FRUIT OF MELOCANNA BAMBUSOIDES. 417 
correspond to the inner angle of the embryo proper, the broad back of the cylinder 
fragment to the ideal contaet-plane with the scutellum, whilst only a small portion of 
the cylinder surface, a triangular piece on each side, is left exposed ; but even of these 
flanks, a considerable portion or the whole is overlapped by the deflexed rim of the 
scutellum. "The epidermis of the hypocotyl resembles that of this rim. Below it there 
is parenehyma containing starch in quantities inereasing in the direction towards the 
scutellum. The centre of the hypocotyl is, however, occupied by a plate (Pl. 47. figs. 40, 
48, 51) consisting of a perfect tangle of vascular strands recalling the structure of 
mazerwood (Pl. 47. figs. 52, 53). This plate extends close to the outer base of the coleoptile 
on one side, and to near the root initials on the other. It is convex and well-defined 
towards the scutellum, but breaks up towards the plumule into numerous procambial 
strands going to the coleoptile and the leaf-primordia. Another set of vascular strands 
enters the scutellum, starting abruptly out of the dorsal surface of the plate. The 
structure of the strands is on the whole plain enough (Pl. 47. fig. 54). They consist of 
procambium with usually simple threads of tracheids. Only those which are situated 
towards the ends of the plate, that is near the outer base of the coleoptile and the root- 
primordia, show a higher differentiation, containing rows of wide cells with grate-like 
perforations (a kind of short-jointed vessels), thin-walled parenchyma more or less rich 
in starch, and subprosenchymatie portions with lignified thick walls and full of starch, 
and finally a slender strand of delicate phloem. The space between the tangled strands 
of the axil plate is occupied by parenchyma containing starch. This concludes the 
description of the internal structure of the fruit. 
VII. GERMINATION, 
I have already quoted Kurz’s and Robinson’s remarks concerning the germination of 
the Melocanna fruit on the tree (p. 402). There is no more recent account of this 
phenomenon, which is, with the exception of JJelocalamus, apparently quite unique 
among grasses; their statements, more especially Robinson’s, are, however, so precise 
and definite, that there can be no doubt that we have in Melocanna a remarkable 
example of vivipary. All the fruits of Jlelocanna which were at my disposal, with 
the exception of those from the Garo Hills, possessed a well-developed embryo, and in 
about half of them the radicle and plumule had already pierced the pericarp, the 
former having in several cases produced a bundle of secondary roots. But no fruit was 
gone so far as some observed by Kurz, who says of Melocanna bambusoides (‘Indian 
— Forester,’ i. p. 268): “the fruits drop from the plant after they have made shoots some 
6 inches long." | 
The first outward sign of beginning germination is that the embryotegium turns brown 
or black, and is lifted up like a lid by the tips of the plumule and radicle which appear 
simultaneously. They soon push the embryotegium aside; but it may still remain 
attached to the fruit for some time. Then the plumule (Pl. 45. figs. 6, 7; P. 47. figs. 41,51) 
grows out; its young leaves burst through the coleoptile, and push out like the joints of a 
telescope. A small shoot is formed, densely covered by firm, rather thick, puberulous 
SECOND SERIES.— BOTANY, VOL. VI. 30 
