337 
synonymy was further complicated by G. C. Nees, who named 
some specimens in Wight's herbarium (No. 1702) Andr<>p„g<. n 
nir, ! id's, and in 1841 published a description of Andropogon wir- 
doides* at the same time reducing Trinius's Andropogon pt/r/modrs 
to it as a synonym. There were thus four names in the field, 
more or less definitely connected with the t'raaiant Ximar -ra>s : 
Andropogon Martini (1820), A. pachnode* (1833), .1. Cubanns 
ammrttinis i.ISIO), and A. nardoides (1841). They were all set 
aside, when in 1862 Munrof pointed out that the ' type ' of Andro- 
pogon Schoenanthus in Linnaeus's herbarium was Roxburgh's 
Andropogon Martini. I have in another place shown the value 
of that * type ' and explained how little it has to do with the Rusa- 
grass. However, the fact was accepted as implying that Linnaeus 
had this grass in view when establishing his Andropogon 
Srho, nan thus, and consequently Fliickiger and HanburyJ (in 1S74) 
put '■Andropogon Schoenanthus, L.,' down as the source of the 
Rusa-oil, an assumption which has since then remained unchal- 
lenged. From what I have said, it is, however, perfectly clear 
that the Rusa-grass is actually identical with Andropogon Martini 
of Roxburgh, and has to stand as such, or, if transferred, to 
Cymbopogon, as C. Martini. 
Area and Variation.— The area of C. Martini extends in 
India from the Rajmahal Hills on the bend of the Ganges to the 
Afghan frontier, and from the subtropical zone of the Himalaya to 
about 12° N., leaving out the desert and steppe region of the Panjab, 
the outer slopes of the Western Ghats, and, as would appear, a 
great portion of the Northern Carnatic. From the collectors' and 
writers' notes it appears to be locally very common, and a con- 
spicuous feature, particularly in the late autumn when the panicl. s 
change colour and impart their rnh brown-red tints to the hill sides. 
Bo Striking is this colour effect that one is tempted to suggest that 
the two commonest vernacular names for the grass 'rusa' with its 
numerous variations, and ' mirchia gand,' take their origin from It : 
'rusa' being possibly derived from 'run,' Sanscrit for "to be red," 
and ' mirchia gand,' having reference at the same time to the colour, 
the red of i, tnd to the scent (gandha). Within 
the greater | - h very uniform in the 
structure of ti eculiarly soft and delicate t< 
of the lea -hie in stature and in the dime 
of the leaves. From less than 1 m. it grows to a size which is 
described as ' gigantic,' whilst the leaf -blades range from -25 to 
probably quite 1 m. in length and from 8 (in extreme cases 5) mm. 
t'> 30 niiii. in width. One of the most characteristic features of the 
leaf of G. Martini is that the greatest width is generally near the 
base of the blade, which is rounded off and suddenly constricted 
and frequently clasps the culm. This form is for example 
illustrated in Trinius's and Royle's figures and might be called the 
' pachnodes ' type. In Roxburgh's type specimen the shape ot the 
blade is somewhat different in so far as the width is almost the 
same for a verv considerable distance from the base upwards, whilst 
the base itself is less constricted and not stem-claspi ng, l he same 
# Nees Fl Afr Austr. (1841), p. 116. 
t Munro in Journ. Linn. Soc. vol. vi. (1862) P- 52. 
X Fliickiger and Hanbury, Pharmacograpbia (1874;, p. w. 
