96 Researches on the Situation of Zones without Rain, &e. 
by an absolute dryness (Sahara, Agoa, Lower California, 
and the Peruvian coast). 
5th, That a great elevation of the surface, in the form of a 
plateau, may produce the approximation of the two regions of 
estival and hiemal rains in such a manner that they mani- 
fest themselves consecutively in one and the same country 
(the northern portion of the Mexican plateau). 
6th, Lastly, that a great irregularity of the surface may 
completely disturb the normal arrangement, by causing rains 
out of the usual season, even between the tropics (the coast 
of Brazil, New Orleans, &c., &c.). 
Before concluding this memoir, it may be useful to explain 
more particularly the meaning of some expressions we have 
employed, or that are made use of by the inhabitants of coun- 
tries more or less resembling deserts. 
The words savannahs and pampas are employed, the one in 
the south of South America, the other in the south of North 
America, to designate slightly undulating and for the most 
part grassy plains. They are great prairies ; but the pampas 
correspond more exactly to dry savannahs, and for wet sa- 
vannahs there is an equivalent term, viz., Canadas. The 
steppes of the Russians, the yai/a of the Persians, in a like 
manner, designate flat plains which are dry and at the same 
time grassy ; and the Wanos of the north of South America, 
as well as the karroo of southern Africa, only differ from 
them by being liable to become more completely arid in the 
seasons of the droughts. 
The gobi or cobi of the Mongols are sandy deserts ; but 
the term is applied generally, in northern Asia, to all steppes 
devoid of water, while the name of Khangai is given to the 
portions which are watered, and are covered with vegetation. 
The cha-mo of the Chinese is, properly speaking, the sea of 
sand, a true lande; but this expression is not applied to the 
portion beyond Hami, so that the preceding distinctions are 
sufficient to shew that the words cobi and cha-mo, taken in 
a collective sense by geographers, ought not to be applied to 
all the space ordinarily designated in this manner, because 
its different portions have received different names, accord- 
ing to their characters. 
6 
a 
