My first principle in usage has been, simply, to make the story clear. I wanted if possible to avoid the irritating expedient of the glossary, where the reader has to interrupt himself to look up the place-names, or decide to read straight on and lose himself mentally. And non-British readers suffer further; they look up Calleva in the glossary, find it is Silchester, and are none the wiser until they consult a map. Either way the story suffers. So wherever there was a choice of names I have tried to use the one that will most immediately put the reader in the picture: for this I have sometimes employed the device of having the narrator give the current crop of names, even slipping in the modern one where it does not sound too out of place. For example: “Maesbeli, near Conan’s Fort, or Kaerconan, that men sometimes call Conisburgh.” Elsewhere I have been more arbitrary. Clearly, in a narrative whose English must be supposed in the reader’s imagination to be Latin or the Celtic of South Wales, it would be pedantic to write of Londinium when it is so obviously London; and I have used the modern names of places like Glastonbury and Winchester and Tintagel, because these names, though mediaeval in origin, are so hallowed by association that they fit contexts where it would obviously be impossible to intrude the modern image of (say) Manchester or Newcastle. These “rules” are not of course, intended as a criticism of any other writer’s practice; one employs the form the work demands; and since this is an imaginative exercise which nobody will treat as authentic history, I have allowed myself to be governed by the rules of poetry: what communicates simply and vividly, and sounds best, is best.

The same rule of ear applies to the language used throughout. The narrator, telling his story in Fifth Century Welsh, would use in his tale as many easy colloquialisms as I have used in mine; the servants Cerdic and Cadal would talk some kind of dialect, while, for instance, some sort of “high language” might well be expected from kings, or from prophets in moments of prophecy. Some anachronisms I have deliberately allowed where they were the most descriptive words, and some mild slang for the sake of liveliness. In short, I have played it everywhere by ear, on the principle that what sounds right is acceptable in the context of a work of pure imagination.

For that is all The Crystal Cave claims to be. It is not a work of scholarship, and can obviously make no claim to be serious history. Serious historians will not, I imagine, have got this far anyway, since they will have discovered that the main source of my storyline is Geoffrey of Monmouth’s History of the Kings of Britain.