[495] The various versions of the letter are as follows: Ulloa (Historie, 1571, ch. 8). Dalla città di Lisbona per dritto verso ponente sono in detta carta ventisei spazi, ciascun de’ quali contien dugento, & cinquanta miglia, fino alla ... città di Quisai, la quale gira cento miglia, che sono trentacinque leghe.... Questo spazio e quasi la terza parte della sfera.... E dalla’ Isola di Antilia, che voi chiamate di sette città, ... fino alla ... isola di Cipango sono dieci spazi, che fanno due mila & cinquecento miglia, cioè dugento, & venticinque leghe.
Barcia. Hallareis en un mapa, que ai desde Lisboa, à la famosa ciudad de Quisay, tomando el camino derecho à Poniente, 26 espacios, cada uno de 150 millas. Quisai’ tiene 35 leguas de ambitu.... De la isla Antilla hasta la de Cipango se quentan diez espacios, que hacen 225 leguas.
Las Casas: Y de la ciudad de Lisboa, en derecho por el Poniente, son en la dicha carta 26 espacios, y en cada uno dellos hay 250 millas hasta la ... ciudad de Quisay, la cual etiene al cerco 100 millas, que son 25 leguas, ... (este espacio es cuasi la tercera parte de la sfera) ... é de la isla de Antil, ... Hasta la ... isla de Cipango hay 10 espacios que son 2,500 millas, es á sabre, 225 leguas.
Columbus’s copy: A civitate vlixiponis per occidentem indirecto sunt .26. spacia in carta signata quorum quodlibet habet miliaria .250. usque ad nobilisim[am], et maxima ciuitatem quinsay. Circuit enim centum miliaria ... hoc spatium est fere tercia pars tocius spere.... Sed ab insula antilia vobis nota ad insulam ... Cippangu sunt decem spacia.
[496] Cf. “Les îles Atlantique,” by Jacobs-Beeckmans in the Bull. de la Soc. géog. d’Anvers, i. 266, with map.
[497] Of these collections, those of Kunstmann and Jomard are not uncommon in the larger American libraries. A set of the Santarem series is very difficult to secure complete, but since the description of these collections in Vol. II. was written, a set has been secured for Harvard College library, and I am not aware of another set being in this country. The same library has the Ongania series. The maps in this last, some of which are useful in the present study, are the following:—
1. Arabic marine map, xiiith cent. (Milan); 2. Visconte, 1311 (Florence); 3. Carignano, xivth cent. (Florence); 4. Visconte, 1318 (Venice); 5. Anonymous, 1351 (Florence); 6. Pizigani, 1373 (Milan); 7. Anon., xivth cent. (Venice); 8. Giroldi, 1426 (Venice); 9. Bianco, 1430, (Venice); 10. Anon., 1447 (Venice); 11. Bianco, 1448 (Milan); 12. Not issued; 13. Anon., Catalan, xvth cent. (Florence); 14. Leardo, 1452; 15. Fra Mauro, 1457 (Venice); 16. Cantino, 1501-3 (Modena). This has not been issued in this series, but Harrisse published a fac-simile in colors in connection with his Les Corte-Real, etc., Paris, 1883. 17. Agnese, 1554 (Venice). The names on these photographs are often illegible; how far the condition of the original is exactly reproduced in this respect it is of course impossible to say without comparison.
[498] The notions prevailing so far back as the first century are seen in the map of Pomponius Mela in Vol. II. p. 180.
[499] Vol. II. p. 36.
[500] Lelewel (ii. 119) gives a long account of Sanuto and his maps, and so does Kunstmann in the Mémoires (vii. ch. 2, 1855) of the Royal Bavarian Academy; but a more perfect inventory of his maps is given in the Studi biog. e bibliog. of the Italian Geographical Society (1882, i. 80; ii. 50). Cf. Peschel, Gesch. der Erdkunde, Ruge, ed. 1877, p. 210. Sanuto’s map of 1320 was first published in his Liber Secretorum fidelium crucis (Frankfort, 1811. Cf. reproduction in St. Martin’s Atlas, pl. vi. no. 3). Further references are in Winsor’s Kohl Maps, no. 12. It is in part reproduced by Santarem.