Part 11

The round towers, with one or two exceptions, are detached buildings, though situated within a few feet of churches, have their entrances at

several feet from the ground, and have jambs inclined towards each other upwards. They have other differences of construction, which, however, may be in part accounted for by the difference of the available materials.

The Irish tower which most resembles Egilsey is the smaller one of the two at Clonmacnoise. This is an integral part of the church, and joined to the south-east angle of the nave. The entrance to it is on the ground, from the chancel. To this church Dr. Petrie assigns the date of some years before 1000 (p. 271). In many particulars the tower at Egilsey approximates more closely to the round towers of Norfolk, which are of the twelfth and thirteenth centuries. There is no evidence of another church in another place having existed in the island; nor account of the _building_ of any church in the island.

These, however, are very slight evidences of the antiquity of the present church.

There does not appear to be positive evidence that the Christian priests were exterminated by the Norse conquest; though it is probable that they would fly westward. If, however, we give up the idea of its Norse origin, we ought to find resemblances between it and the ancient Irish churches of the eighth, ninth, and tenth centuries. These we do find. First, the size of Egilsey is close on the authoritative size of the more important of the ancient churches, presuming the present foot to be about equal to the ancient Irish foot. The authorised Irish size is 60 feet by 27 feet; Egilsey is 62 feet 9 inches and 21 feet 7 inches. The lowness of the chancel, the chamber or “croft” over it, the second chamber in the tower with an arch from it into the nave over the tower arch, the number and size of the windows, the peculiarity of their splays, having no exterior chamfer, the character of the masonry, all resemble the early Irish churches. On the other hand, three great criteria of an early Irish church are not found in Egilsey--the approximation of the jambs of doors and windows towards their imposts, the horizontal heads to the doorways, and the east window. It seems, however, probable that the difference of the materials induced one of these differences. It was difficult to get in Orkney lintels strong and long enough for heads of doorways, though we find them in the broughs. The rebates in the doorways are again of a very early date.

It seems on the whole fair to suppose Egilsey to have been built after the traditional Irish form, but with modifications; and soon after the reconversion of the islands to Christianity in 998. If built before that time, we must refer it to the beginning or middle of the ninth century.

Unfortunately Neale, when in Orkney, was unable to see this church, and has given a cut, professing to be an approximate likeness of the tower, which is very erroneous in several particulars. The etching in Muir’s “Sketch” is also wrong in proportion.

In the _Orcadian_ of January 1855 is an account by Mr. Petrie of the finding the remains of Bishop William “the Old” in the cathedral in 1848. His bones, and the chest containing them, were moved when the cathedral was reseated in 1856. The leaden plate bearing the inscription, and a bone article (doubtless the cross handle of his walking stick) found in the chest, are now in the Edinburgh Museum.

CHURCH ON BROUGH OF BIRSAY.

The brough contains about 40 acres, and is separated on the east from the mainland by a rocky channel, which is about 150 yards wide, and dry at low water. The surface of the brough slopes down from a high cliff on the west to a cliff of about 20 feet high on the east. The chapel is about 50 yards from the shore at the point nearest to the mainland. (Fig. 105.) It is enclosed in a yard about 33 yards east and west, by 27 yards north and south, of which the wall is destroyed. At the edge of the cliff are traces of a wall. The chapel consists of nave, chancel, and apse, all well defined, and all apparently built at the same time. (Fig. 106.) The material is grey whinstone, and no traces of freestone dressings appear. It stands nearly east and west, but the west end facing a little to the north of west.

The extreme exterior length is 57 feet, and extreme width 21 feet 3 inches.

The west wall is 3 feet 8 inches thick; the north and south walls of nave and aisles 2 feet 9 inches; the wall of apse 2 feet 4 inches. There were no buttresses.

The interior is filled with debris to the depth of about 2 feet 6 inches above the floor, and the exterior to about the same depth. Partial excavations were made for this plan. Of the west end, only 3 feet in height of wall remains (Fig. 107); of the north wall (Fig. 108) a little more, and of south wall a little less; of the north wall of chancel, 8 feet 6 inches; of south wall of chancel, about 2 feet; of the apse, about 2 feet 4 inches. The nave is 28 feet 3 inches by 15 feet 6 inches inside. The only entrance to the church is in the west end, 3 feet 8 inches wide. The jambs are parallel, without any chamfer, and there is no rebate for a door nor appearance of hinges; and original plaster remains on the jamb down to the sill and all across it. There is no appearance of the insertion of a wooden frame. This mode of making jambs of doorways is to be seen at Lybster in Caithness, Wyre, Lynton in Shapinsay, Uya in Shetland, and in some of the oratories in Ireland, &c.[131] Were there doors in these doorways,

and, if so, where placed and how hung? It is known that in many cottages in old times the door was an animal’s hide hung across the opening, and probably this may have been the case in these unrebated church entrances. Possibly this plain jamb may be an evidence of antiquity; though Egilsey and Kirk of the Ness in Yell have the usual form of jamb. Probably this doorway was surmounted by a semicircular arch, though many of the ruder chapels had square-headed doorways with a large stone lintel, as at Lybster, Caithness. There are very few instances of the passage between a nave and chancel being square-headed. Where there is a semicircular arch on plain jambs, as in the doorways beforementioned, and in some chancel arches, as at Linton, Shapinsay; Kirk of the Ness, Yell; Wyre, &c., there is generally a peculiarity which may be mentioned here--the feet of the arch are set back on the jambs at the imposts two or three inches on each side. This was probably to support the centre on which the arch was built, instead of supporting it by props from the floor.

In the north-east and south-east corners are two circular spaces 5 feet 6 inches in diameter (see Fig. 108), the south one of which contains a freestone stair step, and the north one some broken stone. Probably both were staircases; but how high they reached, to what they led, and what was the superstructure we can only conjecture. One or both may have led to turrets, or to priests’ rooms over the chancel arch, or chancel or nave. We cannot suppose a rood-loft to have existed here. The support of the west side of the stairs is now gone, but we must suppose that a block of masonry existed which included the circle of the staircase. Against this, the _seat_ on the south apparently runs past the spot at which it should have stood, and there are no decisive traces of a junction in the north and south walls. There is no other solution of the difficulty. This very massive separation between nave and chancel is probably unique. M‘Cormac’s chapel at Cashel has a turret and a chamber in nearly the same positions as these staircases; but the construction in that case is clear.[132]

A stone seat, 1 foot 2 inches high and 1 foot 2 inches wide, ran all round the nave--at least, it was traced at west doorway and at the east end of the south wall of the nave. The roof was probably of tie beam construction, and covered with stone slabs.

The entrance to the chancel is 4 feet 3 inches wide, of which 4 feet in height of the jambs remain. They are 3 feet 7 inches thick from west to east, but if the construction was that suggested above, the jambs were 7 feet from west to east, including an entrance into each turret 2 feet wide. (Fig. 109.) The entrance to the chancel was probably surmounted by a semicircular arch of whinstone, plastered, about 8 feet or 9 feet high to the apex, with string-course caps 4 or 5 inches deep. The angles were not chamfered.

There is a step in this entrance of only 2 inches, and from that to the east end of the apse was flush originally. The pavement is of flagstones. This low position of the original altar is peculiar.

The chancel is 10 feet 9 inches east and west, and 10 feet 3 inches inside. Only one window remains, which is in the north wall of chancel, and this has lost its lintel (Fig. 110). The clear opening was 3 feet by 10½ inches, with, as stated, a square head. The window has an internal and external splay, each widening to 1 foot 10½ inches. This form is peculiar. There was probably a frame, fixed in the narrow part of the opening, containing glass. In this wall, just east of the window and lower, is a square ambry, 3 feet high by 2 feet 8 inches wide, and 1 foot 11 inches in recess. The use of so large an ambry is not evident. The bottom of it is only 2 feet above the floor. There is no appearance of its having had doors. In the north-east part of the nave of Enhallow Chapel is a similar recess, and one in the same position as this in the chapel on the Brough of Deerness. It is possible, but not probable, that these were Easter sepulchres. In the Kirk of the Ness, North Yell, is a larger one. Of the south wall of the chancel so little remains that nothing can be stated as to window piscina or sedile.

It is remarkable that the floor was originally level to the end of the apse, and no elevation given to the altar, which was probably at the chord of the apse; but in later times they built a reredos which _blocked off_ the apse, and then they appear to have made steps to the altar. Some part of the altar remains. It appears to have been 4 feet 1 inch by 2 feet 7 inches. Possibly the reredos did not reach many feet high, so that the upper part of the apse appeared over it.

As there were no buttresses, it is not likely that the chancel was vaulted. The roof was probably of tie-beam construction, or of rafters coupled half-way up, as in cottages, and covered as the nave roof. It is probable that the apse was vaulted, and the arch into it may be conjectured to have been 10 feet high, and semicircular, and to have formed part of a plain vault, as at Orphir and St. Margaret’s Chapel at Edinburgh. The ground plan is somewhat in the form of a horse-shoe internally. Probably there was a small window in east end, and this must have been at a lower level than the one on the north.

The interior of the nave is of the proportion of the vesica piscis. The width of the chancel is less than of the nave by the thickness of the side walls. The interior of the chancel is square. It is stated in Barry[133] to have been dedicated to St. Peter, and to have been a place of pilgrimage. The writer was informed that many years ago a skewer was found, with a skull, outside the chapel. This was probably the pin of a winding sheet.

There can be little hesitation in assigning this church to about 1100. It is recorded that Earl Thorfin built Christ’s Church in Birsay, and that the body of St. Magnus was buried there. It appears probable that if St. Peter’s had been built as long after his death as his canonisation, it would have been dedicated to him. Even if it had been built in memory of St. Magnus before his canonisation, and dedicated to some other saint, probably the dedication would have been changed after his canonisation. As it is recorded that Thorfin built Christ’s Church, it is probable that if he had, St. Peter’s would have been recorded also. It is more probable that it was built by his second son, Erlend, the father of St. Magnus.

This church bears considerable resemblance to Old Bewick, Northumberland, figured in Muir’s “Sketch”; to Moccas and Kilpeck in Herefordshire, and to St. Margaret’s Chapel, Edinburgh.

Mr. Leask of Boardhouse, in 1867, made an excavation in the church, to determine, if possible, whether foundations exist of the walls supposed to have enclosed the west sides of the circular recesses. He found none and found plaster on the north wall of the nave, where the supposed wall should have joined. Possibly the supposed walls were taken down during the time the chapel was used. He found in the north wall of the nave, at 5 feet west of the angle of the recess, an entrance 2 feet 4 inches wide. The position is very unusual. Planned 1866.

CHURCH AT ORPHIR, ORKNEY.

This highly interesting fragment stands near the east end of the parish church, and probably the reason why it has not obtained that notice which it deserves is, that the larger and most interesting part of it was destroyed before 1758, to build or enlarge the present parish church. It consisted originally of a circular nave, and apsidal chancel added to its east part. (Fig. 111.) The chancel remains, but only 9 feet on each side of it of the circular nave. Fortunately we have a short record of its size and form in Sinclair’s _Statistical Accounts_, xix. 417, quoted in Wilson, _Prehistoric Annals_, 598. It is there stated:--“In the churchyard are the remains of an ancient building called the Girthhouse, to which great antiquity is ascribed. It is a rotundo, 18 feet in diameter and 20 feet high, open at top; and on the east side is a vaulted concavity, where probably the altar stood, with a slit in the wall to admit the light; two-thirds of it have been taken down to repair the parish church,” &c. In the translation of Torfœus by the Rev. A. Pope is this note by the translator (p. 108), who visited Orphir in 1758:--“The temple of Orphir, or Gerth-house, was a rotundo 22 feet in diameter, and 61 feet perpendicular wall above ground. The cupola, with the open for the light, was of an elegant cast, and the light was all from the open, and lighted the house sufficiently. There was a small slit in the east side for light to the priest, who stood in a niche elegantly done. The work was very firm; and though at that time there was a breach made in order to get stones to repair the parish kirk, yet the stones crumbled to pieces before they could be loosed,” &c.

In the _Orcadian_, July 1861, is an account of this church by Mr. G. Petrie. It is there stated that the present parish church was erected in 1829. (See also his notice of the ruins, in the _Archæological Journal_, 1861, No. 71, p. 226-230.)

It is evident that the east wall of the parish church could not have co-existed with the old nave, for the new church stands on part of it.

From these several statements we must infer that part of the old church was pulled down before 1758 to build or enlarge, not to repair, the parish church, and possibly a farther portion pulled down in 1829.

We must understand the “cupola” to be a conical roof, and “the open for the light” to be a glazed lantern on it, which, however, was probably not original. (Fig. 112.) “Open at top,” in Sinclair, must refer to this lantern. The framework of it must have been of wood, as at Ely.

The curvature of the two parts of the nave wall still remaining gives an interior diameter of about 19 feet, thus corroborating the diameters given by Pope and Sinclair. Neither of these persons had seen the church perfect; and the differences in the measures given by them show that their informants did not speak from exact data. The “61 feet perpendicular wall” is a gross mistake either of printer, or translator, or his informant. Even the “20 feet high” seems over the mark, though this may be a loose approximation to the height of the side walls.

Supposing the preceding conjectures as to the shape of the roof to be correct, the top of the “fleche” surmounting the lantern would be about 40 feet. Then, supposing that Pope was told that the height was 41 feet, and the height of the side walls 20 feet, he might have added them together.

The diameter was less open to error than the height. One writer, however, might give the outer diameter and the other the inner.

It is built of yellow Orphir freestone. The nave walls are 3 feet 9 inches thick, well built; and probably the entrance was at the west, as in the other round churches.

It is highly improbable that Pope is right in stating that “the light was all from the open”--that is, that there were no windows in the nave walls. Probably there were four single lights.

The chancel, which is little more than the apse, is 7 feet 2 inches wide and 7 feet 9 inches deep, or long, with wall 2 feet 8 inches thick. The arch into it is semicircular, and forms part of a plain unribbed vault, as at St. Margaret’s Chapel at Edinburgh. (See Fig. 111.) The impost of the vault arch is 6 feet 5 inches above what appears to have been the floor, and the top, consequently, 10 feet 6 inches high. Outside the vault was originally probably a solid stone roof, the apex of which was about 14 feet

from the floor. This height (supposing also that there was a step at the chancel arch) would oblige the walls of the nave to be about 15 feet high. The chancel has no buttresses. There is one window in the chancel, in the east end, 2 feet 5 inches by 10½ inches clear, opening with jambs splayed inward to 1 foot 8 inches wide. (Fig. 113.) The outer edges are chamfered, and the head semicircular. The impost is at the same level as of vault. It has a groove for glass.

A stone lying down appears to be part of a stoup.

The exterior width of the chancel is half the exterior width of the nave.

On the south side of the nave is the trace of a building in the form of a parallelogram, and other indications of buildings. Pope states that before his visit in 1758, large foundations had been found in digging earth for the Bow of Orphir, near the Gerth-house.

The “four round churches” of Britain are--Cambridge, consecrated in 1101; Northampton, probably shortly before 1115; Maplestead, 1118; and London, 1185.[134] To these must be added the small Norman chapel in Ludlow Castle, and we may safely add, as a sixth, Orphir. The church of the Holy Sepulchre at Jerusalem was the type of all, and the Crusades were the means of importing the form into Britain. In conjecturing the date of Orphir, we may take into account that there was a palace of the Earl Haco (son of Paul I.) at Orphir, according to Torfœus. He went to the Holy Land and back, and died in 1103. The same writer describes the palace of Earl Paul II. at Orphir, and states that a temple stood opposite the wall of the palace. This earl died about 1136. If this statement is to be believed, this must be the church mentioned, and hence we get 1090-1137 as the limits within which this church was built.

Earl Ronald founded the cathedral about 1138, and soon after went to the Holy Land, accompanied, of course, by some of his high officers. He died in 1158. We may fairly conclude that this round church had some connection with one of these expeditions. As Earl Ronald needed all his money for the cathedral, we cannot suppose that he had any hand in building Orphir; but it is not improbable that some one of his wealthier followers built it, and we may, from the dates of the other churches and these facts, fix on 1090-1160 as limits of the date of its erection.

In the Edinburgh Museum is a stone article, presented by D. Balfour, Esq., found at this church. It is circular, 4¼ inches diameter, ¾ inch thick, with a small square hole in centre. Its use is not evident. Planned 1855.

CHURCHES IN SHETLAND.

THESE ARE ALL CHANCEL CHURCHES EXCEPT CULBINSBROUGH, WHICH IS CRUCIFORM.

CANONS ASHBY, _Feb. 21, 1868_.

SIR,--Last year you inserted in the _Orcadian_ accounts by me of some of the ruined churches in Orkney; I now send you accounts of some of those in Shetland. Every year destroys some portion of these ruins, and of the most complete in Shetland a considerable portion has fallen since I made the plans of it in 1855. Though the owners of these humble ruins do not value them, they have a value with historians and antiquaries as examples of the ecclesiastical architecture of the North in its least elaborate state, and evidences of the religious condition of the islands in past times. We find no St. Magnus and no Egilsey in Shetland; but still the ruins which exist there are worthy of record.

The number of churches which once existed in Shetland is extraordinary. For instance, Unst, which is about 6 miles by 3 miles, had at least twenty-four. Of most of the Shetland churches every trace but a name or a tradition is gone; of others only a heap of stones remains; of others a few fragments of wall enable us to make approximate plans, and one only is tolerably complete. A little labour with pick and spade would enable us, in many instances, to determine more about them than is possible in their present state. No people ought to be more impressed with religion and the uncertainty of life than the Shetlanders. The effects of the storms on sea and land, the terrible tideways among the islands, and the barren nature and exposed position of much of the land, make life a continued struggle for safety or for food. Doubtless the tradition concerning nearly every church was true of some--that they were built as thankofferings for escape from an ocean grave. The churches of which I send an account are Culbinsbrough, Bressay, Noss, West Sandwick, Yell, the Ness (Yell), Uya, and Norwick, Haroldswick, Kirkaby, and Colvidale, all in Unst.

I have to acknowledge the assistance of Mr. J. T. Irvine, formerly of Yell, and of Mr. Sandison of Uya, and to thank many friends in Shetland for hospitality and information.--I remain, your obedient servant,

H. DRYDEN.

CHAPEL OF NOSS, BRESSAY.

This ruin is near the sound which separates Noss from Bressay, on an elevation of about 20 feet or 30 feet above the water. The dedication is unknown.

It consisted of a nave and chancel (Fig. 114), of which only a fragment

of the north wall of the chancel remains, about 4 feet or 5 feet high. The nave was about 18 feet 6 inches by 14 feet, and the chancel about 12 feet east and west, by 10 feet north and south, as shown by the ridge in the turf.

Nothing certain can be stated of the door or chancel arch. Planned in 1852.

KIRKABY, WESTING, UNST.

This is in the west part of the island, in a walled churchyard, about 80 feet north and south, and 60 feet east and west. By remains outside the south-east part of the wall, it is evident that the church has been