The scholarship of the day was not always accurate in discriminating between ancient systems of religion, and Gray, in his letters to Mason in 1758, when "Caractacus" was still in the works, takes him to task for mixing the Gothic and Celtic mythologies. He instructs him that Woden and his Valhalla belong to "the doctrine of the Scalds, not of the Bards"; but admits that, "in that scarcity of Celtic ideas we labor under," it might be permissible to borrow from the Edda, "dropping, however, all mention of Woden and his Valkyrian virgins," and "without entering too minutely on particulars"; or "still better, to graft any wild picturesque fable, absolutely of one's own invention, upon the Druid stock." But Gray had not scrupled to mix mythologies in "The Bard," thereby incurring Dr. Johnson's censure. "The weaving of the winding sheet he borrowed, as he owns, from the northern bards; but their texture, however, was very properly the work of female powers, as the art of spinning the thread of life in another mythology. Theft is always dangerous: Gray has made weavers of the slaughtered bards, by a fiction outrageous and incongruous."[6] Indeed Mallet himself had a very confused notion of the relation of the Celtic to the Teutonic race. He speaks constantly of the old Scandinavians as Celts. Percy points out the difference, in the preface to his translation, and makes the necessary correction in the text, where the word Celtic occurs--usually by substituting "Gothic and Celtic" for the "Celtic" of the original. Mason made his contribution to Runic literature, "Song of Harold the Valiant," a rather insipid versification of a passage from the "Knytlinga Saga," which had been rendered by Bartholin into Latin, from him into French by Mallet, and from Mallet into English prose by Percy. Mason designed it for insertion in the introduction to Gray's abortive history of English poetry.