majority of settlements established by the saints in Wales are in low- lying parts of the country, on the lower valley-slopes, and on the valley- floors'.3 In addition, Sian Victory has reminded us that those churches standing today whose sites originated as early Christian burial-grounds can be recognised by the following features: their graveyards will be built up by layer upon layer of burials, and will be circular, or at least rounded in shape; early memorial-stones or stone crosses may have been found in the vicinity; and the churches will probably be dedicated to local, Welsh saints.4 Further, she writes, 'monks and clerics and anchorites would naturally choose to settle where fresh water was to hand.'5 In most respects Kilgwrrwg fits the pattern outlined by Bowen and Victory. The church, on its low eminence, is in the floor of the valley. The 'new church' [Newchurch] is high on the ridge above it. The churchyard, though not noticeably built up, is certainly rounded in shape, and still contains a pre-Norman cross, crudely carved from a solid monolith of local conglomerate, and which it is commonly accep- ted dates from the earlier part of the eleventh century. There is water, too; it is necessary to cross the stream which runs along the valley floor to reach the church, though there does not seem to be any local tradition of a 'holy well' associated with the building. The continuing isolation of the church re-inforces the argument for a Celtic foundation. Professor Bowen commented nucleation of settlement around a Celtic church is normally very restricted, or almost non-existent The chief characteristic of the Celtic churches is the isolated position occupied by a very large number of them in the settlement pattern of Rural Wales down to the present time.6 The documentary evidence for a Celtic foundation is inconclusive. E. T. Davies was content to accept the identification of the Cumcerruc mentioned in the Book of Llan Ddv as Kilgwrrwg.7 Wendy Davies, in her more recent scholarly analysis of the early Llandaff charters,8 is more cautious, though acknowledging the probability.9 She provision- ally dates the grant of three 'unciae agri pleni in medio Cumcerruc id est uillam quae fuit Guroc' to Bishop Berthwyn by King Ithel and his sons Ffernfael and Meurig as circa 722.10 No church is mentioned in this grant, which was of a substantial estate. Three unciae Wendy Davies calculates at approximately 1,500 acres," when the standard unit was some three modii, approximately 125 acres. If this estate was Kilgwrrwg, then Llandaff could be expected to have provided a church for the population there soon afterwards. A foundation date for Kilgwrrwg at some time within the eighth century is therefore not unlikely. Although Sian Victory commented that such foundations were often 'dedicated to local, Welsh saints',12 the original dedication of Kilgwrrwg church is unknown. There is absolutely no indication of the identity of the patron saint in either pre- or post-conquest documentary evi-