We will take a step forward, and turn from local to personal attachments of tradition. There is a whole class of traditions attached to personages about whose historical existence there can be but little doubt, and just because of the accretion of tradition round them their historical existence has oftentimes been denied. The most famous example in our history is of course King Arthur, and so great an authority as Sir John Rhys is obliged to resort to a special argument to account for the problems he is faced with. He argues, and argues strongly, for an historic Arthur--an Arthur who was the British successor of the Roman emperor after Britain had ceased to be a part of the Roman Empire.[40] But because of the myths which have grown round him, he suggests that there must also have been "a Brythonic divinity named Arthur," and we are thus introduced to a dual study of history and myth which does not appear to me to take us very far, and which, in fact, just separates history from myth, instead of showing where they join hands. This dual conception of myth is indeed a rather favourite resort of those scholars who cannot appreciate the evidence that proves a character in a mythic tradition to be an actual historical personage. It is the basis of the famous Sigfried-Arminius controversy. It does duty in many less important cases,[41] and most frequently in connection with northern mythology, where the line between mythic and historical events gathering round a hero is generally so finely drawn as to be almost imperceptible. But it is so obviously a piece of special pleading on self-created lines that other explanation is needed. And another explanation is to be obtained if only students will rely upon the evidence of tradition itself instead of appealing to every fancy derived from sources which have nothing to do with tradition.