l6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 59 



and his companions may have brought the holy image from Ireland ; 

 but in view of the great gap of time to be accounted for, I incline 

 rather to the entertaining Father Espinosa's artless declaration that 

 angels brought it straight from Heaven. 



In all this there is not much to be fairly called corroboration of 

 the internal evidence of the medieval voyage-narrative ; but it is 

 certainly interesting to find the sixteenth century Spaniards of the 

 Canaries well up in the legends of St. Brandan and St. Malo, and 

 confident of their visit to those islands a thousand years earlier. 



3. THE MYTHICAL ISLANDS OF THE ATLANTIC 

 The only place where one can still see St. Brandan is on Pizigani's 

 map 1 of 1367, bestowing his benediction, in medieval portraiture, 

 on his " Fortunate Islands," thus named collectively in the map-le- 

 gend, but individually as Ysola Caporizzia, Ysola Canaria, and Ysole 

 clouer Sommart. Possibly they were borrowed from Dulcert 1339 

 of Genoa, who calls the first-named island Capraria and the last 

 Primaria. 2 The site of the latter is identical in both maps and approx- 

 imately occupied by a cluster of rocks in a more modern one. S6m- 

 mart (somma) is, however, more likely to indicate the peak of Pico; 

 and the plural form Ysole may convey a sense of its less lofty Azorian 

 companions. Whatever the explanation of this item, the cartographer 

 of the Atlante Mediceo or Gaddiano map (1351) thought best to 

 omit it ; as does also the Catalan map of 1375. They substituted, how- 

 ever, for Caporizzia, Legname or d'Legname (Markland, forest-land) 

 because of the great woods " de haute futaie " (D'Avezac) 3 with 

 which the early visitors found it covered, also the companion island 

 becomes Porto Santo, as now, and Las Desertas have already taken 

 their name as Insulse Desertae. Zuan da Napoli, whose map that is, 

 the Venetian one uncertainly attributed to him is given by Kohl 

 approximately the date 14 (perhaps of 1440 or later) translates 

 Legname into Madera, its Portuguese equivalent, which, with a 

 little change in spelling, still remains. It seems pretty clear that 

 Madeira is the original Markland of Atlantic voyagers; also that 

 it and its neighbor, Porto Santo, with or without some lesser com- 



1 Kohl's collection of maps in Library of Congress. Also Jomard's Atlas. 



2 Nordenskjold's Periplus, pi. 8, also K. Kretschmer : The Discovery of 

 America (Die Entdeckung Amerikas), Atlas, Tafel I, pi. 2. Benincasa 1482 

 and others also show the Madeira group as three islands ; but consider Las 

 Desertas one of them, omitting Primaria or Sommart. 



3 Marie D'Avezac : Discoveries of the Middle Ages, pp. 7, 8. The best repro- 

 duction is in Fischer's Sammlung. There is also a good one in Benzley's The 

 Dawn of Modern Geography and an incomplete facsimile in Nordenskjold's 

 Periplus. 



