298 
The botany of the oil-grasses is only slightly touched upon in the 
work. ‘The definitions and the nomenclature of the species are, 
on the whole, those of Hackel’s monograph of Andropogoneae and 
of Hooker’s elaboration of the grasses of India. The origin of the 
several oils from the species mentioned is stated somewhat 
apodictically, and there is nothing to suggest the difficulties which 
from time to time have arisen owing to the unsatisfactory state of 
our knowledge of the “botany”’ of those grass-oils. This condition 
has become more accentuated with the increased interest in the 
grass-oil industry during the last few years and with the attempts 
at reorganising and extending it in its old homes and at intro- 
ducing it into other tropical countries. The ‘Semi-Annual 
Reports’ published by Schimmel & Co. contain not a few 
direct and indirect references to the uncertainty of the taxonomy 
of the grasses involved : I quote only two passages. 
In the April-May Report for 1903, p. 23, the authors say :—“ It 
has repeatedly attracted our attention, that when it is a question 
of their origin, the Andropogon grasses are frequently confounded 
with each’other . . . there are some exactly defined species 
indicated as the mother plants of oils which, according to our 
information, could not possibly be produced from them”; and 
again in the October--November Report for 1905, p. 52: “On a 
previous occasion we have already pointed out that the Andro- 
pogon grasses, where their origin is mentioned, are frequently 
mistaken one for the other. This inconvenience is ail the more 
felt as the uncertainty of the botanical nomenclature also exists 
in scientific work.” 
At Kew the experience has been the same. The incongruous 
application of the names ‘Andropogon MSchoenanthus’ and 
‘Lemon-grass’ and the obscurity of De Candolle’s Andropogon 
citratus have been among the principal sources of trouble. To 
them has had to be added, more recently, uncertainty as to the 
origin of the Ginger-grass oil. Even the comparatively well- 
known Citronella grass has been suspected to be a “composite ” 
species, or at any rate to include two varieties, the distinctive — 
morphological characters of which were still to seek. 
Under the circumstances a thorough overhauling, from the 
taxonomic standpoint, of the grasses involved was essential in the 
interests of the grass-oil industry. Having been entrusted with 
this task, I revised in the first place the material in the Kew 
Herbarium. Rich as it was in some respects, it was sadly lacking 
in others. <A fine collection of oil-grasses made in Southern 
India at the instigation of Mr. C. A. Barber, Government Botanist 
at Madras, went a long way to fill the gaps, so far as the Madras 
Presidency and Travancore were concerned, whilst Dr. Lotsy, of 
the Rijk’s Herbarium at Leiden, and Dr. Treub, of Buitenzorg, 
supplied useful material from Java. Convinced that no scientific 
problem should be approached without due consideration of its 
historical development, I have endeavoured to get a good grasp | 
of the history of the subject. This has entailed a great deal of 
library work and search for original specimens, as documentary 
evidence, in the older herbaria. Professor Urban of Berlin, — 
Professor Mattirolo of Turin, and Dr. Briquet of Geneva, have 
assisted in this direction by the loan of specimens; but the 
