vast shoals, that check and appall the mariner. Aristotle thought that beyond the columns the sea was shallow and becalmed. Plato utilized the common idea of the mudbanks and shoal water of the Atlantic in accounting for the disappearance of Atlantis. Scylax reported the ocean not navigable beyond Cerne in the south, and Pytheas heard that beyond Thule sea and air became confounded. Even Tacitus believed that there was a peculiar resistance in the waters of the northern ocean.[316] Whether the Greeks owed this dread to the Phœnicians, and whether the latter shared the feeling, or simulated and encouraged it for the purpose of concealing their profitable adventures beyond the Straits, is doubtful. In two cases, at least, it is possible to trace statements of this nature to Punic sources, and antiquity agreed in giving the Phœnicians credit for discouraging rivalry by every art.[317] To an age averse to investigation for its own sake, ignorant of scientific curiosity, and unimpelled by economic pressure, tales like these might seem decisive against an attempt to sail westward to India. Rome could thoroughly appreciate the imaginative mingling of science and legend which vivified the famous prophecy of the poet Seneca: Venient annis saecula seris Quibus Oceanus vincula rerum Laxet, et ingens patebit tellus Tethysque novos deteget orbes Nec sit terris ultima Thule.[318] But even were it overlooked that the prophecy suited better the revelation of an unknown continent, such as the theory of Crates and Cicero placed between Europe and Asia, than the discovery of the eastern coast of India, mariners and merchants might be pardoned if they set the deterrent opinions collected by the elder Seneca above the livelier fancies of his son.[319] The scanty records of navigation and discovery in the western waters confirm the conclusions drawn from the visions of the poets and the theories of the philosophers. No evidence from the classic writers justifies the assumption that the ancients communicated with America. If they guessed at the possibility of such a continent, it was only as we to-day imagine an antarctic continent or an open polar sea. Evidence from ethnological comparisons is of course admissible, but those who are best fitted to handle such evidence best know its dangers; hitherto its use has brought little but discredit to the cause in which it was invoked. * * * * *