centuries before him, when reasoning from the assumed sphericity of the earth, was really the first to point out that the west coast of Spain was the fittest point of departure for India.
The latitudes were reckoned in stadia from the Equator to Syracuse, the stadium being about two hundred and one yards and one foot. The determination of the longitude was, however, a far more difficult problem; as the only phenomena whereby men could readily determine the distance between any two places, viz. eclipses of the moon, would have been of no practical value in calculating a ship's position at sea; moreover, it would not be easy to secure certainty in such observations, nor could they easily be repeated. Hence the ancients were led to depend either on actual survey, or on the vague information obtainable from the reckonings of sailors, or on the itineraries of travellers. We need not, therefore, be surprised when we see how Ptolemy and the greatest of ancient geographers have erred, owing to the impossibility of fixing with even tolerable accuracy the longitudes of different places. It is likely that their practice of constantly landing might have in some degree supplied their deficiency in this particular; but we have now no record of any astronomical observations which were made at sea, by even the most skilful of ancient navigators. A sort of dead-reckoning—an observation of the position of the sun during the day, or of certain stars during night—was the haphazard mode by which their positions at sea were chiefly ascertained. If they
