^{page 55 note 1} Archaeologia, lxxii, 51.

^{page 55 note 2} For this road, see Dawkins, W. Boyd, Archaeological Journal, lxi, 316, andGoogle ScholarCole, E. Maule, East Riding Antiquarian Society's Transactions, vii, 43Google Scholar.

^{page 55 note 3} Appleton-le-Street and Hovingham.

^{page 56 note 1} On the plan, original work which still remains is shown black. The lighter shading shows original walls which have been destroyed or rebuilt. Later works are dotted and unshaded.

^{page 56 note 2} Associated Architectural Society's Reports, vi, p. cxviGoogle Scholar. My friend, Mr. Walter H. Brierley, F.S.A., has kindly lent me the original drawings of the church by Messrs. T. B. and Wm. Atkinson, for Lord Middleton, dated July 1862. From these drawings it would appear that the rebuilding of the chancel was not contemplated when they were made. The works then carried out included, besides the rebuilding of the chancel and the addition of the south porch, the opening out of the western arch of the nave arcade and of the tower arch, which had been blocked up, the insertion of a window in the blocked west doorway, and new roofs.

^{page 57 note 1} The Northumberland towers of Corbridge (originally a porch), Bywell St. Andrew, Warden, and Whittingham are of similar size, and the walls of the towers of Corbridge, Bywell St. Andrew, Bolam, and Whittingham are 2 ft. 7 in. in thickness (Hodges, C. C., The Pre-Conquest Churches of Northumbria, in The Reliquary, vii (new ser.), pl. iv, opp. p. 85)Google Scholar. The tower of Restennet (Forfar) measures externally 15 ft. 6 in. from east to west by 15 ft. 11 in. from north to south, with walls 2 ft. 8 in. thick (Macgibbon, and Ross, , The Ecclesiastical Architectureof Scotland, i, 178)Google Scholar. The towers of several Wold churches (later than Wharram) have very similar internal dimensions, but, as their walls are thicker, their external bulk is greater. Frequently these towers are not quite square, but their greater dimension is not always in the same direction.

^{page 57 note 2} I use the term ‘Norman’ here to indicate the Norman Romanesque from the time of the Conquest to the end of the eleventh century.

^{page 57 note 3} There are capitals of this type in the west doorway of Kirkdale (Yorkshire), and in the tower of Broughton (Lincolnshire), illustrated in Brown, G. Baldwin, The Arts in Early England, ii, 98Google Scholar, fig. 49 b, and 213, figs. 128 and 129. The late J. T. Micklethwaite attributed Broughton to about 1050-60 (Archaeological Journal, liii, 335)Google Scholar, but the characteristically Norman base profile of a double hollow suggests a post-Conquest date.

^{page 58 note 1} On the side next the nave the abaci were originally returned a short distance beneath the outer order of the arch. On the south side the projection has been cut away (indicated by the dotted lines on figs. 2and 3); on the north side the projection remains.

^{page 58 note 2} i.e. not Norman of the Conquest period. The profile occurs however in one of the earliest churches in Normandy, in the impostsof the ground-story arcade of Saint-Pierre, Jumièges, which is believed todate from c. 940 (Lasteyrie, R. de, L'architecture religieuse en France à l'époque romane, p. 153Google Scholar and fig. 137; Gard, Roger Martin du, L'abbaye de Jumièges, pp. 26Google Scholar and 195, and fig. 25).

^{page 58 note 3} M. de Lasteyrie (op. cit., 319) speaks of the horseshoe arch as a survival of Carolingian practice, and adds—‘la Normandie est la province qui semble en fournir le plus d'exemples’. My friends Dr. Coutan and M. Louis Régnier, however, confirm myown observation that the horseshoe arch is by no means frequent in Normandy. It is true that the form is found at Bernay (crossing-arches, etc.), where however itseems to be simply a variety of stilting, as in some of the arches of the main arcades and triforium of Ely cathedral, and is much less pronounced than in the tower arch at Wharram.

^{page 58 note 4} The arch of the west doorway is constructed in the same manner (fig, 4).

^{page 58 note 5} Cf. G. Baldwin Brown, op. cit, ii, 127.

^{page 58 note 6} So also in the tower arch of Restennet (Forfar).

^{page 58 note 7} There is now no floor between the second and third stages.

^{page 59 note 1} The hollow here is only about 10 inches high.

^{page 60 note 1} The jambs of this doorway are not rebated, and its clear width is 2 ft. 9½ in.

^{page 60 note 2} I am indebted to my friend Mr. C. R. Peers for pointing this out to me, and also for kindly examining Wharram church with me.

^{page 60 note 3} Collingwood, W. G. in the Yorkshire Archaeological Journal, xxiii, 260Google Scholar.

^{page 61 note 1} Illustrated in Rickman (5th ed.), p. xxxix, and in the York volume of the Royal Archaeological Institute (1846), p. 18. Also by C. C. Hodges, op. cit., Reliquary, viii, 202. This tower shows no specially Norman characteristics, and may well have been built in the time of Edward the Confessor.

^{page 61 note 2} The north wall at the east and west responds of the arcade and the east wall at the chancel arch measure 2 ft. 9 in. in thickness including the plastering on each side.

^{page 61 note 3} At the north-east angle, the plinth of the fourteenth-century aisle is continued up to the north wall of the chancel.

^{page 62 note 1} There is no trace of this window in the rebuilt wall, but it is shown on Messrs. Atkinson's drawings of 1862 as a single narrow light, about 6 in. wide and 2 ft. 10 in. high including its semicircular head, with chamfered external jambs, and with wide splays internally to the jambs, arch, and sill. The position of the window is indicated on the plan (fig. 1).

^{page 62 note 2} Part of the quirk on the lower edge of the angle roll is visible over the east jamb.

^{page 62 note 3} In the Wold churches, the chevron seems to make its first appearance as decoration of a flat face, before it was used on the angles of arch-orders, etc.

^{page 64 note 1} Page, Wm., Some Remarks on the Churches of the Domesday Survey, in Archaeologia, lxvi. 101Google Scholar.

^{page 64 note 2} Domesday Book (Rec. Com.), i, 307 a.

^{page 65 note 1} It has generally been assumed that Nigel died shortly after the completion of the Survey, and that it was his son Robert who became tenant-in-chief of the lands which his father had held under the count of Mortain; but Dr. Farrer has shown that there is evidence which indicates that Nigel survived until about the year 1120 (Early Yorkshire Charters, ii, p. 326; cf. p. II).

^{page 65 note 2} Farrer, , Early Yorkshire Charters, i, nos. 26, 27Google Scholar; iii, no. 1439.

^{page 65 note 3} Ibid., ii, no. 1012. The grant, which is not later than 1129, is attested by William the treasurer (son of Herbert the chamberlain), the future archbishop and saint.

^{page 65 note 4} The name appears in various forms: Athelulf, Adalulf, Adelulf, Aiulf, Adelwold, Adelwald, etc. Freeman (Norman Conquest, v, 230) calls him Æthelwulf, and says that the name is sure proof of his English birth. Mr. W. H. Stevenson however kindly tells me that names in Adal- and in -ulf were fairly common in France in the twelfth century, and he thinks that the prior was more probably a Norman or a Frenchman, as he had been the confessor of Henry I (Robert de Torigni, in Chronicles of Stephen, etc., ed. Howlett, , iv, 123)Google Scholar.

^{page 65 note 5} Neve, Le, Fasti (ed. 1854), iii, 177Google Scholar. In February 1123, when William of Corbeil was elected archbishop of Canterbury, Athelulf was with archbishop Thurstan at the Gloucester court, and they spoke favourably to the king of the archbishop-elect; the messengers whom Thurstan sent to William about his consecration were the abbot of York, prior Athelulf, and ‘certain of our archdeacons and canons’ (Hugh the Chanter, in Historians of the Church of York, ed. Raine, , ii, 199)Google Scholar. Is it possible that the Bramham prebend had already been constituted, and that Athelulf was in attendance on Thurstan as one of his canons?

^{page 65 note 6} Mon. Angl., vi, 208, no. 1; Chartulary of Rievaulx (Surtees Soc., 83), nos. 216 and 347.

^{page 65 note 7} Mon. Angl., v, 280, no. 2.

^{page 65 note 8} Nigel Fossard and Walter Espec attested archbishop Thurstan's town charter to Beverley [115-28] (Farrer, , Early Yorkshire Charters, i, no. 95)Google Scholar. Walter Espec attested Henry I's general confirmation to the canons of Nostell [1121-27] (ibid., iii, no. 1428, a doubtful charter in its present form), as well as another confirmation [1123 or 1127] (ibid., iii, no. 1435). Archbishop Thurstan, Athelulf prior of Nostell, and earl David attested the king's letters of protection to Walter Espec's new foundation of Austin Canons at Kirkham [1122] (Farrer, , Itin. Hen. I, no. 478)Google Scholar. Athelulf prior of Nostell, and William treasurer of York and archdeacon of the East Riding, attested two of archbishop Thurstan's charters to the Austin Canons' house of Bridlington [c. 1125-33 and c. 1130-33](Early Yorkshire Charters, ii, no. 1151; iii, no. 1367). Walter Espec attested a grant by Henry I for a prebend for William, treasurer of York [1133] (ibid., i, no. 132). The dates within square brackets are those assigned by Dr. Farrer.

^{page 66 note 1} For more detailed description and illustration than is attempted here, see Macgibbon, and Ross, , The Ecclesiastical Architecture of Scotland, i, 185Google Scholar(plan, fig. 153, p. 186); Fleming, D. Hay, Handbook of St. Andrews (1910), 87Google Scholar; and Rivoira, G. T. in The Burlington Magazine, April, 1912Google Scholar. For illustrations, see also Sir Scott, G. G., Lectures, ii, figs. 203-5 (PP. 24–5)Google Scholar; and Walker, J. Russell, Pre-Reformation Churches in Fifeshire (1895)Google Scholar.

^{page 66 note 2} Professor W. Brown's description of St. Rule's in 1787 states that the foundations then visible of the ‘choir’ (as he calls it) indicated an internal length of 24 feet (Bibliotheca Topographica Britannica, no. xlvii, 206). The position of some foundation masonry discovered in digging a grave in 1892 (noted by Dr. Hay Fleming) has been thought to indicate an apse, but the evidence is scarcely decisive, and one would rather have expected a square east end.

^{page 66 note 3} As at Wharram.

^{page 66 note 4} The walls of the lowest stage appear to be 3 feet thick, which makes the internal measurement about 14 ft. 6 in. square, but the insertion of the modern staircase makes it impossible to measure this precisely.

^{page 66 note 5} Fleming, D. Hay, Proc. Soc. Antiq. Scot., 5th ser., i, 232Google Scholar

^{page 67 note 1} The inner order of this arch has the Norman profile of the outer order of the east arch of the choir, but the blocking of the opening makes it impossible to say whether it had the sunk soffit. The unmoulded outer order of shallow projection is found in some Yorkshire churches of the ‘overlap’ period: e.g. in the west doorways of Hovingham and Kirkby Grindalythe.

^{page 67 note 2} The absence of any window on the western face of the second and third stages might be taken to indicate that the addition of a nave had already been undertaken before these stages of the tower were built, but for the fact that each of the four stages above the ground story is lighted by one window —on the south to the second and third stages, on the west to the fourth stage, and on the east to the fifth stage. These four stages have no window to the north.

^{page 67 note 3} As at Wharram.

^{page 68 note 1} These nook-shafts are partly detached, and partly attached and bonded.

^{page 68 note 2} The profile of this order and the jamb plan are distinctively Norman, and prove that the church cannot have been built before the time when Norman influence began to affect the architecture of Scotland. These details alone sufficiently disprove the preposterously early dates to which some writers have assigned the church.

^{page 68 note 3} Some of the voussoirs are jointed in the normal way, but others have the peculiar jointing in vertical rings noticed at Wharram.

^{page 69 note 1} Dr. Joseph Robertson in 1849 seems to have been the first modern antiquary to date St. Rule's correctly (Quarterly Review, Ixxxv, 120).

^{page 69 note 2} In 1114, according to Bower's Scotichronicon (ed. Goodall, ), i, 286Google Scholar. In 1115, according to the Chronica de Mailros (Bannatyne Club), 65.

^{page 69 note 3} Liber Ecclesiae Sande Trinitatis de Scon (Bannatyne Club), I; Lawrie, Early Scottish Charters, no. 36. Lawrie (p. 280) regards this charter as spurious, though he admits that Alexander's foundation of 1114-5 is probable; but one of his reasons for rejecting it is certainly unsound, viz. that Athelulf did not become prior of Nostell until 1128, and that Alexander cannot have brought canons from Nostell so early as 1115 because it was not then a house of canons regular. See the Rev. James Wilson's article on the Foundation of the Austin Priories of Nostell and Scone, in The Scottish Historical Review, vii, 141, with his conclusion that there need be no hesitation in accepting king Alexander as the founder of Scone, and acknowledging prior Athelulf's co-operation with him in establishing the Augustinians in Scotland. The notification by archbishop Thomas II of an agreement between the monks of La Charité and the canons of St. Oswald (Farrer, , Early Yorkshire Charters, iii, no. 1465)Google Scholar proves that canons regular were established at Nostell before the archbishop's death in 1114.

^{page 70 note 1} Fordun (Chronica gentis Scotorum, ed. Skene, W. F., i, 227)Google Scholar simply says that canons were brought from Nostell. Bower's, Scotichronicon (ed. Goodall, , i, 316)Google Scholar says that six canons were brought from Nostell, with the consent of prior Athelulf, and that Robert was one of them.

^{page 70 note 2} Four months before his death, according to Symeon of Durham (Historia Regum, ed. Arnold, , ii, 275)Google Scholar, probably after the news of Eadmer's death (13th January 1124) had reached the king (Dowden, , The Bishops of Scotland, 4)Google Scholar.

^{page 70 note 3} Not later than 17th July 1127, when Robert issued his charter to Coldingham (Raine's, North Durham, appendix, no. 446Google Scholar; Lawrie, , Early Scottish Charters, no. 73)Google Scholar. The name of ‘Adulfus’ or ‘Adelof’ prior appears among the witnesses to the ‘without prejudice’ letters which archbishop Thurstan and king David exchanged at the time of Robert's consecration (Historians of the Church of York, ed. Raine, , iii, 51, 52)Google Scholar.

^{page 70 note 4} Chronicles of the Picts, Scots, etc., ed. Skene, W. F., 191Google Scholar.

^{page 70 note 5} ‘Ordinatus igitur episcopus, atque ad sedem propriam reversus, quod anhelabat in pectore, exercere studebat in opere, utecclesia, viz. ampliaretur, et cultui divino dedicaretur.’

^{page 71 note 1} Athelulf only resigned his priorate of Nostell shortly before his death in 1155 or U56 (Farrer, , Early Yorkshire Charters, iii, nos. 1473 and 1474)Google Scholar.

^{page 71 note 2} Liber Cartarum Prioratus Sancti Andree in Scotia (Bannatyne Club), 122; Lawrie, , Early Scottish Charters, no. 162Google Scholar.

^{page 71 note 3} See note 8, p. 65 supra.

^{page 71 note 4} Farrer, , Early Yorkshire Charters, iii, no. 1427 [1121-2]Google Scholar.

^{page 71 note 5} Mon. Angl., vi, 152, no. 5 (attested by Walter Espec).

^{page 71 note 6} Farrer, , Early Yorkshire Charters, iii, no. 1464Google Scholar(attested by Athelulf, bishop of Carlisle).

^{page 71 note 7} Powicke, F. M., Ailred of Rievaulx, 53Google Scholar.

^{page 71 note 8} Freeman, , Norman Conquest, v, 257Google Scholar.

^{page 72 note 1} I have to express my thanks to the friends who have kindly contributed the photographs which illustrate this paper; to Mr. J. V. Saunders for those of Wharram-le- Street, except the chancel-arch capitals, which I owe to Mr. J. W. Leedham; and to Professor G. Baldwin Brown for two photographs of St. Rule's. I have also to thank Dr. D. Hay Fleming for kindly giving me the benefit of his intimate knowledge of St. Rule's, and the Rev. William Dale, vicar of Wharram, for his kindness in affording me every facility for the examination of his church.