414 DR. OTTO STAPF ON THE 
cells (Pl. 46. figs. 30, 31) constituting the tissue, which in this condition may measure as 
much as 0-75 mm. in diameter. This means that, in its crushed state, it was compressed 
to Ae to 4y of its original thickness. In expanding this tissue, it always struck me 
that the outer and middle strata yielded much more readily to the needie than the 
innermost, which often remained in their compressed condition. The cells of the 
outermost two or three strata are the smallest and nearly isodiametric. They contain 
the scanty remains of the plasmatic wall-lining and of the nucleus. Farther inwards 
the cells increase rapidly, with their greatest diameter in the direction of the radius. 
They are completely empty or contain the vacuolate skeleton of a nucleus. All the 
cells of this tissue possess extremely delicate walls giving typical cellulose reaction, 
and are destitute of pits or any other sculpturing. What appears at first glance as 
tangential striation is nothing but the delicate folding of the completely erushed radial 
walls, and corresponds to the equally delicate undulation they exhibit in transverse 
sections (Pl. 46. fig. 32). 
C. The Embryo. 
'The general strueture of the embryo on the whole agrees with that of the typical 
grass fruit, with the exception of the enormous development of the scutellum, which 
either completely fills up the cavity of the pericarp or, towards the top, leaves a portion 
of it empty. We may distinguish the following parts in the embryo (Pl. 47. figs. 36, 44, 
51):—(1) the plumule; (2) the coleorrhiza with the primary and several secondary roots 
within it, the whole corresponding to the “ radicle " of descriptive botany ; (3) the curiously 
modified and enlarged scutellum ; and (4) the remaining portion of the embryo, con- 
stituting the hypocotyl. The most striking feature of the embryo next to the scutellum 
is the curvature of the embryo, which is so strong that the inner faces of the plumule 
and the radicle are contiguous. This is unique in grasses, although an approach to it 
may be seen in the curved embryos of Oryza and Leersia. It is evidently due to this 
condition that no trace of an epiblast is seen. Its place would be the inner angle of the 
bend, where there is no room for such an organ. These being the main features of 
the embryo, I will now proceed to describe the different parts more in detail. 
L The Plumule (Pl.45.fig. 3, pl.; Pl. 47. figs. 36-40, col.).—The plumule has the 
shape of a short, somewhat oblique cone, and consists of a coleoptile with a bud in its 
axil, and about seven distinct young leaves and the rather flat growing point. The 
coleoptil (Pl. 47. figs. 44, 45) is open on the top and along a slit descending on the 
inner face which is contiguous with the radicle, the margins being minutely ciliate. 
According to Schlickum (“ Morph. u. anat. Vergleich d. Keimpflanz. d. Monokotyl.,” in 
. Bibl. Botan. vi. (1896) p. 68), the coleoptile of the grasses differs essentially from the 
leaves following it in the absence of stomata and the presence of only two vascular ` 
bundles. This is not the case in Melocanna, the coleoptile (Pl. 47. fig. 46) in these ` 
respects entirely resembling the leaves nearest to it. It possesses not only stomata 
_ and numerous vascular bundles, of which about 17-19 are of the first order, but the 
T structure of the outer epidermis is much more complicated than might be expected. 
