496 
GEOGRAPHY. 
BOOK V. 
lines drawn through those points where the fractions are 
reduced to the smallest denominator) will be isothermal 
lines. If we divide the globe into lines of longitude, and 
compare the numerical proportions of those lines under 
similar isothermal latitudes, the existence of different sysf^s 
of grouping will at once be evident. From such systems can 
be distinguished, even in the present imperfect state of our 
knowledge, those of the new world, of western Africa, of 
India, and of New Holland. As we find that, notwithstanding 
the regular increase of heat from the equator to the poles, the 
maximum of heat is not always identical in different countries 
in different degrees of longitude ; so there exist places where 
certain families attain a greater degree of developement than 
elsewhere ; as is the case with Compositae in the temperate 
region of North America, and especially at the southern ex- 
tremity of Africa." 
Now follow tables of the different numerical proportions of 
certain extensive families and divisions of plants, as far as 
they have been ascertained. I give them in Humboldt's 
words, with a few interpolations, which are distinguished by 
being included within crotchets [ ]. 
" ACOTYLEDONES. 
" Cryptogamic plants (fungi, lichens, mosses, and ferns) ; 
cellular and vascular Agamae of De Candolle. Taking the 
plants of the plains along with those of the mountains, we 
have found, under the^ tropics, ^ ; but their number ought 
to be much greater. Brown has shown that it is probable 
that, in the torrid zone, the proportion is for the plains, 
and ^ for the mountains. In the temperate zone crj^toga- 
mous plants are generally to phaenogamous as 1 to 2 ; in the 
frozen zone they maintain as large a proportion, and often 
much surpass it. [In Melville Island the numbers are 
58 crypt, to 67 phaenog., or nearly equal : in Sweden, ac- 
cording to the computation of Wahlenberg, they are some- 
thing less than 4 to 1 ; and it is probable that this is a near 
approximation to the true proportions of Sweden, the crypto- 
gamic flora of that country having been more accurately 
investigated than that of any other part of the world.] 
