STRUCTURE OF BAMBOO LEAVES. Ji 
flat thin-walled cells can readily be isolated by boiling in diluted nitric acid with 
potassium chlorate. Karelstschicoff, (3) t. 3. fig. 3 (a), shows them in Dendrocalamus 
strictus, and on Pl. 14. figs. 40-43 of the present paper they are represented in Sasa 
paniculata, Makino & Shibata, in ‘ Tokyo Botanical Magazine,’ xy. (1901) p. 25 
(=Arundinaria Kumasasa, Hort. Kew); Cephalostachyum capitatum, Munro; Melocanna 
bambusoides, Trinius ; and Ochlandra Beddomei, Gamble. They are exceedingly uniform 
in appearance, the sharp points on the outside of the walls indicate the points inserted 
between the cells of the mesophyll and between the indentations of the cells constituting 
the parenchyma-sheath of the vascular bundle. 
These flat thin-walled cells may also be studied on flat sections made parallel with the 
surface of the leaf, as represented in Pl. 14. fig. 35 for Melocanna bambusoides, the 
section passing through the longitudinal nerves and the chlorophyll-tissue, which 
separates two apparent cavities and which is situated below the bulliform cells. In 
pieces of Bamboo leaves boiled in water and afterwards placed in glycerine, the flat thin- 
walled cells appear under the microscope as distinct transverse lines (JMelocanna bam- 
busoides, Pl. 18. fig. 26). They are conspicuous when the leaf has been kept in a concen- 
trated solution of phenol, Pl. 11. fig. 1 (Arundinaria falcata, Nees) and PI. 13. fig. 27 
(Melocanna bambusoides). They have been omitted in the drawing of Cephalostachyum 
pallidum, Munro, Pl. 13. fig. 21, so as not to confuse the picture. 
The transverse section of a bud of Arundinaria japonica, Sieb. et Zuce. (Pl. 11. fig. 2), 
shows that at an early stage, when the epidermis and vascular bundles have been formed, 
but long before the bulliform cells of the upper epidermis have developed, the tissue of 
the mesophyll is gradually differentiated, with the result that the apparent cavities 
appear. ‘The earliest development of the flat thin-walled cells filling up these cavities 
will doubtless form the subject of an interesting histological study, which I must leave 
to younger heads. 
It is a remarkable fact that in a transverse section the apparent cavities here described 
have the appearance of being empty, although, even in the thinnest section, there must 
be several layers of these thin-walled cells, either (in the case of a very young leaf) 
turgescent or (in the case of older leaves, when the walls have collapsed) separated by 
wide intercellular spaces, probably filled with air, which, however, is readily absorbed 
by the fluid in which the section is mounted under the microscope. The explanation of 
this remarkable appearance I must leave to others. It may be useful to mention that 
chlor-zinc-iodine gives these apparent cavities in transverse sections a very faint bluish 
tinge, and that in longitudinal sections the walls of the flat thin-walled cells are 
exceedingly bright under the polariscope. 
From what has been here stated it will be understood that the tissue of flat thin-walled 
cells here described, which fills the apparent cavities in Bamboo leaves, has a structure 
entirely different from the tissue of short-armed cells which in young leaves of Glyceria 
aquatica and other aquatic grasses fills the air-spaces between the longitudinal nerves 
(Duval-Jouve, * Agropyrum,’ t. 16. fig. 13). 
The apparent cavities here described are fairly uniform in size in a transverse section 
N2 
