But on shipboard the variation is still great, though the last fifty years has largely reduced the error. Professor Rogers, of the Harvard College Observatory, in examining one hundred log-books of Atlantic steamships, has found an average error of three miles; and he reports as significant of the superior care of the Cunard commanders that the error in the logs of their ships was reduced to an average of a mile and a half.
[395] Lelewel, ii. 130.
[396] Humboldt, Examen critique, ii. 210.
[397] The breadth east and west of the Old World was marked variously,—on the Laon globe, 250°; Behaim’s globe, 130°; Schöner’s globe, 228°; Ruysch’s map, 224°; Sylvanus’ map, 220°; and the Portuguese chart of 1503, 220°.
[398] This sea-chart was the first which had been seen in England, and almanacs at that time had only been known in London for fifteen years, with their tables for the sun’s declination and the altitude of the pole-star.
[399] Cf. Atti della Società Ligure, 1867, p. 174, Desimoni in Giornale Ligustico, ii. 52. Bartholomew is also supposed to have been the maker of an anonymous planisphere of 1489 (Peschel, Ueber eine alte Weltkarte, p. 213).
[400] Strabo, i. 65. Bunbury, Ancient Geography, i. 627, says the passage is unfortunately mutilated, but the words preserved can clearly have no other signification. What is left to us of Eratosthenes are fragments, which were edited by Seidel, at Göttingen, in 1789; again and better by Bernhardy (Berlin, 1822). Bunbury (vol. i. ch. xvi.) gives a sufficient survey of his work and opinions. The spherical shape of the earth was so generally accepted by the learned after the times of Aristotle and Euclid, that when Eratosthenes in the third century, B.C. went to some length to prove it, Strabo, who criticised him two centuries later, thought he had needlessly exerted himself to make plain what nobody disputed. Eratosthenes was so nearly accurate in his supposed size of the globe, that his excess over the actual size was less than one-seventh of its great circle.
[401] There is a manuscript map of Hispaniola attached to the copy of the 1511 edition of Peter Martyr in the Colombina Library which is sometimes ascribed to Columbus; but Harrisse thinks it rather the work of his brother Bartholomew (Bibl. Amer. Vet., Add., xiii.) A map of this island, with the native divisions as Columbus found them, is given in Muñoz. The earliest separate map is in the combined edition of Peter Martyr and Oviedo edited by Ramusio in Venice in 1534 (Stevens, Bibliotheca geographica, no. 1,778). Le discours de la navigation de Jean et Raoul Parmentier, de Dieppe, including a description of Santo Domingo, was edited by Ch. Schefer in Paris, 1883; a description of the “isle de Haity” from Le grand insulaire et pilotage d’André Thevet is given in its appendix.
[402] Cosmos, Eng. tr., ii. 647. One of these early engravings is given on page 15.
[403] Navarrete, i. 253, 264.