22 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 59 



success in 1497. His small ship Matthew won through the storm- 

 belt to the region about the Gulf of St. Lawrence ; and he evidently 

 felt that this was Brazil, for he uttered hopeful forecasts of finding 

 silk and brazil-wood. Since it was midsummer, the extravagance of 

 this would be hidden ; besides, his ideas were no doubt colored by 

 acquaintance with lovely, dye-yielding, forest-clad, fortunate islands 

 of the eastern Atlantic, and his words were perhaps meant chiefly 

 for more southward points than his first landfall, since he may have 

 voyaged a considerable distance that way. 



There are certain features of this Brazil most naturally explained 

 as imperfect 'delineations of that out jutting elbow of North America 

 which includes the Gulf of St. Lawrence, although no one seems to 

 have noticed what they indicate. Thus the Catalan atlas of 1375 

 shows Brazil not as a solid land, but as enclosing a sheet of water in 

 which several isles appear. Nordenskjold x says they are seven in 

 number, and reads them as derived from the legend of the Island of 

 the Seven Cities, giving no authority except his own fancy. But this 

 Brazil is too far north for the Spanish story, which most likely had 

 to do with one of the Azores or Madeira, being perhaps an exaggera- 

 tion of some real migration of escape, such as would be nearly certain 

 to occur at the height of the Moorish conquest. Besides, seven towns 

 do not require an equal number of islands in a great lake or an inland 

 sea. The Spaniards themselves felt no incongruity in hunting for 

 those cities, in 1539-40, among the deserts and mesas of Xew Mexico. 



Again, several maps, for instance Prunes's 2 1553 and Mercator's 

 1595, show Brazil as divided into two islands by a passage or channel. 

 For this also we have a mythological explanation (by Dr. Nansen 3 ) 

 namely the " river of death." But again the conjecture is quite 

 unsupported. Yet again, in several maps, Brazil has a space marked 

 on it after a quaint early fashion of indicating mountainous regions 

 and other natural features, and this bears the inscription Montorius 

 or Mont orious, apparently meaning at least, that a portion of Brazil 

 was mountainous. But the map of Dalorto 1325 or 1330 gives its 

 name in full as Insula de montonis siue de brazile. 4 (See note 3, p. 



If, now, we apply these several distinctive features to the region 

 reached by Cabot, we find this out jutting corner of America sur- 

 rounding the Gulf of St. Lawrence, which contains Prince Edward's 

 Island and the Magdalen Islands, Brion Island and others. Its east- 



1 Periplus, p. 164. 



2 K. Kretschmer: The Discovery of America, Atlas, Tafel 4. map 5. 



3 In Northern Mists, vol. 2, p. 228. 



