8 INTRODUCTION.
to the endosperm. The endosperm is penetrated, especially in the uppermost portion,
by deep, dark-coloured canals convergent on the centro. These canals are formed of
cells, elongated in the direction of the canal itself, which are also saturated with a
tannic substance, but which is, however, not collected in special cavities or sacs.
Two or three layers of very small cells situated under the tissue in which the
canals originate, or in immediate contact with the endosperm, constitute the testa
of the seed.
The endosperm offers nothing remarkably different from that of a Calamus. It
is horny or bony and perfectly solid even to the centre. From what has been
stated above, it would seem that the rumination of the seed of the Daemonovops is
not entirely due to the penetration of the outer integument within the endosperm,
since the extenial tissue which penetrates into the canals does not appear to be
derived from it, but from a layer of cells placed under the integutneat, a layer
which belongs apparently only to the nuoellus.
Together with the ripe fruits, I was able to examine ¡úso some ovaries of
D. JenJdnsianus also preserved in alcohol, which had already attained a diameter
of 4-5—5 mm. (PI. ), f. 7, 8). In ovaries which have attained this degree of
development, it is clearly seen that the galatinoua-tanniferous fluid, previously
mentioned, transudes from the exceedingly acute angle formed by the insertion of
the funiculus with the ovarian cavity, and accumulates unevenly around the ovule
durinc its transformation into a seed. This uneven accumulation produces depressions
hero °and there upon the surface of the young seed which vanish as the seed
arrives at complete maturity.
In the ovule arrived at the degree of development just stated {i.e., of -1—5 mm.
diameter), there may be observed, in a longitudinal section following the raphe and
passmg by the micropyle (PI. I, f. 8), a considerable cavity in the central part, of
irregular shape (the embryo sae) which starts from the micropyle (PI. I, f. 8), and
extends in the direction of the periphery at the point which corresponds to the
chalaza (PI. I, f. 8, </); this cavity is lined with a delicate layer of endosperm in
a state of formation, and its surface, at the point nearest to the chalaza, is
rendered uneven by tiny protuberances which are no other than the beginnings of
the canalicuU which will later bring about rumination.
Already at this stage of development, in the fleshy mass which later foms
the integument of the seed, the tanniferous sacs can be seen in very considerable
numbers These sacs probably draw by endosmosis out of the mass the tannic
substance which exudes between the ovule and the ovarian wall. The integument
of the seed becomes, in the ripe fruit, fleshy and is frequently acidulate or
lX.~The Seed.
The seeds of Daemonovops are for the most part globular, though often of
irregular shape. When the integument (which, as it dries, adheres closely to the
seed) is stx-ipped off, the surface is seen to be finely pitted all over. These tiny
openings indicate the beginnings of the canals, full of the tanniferous substance,
which penetrate the interior, and are the cause of the rummation of the
Sometimes the seed is more or less compressed or flattened, but more often it
show.'? a prominence or swelling on the side of the raphe {D. melanochaetes, D.
Pierreams D. anqustifolius). In Oalamm the depression of the chalaza or chalazian fovea,
is very strongly marked on the raphal side, but in Daemonovops it is most often
hardly to be made out, beiag quite superficial and punctiform ; however, in the
species belonging to the D. Draco group, it is indicated by a very narrow furrow-
Only in the seeds of D. CalipparMs among all those of the genus Laemonorops
which I have cut open, have I found it very distinct, pit-like and running almost
into the centre of the albumen. IQ D. bptopus the rumination of the seed is
only very slightly marked. The embryo in Daemonoropn is, without exception, basal
or nearly so.
X.—Floral and extrafioval Nectaries.
As in Calamus, so in Deamonorops, one meets with organic parts which bear
the appearance of extra-iiuptial nectaries. Also in Daemonorops the aforepaid parts
sometimes are more evident than in the other genus, although situated in the
same places, namely, in the axilla of the leaflets, and at every division of thg
spadix at the junction of the branches and of the branchlets, as well as at the
axilla of the appendicular organs of the spikelets. From observations subsequent to
thnse published by me under Calamus, I have been enabled to observe how very
numerous are the palms which exhibit tiimefactiona or callosities at the bases of
the segments of pinnated fronds; these tumefactions are formed of a tissue which
even on purely superficial examination, appears to differ entirely from that of therachis
and of the leaflets. In the palms most commonly cultivated in our gardens
in Europe, such tumefactions are extremely plain in Phoenix, but they are so aboveall
others in the^Cfc;« of the 0. capitata group, in which there exists, in the superior
axil of the leaflets, and also in the hollow formed by the bending down of their
limb on the under-aurface, a fissure having more or less tumified edges of a light
colour..
In Gaussia splend'-ns, a Cuban palm, these callosities or tumefactions at the
base of every segment are developed in a very unusual degree. In PUoga polystacMa
Noronha [Diipsis nodifera Mart.), a palm of Madagascar, the callosities in the axilla
of the leaflets are very large, and appear to be frequently visited by certain insectswhich
transform them into quite conspicuous galls.
In the Cocos mentioned, I have not observed that the callosities were ever
visited by insects of any land, nor have I ever noticed any nectar flowing out oi
the fissure; while, oa the contrary. I have seen insects, and especially bees, in great
numbers upon their flowers. It is therefore possible that those callosities of Phoenix
and Cocos act in both cases purely as supports to their respective leaflets, but it is
also possible that in certain palms tbey may act as lures to particular kinds of
insects, such as ants, for 1 have observed that in Phoeniz sylvesiris and Cocos capitata
and related species, many of the cells of which these callosities are composed
are saturated with a saccharine fluid.
I have since been able to satisfy myself that the same thing occurs in the
hard callosities found at the axils of the male spikelets in B. verticiUaris which
ANN. ROT. BOT. QAKD, CALCUTTÌ, VOL. X I I .