The existence of these mounds in distinct Welsh territory is very curious and requires explanation. That this form of dwelling was in common use among the Welsh is certainly not the case. Where moated mounds occur in Wales it is usually on the border, or near the sea-coast, or in or near the open valleys accessible to the English, which the English or Northmen are known to have invaded in the eighth and ninth centuries. The mound near Llanidloes is an exception, being distinctly within the hills. But that of Tafolwern, from which the Welsh princes dated several charters, is near the open valley. That of Talybont, whence Llewelyn dated a letter to the Archbishop of Canterbury in 1275, and which was afterwards visited by Edward the First, is on a plain within easy reach of the sea. Hên Domen and Rhos Diarbed on the Upper Severn are also good examples. Still, as the Welsh princes intermarried and had frequent communication with the English, they must have been familiar with a form of fortification very simple and easy to construct, and yet very capable of being held against a sudden attack. It must be observed, also, that the English hold upon the Welsh border was of a very fluctuating description, and the Welshmen must not only have been perfectly familiar with the English method of construction, but from time to time have been actually in possession of their strongholds. That the Welsh used timber for defensive purposes appears from their law by which the vassals were to attend at the lord’s castle for its repairs or for rebuilding, each with his axe in his hand. In some cases in these Border works there is scarcely any mound, at others the mound is low and hollow in the centre. Caer Aeron and another small earthwork near Builth seem to have been the earthen bases of a mere circular wigwam. Caer Aeron cannot have been a mere temporary structure, as the circumscribing ditch has been cut with considerable labour in the rock. It is very evident, both from the existence of Offa’s dyke, and from the immense number of these moated mounds thrown up along its course, that the English had early and long possession of large tracts of the border territory. Offa ruled over Mercia from A.D. 757 to 796, and his dyke extends from the mouth of the Wye to that of the Dee. At its northern part, for about forty miles, is a second work, known as Wat’s Dyke, a little in its rear, and thought to be a somewhat earlier work, also by Offa. Before the actual line of a work so galling to the spirit of a turbulent people could have been decided upon there must have been many years of contest along the border, and the English must have had something like permanent possession of the land on either side, and have held estates of which the mounds still existing were the “capita” or chief seats. The dyke, it should be remembered, was rather a civil boundary than a military defence.