ApKir. 29, 1904.] 



SCIENCE. 



687 



Not having at hand the oldest of these 

 maps (that of Gaspar Viegas of 1534*) , we 

 take as presumably the most complete and 

 accurate of the reproductions of this proto- 

 type, that of Diogo Homen for the nomen- 

 clature, and of Pierre Desceliers for the 

 topographical design of the neighborhood 

 of Maranhao. This last map is also inter- 

 esting as showing that notwithstanding its 

 date posterior to the voyage of Orellana 

 down the Amazonas, the drawing was made 

 before this event, the river being inter- 

 polated (and erroneously) in a design that 

 figured a continuous coast line to the north- 

 ward of the gulf of Maranhao. The same 

 interpolation, but more artfully made, is 

 to be noted in the Desliens map, while 

 Alonzo de Santa Cruz (and undoubtedly 

 also Alonzo de Chaves in his lost Padron 

 Real), not knowing the Amazonas of Orel- 

 lana, adjusted the new prototype to the 

 old maps by suppressing the Amazonas and 

 identifying the 'Rio de la Mar Dulce' of 

 Pinzon (to which the Spaniards had ap- 

 plied the name of 'Maranon') with the gulf 

 of Maranhao of the Italian cartographers 

 and of the new Portuguese explorer. 

 Other Spanish cartographers (Cabot and 

 Gutierrez) resolved the difficulty by hash- 

 ing the drawing and nomenclature of the 

 new prototype to the west of Maranhao. 

 From this resulted the confusion and dis- 

 cordance of the early maps that figured so 

 largely in the discussion of the limits be- 

 tween Brazilian and French Guyana, with- 



* This map preserved in the National Library 

 of Paris is, according to the note by Harrisse 

 (op. cit., p. 599), a nautical chart representing, 

 in the Brazilian part, the coast from two or three 

 degrees to the west of Maranhao to about two 

 degrees to the south of the La Plata estuary. 

 With reference to it Ferdinand Diniz (cited by 

 Harrisse) says: "Captain Mouchez who had been 

 charged by the French government with the con- 

 tinuation and improvement of the work of Admiral 

 Roussin (the marine chart of the coast of Brazil) 

 was like myself astonished by the relative ac- 

 curacy of this geographical monument." 



out their origin being discovered on ac- 

 count of the loss of the Padron Real of 

 Alonzo de Chaves from which they pro- 

 ceeded. The recent publication in Sweden 

 of the map of Alonzo de Santa Cruz, which 

 is essentially a reproduction of the Padron 

 Real, clears up the matter perfectly, show- 

 ing that in a group of maps antedating 

 the voyage of Orellana there was a total 

 suppression of the Amazonas, which had 

 afterwards to be restored confusedly. 



In view of the circumstances above indi- 

 cated we have selected for reproduction the 

 maps of Diogo Homen, Desceliers and 

 Alonzo de Santa Cruz. The others give 

 more or less diversified variances of the 

 same theme. 



In studying this prototype one is im- 

 pressed with the relatively minute and 

 accurate representation of the hydrography 

 of the basin of the gulf of Maranhao (far 

 superior, for example, to the representation 

 given by the famous Sebastian Cabot to the 

 River Plate basin where he had been for 

 four years), and by the introduction of a 

 group of Indian names of which some 

 (Itapicuru and Pindare) have been pre- 

 served to the present day. This seems to 

 indicate that the explorer found here an 

 European domiciled amongst the Indians 

 and well acquainted with the topographical 

 details of the region. This supposition is, 

 to a certain extent, confirmed by the fact 

 that the only other indigenous names are 

 found grouped between Pernambuco and 

 Cape S. Roque, where the Europeans had 

 been for some years in contact with the 

 Indians so as to have become somewhat 

 acquainted with their language. 



The topographic design of the reproduc- 

 tions of this prototype in the maps of Diogo 

 Homen and Desceliers is sufficiently de- 

 tailed to permit the identification of some 

 of the more salient features, and, based 

 upon these, we can attempt that of the 

 names as follows, taking them from the 



