Letter written by Sid Bradley to Don Brothwell after the first summer’s dig in 1968 and distributed to the 1969 cohort as background information on the Newark Bay site. Edited and reproduced here with his permission. Tentative conclusions are that we have found the site of a fortified manor-house built in the 16th century or earlier and deserted in the latter half of the l8th century. The house was built over a forgotten burial ground which was almost certainly that of the pre-13th century chapel of the administrative districts of Watland and Southend, and which would probably have stood to the north of the burial ground. The Chapel and Burial ground The Royal Commission reported: “Chapel, Newark. Local information is to the effect that a chapel and burial ground formerly existed close to the beach at the farm of Newark. Some confirmation of the tradition was observed in the banks along the shore, where human bones are constantly being exposed. As yet, however, no traces of a building have emerged.” (1930) The assumption that there was a chapel for Watland (the still current name for the wet land around Newark) in the early middle ages is made in the Introductory Notices to the (modern) edition of the Statistical Account 1791-99. There it says that as late as the 17th century the parish of Deerness was divided into six eyrislands which probably represented the divisions made in Norse times, for purposes of taxation. It notes that curiously enough, though these districts were so well defined, the small ancient chapel which usually stood one, or occasionally two, in each ursland, only numbered two altogether, Kirbister and Newark, apart from the parish kirk in Sandwick. (This finding is accredited to “the late Mr Magnus Spence - the best authority possible”. I haven't yet found out anything about him.) It is then assumed by the editor that the chapel in Watland must therefore have served two urslands, those of Watland itself and of Southend (which includes Delday, the farm next to Newark, along the shore towards Kirkwall). The editor takes it as a fact that the chapel stood on the site at Newark. In The Northern Isles, Wainwright is cautious about accepting the general proposition made by Marwick et al.,that the medieval parishes were established on the ancient boundaries of the Pictish land-divisions, which the Norse earls had simply taken over as the boundaries for their tax-areas. But he does agree that there is sufficient evidence, both in Orkney and in other Viking settlement areas, to accept the argument that every ursland (Wainwright calls them ouncelands) had its own chapel. He says, following Marwick: “The ounceland chapels should represent the first spread of Christianity in the islands in the late eleventh and early twelfth centuries under the influence of the minster established at Birsay by the earl himself; they were erected by the local landholders to serve their estates.”(p.181) Furthermore, some of these chapels would no doubt have Celtic origins, as a few of those already excavated can be shown to have. The chapel on the Brough of Deerness is Celtic in origin, as is demonstrated by the presence of monastic buildings, upon which Norse building has been superimposed as at Birsay. The chapel is a simple single-chambered building, of short proportions. Of a similar chapel at Marwick in the parish of Birsay, Wainwright says: “Its layout may be regarded as typical of the original form. The small rectangular chapel, measuring 17 feet by 12 feet internally, has the typical short proportions of the Celtic period ... ” (though the presence of mortar and other features suggest that the chapel is actually Norse built in the style of the existing Celtic chapels). St Tredwell's Chapel on Papa Westray is another example, probably a genuine Celtic chapel with Norse additions to the site. “Simple sites of this type cannot be closely dated without excavation, but the type does not belong to the full medieval period, when the parish church became the normal type of ecclesiastical building. Moreover, a simple chapel of the type revealed at Marwick is typologically older than developed Romanesque buildings like the chapel on Wyre and others which belong to the middle or second half of the twelfth century.” (Wyre chapel is of longer proportions, and has nave, chancel, chancel screen and west door in the Romanesque style). If the site at Newark is indeed that of an early chapel, then, there are two possible termini a quo: the period of the first Norse Christianisation, and the period of Celtic Christianity - that is, about 1100, or any date from about 575 onwards. Further excavation could settle this. If, for example, the wall in section 11-14 (which is sited to the north of the burials so far found - the position in which a chapel would be generally expected to stand) which is below the level of the main building turns out to be a separate building, its shape and orientation would give some idea of its eligibility as a chapel, and its manner of building (which is strikingly reminiscent of the late Celtic additions to the Broch of Gurness, consisting of slabs laid on their edges with infilling of freestone) could determine whether it is typical of Norse or of Celtic building. We should look out for traces of any further building which might represent the monastic complex typical of Celtic sites, and for enclosure walls, whose shape might also offer information. (“The rectangular enclosure of the church-yard at Birsay shows the Norse type of cemetery. A rectilinear layout is foreign to the earlier Celtic church.” Wainwright, p.180.) The rectangular wall at the Brough of Deerness, like that at Birsay, can be shown to have been superimposed upon the original Celtic layout by the later Norse occupants. The terminus ad quem may be tentatively fixed in relation primarily to the reorganisation of ecclesiastical administrative areas in Orkney, which took place around 1200, and secondarily to the erection of the manor house of Newark over the burials during the 16th century. About 1200, the parish system was introduced in Orkney, replacing the ursland-based system of chapel distribution. According, to Marwick it was the usual process for the chapel of the ursland in which dwelt the most prominent family to be elevated to the status of parish church. In Deerness the chapel serving Sandwick, situated at Skaill, was the favoured one. The chapels thus superseded would have fallen gradually into disuse and disrepair, though Marwick suggests that the burial grounds might well have continued in use for some time. But it is fairly obvious that no-one would knowingly build a manor house over a consecrated burial-ground, and one must assume that by the date of the building of Newark the existence of the chapel and burial-ground on that site must have been forgotten. However, the field adjacent to the site, on the opposite side of the cart-track still preserves a memory of the chapel in its name Kirkady, the kirk marsh. Local tradition also has it that the site is that of an old chapel or church, though how far this tradition is genuine or whether it stems from relatively recent, antiquarian writings is hard to establish. Part of it may be due to the appearance of bones in the cliff face of which we heard instances quoted dating back forty years or so. But the elaborate story about the intention of building a new church on the site and the storm which carried the raft of timber round to Skaill where the church was finally built suggests some tradition more substantial. The tradition of a shipwreck, or alternatively of a great sea-fight off the coast, accounting for the bodies, may represent a rationalisation of the appearance of bones in the cliff-face after storms, influenced by the story of the battle off Deerness between Thorfinn and the king of the Scots as described in the Orkneyingasaga. It is possible that the appearance of masonry in the cliff face also influenced local tradition, though this appearance is a recent development, since the Royal Commission reported in 1930 that no signs of any building were to be seen near the shore at Newark, and Mr Delday the owner of Newark says that he can remember when there was space enough between the wall of his field (most of which has now fallen to the beach) and the cliff-edge for carts carrying sea-weed to drive along. He agreed that there was no obstruction in this pathway, and so it seems that the existing masonry, which projects above the level of the field, must have been incorporated in the wall, - and it is indeed in line with the surviving remains of the wall. To sum up on the chapel and burial-ground: tentative conclusions are that the excavated skeletons are those of the inhabitants of Watland and Southend up to perhaps, 1300.  The chapel which seems to have stood on the site may have been a Norse foundation or a Celtic foundation, dating, that is, from 1100 or earlier. Examples of both kinds are available for comparative study in Orkney, but further investigation of both types is needed. Some speculations: interesting features of the area not so far tied in with the idea of a chapel (apart from the masonry of the manor house) are the tunnel leading from the cliff face into the bank to the west of the site, the mound at the top of the hill slope, near the farm buildings, and the stone patterns in the sand quarry to the north of the farm buildings. Two other notable facts are that the site lies in an outcrop of heavy boulder-clay in an area otherwise sandy or boggy and that the foundations of the manor house are buried surprisingly deeply under black earth, stone and sand. If, as seems most likely, the burial ground predates the house, and the tunnel predates the burial ground, at least three inhabitation periods are postulated: four, if the chapel should prove to be Celtic. This argues a certain suitability of the site for human occupation, and the features attracting habitation may be the strategic advantage afforded by the natural defences of the site - on one side, the sea-cliff; on two others, the marshes which gave rise to the still current names Watland, Kirkady, etc. Examples exist in Orkney of the use of similarly protected sites: the Brough of Deerness, St Tredwell's chapel on Papa Westray, and of course Birsay. In the Norse period and afterwards, habitation sites were partly chosen for their security against surprise attack. The choice of the Newark site for a 16th century fortified house (see below) suggests its suitability for military defence; so perhaps it can be assumed that the site would also have commended itself for the same reasons to Norse settlers. In the Celtic period, similar geographical situations wore sought - but not solely for physical protection against attackers. “On the Brough of Deerness ... there is an excellent example of a Celtic Monastery ... The site lies on a projecting headland, separated from the main plateau by a deep cleft or geo. The only approach is by a steep path up the landward side of the headland where the slope is crowned with a ruined wall, which isolates the flat top. This wall forms the vallum monasterii, so often recorded in the texts; it served not only to enclose the monastic city, but as a symbolic barrier, excluding the world.” Wainwright, p.166. Even more suggestive of features we might look for in future digging is Wainwright's description of the settlement pattern at St Tredwell's Chapel. He says that there is a standard type of religious enclosure known in Britain from Celtic times, consisting of a burial-ground marked as hallowed only by a cross, having no chapel. These date back to the earliest period of the conversion. Later, chapels were built in them, often over the top of earlier graves, and later still some of them became the sites of the medieval parish churches. He goes on: “St Tredwell's Chapel ... is a good example of the type. The site is a small mound, now a peninsula; it was probably an island in early. Christian times. The oldest structure, a curved wall with a pronounced batter, is probably the base of a broch; and the souterrain, now largely blocked, must be associated with this secular occupation. Later than these structures are traces of a wall apparently surrounding the base of the mound. Within this enclosure are two small rectangular buildings. The more prominent, measuring 29 feet by 22 feet overall, is stated to have been dry-built with lime pointing on the faces of the walls. It closely resembles small chapels of the early Norse period, such as that at Marwick... To the northwest and occupying the summit of the mound is a rather smaller more ruined and apparently older building with no recognizable detail... Careful examination of the ground might produce further examples, but remains of certainly Celtic type are rare. The explanation is probably historical... Norse chapels were built over older, Celtic foundations ...It is probable that many of the Orkney chapels, though they have lost the Celtic names, are nevertheless Celtic in origin.” (p.171-2). By this example, one could speculate whether the Newark tunnel is in fact the entrance to a souterrain of Pictish origin formerly associated with a broch (the mound at the farm? the profusion of stones in the ground?), occupied in Celtic times (the style of the wall in section 11-14?) and made the site of a burial ground and chapel, whether the depth of earth above the house walls and above the tunnel (NB the depth of the burial in the tunnel much deeper, apparently, than those under the house - unless the ground level has subsequently been raised) may have resulted from a mound (the roof of the souterrain? ruins of the broch?) being ploughed flat after the ruin of the house, whether the traces of building in the sand-pits may have a monastic origin - cells, or remains of the vallum monasterii? Only speculation, but an attempt to relate what has been found, and to guide next year's digging programme. An aerial photograph of Newark would probably be illuminating. The Manor House of Newark Evidence enabling location for Viking settlements in the Orkneyingasaga is scanty as far as Deerness is concerned. Deerness is named in the account of the sea-battle in which Earl Thorfinn defeated the Scots-king in the mid-11th century, and in the same account it is implied that Sandvikr (Sandwick), where Thorkell fostri lived, was by Deerness (which seems to be the name of a particular headland under which Thorfinn anchored). Thorkell's home is earlier described as being “í Hrossey í Sandvík á Hlaupandanesi”, on the Mainland, at Sandwick on Hlaupandaness. It is here he murders earl Einarr. This is the place hopefully identified with Skaill and the Birmingham dig, on the assumption that Hlaupandaness refers to that part of the modern Deerness coastline. Only one other farm is mentioned in the context of Deerness, Skeggbjarnarstadir, home of a man called Gauti. The name has been lost, and the site remains, so far as I know, unidentified. The name Newark is typical of place names given in the 17th century in Scotland, but the site has an older Norse name Maill, meaning sand. It is found in various spellings, Maile, Male, Meal (and Mel- as in Melsetter). In 1591 Earl Patrick Stewart granted lands in Deerness to Lord Lindores, including Meal, with “the new houses called the Newark”. There may be an implication here that Meal was an older building or complex, to which the “new houses” were added in the latter 16th century. Anyway, the house seems to have had some importance in its time. It was strongly built, as implied by the description in a charter of 1673 of the “tower, fortalice, and manor place of Newark”; and it was occupied in 1653 by Colonel John Stewart, one of the chief Deerness landowners of the time. The house was still standing in 1774 when the Rev George Low made his tour through Orkney. He wrote: “The only Gentleman's seat now in this parish [ie Deerness] is that belonging to Mr Covingtree of Newark, built by John, Lord Kinclaven, Earl of Carrick. Several other seats, now in ruin, are yet to be seen...” A sketch-map in his book shows the house situated near the shore in the eastern half of the Bay of Newark, roughly on the site of the existing farm. Towards the end of the century the house was used as the meeting place for the kirk council whilst the church at Skaill was in a state of serious disrepair. But by about 1790 the house had become a ruin. The Statistical Account (pub. 1791-99) says: “There were formerly several gentlemen's seats in the parish, but they are all now in ruins. One of them, which belonged to the Coventries of Newark, was built by John Lord Kinclaven, Earl of Carrick”. The ruins were still visible in the middle of the 19th century, when they were plotted and marked as “An Old Ruin” on a map of the Balfour and Zetland estates in Deerness, which showed roads, field boundaries, shore line, etc, on a large scale. This map also shows buildings on the site of the existing sand pits to the north of the farm. This map is owned by Mr Skea, a farmer in Sandwick parish. On the site of the existing farm of Newark the map outlines a building forming three sides of a square. The south wing corresponds in position with the masonry uncovered in the excavation, the north wing lies on the other (north) side of the existing chicken houses, and the connecting wing runs a few feet inside the existing field wall, along the cart-track. The plan so far uncovered in excavation does not agree with the simple rectangular shapes of the plan on the Skea map. Probably the position and orientation of the map-plan can be trusted, but it is quite probable that the ruins were too tumbled to reveal the exact ground plan to the map-maker. It may be worth noting that one of the earliest styles of 'tower-house' (to which category Newark evidently belonged) used in Scotland is the so-called 'Z-plan' house, dating from the early 16th century. An example survives in Noltland Castle, Westray. If the Newark cliff-face is imagined cutting across the plan below at the dotted lines, it will be seen that a profile reminiscent of (though not identical with) the exposed masonry's profile is produced. The castle at Noltland is several storeys high, and is fortified, as is typical of the Z-plan tower house. Well, that's all my information and ingenuity expended. We cleared up the site as much as we could. The new burials we found in the last stages were located rather inconveniently beneath the walls and under the pavement etc, and we had to leave some bones in situ, taking care to cover them in the hope that neither winter weather nor curious visitors would expose them. The tunnel which Malcolm so splendidly masoned up was already open again when we looked over the site for the last time, though the hole made by some inquisitive person was not big enough to crawl through. We filled it up again and did our best to disguise it. All along the cliff face below the walls we built a huge sea-wall which took us most of one day in pouring rain to erect. While we were there on the last day MacGillivray turned up with Cruden, playing truant from the Quincentenary lectures. Cruden said he thought we should cover the whole site for the winter, but MacGillivray persuaded him this wasn't necessary. Cruden was completely non-committal about the site's meaning, and simply listened to what I had read up. But he said he was very glad that it was being excavated and advised that we should proceed cautiously, not opening up too much at once, because that led to conservation problems - which, he added, would be his problem, not ours; whatever implications you can find in that. Val brought all the remaining skulls back in the Landrover, together with a few bags of other bones. We put the rest in polythene bags and hung them in great clusters from the roof of one of the small sheds in the school yard, which is dry and well ventilated, and left the key with the architect so that if you need any of the bones we can get in touch with him. We parted from Mr Delday of Newark on very good terms. He was charmingly reluctant to take any kind of “rent” for the field in respect of our “occupation” of it for the coming year, but I thought it politic to press a couple of quid on him, just to ensure that there won't be any snags about our returning to it next summer. He seemed well satisfied, and said he'd keep an eye on weather and visitors and put aside any skeletons, pots of gold or silver hoards which the spring tides might wash out! It was quite gratifying to gather a few compliments during our leave-takings. The school clerk said it had been a pleasure having such nice young people as our students around the place and the inhabitants were sorry we were leaving. Norse courtesy for you! Another came out more bluntly with thanks for the business we'd brought them! Certainly the news of our diggings had spread around a lot, and a lot of people we met had made a trip down to look at the site. Thanks most sincerely for your excellent leadership of the dig. You left behind a zealous little body of budding archaeologists! The sources for information I have so far used are all fairly obvious ones: Hugh Marwick, Orkney Farm-Names, 1952. Hugh Marwick, Orkney, 1951. 12th Report of the Royal Commission on the Ancient Monuments of Scotland. The Statistical Account of Scotland, 1791-99. Peterkin's Rentals of the Ancient Earldom and Bishoprick of Orkney, 1820. George Low, A Tour through Orkney and Schetland (undertaken in 1774), J.G. Dunbar, The historic architecture of Scotland, Batsford, 1966. F. Wainwright (ed.), The Northern Isles, Nelson, 1962. S.A.J. Bradley 26th August 1968 Home page

Click to enlarge