the Vegetation of a Country, 441 



which may be calculated from Floras, but which is not evident in 

 nature. I defy the most practised botanist to determine at the 

 first glance what is the proportion of the two classes, even in a 

 limited district. On the contrary, it is easy, at first sight, 

 to say whether the Compositse, the Leguminosse, or evergreen 

 plants predominate in a region, because these groups are more 

 homogeneous, more easily seized in their totality and compared 

 to one another. It would at least be necessary, to give any 

 importance to the proportion of the two great classes, that 

 the composition of each should be added, particularly in the 

 case of the Monocotyledones, the forms of which are so very 

 difi'erent. 



Proportions of the Natural Groups superior to the Families, 

 but inferior to the Classes. — Botanists have endeavoured to asso- 

 ciate the families in groups inferior to the great divisions of the 

 vegetable kingdom, but still founded upon positive characters ; 

 but these attempts are as yet too recent and too imperfect to be 

 capable of employment in botanical geography. It would be 

 premature to calculate the proportions of the species in these 

 groups, which are only provisional, or at all events ill-defined. 

 Other associations, of rather small botanical value, but which 

 still repose upon very apparent characters, merit more of the 

 attention of the geographical botanist. 



The proportions of the woody and herbaceous species, or of 

 annual, biennial, perennial and woody species, whether mono- 

 carpous or polycarpous ; the proportion of species with fleshy 

 leaves or stalks, or succulent plants ; that of the species with 

 compound, or with persistent and deciduous leaves, — these are 

 elements that should be ascertained in every assemblage of 

 plants. Each of these groups includes plants of various families 

 or classes; but their importance in nature is evident. The 

 number of woody species, trees especially, has an actual value, 

 by reason of the aspect of forests and their positive action upon 

 herbaceous plants. In this point of view, statistical observa- 

 tions upon the extent of the forests in a country are by no means 

 without value. I may even say, that a statistical table, showing 

 the proportion of forests, cultivated land, meadows, marshes, &c., 

 will give more information regarding the general vegetation of a 

 country, than certain Floras which exhibit a great deal of learn- 

 ing, and which are in high estimation amongst botanists. 



Some attempts have been made to arrange the forms of plants 

 in certain categories, answering to the aspects they hear in 

 nature. Von Humboldt*, and after him Meyenf, have distin- 



* Essai sur la Geograpliie des Plantes, 4to, p. 31, and Tableaux de la 

 Nature, 1851, ii. p. 22. 



t Grundriss der Pflanzen^eogr. iii. p. 117. 

 Ann. ^ Mag. N. Hist. Ser. 2. Vol.XY. 29 



