Between these two schools of opinion it is necessary for us to pick our way, and in doing so I propose to devote the largest part of my attention to the arguments of Dr. Nansen, which set out most skilfully, and with a wealth of research which it would be difficult to equal, the point of view which is most directly opposed to my own. _Admixture of the Supernatural._ Of course in the writings of so primitive and superstitious an age, based upon oral traditions of an even earlier date, we cannot expect to find a standard of historical accuracy equal to that of the present day. The authors, however truthful in intention, had not reached a stage of enlightenment enabling them to winnow fact from myth, both elements appearing to them to be equally credible. As Livy candidly postulated in the case of Rome, some licence must be conceded to antiquity in the dressing-up of early history by an admixture of superstition with the facts it seeks to record. ‘To suppose’, says Dasent, in his admirable introduction to the Njál Saga, ‘that a story told in the eleventh century, when phantoms, and ghosts, and wraiths were implicitly believed in, and when dreams and warnings and tokens were part of every man’s creed, should be wanting in these marks of genuineness, is simply to require that one great proof of its truthfulness should be wanting.’ In other words, one would be entitled to regard the authenticity of any history alleged to be early with great suspicion, if no traces of the supernatural were to be found in it. Such things are to be seen in contemporary chronicles of early times no less than in histories written long after the events described; the evidence might not be sufficient to satisfy a member of the Psychical Research Society, but it was good enough for those who lived in primitive and credulous times. The ghosts and miracles of such history, not in Iceland alone but everywhere, are not conscious inventions on the part of the historian, and do not really damage his credit.