182 THE BAMBOO GARDEN CHAP. VI 
NITIDA are of a purple-black colour, whereas those of 
A. ANCEPS, purple at first, ripen to a greenish yellow or brown ; 
2nd, The leaf sheath in A. ANCEPS has a circular fringe of 
short bristly white hairs at the insertion of the leaf: this 
is a feature not found in A. NITIDA, whose leaf sheaths are 
hairless ; 3rd, The petiole and edges of the leaves lack the 
purple tint which is conspicuous in those of A. NITIDA. 
The culms of ARUNDINARIA ANCEPS are tall, round, smooth, 
very slender and rush-like; I have seen them as much as 7 
feet high, and shall be surprised if they do not attain a far 
ereater development. The cavity in a 4-foot stem is very 
small indeed, not more than half the diameter of the wall. 
The nodes are fairly salient, the lower rim sharply defined, 
the upper rim more rounded and protuberant ; the colour of 
the node is purple. The branches are purple in colour, very 
slender, with comparatively long internodes. In the leaf 
sheaths the ligule is hardly to be detected, but the little halo 
of hairs at the insertion of the leaves is very persistent and 
distinct, and markedly different from all the Arundinarias of 
anything like kindred habit, such as FALCATA, SPATHIFLORA, 
NItTIpA, NOBILIS, and FALCONERI. 
The tender leaves—about 2 inches long by about a quarter 
of an inch wide, more or less—are of a very brilliant green, 
rather more glaucous on the lower face, linear-lanceolate, 
ending in a sharp point, and attenuated to a rather short 
petiole. The tessellation is very minute and perfect. There 
are from two to three secondary nerves on either side of the 
midrib. Such serration as exists, and it requires a lens to see 
it, is more on one side than on the other. 
