32 STEPPES AND DESERTS. 



ture, described the Deserts of Northern Africa, of Yemen, of Ker- 

 i man, and Mekran (the Gedrosia of the Greeks), and even as far as 

 | Moultan, as forming a single connected sea of sand. ( 21 ) 

 P^In addition to the action of these hot winds, there is (so far as 

 we know) an absence or comparative paucity in Africa of large 

 rivers, of widely extended forests producing coolness and exhaling 

 moisture, and of lofty mountains. Of mountains covered with per- 

 petual snow, we know only the western part of the Atlas, ( 22 ) whose 

 narrow range, seen in profile from the Atlantic, appeared to the an- 

 cient navigators when sailing along the coast as a single, detached, 

 lofty, sky-supporting mount. The eastern prolongation of the chain 

 extends nearly to Dakul, where Carthage, once mistress of the seas, 

 now lies in mouldering ruins. As forming a long extended coast- 

 chain, or Gaetulian rampart, the effect of the Atlas range is to inter- 

 cept the cool north breezes, and the vapors which ascend from the 

 Mediterranean. 



The Mountains of the Moon, Djebel-al-Komr, ( ffl ) (fabulously re- 

 presented as forming part of a mountainous parallel extending from 

 the high plateaux of Ilabesh, an African Quito, to the sources of the 

 Senegal,) were supposed to rise above the limit of perpetual snow. 

 The Cordillera of Lupata, which extends along the eastern coast of 

 Mozambique and Monoinotapa, as the Andes along the western 

 coast of Peru, is believed to be covered with perpetual snow in the 

 gold districts of Machinga and Mocanga. But all these mountains, 

 with the abundant waters to which they give rise, are far remote 

 from the immense Desert which stretches from the southern decli- 

 vity of the Atlas to the Niger. 



Possibly, however, all the causes of heat and dryness which have 

 ' been enumerated may have been insufficient to transform such con- 

 siderable parts of the African plains into a dreadful desert, without 

 the concurrence of some revolution of nature, such, for instance, 

 as an irruption of the ocean, whereby these flat regions may have 

 been despoiled of their coating of vegetable soil, as well as of the 

 plants which it nourished. Profound obscurity veils the period of 

 such an event, and the force which determined the irruption. Per- 

 haps it may have been caused by the great " rotatory current" ( M ) 

 which sends the warmer water of the Mexican gulf over the banks 



