BOOK V. 
GEOGRAPHY. 
491 
tribution of temperature with respect to vegetation. It will 
have been seen that little is known of the proportion of 
humidity in the atmosphere of different climates, and that the 
amount of light in various latitudes has scarcely been noticed. 
That the effect of both these agents upon vegetation is most 
important, cannot be doubted ; especially of the latter, upon 
which the most material vital functions of vegetation mainly 
depend : but, unfortunately, there are no data from which the 
precise amount or action of light in different latitudes can be 
appreciated. 
I shall now proceed to state what is known or conjectured 
of the distribution of the different orders or divisions of vege- 
tables over the surface of the globe. In doing this, I shall 
merely translate a portion of the very valuable essay of 
Humboldt upon the subject, as published in the Dictionnaire 
des Sciences Naturelles, vol. xviii. p. 422., in which is compre- 
hended the sum of all that is known of the laws that are 
observed in the distribution of the various forms of vege- 
tation. — " The numerical relations of the forms of vegetation 
are capable of being investigated in two very different modes. 
Supposing that the natural families of plants are studied 
without reference to their geographical distribution, the ques- 
tion will arise as to which type of organisation it is after 
which the greatest number of species have been created. Are 
there most Glumaceae (Cyperaceie, Graminese, and Junceae, 
are so called by Humboldt), or Compositae in the world? 
Do these two tribes together constitute a fourth part of phae- 
nogamous vegetation ? What proportion is borne by Monoco- 
tyledones to Dicotyledones ? Questions of this kind refer 
rather to the science of vegetable organisation and of mutual 
affinities. But if, instead of studying natural groups of spe- 
cies in this abstract manner, we view them with reference to 
the relations they bear to climate or to the distribution over 
the surface of the globe, other questions of a much more 
varied nature will arise. Which families, for instance, are 
more predominant in the torrid zone than in the polar circle ? 
Are Compositae more numerous in the same parallel of latitude 
or in the same isothermal line in the old world or the new ? 
Do those forms which are found to diminish in retreating from 
