ERRORS IX THE BEST MAPS. 29 



and the local circumstances, such as they can now be des- 

 cribed, it will be easy to conceive how the different hypo- 

 theses recorded on our maps have taken rise by degrees, and 

 have modified each other. To oppose an error, it is suffi- 

 cient to recall to mind the variable forma in which we have 

 seen it appear at different periods. 



Till the middle of the eighteenth century, all that vast space 

 f land comprised between the mountains of French Guiana 

 and the forests of the Upper Orinoco, between the sources of 

 the Carony and the River Amazon (from to 4.of north lati- 

 tude, and from 57 to 68 of longitude), was so little known, 

 that geographers could place in it lakes where they pleased, 

 create communications between rivers, and figure chains of 

 mountains more or less lofty. They have made full use of this 

 liberty ; and the situation of lakes, as well as the course and 

 branches of rivers, has been varied in so many ways, that it 

 would not be surprising, if among the great number of maps 

 some were found that trace the real state of things. The field of 

 hypotheses is now singularly narrowed. I have determined the 

 longitude of Esmeralda in the Upper Orinoco; more to the east, 

 amid the plains of Parima (a land as unknown as Wangara 

 and Dar-Saley, in Africa), a band of twenty leagues broad has 

 been travelled over from north to south along the banks of 

 the Rio Carony and the Rio Branco, in the longitude of 

 sixty-three degrees. This is the perilous road which was 

 taken by Don Antonio Santos in going from Santo Thome 

 del Angostura to Rio Negro and the Amazon ; by this road 

 also the colonists of Surinam communicated very recently 

 with the inhabitants of Grand Para. This road divides the 

 terra incognita of Parima into two unequal portions ; and 

 fixes limits at the same time to the sources of the Orinoco, 

 which it is no longer possible to carry back indefinitely 

 toward the east, without supposing that the bed of the Rio 

 Branco, which flows from north to south, is crossed by the 

 bed of the Upper Orinoco, which flows from east to west. 

 If \ve follow the course of the Rio Branco, or that strip of 

 cultivated land which is dependent on the Capitania General 

 of Grand Para, we see lakes, partly imaginary, and partly 

 enlarged by geographers, forming two distinct groups. The 

 first of these groups includes the lakes which they place 

 between the Esmeralda and the Rio Branco; and to the 



