Four years later Hanbury received the same plant from 
E. G. Waring, with this note: "Andropogon (?) which yields 
the Lemon-grass of Travancore— abundant on the plains— is not 
cultivated." The statement in the Pharmacographia Indica, vol. iii. 
(1893), p. 565, that the oil is distilled in Travancore from 
Anjengo northwards, and that the grass is burnt at the end of ! 
the dry weather, no doubt also refers to C. flexuosus, and not to 
C. ctoratiis, as the authors of that work believe. It is probably 
due to this confusion that Gildemeister and Hoffmann say, quoting 
Dymock, Warden, and Hooper ae it ''the grass 
/ on a large scale only on the Malabar coast in Travan- 
core, on the western slope of the mountains, north of Anjengo." 
Mr. T. F. Bourdillon writes quite recently from Quilon that only 
within the last year or two extensive areas have been planted up 
with the Malabar grass. As the Travancore grass oil is not, in 
y distinguished from the oil of C. citrdtus, 
both being sold as ' lemon-grass oil,' it would be interesting 
to know how far the analyses of 'lemon-grass oil' refer to the 
one or the other. Certain discrepancies in the results obtained 
by chemists may have their origin in the indiscriminate use of 
the term. 
Assuming that the whole of the 'lemon-grass oil' exported 
from the Malabar Coast is referable to C. flexuosus, the figures 
for the export of that oil were, for 1896-97, 270,000 kilos, or 
595,080 lbs. 
6. Cymbopogon coloratus, Stapf. 
(Andropogon coloratus, Nees, ms.) 
Under the name of 'Andropogon coloratus, N.E.,' Wight 
distributed a grass (numbered 1703) which although similar to 
C. flexuosus differs from it distinctly in its much smaller stature, 
narrow blades and leaf-sheaths, dense and erect panicles, more 
conspicuously bearded rhachis-joints and pedicels— the white 
hairs contrasting vividly with the brownish spathes— and much 
swollen pedicels at the base of the racemes. No description of it 
as a species was published ; but in 1896 Hooker* distinguished it 
as a variety of Andropogon Nardus. A similar form, but taller 
with longer, stiff and dense panicles and paler spathes was issued 
by Wight at the same time an •• Andr'^,,,,,,,, ['',/,„/,.. r,i^>»< 
yelatior,cuhn,, ,/•,,/„ tin,,:." No. 1700c and in a diseased form 
under No. 170Gtf, the panicles of the latter being infested with an 
represented by No. 1703 was also 
eoUected by Klein on the 9th of July, 1808, but where, is not 
stated ; the taller form both in its normal and diseased states has 
been repeatedly gathered throughout the Carnatic from the 
extreme south as far as the Cuddapa District, and from the 
Tinnivelli hills to the Anamallais. This is also almost certainly 
the plant which, as I shall have to point out in another place, 
Roxburgh had livrinvd (No. K ,< - , "> in his duplicate collection of 
drawings, now at Kew), and erroneously i lent! tied with his own 
Andropogon Martini. It is very likely that the original drawing 
was made at Samulcotta which might suggest an extension of the 
a _ rea of C. co loratus towards th e Circars. 
* Hook., Fl. Brit. Ind., vol. vii. (1897), p. 206. 
