147 



Portuguese and Spanish possessions ; a boundry that would 

 accompany the Oroonoko, the Cassiquiare> the Rio Negro, 

 the banks of the Amazon, for a distance of twenty leagues, 

 the Rio de la Madeira, the Guapore, the Aguapehi, the 

 Jauru, the Paraguay, and the Parana, or Rio de la Plata, 

 and would form a line of demarcation of more than eight 

 hundred and sixty leagues. On the east of this boundary 

 the Spanish Americans possess Paraguay, and a part of 

 Spanish Guyana; and on the west, the Portuguese Ameri-^ 

 cans have occupied the country between the Javary and the 

 Rio de la Madeira, and between the Putumayo and the 

 sources of the Rio Negro. It is not from the coasts of 

 Brazil and Peru only, that civilization has advanced toward 

 the central regions ; it has penetrated them also by three 

 other roads, the Amazon, the Oroonoko, and the Rio de la 

 Plata 5 and has ascended the tributary streams of those three 

 rivers and their secondary branches. From the increase of 

 these routes, and their various directions, a configuration 

 of territory and a sinuosity of frontier have resulted, no less 

 difficult to determine astronomically, than disadvantageous 

 to inland trade. 



To the two causes of uncertainty in the estimation of sur- 

 faces, which we have just analyzed, namely, the errors of 

 astronomical geography, and the discussions of limits, may 

 be added a third, the most important of all. When we speak 

 of the area of Peru, or of the ancient Capitania-gencral of 

 Caraccas, it may be doubted whether these names denote 

 only the country in which the Spanish Americans have made 

 settlements, and which consequently depend on their po- 

 litical and religious hierarchy ; or whether we should join 

 to the country governed by the whites (by corregidors, chiefs 

 of military posts, and missionaries), the forests and savan- 

 nahs partly desert, and partly inhabited by savages, that is 

 by native and free tribes. We have seen above, mat in the 

 interior errors easy to suppose of 1 0 of latitude, or 2° of 



