7^ ON THE GEOGRAPHY OF ANIMALS. 



different features in different localities, and which the 

 Brazilians distinguish by appropriate names. The 

 Campos are vast plains similar to those on the banks of 

 the great Rio St. Francisco, covered with coarse grass, 

 and destitute of trees. They are scorched during sum- 

 mer, and present little other vegetation during the rainy 

 season. The Campos appear, in fact, to be a continu- 

 ation of the Pampas of Paraguay and the Rio de la 

 Plata, and are analogous to the interior deserts of 

 Africa : water, excepting in the great rivers, is equally 

 scarce ; and in dry seasons, hundreds of cattle pe- 

 rish, and whole villages migrate. These dreary plains 

 are frequently elevated ; but in such situations, the 

 coarse and scanty herbage is generally intermixed with 

 stunted trees, growing at short intervals, as in a park : 

 clear of underwood, and open to the route of the tra- 

 veller in every direction, such tracts are termed Tabu- 

 laras, or table-lands, since they are almost always 

 raised a few hundred feet above the level of the sea. 

 Lands of this description are frequently broken by 

 narrow valleys, or gentle hollows, wherein the trees 

 become higher, and acquire a more flourishing growth, 

 thus forming woods ; yet they are so matted together 

 by a thick underwood of Cacti, Bromelice, and other 

 spinous plants, intermixed with thickets of coarse-leaved 

 flowering shrubs, as to be almost impassable to any but 

 the hunter : these are the Catinga woods of the 

 Brazilians; and it is here that the numerous and 

 splendid family of Epidendrum, and other parasitic 

 plants, few of which are yet known to botanists, 

 root round the bark, or spring from the stems, of 

 the larger trees. The general character of the soil, in 

 all the localities here described, is more or less sandy ; 

 and although never destitute of vegetation, the plants 

 have almost always a parched, stunted, and withered ap- 

 pearance, except, as before observed, during the rainy 

 season. These observations, apparently foreign to our 

 present subject, are nevertheless so closely connected 

 with it, that, without them, it would be impossible to 



