We will now turn to consider the sources of his History of the Kings of Britain. Geoffrey says: "In the course of many and various studies I happened to light on the history of the Kings of Britain, and wondered that, in the account which Gildas and Bede, in their elegant treatises, had given of them, I found nothing said of those kings who lived here before Christ, nor of Arthur, and many others; though their actions were celebrated by many people in a pleasant manner, and by heart, as if they had been written. Whilst I was thinking of these things, Walter, Archdeacon of Oxford, a man learned in foreign histories, offered me a very ancient book in the Britannic tongue, which, in a continued regular story and elegant style, related the actions of them all, from Brutus down to Cadwallader. At his request, therefore, I undertook the translation of that book into Latin." At the end of his history he adds: "I leave the history of the later kings of Wales to Caradoc of Llancarven, my contemporary, as I do also the kings of the Saxons to William of Malmesbury and Henry of Huntingdon. But I advise them to be silent concerning the kings of the Britons, since they have not that book written in the Britannic tongue, which Walter, Archdeacon of Oxford, brought out of Britannia."
There has been a good deal of controversy as to whether this very ancient book was in Welsh or Breton, but the first question is, Did it ever exist? Was Geoffrey a translator, or an inventor, or a collector of oral traditions current in Wales or Brittany during his time?
There can be little doubt that the conclusion of Thomas Stephens, in the "Literature of the Kymry," is correct--that "Geoffrey was less a translator than an original author." It is very doubtful whether the Britannic book ever existed, whether it was not a mere ruse, such as was often resorted to by mediæval romancers, and is still a favourite method with modern historical novelists--to give their works an appearance of genuineness. It has been argued against this, that in that case, Archdeacon Walter must have been a party to the fraud--which is incredible. Such an argument implies a large ignorance of the archdeacons of the twelfth century--when it was a question solemnly discussed among the learned--whether an archdeacon could possibly be saved. It would be well if there were nothing worse to bring against them than such an innocent fraud on the public as this. But the strongest argument against the existence of the Britannic book is (not that it is not extant now, but) that the historians of the next generation never saw it. Geoffrey's History at once created a tremendous stir in the literary world--nor was it accepted on trust--but received with suspicion and incredulity. Thus William of Newburgh, in the latter part of the twelfth century, calls Geoffrey roundly, "a saucy and shameless liar." William, of course, did not know Welsh, and could not have made anything out of the Britannic book, even if he had seen it. This objection does not apply to Giraldus Cambrensis; his knowledge of Welsh was indeed slight--but he had plenty of Welsh-speaking relatives and friends, and he was himself a collector of manuscripts. Gerald refers to "the lying statements of Geoffrey's fabulous history," and implies in a much-quoted passage that he regarded Geoffrey's history as a pack of lies. Speaking of a Welshman at Caerleon who had dealings with evil spirits, and was enabled by their assistance to foretell future events, he goes on: "He knew when any one told a lie in his presence, for he saw the devil dancing on the tongue of the liar. If the evil spirits oppressed him too much, the Gospel of St. John was placed on his bosom, when like birds they immediately vanished; but when the Gospel was removed, and the History of the Britons by Geoffrey Arthur was substituted in its place, the devils instantly came back in greater numbers, and remained a longer time than usual on his body and on the book." Geoffrey may very probably have used some Britannic manuscript, but it could not have been very ancient; and he certainly did not translate it, but used it as he used Gildas and Bede and Nennius--sometimes quoting their statements, more generally amplifying them almost beyond recognition.