1899] ON THE STUDY OF PLANT ASSOCIATIONS 117 



moisture is on the whole the most important condition in determining 

 any particular association, and he employs this as the chief basis of his 

 classification. He groups the plant associations as follows : — 



Hydrophyte vegetation. — This requires a substratum with at least 

 80 per cent of water. It includes associations of aquatic and 

 marsh plants. 



Xerophyte vegetation. — This is the opposite extreme from the 

 previous class, requiring less than 10 per cent of water. It 

 includes associations of rock, desert, moor plants, etc. 



Halophyte vegetation. — Also an extreme vegetation, allied in many 

 of its adaptations and habits to the Xerophyte vegetation, 

 but characterised by requiring the presence of much salt, for 

 instance the strand and the dune associations. 



Mesophyte vegetation. — An intermediate class adapted to medium 

 conditions of moisture, and not requiring a special amount of 

 salt, for instance the associations of beech and oak woods. 



For further subdivisions of these four classes we refer to Warming's 

 original work. 



In employing this classification it is essential to remember that 

 variation in humidity is not the only acting condition, — the species 

 are very differently adapted for the struggle with each other, even if 

 the acting conditions are uniform ; the nature of the food supply in 

 the soil varies from place to place, and each variation may more or less 

 affect the association ; local conditions of light and shade, e.g. in the 

 woods, modify the minor grouping of the species ; and further, all the 

 species are dependent upon the wider conditions of light and of 

 temperature of the particular region of the globe. A similar physio- 

 logical method of grouping plants to this of Warming was suggested 

 by Alphonse de Candolle (1874), based upon the combined con- 

 ditions of heat and moisture. Its application to the graphic repre- 

 sentation of the vegetation of the globe will be seen in maps by Engler 

 (1879) and Drude (1887). 



The methods for the detailed mapping of the vegetation of a 

 country have recently been greatly advanced by the labours of Pro- 

 fessor Flahault of Montpellier (1897 b), who has shown the possibility 

 of a very exact botanical survey, applicable to the consideration of 

 questions of botanical geography, meteorology, geology, forestry, and 

 agriculture. Briefly stated, the method consists in recording, by 

 means of a large-scale contour map, the exact range over the country 

 of certain representative plant associations. The selected associations 

 are naturally as large and well defined as can be found ; in this case, 

 where Southern France has been studied, the dominant trees have 



