26 PROCEEDINGS COTTESWOLD CLUB 1910 



volume became augmented — overflowed, and like most such lakes, found relief 

 by forming an overflow channel between the moraine-dam and the hard rock- 

 bank on the down-stream side. The outflowing water soon commenced 

 cutting into the slaty rocks of the Ludlow, filing its way downwards, and 

 soon grading with the Wye. Having established a passage, the Edw did not 

 attempt to revert to its pre-glacial channel. Now, differential denudation 

 has made its influence felt more on the morainic material than on the hard 

 crag between it and the Aberedw Rocks, with the result that there is seen in 

 a section from south to north (i) the Aberedw Rocks, (2) the deep gorge with 

 the swift-flowing Edw in its bottom, (3) the bailey-castle-crowned crag, (4) 

 the broad meadows on the morainic material, and (5) the hill-slope again 

 composed of solid rocki (Plate V.) 



It was unanimously agreed that the explanation was in accordance with 

 the facts, and after a few questions had been asked and answered, a discus- 

 sion ensued on the probable history of the mound upon which they were 

 stood. Lieut. -Col. Duke thought it was a tumulus, and Mr J. M. Dixon 

 a bailey-castle similar to that at Upper Slaughter, near Bourton-on-the- 

 Water, only smaller. 



The quaint old church, which is of a very usual Welsh type, having the 

 nave and chancel of the same width, came in for much attention, particularly 

 the old porch. 



Next Llywelyn's Cave was visited (Plate IV., fig. i). It is not easy to find, 

 and when found turns out to be but a small room-like cave cut out of the 

 Ludlow rocks. Its historic associations are such, however, as to add zest 

 to the investigations of Llywelyn's connection with these parts. 



It appears that William, son of William de Breos (the son of Philip 

 de Breos, who probalily built the first Norman castle at Builth), sided 

 with the Barons against King John, and as a result was deprived of his pos- 

 sessions and attainted. Some, however — including the lordship of Builth — • 

 were restored to William's second son, Giles, Bishop of Hereford. On 

 his death it passed to his brother Reginald. Reginald and Llywclyn for some 

 time fought against John, but later Reginald forsook Llywelyn and sided 

 with Henry III. In 122 1 Llywclyn besieged Reginald in the Castle at 

 Builth, and was only caused to desist by the arrival of a strong force sent by 

 Henry from Gloucester. Reginald hacl three daughters — Eleanor, Eva, and 

 Maud, and when in 1230 he was summarily hanged for intriguing with 

 Llywelyn's wife, his Builth and Radnor possessions passed to his third 

 daughter, Maud. This Maud married Roger de Mortimer, but it would seem 

 that ever since the attainder of the first William de Breos the King of 

 England kept Builth Castle in his own hands, and the Mortimers were mere 

 custodians. In 1260 Llywelyn captured the castle by night, and Mortimer 

 was only with difficulty acquitted of connivance. Rhys ap Griffin acted 

 as custodian of the castle for Llywelyn for a while, but was much molested 

 by the Mortimers. When Edward I. came to the throne, Llywelyn partly 

 destroyed Builth Castle and abandoned it. Edward I., however, rebuilt it, 

 and made it, for its size, the strongest in the country. 



John Giffard — the same man who had fought with Llywelyn at Lewes on 

 the side of the Barons against Prince Edward, but had afterwards deserted 

 them and joined the King — was placed in command. Soon Llywelyn was 

 engaged in a fierce struggle against Edward I. in North Wales. In order to 

 capture Llywelyn, Giffard pretended to have deserted the King, and invited 

 Llywelyn to Builth. Llywelyn, not suspecting treachery, hastened to 

 Aberedw Castle with only eighteen picked men. There, however, Robert 

 I'Estrange and Edmund and Roger Mortimer made an unsuccessful attempt 

 to surprise and capture Llywelyn that night. He escaped, and is supposed to 



I The glacial features here have been described in detail in the Geological Magazine 

 for November, 1909. 



