.INDIAN BAMBUSEÆ; GAMBLE. 53 
but occurs not uncommonly in Guzerat. It is very common in both its small and 
large varieties in Orissa, the Cirears and Carnatic. It is common in the Concan and 
on the Western Ghat Range. In the Deccan it occurs in valleys in the hills as it 
does throughout South India, ascending in the hill ranges, as in the Nilgiris, to 3,000 
ft. and higher occasionally. In Ceylon it occurs in the warmer parts of the island on the 
margins of rivers and streams (Thwaites). In Lower Assam it is found, but infrequently, 
in Gauhati and Nowgong, also in Sylhet. It is rather scarce in Eastern Bengal and 
Chittagong, but becomes more common in Burma, all over Pegu and Martaban down 
to Tenasserim, It is very largely cultivated everywhere, as in Dehra Dún and in 
places at the foot of the Punjab Himalaya. It is probably found in its largest size 
and finest condition in the hills of the Circars, especially about the Godavari, on the hill 
ranges of the eastern and southern scarps of the Mysore plateau, and in the 
Nilgirs. The finest clumps I have seen are those in the Rumpa country, north of 
the river Godavari. Those of Gumsur are also very good. It has been very often 
collected, especially at the rare seasons of its flowering. It has been introduced into 
the West Indies. 
It will be seen that I have included in this species all the three described by 
Nees and Roxburgh and acknowledged by Ruprecht, Munro and Beddome, viz, В. 
arundinacea, В. spinosa, and В. orientalis ; also the В. Arundo admitted by Ruprecht. 
_ I have examined a great series of specimens and can find по real specific difference 
between them. Munro gives the following as characters for separating the three:— 
(1) P. arundinacea (including В. Arundo).—Rachis very glabrous, shining, hard; 
spikelets, few, long, 6- to 12-flowered; rachilla hirsute, visible; leaves 
smooth; leaf-sheaths hairy. 
(2) В. spinosa.—Rachis striate, not shining, hard; spikelets many, shorter, 4- to 
6-flowered; тасы а hardly visible; leaves glabrous above, hairy beneath; 
leaf-sheaths hairy, then sub-glabrous. 
(8) В. orientalis.—Rachis glaucous green, angled, almost soft; spikelet membran- 
ous, 5- to 8-flowered ; leaf-sheaths hairy with white cili; petiole hairy. 
From these, it appears that В. arundinacea and B. spinosa differ very little indeed 
except in the rachis of the panicle and the number of flowers, but the former character 
is not, I think, constant, and the latter is probably the result of differences in climate 
aud soil; while В. orientalis seems more nearly a separate species, and indeed the rachis 
is remarkable in the specimens | have seen, and justifies its admission as a variety. 
. Var. orientalis—rachis of the panicle green, angled, almost soft; spikelets mem- 
branous; leaf sheaths hairy, white ciliate; leaf petiole hairy. 
But except these, the characters are by no means constant, and I feel that without 
better information I am right in thinking that we have in India proper only one thorny 
Bambusa, and that that widely-spread species merely shows, as does the equally universal 
Dendrocalamus strictus, an amount of variation such as is fully accounted for by the 
variations of climate and soil. Both Brandis and Kurz considered that there was only 
one species, and I fully agree. Were I to attempt to separate it into varieties, I 
should make a different division te that adopted by Munro. All the three have practi- 
cally the same culm-sheath—a character which I believe Kurz, whose knowledge of, 
and iuterest in, bamboos was so great, considerel to settle the matter. 
