300 
generally. Having examined the leaf-structure of more than one 
fourth of all the Andropogoneae described in his monograph to 
see how far the anatomical characters coincide with the morpho- 
logical and can be used for the definition of the natural groups, 
he says : “ The result 1s absolutely negative. Neither is it possible 
to distinguish anatomically the Andropogoneae from the allied 
tribes or even the remote tribe of Chlorideae, nor is there any one 
character or combination of characters which is confined to one 
genus. Hven the sub-genera possess a uniform leaf-structure only 
in some cases when they comprise less than ten species. The 
species, however, are, with the exception of such as are very poly- 
morphous, mostly well characterised by their anatomy.” It must, 
however, be remarked that the leaf is probably more plastic than 
any other organ in grasses and might therefore a@ priori be 
expected to exhibit the phenomena of epharmosis in a prominent 
degree. The study of the anatomy of the glumes and fruits would 
possibly yield a different result. However this may be, an in- 
vestigation into the anatomy of the oil-grasses, and particularly 
into the development,and distribution of the oil cells, is highly 
desirable. With the exception of a very valuable description of 
the oil cells of “ Andropogon Schoenanthus”’ by Professor F. von 
Hohnel,* nothing is known in this direction. Yet it is quite 
obvious that to know the seat of the: oil-yielding tissues, their 
properties, the time and conditions of their formation and the 
changes they subsequently undergo, must be of considerable 
importance for the rational development of the grass-oil industry, 
just as it is, from the standpoint of pure science, necessary for the 
complete understanding of the organisation of those grasses. 
In so faras organisation means correlation of structure and 
function, new problems await us on that ground, but they are 
problems for the physiologist. Some are of ageneral nature, as the 
question of the genesis of the grass-oils and the place of these in the 
economy of the plants which produce them; others are more directly 
connected with the practical side of the subject, such as the problems 
of the changes in the yield of oil according to the season, its reduc- 
tion in old plants, the variation in the chemical constitution of the 
oils in morphologically indistinguishable races, and the apparently 
capricious limitation of some forms, particularly suitable for in- 
dustrial exploitation, to certain geographical areas. Remote as the 
relations of physiology to the taxonomy of the oil-grasses may 
appear to be, there is one problem which touches the latter directly. 
This is the question of purely physiological races: how far they 
actually exist, what they are, and what place they ought to be 
given in the ‘system.’ Other physiological problems are inti- 
mately connected with ‘variability, and so have a distinct 
bearing on taxonomy. Beyond this it is at present probably 
impossible to indicate in detail the help which in this, as in similar 
* F. von Hohnel, in Sitz. Ber. Akad. Wissensch. Wien, Mathem-Naturhist. CL, 
vol. Ixxxix., part. 1. (1834), pp. 14, 15.—I have put the name Andropogon 
Schoenanthus between inverted commas because the author obviously intended 
to deal with the anatomy of the grass yielding the Palmarosa oil, viz., the 
Andropogon Schoenanthus of most Indian botanists, whilst he actually described 
she structure of the original Linnaean Andropogon Schoenanthus or the A. laniger 
of Desfontaines. Such are the pitfalls of a confused nomenclature. 
