84 Professor Fournet’s Researches ov the 
of moving sand mixed with cultivation, and this whole series 
indicates not a state of absolute aridity, but simply the facility 
with which the savannahs may assume the physiognomy of 
deserts. 
To recapitulate ; the New World may be regarded as divided 
by the Andes into two systems, characterised by their con- 
figuration, as well as by their meteorology. The western 
portion is very narrow, but of a simple structure; whereas 
the other is broad, deeply articulated, and irregularly ele- 
vated ; the first presents all the great phenomena which may 
be considered as the immediate results of the solar influence 
on the second: they tend to become effaced in the series of 
partial causes, and especially in the effects of the superposi- 
tion of periodical rains. The most important consequence 
of this irregularity is the annihilation of absolute deserts, 
and it is thus that an extreme uniformity of surface, which, 
at the first glance, would seem to be an element of prosperity, 
from the facility it affords for communication, becomes, on 
the contrary, one of the most formidable obstacles which na- 
ture interposes to civilization. 
By shewing that an absolute want of rain is necessary to 
constitute the absolute aridity of a desert,—that the latter is 
nothing else but the reflection of a dry atmosphere,—we have 
only answered one part of the question. We see, indeed, 
that the cause of which we are in search is essentially to be 
sought for in the atmosphere ; but we have not explained 
why it does not rain between the two intertropical and sub- 
tropical zones ; we must, therefore, enter into some details - 
on this subject. 
The intertropical rains commence at each place at the time 
the sun reaches its greatest altitude, because then, under the 
influence of ascending columns of air, the breezes of the 
trade-winds become uncertain; and for them there are sub- 
stituted calms interrupted by the winds which blow from the 
heteronymous pole. There is thus produced, at that time, 
an unequal distribution of heat, the result of which is the 
condensation of the aqueous vapour dissolved in the air. 
The hiemal rains take place, on the other hand, in the 
corresponding zone whenever, in consequence of the increased 
