^{page 283 note 2} Cf. Raine, , op. cit. 217.Google Scholar

^{page 283 note 3} Even in the translation of 1104 only the ends of the stole had been exposed: cf. the evidence of Reginald of Durham ( Raine, , op. cit. 88)Google Scholar.

^{page 283 note 4} I cannot regard it as probable that the eleventh-century sacrist, Alfred Westou, put the cross into the coffin; for this man see Raine, , op. cit. 58Google Scholar.

^{page 284 note 1} Only a tiriy portion of the original shell is now visible; the exposed ring has, however, been completed with dental plaster.

^{page 284 note 2} I have to thank Mr. Herbert Maryon of Armstrong College, Newcastle-on-Tyne, for much help in connexion with the technical description of this cross. Mr. Maryon very generously made special visits to the British Museum, when the cross was in our keeping, in order to study the jewel.

^{page 285 note 1} Dr. Kitchen, says (V.C.H. Durham, i, 254)Google Scholar that under the existing loop ‘may be discerned a thin loop in gold wire, worn through and replaced’. He was, no doubt, referring to the torn ends of the beaded wire on the arm of the cross.

^{page 285 note 2} It was presumably necessary to remove one or two of the garnets in order to fix these rivets.

^{page 286 note 1} It seems to have been provided with a shank that passed through a hole punched for the purpose (b on fig. 2) and was then hammered down. There was probably a corresponding hole (e on fig. 2) that was made for the repairs to the damaged cloison on the other side.

^{page 286 note 2} Dr. Raine says that the arm was ‘found broken off’ and that ‘upon examination it appeared to have been broken once before, as there were evident proofs that it had been repaired by means of small rivets, some of which were remaining’. I take it to be quite clear from this that the binding-strip was not in position when the cross was found, as otherwise it would have been mentioned as the most obvious sign of the repair to which he alludes. Raine's engraving shows the cross unbroken. There is, however, in the Conway Library another very old engraving that shows the strip in position. I have not been able to trace the work in which this appears, but it is certainly not much later than Raine's engraving and suggests that the modern repair took place very soon after the discovery in 1827.

^{page 288 note 1} An example in London of the ‘Mediterranean Gothic’ jewellery to which I am referring is a fifth-century gold buckle from Spain in the British Museum, Zeiss, H., Grabfunde aus dem spanischen Westgotenreich, 1934, pl. 7, 3.Google Scholar

^{page 288 note 2} In the Continental work of the type of the buckle to which I referred in the preceding note the rivets function by fastening the base-plate to the main body of the jewel. But in a handsome jewel of the same cloisonné group that was found at Milton, Kent, the base-plate and sides are made in one piece and the projecting rivets are purely ornamental, British Museum Guide to Anglo-Saxon Antiquities, 1923, pl. III, 3Google Scholar . So far as I know at present it is only in England that we find this use of the cylindrical projecting rivet as a vestigial ornament.

^{page 288 note 3} e.g. Allen, Romilly, Early Christian Monuments of Scotland, Edinburgh, 1903 fig. 532.Google Scholar

^{page 289 note 1} Journ. Brit. Arch. Assoc. viii (1852), 139Google Scholar . Note that it is there erroneously said to have been found at Lakenheath.

^{page 290 note 1} Under the guard on the hilt of Childeric's sword there is gold with a similar incised zigzag pattern.

^{page 291 note 1} I have to thank Mr. Leeds and Mr. Harden for giving me the opportunity of making a close examination of this jewel.

^{page 291 note 2} An excellent example of a mutilated jewel derived from an earlier archaeology; see Smith's, Roachcogent remarks, Collectanea Antiqua, iv (1857), 163Google Scholar.

^{page 291 note 3} Leeds, , Early Anglo-Saxon Art and Archaeology, Oxford, 1936, pl. xxxiii, 4.Google Scholar

^{page 292 note 1} A most important matter that is usually overlooked. Lombard archaeology, for instance, would be better understood if we took into account the fact that both the ‘fine style’ cloisonné brooches from Castel Trosino (Graves, 168, 175) have lost all their stones. Similar instances occur in Kent, cf. Antiquity, vii (1933), 431, and pl. v, opp. 448.Google Scholar

^{page 293 note 1} Acta Archaeologica, vii (1936), 63, note 17Google Scholar . It is impossible to argue with any one who honestly thinks that the Kingston brooch must be of the same date as these pieces in the Wieuwerd hoard. Dr. Werner is an astute and experienced archaeologist; so there is nothing for me to do but to record that he does consider the best Kentish work and these miserably degenerate things to be stylistically identical

^{page 293 note 2} The two cloisonné pendants in the Wieuwerd hoard can be matched by Kentish pendants which are different from and inferior to the ‘Kingston’ type of cloisonné. We know on unassailable evidence that the ‘Kingston’ cloisonné cannot be put later than the Wieuwerd type pieces, because we have examples of the subsequent work and can see that it is of other kinds. Therefore the better work must be earlier (as, indeed, the pattern-sequence suggests); which brings us to the early sixth century.