readily conceives that, if one sailed straight westward out of the Mediterranean through the Straits of Gibraltar, he would ultimately come back round the world by way of the East—­that is to say, by India.  It was not left for Columbus to invent that doctrine.  It is true that in calculating the circumference of the earth they had made it as much as one-seventh too large, but the wonder is that they came so near as they did.  In regard to the distance of the moon they were not more than 1/12th from the modern estimate.  The possibility of error in dealing with the sun was much greater, and their 51,000,000 miles is little more than half of what it should have been.  Exactly how far this doctrine of the sphericity of the earth was popularly entertained we cannot tell; it was probably almost confined to those directly interested in the question.  A theory, anticipating Galileo, that it is the earth which moves round the sun, had been mooted, but certainly had very little currency.  Nor was speculation confined to such astronomical conclusions.  In the region of physical geography rational attempts were made to account for various phenomena, such as the existence of deltas or the risings of the Nile, or the appearance of sea-shells high on dry land.  Strabo, in dealing with the Black Sea, has his theories of the elevation or subsidence of land.  He also suggests previous volcanic conditions of certain districts which had been quiescent from before the memory or tradition of the inhabitants.

[Illustration:  FIG. 113.—­WORLD AS CONCEIVED ABOUT A.D. 100.]

Sound methods of discovering latitude and longitude were not yet in use, and therefore a map of the world according to ideas current in the first century would present a strange aspect to us.  There is much error in the placing of towns or districts upon their parallels; and coasts or mountain ranges, particularly, of course, on the outskirts of the empire or in the less familiar lands beyond its bounds, are perhaps made to run north instead of north-west, or east instead of south-east.  It follows that measurements of distances especially across the wider seas, were often very inaccurate, although within and about the Mediterranean there was so much traffic and such close observation of the stars that the errors were gradually reduced.  The mariner, when he did not follow the coast and guide his course by familiar landmarks, steered by the stars, but of these he had a very intimate knowledge, to which he joined a close observation of the prevailing direction of the winds at the various seasons.  There was a well-ordered system of lighthouses, and charts and mariners’ guides were not wanting.  In the winter months navigation over long distances was regularly suspended, and ships waited in port for the spring.