4 6 
C. E. Moss. 
regard to the group of mesophytes in general. This appears a some¬ 
what formidable indictment; but the fact remains that, at the time, 
Warming’s grouping was the best constructive effort which had been 
put forward in the direction of a natural and scientific classification 
of plant communities. 
Schimper’s (1898) division into hygrophytes, xerophytes, and 
tropophytes was in several respects superior to the earlier one put 
forward by Warming. 
Schimper’s “ climatic formations ” are doubtless, in general, 
groups of formations; whilst others may really be formations. As 
Clements (1904: 27), Smith (1904: 620), and Warming (1909: 132) 
have maintained, it is undesirable to separate those habitat factors 
depending on climatic conditions from those depending on edaphic 
conditions. Still, with regard to certain districts, such as the rainier 
portions of several tropical countries, whose vegetation and flora, 
partly from lack of exploration and partly owing to certain inherent 
difficulties, are little known, such “climatic formations” as tropical 
rain forest are at present the only practicable vegetation units. 
Kurz (op. cit.) has, however, indicated that the classification of even 
tropical forests will yield ultimately to more intensive treatment; 
and it seems clear that many “ climatic formations” must ultimately 
be analysed into formations. With regard to well-worked countries, 
there is, from the standpoint of an ultimate and natural grouping 
of plant communities, nothing whatever to be said in favour of 
regarding as “ formations ” such complex groups of associations as 
deciduous forests, coniferous forests, grassland, and so on. For 
example, it can scarcely be contended that Warming’s “ formation 
of deciduous dicotyledonous forest” (1909: 144 and 329) is of the 
same rank as his “ low moor formation ” (p. 196) or his “ high moor 
formation ” (p. 200); and the real difficulty is only temporarily 
shelved by subdividing highly heterogeneous “formations” into 
“subformations” of little less heterogeneity. 
The new classification of formations into higher groups by 
Warming and Vahl (1909) is much more involved, and much less 
simple and clear, than Warming’s earlier one; and it is very 
doubtful if the changes have resulted in any gain in accuracy. 
Already it has met with some adverse criticism by Cowles 
(1909: 151), Tansley (1909: 221), and Moss (1909: 351) in their 
respective reviews of Warming’s new book (1909). In particular, 
the “ ecological class ” of coniferous formations has been singled 
out for criticism. “ Whilst this group,” says Cowles (loc. cit.), “ is 
