STRUCTURE OF BAMBOO LEAVES. 81 
parenchyma, which alternate with bands without stomata (Pl. 13. fig. 28, Melocanna 
bambusoides; Pl. 14. fig. 37, Ochlandra Beddomei, Gamble). On the upperside the stomata 
are found sparingly between the longitudinal nerves and the bands of bulliform cells (P1. 11. 
fig. 4, Arundinaria japonica, and fig. 9, Bambusa vulgaris). The stomata of Bamboos are 
often surrounded by remarkable excrescences, fully described by Wendehake and shown 
in figs. 2 & 4. These excrescences are employed by him to distinguish several species 
of Arundinaria and Phyllostachys. Into this subject I have been unable to enter. 
I have already explained that in Bamboos the bulliform cells are frequently filled 
with silica. So are, as described in detail by Grob, many of the short cells (Pl. 13. fig. 23, 
Dinochloa andamanica; fig. 21, Cephalostachyum pallidum). In addition to these, many 
of the linear or oblong epidermis-cells of the upper, as well as of the under side are 
often completely filled with silica (Pl. 11. fig. 11, Pseudostachyum polymorphum; Pl. 13. 
fig. 21, Cephalostachyum pallidum; fig. 22, C. capitatum; fig. 27, Melocanna bam- 
busoides). On the underside they are not rarely found clasping the stomata (fig. 38, 
Ochlandra Beddomei). This, however, is not a character peculiar to Bamboos. Grob 
figures linear epidermis-cells filled with silica in Leptaspis conchifera, t. 4. fig. 1 (a grass 
belonging to the tribe of Phare, Stapf, in * Flora Capensis, which, as will be explained 
further on, in regard to leaf-structure is closely allied to Bamboos), as well as in two 
species of the tribe Andropogone:e, Andropogon Ischemum, t. 3. fig. 19, and Ischemum 
petiolare, t. 3. fig. 16. They will doubtless be found in many other species. 
It may here be mentioned that the tendency to deposit silica in the cells of Bamboo 
leaves is not limited to the cells of the epidermis, but that cells of the chlorophyll 
parenchyma and even the apparent cavities are often filled with solid bodies of silica. 
Instances are shown in Pl. 12. fig. 15, Dendrocalamus Hamiltonii, where the apparent 
cavities and the chlorophyll-cells with folded walls are filled, and in Pl. 11, fig. 11, 
Pseudostachyum polymorphum, where some of the silica-filled chlorophyll-cells shine 
through the epidermis. 
I have already stated that leaves, while rolled up in bud, contain no silica in the 
bulliform cells. Nor have I found any of the other epidermis-cells filled with silica. 
So far as I have been able to ascertain, the deposit of silica does not take place until 
the leaves have attained their full size and are spread out flat. But, apart from this, 
the amount of solid silica deposited in the cells of the epidermis varies within wide 
limits in leaves of the same species. It must be distinctly understood that the 
deposits of silica in the bulliform or other epidermis cells of Bamboos cannot be used as 
a character to distinguish species, although, as will presently be shown, certain species 
seem to have a stronger tendency than others to accumulate silica in these cells, The 
following is a list of the Indian species of Arwndinaria of which I have been able to 
examine specimens. The procedure, as stated above, is exceedingly simple. Pieces of 
the leaves are boiled in water for some time and then placed in a concentrated solution 
of phenol (100 parts crystallized phenol and 10 parts seen ) for two or — ndm 
The solid contents of the cells then appear with great precision. ]t wil be seen that 
of the 24 species enumerated, 4 have a strong tendency in the direction here indicated 
