224 
Review: 
would be regarded as phases in the development of a formation, 
whose final stage would be represented by the highest type of 
vegetation possible for the given soil. Only if the substratum is 
sufficiently changed, after the development upon it of a closed 
community, by the activity of that community, to initiate a new 
developmental series, which destroys the community in occupation 
and finally gives rise to another, should we speak of the develop¬ 
ment of a new formation. This phenomenon is particularly seen in 
development of new land through the stage of marsh. The term 
“ association ” would find its place in designating each well-defined 
community, whether transitional or final. 
It would be out of place in a review adequately to develop and 
illustrate these ideas, but it seems probable that the concept of the 
formation, resting upon that of the habitat, should be extended in 
in this direction, rather than in that indicated by Professor 
Warming. The conceptions in question owe a great deal to 
Clements’ insistence on the importance of succession (though his 
terminology is different) and are in general harmony with the ideas of 
Cowles and Nilsson quoted by Professor Warming (p. 147) and of 
Hult (p. 364). It is said (p. 148) that “ it does not seem possible 
to use development as the fundamental basis of classification of 
plant-communities : for developmental changes are too dependent 
upon local conditions; a formation does not develop merely in a 
single definite direction, but will modify in one direction at one 
place and in another at another place, according to the prevailing 
conditions.” This is all perfectly true, but it does not in the least 
invalidate the conception of development or change within the 
formation. Rather do these facts lend support to the notion that 
the true fundamental unit is the sum-total of the phases of such 
development and change, under certain defined general conditions of 
soil and climate. 
There will of course be differences of opinion as to the limits 
of certain formations, just exactly as there are as to the limits of 
certain taxonomic species, but the principle described seems to 
offer the surest because the most natural basis for the concept. 
Our plea is for the concentration of attention on the recognition, 
description and study of the natural units of vegetation (formations 
in the sense indicated) in different parts of the world, without at 
present troubling ourselves overmuch about world-wide synthesis. 
The time is not nearly ripe for that, unless we content ourselves 
with the more superficial comparisons. It is possible and useful, 
of course, to recognise the identity of edaphic conditions deter¬ 
mining the occurrence, for instance, of reed-swamp all over the 
world, or the climatic similarities giving rise to the occurrence 
of the sclerophyll type in the Mediterranean region, Southern 
California, Southern Chili, the Cape and parts of Australia. But 
the effective development of ecology during the next twenty years 
will be on the lines of the recognition and the intensive study of 
localised natural units of vegetation. 
It would be impossible to review in detail Professor Warming’s 
and Dr. Vahl’s treatment of the “ systematic ” portion of their 
subject. It must suffice to say that an enormous mass of material 
is dealt with, drawn from all available sources and arranged under 
the heads already given. As a work of reference this will 
