Lord Rhys as the wife of Rhys Sais, because Rhys was first cousin of the Lord Rhys (they were sisters' sons); nor can Mr. Hughes find any confirmation of a second wife of Rhys Sais given in Salusburie; the three sons of Rhys, Tudor Elida, and Iddon, are clearly established in Welsh history as the murderers of Gwrgenan ap Silsilt, Prince of Powys, in 1079.

It should be noted that the Salusburie MSS. gives Grono as a bastard son of Tudor ap Rhys, and Joanes as his illegitimate son; perhaps his fellow-countrymen so treated him on account of his paying tribute to the Earl Roger Montgomery: a fact which would excite their contempt. The dates given, 12 William I, and 1070, are inconsistent with our ideas of the date of Domesday; but probably that great work was not, as our savants suppose, accomplished in a single year, but more probably occupied the full twenty years from the Conquest to 1086, the date assigned by Ordericus for its completion; and, in that case, 12 William I may be the true date of the Shropshire portion of it. It would be of value to collect every date that can be found in connection with it-in which Welsh records may materially assist.

The mother of Rhys Sais is given as Jonet, verch Rivallon ap Cynfyn, Prince of North Wales; apparently he was second husband of Angharad v. Mdd ap Owen, and was married after 1025. The pedigree just given of this lady in vol. ii, p. 25, of Add. C. No. 177, in the Bodleian Library, which gives some dates. Owen, her grandfather, was son of Howell Dda, who died 948, and brother of Angharad, wife of Tudor Trevor; her father was Mdd ap Owen ap Howell, who died 998 ; and her first husband, Lln. ap Silsilt, is said to have died 1023. Looking at the fact that Ranulf Peverel and his son William were coeval with the Conquest, it seems impossible that the mother of Rhys Sais could have been a granddaughter of Cynfyn ap Gwerston. It would seem also that some of the Welsh wives had been selected at random; the males would rank fairly in these Poems, but not necessarily their wives.

Robert Eyton was too much carried away by scorn for pretentious quacks to do full justice to the Peveril pedigree he rashly declared that in no instance did any of the Peverils succeed Ranulf or Wrenock in his possession of Domesday property. This is manifestly inaccurate as regards the Norfolk property, which most certainly descended from Wrenock to the Peverils, as well as to Alan fil. Flaald, who seems to have been chief Lord, and who, under the advice and influence of the Lords of Monmouth, granted interest in Sporle and Mileham, so William Peveril of Dover succeeded, in Herts, Essex, and London, Ranulph Peveril and Robert Gernon (or the "Bearded One"), whose posterity crept into part of the Peveril inheritance at Bakewell. More curiously still the Peverils, under the name of de Hesding, or Hastings, succeeded him in that rape in Sussex Ernulf de Hastings, put to death by King Stephen. Ordericus calls him "Avunculus" of William fil. Alan: which may mean his father's or mother's sister's son, or more probably a female relative through Alan's mother.

It is quite clear, in spite of Eyton's objections, that several of the Peverels succeeded to Wrenoch's Domesday estate, and in so many places, that it could only have been as his heirs by the Welsh or Roman law of gavelkind: Eyton, having been equally misled respecting the family of Warin de Metz, whom he fails to recognise as the direct descendants of Warin the Bald; but a study of their lawsuits with the Peverils, extending over 100 years, proves conclusively that they were asserting the rights of their ancestor Warin against the Welsh usurpers of his estates. But the Fitzwarens also acknowledged the over-lordship of the Fitzalans, whose ancestors might have paid tribute to the Norman lords, as occasion made it necessary, but who generally asserted their descent from the Earls of Mercia, the ancient Lords of Whittington and Oswestry.

The MSS. open to the writer, except those of the Salusburies, gave no help on this subject. Griffith Hughes has no mention of Rhys Sais, nor has Peter

Ellis; and the Bodleian MSS. are also silent. Peter Ellis, indeed, in his pedigree of Tudor Trevor, adds one element of doubt. He, indeed, confirms the Salusburies' account that Ednoved ap Llowarch ap Lluddic ap Tudor Trevor married Jenett verch Rivallon ap Kynoyn, but he makes Rhys Sais not the son but the bother of Ednoved, whom he makes the husband of the impossible wife, Eva verch Griffith Hir ap Griff ap ye Argelwydd Rys; and to Tudor, son of Rhys Sais, he gives a wife, Jane, verch Rys Vychan ap Rys ap M'dd; and to him he gives two sons, Blethen and Grono; but obviously their date precludes the latter from being Ranulf Pever, of the time of the Conquest. Salusburies' pedigree makes Tudor only the father of a bastard son, Grono.

137

CHURCH OF ST. JOHN THE BAPTIST,
NEWTON NOTTAGE, GLAMORGAN.

BY G. E. HALLIDAY, Esq.

THE TOWER.

APART from its archæological interest, the Parish Church of Newton Nottage has a charm of its own, both in itself and its surroundings.

The old churchyard seems merged in the undulating golden-green sandhills stretching away to the mouth of the river Ogmore on the east-and to Porthcawl, a mile or so to the west. Beyond the sandhills lies the sea, with the Devonshire hills far away in the distance.

The church consists of a chancel, nave, western tower, and an unusually large south porch (Fig. 3), containing many good examples of thirteenth, fourteenth, but more especially fifteenth-century work.

There seems no evidence, however, of any remains belonging to the twelfth century being in situ ; although it appears to the writer that the bases of the fourteenthcentury porch entrance arch are in reality Norman capitals turned upside down, to suit the builders of that

time.

The tower, to which the writer more particularly wishes to draw attention, is a massive structure, in all about 54 ft. high-27 ft. from north to south, and 22 ft. from east to west, supported at its four corners by six exceedingly heavy buttresses. From its general appearance, and from the evidence of the early details still remaining, there is little if any doubt but that this portion of the building, at any rate, was used for defensive purposes.

The range of eight massive corbels projecting about 2 ft. from the eastern face of the tower wall, formed in