[APPENDIX D]
THE WORDS “CASTRUM” AND “CASTELLUM”
If we take the chroniclers of the reign of Charlemagne and his successors in the 9th century, we find the word castrum constantly used for places such as Avignon, Dijon, Macon, Rheims, Chalons, Cologne, Andernach, Bonn, Coblenz, etc., all of which are known to have been Roman castra, when there can be no doubt that the city is meant. Take, for instance, the Annales Mettenses (Pertz, i., 326), 737: Karl Martel hears that the Saracens have taken “castrum munitissimum Avinionem” (Avignon); he marches against them, and “predictam urbem obsidione circumdat.” But these cities are not only called castra, they are also called castella. Thus the chronicle ascribed to Hincmar calls Macon both castrum and castellum in the same breath. (Migne, 125, 1298.) The fortifications built by Charlemagne against the Saxons are called castra, castella, and civitates. (Chron. Moissiacense, Pertz, i., 308. Ann. Einhardi, ibid., 196, 204.) The camps of the Northmen, which as we have seen, were of great size, are also called not only castra, but civitates, castella, munitiones, oppida. (Annales Fuldenses, Pertz, i., 397.) The camp built by Charles the Bald at Pistes in 868 is called a castellum, though it was evidently an enclosure of great size, as he measured out quarters in it for his nobles, and formed an elaborate scheme for its maintenance. (Hincmar, Migne, 125, 1242, 1244.) Coming to the 10th century, the following passage from Flodoard will show the vagueness of the words in common use for fortifications: “Heribertus Ansellum Bosonis subditum, qui prædictum custodiebat castrum (Vitry), cum ipso castello recipit, et Codiacum S. Remigii municipium illi cum alia terra concedit. Nec longum, Bosonis fideles oppidanorum proditione Victoriacum (Vitry) recipiunt, et Mosonum fraude pervadunt. At Heribertus, a quibusdam Mosomensibus evocatus, supervenit insperatus, et entrans oppidum, porta latenter a civibus aperta, milites Bosonis, qui ad custodiam loci residebant, ibidem omnes capit.” (Migne, 135, 297.) Here it is clear that castrum, castellum, municipium, and oppidum all mean the same thing, and the one word civibus betrays that it is a city which is meant. Undoubtedly the chronicler thinks it elegant to change his words as often as he can. Munitio is another word frequently used; in classical Latin it means a bulwark, a wall or bank; in the chroniclers of the 10th century it is used indifferently for a town or castle, though certain passages, such as “subversis multarum munitionibus urbium” (Flodoard, i., vi.), show that the right sense is not far from the mind of the writer. The numerous passages in which we are told of monasteries being enclosed with walls and converted into castella, show that the enclosure is the chief idea which the chroniclers associate with this word. The citations made above are not exceptional, but typical, and could be paralleled by countless others.
Since the above was written, I have read Keutgen’s Untersuchungen über den Ursprung der Deutschen Stadtverfassung. He remarks that the Latin words for a town (in the 10th and 11th century writers) are urbs, castellum, civitas, sometimes arx; for a village, villa, oppidum, vicus. This absolutely agrees with what I have observed in these writers, except that I have certainly found oppidum used for a town, as in the passage from Flodoard cited above.
[APPENDIX E]
THE BURGHAL HIDAGE
The Burghal Hidage has been printed by Birch, Cartularium, iii., 671. The manuscript is very corrupt, and several of the places cannot be identified. Those which can be identified are: Hastings, Lewes, Burpham (near Arundel), Chichester, Porchester, Southampton, Winchester, Wilton, Tisbury, Shaftesbury, Twineham, Wareham, Bridport, Exeter, Halwell, Lidford, Pilton, Barnstaple, Watchet, Axbridge, Lyng (near Athelney), Langport, Bath, Malmesbury, Cricklade, Oxford, Wallingford, Buckingham, Eashing (near Guildford), and Southwark. The list thus seems to give an outline of Alfred’s kingdom as it was at his death, or at the beginning of the reign of his son. Dr Liebermann refers it to the latter date. (Leges Anglorum, 9.)
[APPENDIX F]
THELWALL
A writer in the Manchester Guardian a few years ago suggested a new solution of the name Thelwall. He believes that the Thelwall raised by Edward was a boundary wall of timber, stretching from Thelwall to Runcorn. The Mersey, he argues, above Thelwall formerly broadened out into a series of swamps which would effectually defend the frontier towards the east. But westward from Thelwall there were no such obstacles, and it is assumed that Edward made a timber wall from Thelwall to Ethelfleda’s fortress at Runcorn. Some support to this hypothesis is given in the names of places between Thelwall and Runcorn: Stockton, Walton (twice), Stockham, Walford, Wallmore, and Wall-hes. Further, when the bed of the Mersey was delved for the Ship Canal, discovery was made of “a remarkable series of submerged piles, 9 feet long, arranged in two parallel ranks which were 30 feet apart. The intervals between the piles varied, but seem to have averaged 5 to 6 feet. Between the ranks were diagonal rows of upright stakes, each stake about 5 feet long, extending from either rank chevron-wise to the middle and there overlapping, so that the ground-plan of them makes a kind of herring-bone pattern. By this plan, anyone passing through would have to make a zigzag course. In some places sticks and sedges were found interwoven horizontally with the stakes, a condition of things which probably obtained throughout the whole series. The tops of the tallest piles were 10 feet below the present surface of the ground, which fact goes far toward precluding the possibility that this elaborate work may have been a fish-weir. The disposition of the stakes points to a military origin. So arranged, the advantage they offered to defending forces was enormous.” I think it worth while to reproduce this account, especially because of the place-names, but those who are learned in the construction of fish-weirs may perhaps think that the description will apply to a work of that kind.