When, however, we come to consider the physical difficulties which had to be overcome by these great navigators in the accomplishment of their purpose, the greater credit must undoubtedly be awarded to Vasco da Gama. Columbus, trusting as implicitly to the chart and sailing directions of Toscanelli as did Vasco da Gama to those of Dias, and, perhaps, of Pero de Covilhão, shaped a course westward of Gomera; and, having sailed in that direction for thirty-six days, and for a distance of 2,600 miles, made his first landfall at Guanahani, being favoured all the while by the prevailing easterly winds. The task which Vasco da Gama undertook was far more difficult of accomplishment. Instead of creeping along the coast, as had been done by his predecessors, he conceived the bold idea of shaping a course which would take him direct through the mid-Atlantic from the Cape Verde Islands to the Cape of Good Hope. The direct distance to be covered was 3,770 miles, but the physical obstacles presented by winds and currents could only be overcome by taking a circuitous course, and thus it happened that he spent ninety-three days at sea before he made his first landfall to the north of the bay of St. Helena. This first passage across the southern Atlantic is one of the great achievements recorded in the annals of maritime exploration. Once beyond the Cape, Vasco had to struggle against the Agulhas current, which had baffled Bartholomeu Dias, and against the current of Mozambique; and it was only after he had secured a trustworthy pilot at Melinde that the difficulties of the outward voyage can be said to have been overcome. In one other respect Vasco da Gama, or, perhaps, we ought to say his pilots, proved themselves the superiors of Columbus, namely, in the accuracy of the charts of their discoveries which they brought home to Portugal. Accepting the Cantino Chart[7] as a fair embodiment of the work done by this expedition, we find that the greatest error in latitude amounts to 1° 40´. The errors of Columbus were far more considerable. In three places of his Journal the latitude of the north coast of Cuba is stated to be 42° by actual observation; and that this is no clerical error, thrice repeated in three different places, seems to be proved by the evidence of the charts. On that of Juan de la Cosa, for instance, Cuba is made to extend to lat. 35° N. (instead of 23° 10´), and even on the rough sketch drawn by Bartolomeo Columbus after the return from the Fourth Voyage, Jamaica and Puerto Rico (Spagnola) are placed 6° too far to the north.[8] Verily, the Portuguese of those days were superior as navigators to their Spanish rivals and the Italians.