u The Oecology of Plants 
22 1 
C. Soil physically dry. 
6. Lithophytes (on rocks). 
7. Psammophytes (on sand and gravel). 
8. Chersophytes (on waste land). 
D. Climate dry. a 
9. Eremophytes (desert and steppe). 
10. Psilophytes (savannah). 
11. Sclerophyllous formations (bush and forest). 
E. Soil physically or physiologically dry. 
12. Coniferous formations (forest). 
F. Soil and climate favour the development of mesophilous 
formations. 
13. Mesophytes. 
The only one of these classes the contents of which will not, 
in a general way, be familiar to ecologists, is the eighth, the 
Chersophytes. These include communities of xerophytie perennial 
herbs on particular dry soils, often dominated by grasses, but 
including many dicotyledons, such as the vegetation of certain 
“ Meadows ” on the Alps, in Montenegro, in the “ Pontic ” region, in 
Spain and Madeira, and also certain “bushland” in North Europe, 
Chodat’s “ Garide,” and finally “ Fern-heath ” dominated by Pteris 
aquilina in the south of England and elsewhere. It is impossible 
to resist the conviction that this class is a kind of “rubbish- 
heap,” including a number of very various and insufficiently 
studied plant-communities, many of which, at least, have little in 
common, except that they do not fit into any of the other classes. 
But indeed we must regard the whole classification as quite 
provisional, though in some respects it is no doubt an advance on 
previous schemes. The class of Mesophytes is altogether too 
vague in definition. It has to be characterised in terms of itself, as 
existing where the “ soil and climate favour the development of 
mesophilous formations,” which is something like defining civilization 
as that which obtains under civilised conditions of existence. As 
an avowedly vague descriptive term referring to plants existing 
in a medium relation to water, the word “ mesophyte ” has its use, 
but we cannot think it should have a place in a formal scheme 
of classification. On the whole we still prefer Schimper’s 
terms “ hygrophyte ” and “ tropophyte ” in spite of the author’s 
criticism, and admitting that they also do not help us much towards 
constructing a satisfactory formal classification, Coniferous forest, 
i.e. forest of which the trees belong to a definite systematic group 
should surely not be co-ordinated with the other classes. Sclero¬ 
phyllous scrub, in the Mediterranean region at least, is often merely 
the remnants of coniferous forest, existing under exactly the same 
conditions of life, and into which for some reason or another the 
pines fail to get back when once they have been cleared. Then 
again “ psammophilous ” vegetation as limited in this book is surely 
not always, or even generally, independent of the “ground-water,” 
and if we admit this, why should we exclude meadow and forest 
which is also developed on sands and gravels where the water- 
level is high ? 
There is no fault to be found with Professor Warming’s general 
attitude towards his own schemes. He frankly admits that the 
arrangement into groups must be done “ in a very uncertain manner,” 
