE I, fig. Vc, and PLATE II, fig. 4A);
sometimes, however, their tubular part is so short that they look like bracts or ar©
boat-shaped or spoon-shaped with the axis passing exceatrically thi-ough them; the
spathels are rounded on one side and slightly prolonged on the other, the outer, into
a small point.
ILL the axil of each spathel is a flower with its special involucre (PLATE I,
%. li, and I'LATE I I , fig. 4i). The involucre has usually the form of a small cup
(PLATE I, figs. l - 3 i ) , being more or less concave; in a few instances we find in
placc oi the involucre two small scales or bracteoles united by their bases which clearly
the origin of the involucre from two coalescing appendicular organs; for
in C. asperrimus, G. siphonospathus and others.
Authors usually term the involucre a "spatliellula," a name that may easily be
mistaken for that of a " spathella," and one that, as we shall presently see, has not
been always properly and uniformly applied. In the simple spikelets it is always easy
to distinguish the spathels from the other appendicular organs as it is from the axils
of these spathels that the flowers arise; but in the case of compound or branched
spikes it is a spikelet, which has its own spathels, that is situated in the axils of a
spathel; in this case therefore we have primary, secondary, and even tertiary spathels,
according to the degree of division of the primary spike. But all this is of very
slight importance since, as a rule, the secondary and tertiary spathels differ from the
primary only in size. It is sufficient if wo hold that in a male spikelet we mean
by a spathel tlie appendicular organ which clothes the axis of a spikelet—whether of
& primary or a secondary spikelet is of no consequence—and that a spathel of the
ultimate branching protects a flower which in turn has its base enveloped by a special
involucre.
The modifications, peculiarities and diagnostic characters afforded by the spathels
are essentially the same as those of the female spathels to be described further on;
but as a rule the spathels of the male spikelets are shorter, broader, more distinctly
infundibuliform and more closely packed than those of the other sex. The flowers of
the male spikelets are solitary in the axil of each spathel except in a very few
species, such as G. viminalis, G. fseudo-tenui^ and some few others. In G. viminalis I
have counted as many as 8 glomerulate flowers to each spathel; but in this
the glomerules must be considered secondary much reduced spikelets as each flower
is provided with its own diminutive bracteole.
S P I K B L E T S .
Female Sfikclet8.—'T\iQ female spikelets are as a rule more elongated than the male
•ones and though the flowers are bifarious they are more remotely disposed.
The female spikelets also have tlieir axis constantly clothed with tubular or
slightly infundibuliform spathels but these are longer and less distinctly infundibuliform
than those of the male spikelets and are not infrequently represented by membranous
rings round the axis of the spikelets with which in their lower part they are
organically fused (PLATE I, figs. 8 - 1 a n d PLATB 11, figs. 1 and 3/;). It is therefore
often diflicult to distinguish externally the point where the spathels are differentiated
fi'Om the axis.
In some cases the spathels are much reduced in size, are devoid of any tubular
portion and resemble bracts. The morphological structure of the female spikelets does
not differ essentially from that of the male, but in the female the appendicular organs
together with the flowers which arise from the axil of or above each spathe must be
considered in their entirety as a secondary much contracted spikelet. This mode of
interpreting the female spikelets of Calami is in accord with the general plan of
division of the spadices, and assumes that it is supra-decompound as is almost
universally the case with the male spadices. According to this interpretation the female
spikelets would also have been supra-decompound if the small contracted spikelet
existing at each spathel had undergone further development.
The flowers of the female spikelets are inserted, like those of the male ones, in
-or a little above the axils of their respective spathels, but in the female sijikelets
the female flower has two involucres in place of one. And from the descriptive
point of view it is most important to fix the nomenclature of these two involucres.
The organ which remains inside or a little above each spathel and which is
immediately in contact with the axis of the spikelet has been termed by me an
"involucrophorum"; it corresponds with the " spatbellula" of Martins and with the
"bract " or " bracteole " of Griflith.
I have not considered it advisable to retain the name of spathellule for the
involucrophore, because this does not correspond morphologically to what has been
termed the spathellule in the roale spikelets and because it appears to me to be an
organ of axial rather than of appendicular structure. The name spathellule implies a
morphological agreement with the spathel to which it should bear the same relationship
as the spathel does to the spathe or as a secondaiy spathe does to a primary
one. As a matter of fact, however, the "involucrophore" appears to represent the
shortened or contracted axis of a small spikelet, provided with its spathel and bearing
besides its appendicular organ which is the involucre. This structure is very evident
in those species that have the spikelets inserted at the base of their own spathe, as
in C. Btirckianus, C. Zollingeni, 0, GnJfUhianus, etc., when the involucrophore both
emerges from the base of its own spathel and is more or less pedicellate (PLATE
II, iigs. 6-9a).
The precise pTace of insertion of the involucrophore is not always at the base
of its spathel, but is much more frequently just at its mouth or a little above or
below the mouth. When the involucrophore is inserted outside its own spathel, it
generally seems attached to the axis of the spikelet; but in fact it adheres laterally
to the base of the spathel above that in the axil of which it ought theoretically to have