418 ; DR. OTTO STAPF ON THE 
cataphylls. As already remarked, this shoot may attain a length of 15 em. before the 
fruit falls; but, as it seems, it develops scarcely any blades for some time, as Kurz 
says (* Indian Forester,’ i. p. 268) “these shoots attained a height of 15 to 16 inches after 
only ten days that they were planted out in the ground, but had thrown out only 
two leaves:" a statement which is borne out by a drawing of a seedling in Gamble's 
herbarium. 
Simultaneously with the plumule, the root-system develops. The primary and the 
secondary roots break through the coleorrhiza and grow rapidly into a bundle of thick 
root-fibres, among which the primary root is not readily distinguished, if it does not 
altogether die early, as seems frequently to be the case. Plumule and roots thus quickly 
outgrow the depression at the fruit-base from where they started, diverging at the same 
time. The scutellum, however, remains enclosed in the cavity of the pericarp, a neck 
being formed by intercalary growth at the base of the scutellum. This neck may attain 
a length of almost 10 mm. with a diameter of 4-5 mm. (Pl. 45. fig. 7; Pl. 47. figs. 41, 51). 
Longitudinal sections through a germinating fruit show this neck clearly, but they 
scarcely reveal any other macroscopically visible change in the scutellum and the peri- 
carp, except in so far as in some cases certain peripheral portions of the pericarp, chiefly 
in the upper part, appear brown, a sign of decay. The microscopical examination of the 
pericarp, however, shows that the macroscopic changes in the germinating fruit are 
accompanied by the gradual exhaustion of the pericarp. The first sign of this exhaustion 
is the breaking up of the compound starch-grains, after which the starch disappears 
gradually. At the same time the outer lamellze of the thick cell-walls of the parenchyma 
are dissolved, the walls appearing corroded all over, so that finally, in places, not much 
more is left of them than the central lamella. Whilst the depletion of the pericarp 
had already gone far in some of the fruits which were at my disposal, there was no 
change yet noticeable in the scutellum. This organ evidently remains active as long 
as there is any possibility of drawing on the pericarp, and its own exhaustion sets in 
only after the depletion of the pericarp. 
Kurz's and Robinson e accounts of the germination of the Melocanna fruits on the 
tree (see p. 402) give the impression as if this condition was the normal state. No doubt 
it is of frequent occurrence, and may be the rule. Still the structure of the fruit is 
such as to justify the assumption that the young plaat may just as well establish itself in 
the ease of the fruit falling before germination has set in. In fact, Gamble states 
(‘Indian Bambuseæ, p. 120) that he grew young Velocanna plants from fruits which had 
not yet germinated when they were put in the soil. Nor should it be overlooked that 
the chances of injury to the shoot of germinating fruits when falling from a height of 
49-70 feet must be very great, even if they fall into very soft soil. Unfortunately we 
know nothing as to the condition of the Melocanna forest at the critical time when its 
restocking takes place, nor have I been able to ascertain in what season the fall of the 
fruit occurs. So much, however, seems to be certain that the fruit once detached from 
the mother-plant does not stand a long delay, at any rate under the conditions to which it 
. would be naturally exposed on the ground of the forest. There is apparently nothing in its 
