70 SIR DIETRICH BRANDIS ON THE 
longitudinal nerves and the lower and upper epidermis is, as a rule, filled up by thin 
walled chlorophyll-cells, the walls of these cells being folded in a peculiar manner, 
Strands of sclerenchymatous fibres support the vascular bundles, and with them con- 
stitute the longitudinal nerves. There is always an upper and a lower strand between 
the vascular bundle and epidermis. In Bamboos I have never found continuous bands of 
stereome under the epidermis, and in one instance only have I found strands of scleren- 
chymatous fibres below the bulliform cells under the lower epidermis. (The instance here 
referred to is a Bamboo cultivated at Kew as Ochlandra Rheedii. Leaves of O. Rheedii, 
Benth., collected in Cochin and Travancore, have no stereome below the bulliform cells.) 
In other grasses they are not uncommon: Aira cespitosa, Lewton-Brain, in Trans. Linn, ` 
Soc. ser. 2, vol. vi. (1903) t. 40. fig. 67; Agrostis canina, ibid. t. 37. fig. 19; Alopecurus 
geniculatus, A. alpinus, ibid. t. 37. figs. 21, 25. 
I may here mention that the author of each species is, as a rule, given only once, when 
that species is mentioned for the first time. Writers on anatomical subjects often omit 
the authority for a species, so that it is sometimes uncertain which species is intended. 
In such cases I have not been able to supply the deficiency. 
These studies I commenced at, Bonn in 1887, and have had the great privilege ot 
completing them at the Jodrell Laboratory of the Royal Botanic Gardens, Kew, with 
the aid of the excellent library of the Herbarium. I tender my thanks to Mr. L. A. 
Boodle, F.L.S., for the valuable help and advice he has most kindly given me, and to 
Professor Strasburger and Dr. Scott for many important suggestions; to Mr. J. N. Fitch 
for the excellent drawings he has made from my preparations, and to Mr. Martin Cole 
for the admirable sections which he prepared. 
APPARENT CAVITIES. 
As explained by Karelstschicoff, (3) t. 3. fig. 2 (Dendrocalamus strictus, Nees), the 
apparent cavities mentioned above are entirely occupied by a large number of flat thin- 
walled cells, lying upon each other like the leaves of a book. They are, as a rule, filled 
with a colourless sap and the walls show with chlor-zinc-iodine the reaction of cellulose. 
Only in Dinochloa Macclellandii, Kurz, among Bamboos have I found solid contents of 
organie matter, but the study of the dried specimens at my disposal has not enabled me to 
say anything further. Diandrolyra bicolor, Stapf, among Pharee has chlorophyll in the 
thin-walled cells which fill the apparent cavities. "The structure of these flat thin-walled 
cells is best studied in thin longitudinal sections. Plate 12. fig. 17 represents a small i 
piece of a very young leaf, rolled up in bud, of Dendrocalamus giganteus, Munro, grown — : 
in the Palm-house at Kew; the cells are turgescent. In an older state, often before the 
leaves are fully flattened out, the walls collapse and hang between the upper and lower 
strata of chlorophyll parenchyma, and may be compared to empty bags of the finest — 
muslin. This is the stage represented by Karelstschicoff, and that which is see? — 
whenever longitudinal sections of full-grown Bamboo leaves are made. ‘This stage is 
here represented in Arundinaria Japonica (Pl. 11. fig. 3), Phyllostachys bambusoides, ^ 
Sieb. et Zucc. (Pl. 11. fig. 8), and Dinochloa andamanica, Kurz (PL 13. fig. 24). These E 
