« PreviousContinue »
THE SO-CALLED SHELL KEEP
We have three accounts of motte-castles from the 12th century: that of Alexander Neckham, in the treatise De Utensilibus; that of Laurence of Durham, cited in Chapter VII., p. 147; and the well-known description of the castle of Marchem, also cited in Chapter VI., p. 88. All these three describe the top of the motte as surrounded by a wall (of course of wood), within which is built a wooden tower. The account of Marchem says that it was built in the middle of the area. This supports the conjecture in the text. Mr H. E. Malden has shown (Surrey Archæolog. Collections, xvi., 28) that the keep of Guildford is of later date than the stone wall round the top of the motte. Remove this tower, and there would be what is commonly called a shell keep. It would appear, therefore, that it was a common practice to change the bank or stockade round the top of the motte into a stone wall (no doubt as a defence against fire), leaving the keep inside still of wood. Four of the pictures from the Bayeux Tapestry (see Frontispiece) all give the idea of a wooden tower inside a stockade on a motte.
I REGRET that this valuable work did not appear until too late for me to make use of it in my chapter on Welsh Castles. It is worth while to note the following points in which Professor Lloyd's conclusions differ from or confirm those which I have been led to adopt.
Aberystwyth and Aberrheiddiol.-—"After the destruction of the last Aberystwyth Castle of the older situation in 1143, the
chief stronghold of the district was moved to the mouth of the Rheiddiol, a position which it ever afterwards retained, though people still insisted on calling it Aberystwyth” (514). "The original castle of Aberystwyth crowned the slight eminence at the back of the farm of Tan y Castell, which lies in the Ystwyth valley if miles S. of the town. There is the further evidence of the name, and the earthworks still visible on the summit” (426, note).
Carreghova.—I ought perhaps to have included this castle in my list, though on the actual map its site is within the English border ; but as there are absolutely no remains of it [D. H. M.) it does not affect the question I am discussing.
Cardigan and Cilgerran.—“Dingeraint cannot be Cilgerran, because Cilgerran is derived from Cerran, with the feminine inflection, not from Geraint; nor is Cilgerran 'close to the fall of the Teifi into the sea,' as the chronicler says Dingeraint was. The castle built by Earl Roger was probably Cardigan” (401). Professor Lloyd afterwards identifies Cilgerran with the castle of Emlyn (661). This seems to me questionable, as the “New Castle of Emlyn,” first mentioned in Edward I.'s reign, presupposes an older castle, and as I have stated, a mound answering to the older castle still exists not far from the stone castle.
Carmarthen.-Professor Lloyd thinks this castle stood at the present farm of Rhyd y Gors, about a mile below the town ; but I see no reason to alter the conclusion to which I was led by Mr Floyd's paper, that the Rhyd y Gors of the castle was a ford at Carmarthen itself. The fact that Henry I. founded a cell to Battle Abbey at Carmarthen (431) seems to me an additional piece of evidence that the castle was there; castle and abbey nearly always went together.
Dinweiler. — Professor Lloyd assumes Dinweiler to be the same as the castle in Mabudryd built by Earl Gilbert, and to be situated at or near Pencader (501). It should be noted, however, that Dinweiler reads Dinefor in MS. B. of the Brut, in 1158. I am in error in supposing St Clair to be the castle of Mabudryd (following a writer in Archæologia Cambrensis), as St Clair is not in that commote. Professor Lloyd's map of the cantrefs and commotes differs widely from that of previous writers.
Llangadoc.—“Luchewein” should not be identified with this castle ; Professor Lloyd thinks it may refer to a castle at Llwch
Owain, a lake in the parish of Llanarthney, where there is an entrenchment known as Castell y Garreg.
Maud's Castle.-Camden identified "Matildis castrum" with Colewent or Colwyn, but Professor Lloyd is of opinion that “a careful collation of the English and Welsh authorities for the events of the years 1198 and 1231 will make it clear that Payne's Castle and Maud's Castle are the same." This of course does not affect what is said about Colwyn Castle in the text.
Montgomery.- Professor Lloyd deems that the emphasis laid (especially in the Charter Rolls, i., 101) on the fact that the building of Henry III.'s reign was New Montgomery, leaves no doubt that the former town and castle stood elsewhere, probably at Hên Domen. This, if true, would greatly strengthen my case, as Hên Domen is an admirable motte and bailey.
SCHEDULE OF ENGLISH CASTLES KNOWN TO DATE FROM
THE ELEVENTH CENTURY
Name of Castle.
Not given T. R. E.
» 44 acres
Whole area 3 acres
4 acres : First ward 4 acre
Inner wardreand keep } Risen.
Fallen, but rising.
Inner castle about 6 acres
M. and B., O.
Whole area about 3 acres Risen.
M. and B., O.
The Baile Hill, York
M. and B., O.
1 Aldreth and Burton are omitted from this list. 2 M. and B. stand for Motte and Bailey ; K. and B. for Keep and Bailey; O. for Outside the Town.