ART. IV.—An Exploration of "Sunken Kirk," Swinside, Cumberland [Map]¹, with Incidental Researches in its Neighbourhood. By C. W. DYMOND, F.S.A., Hon. F.S.A., Scot. Communicated at Durham, June 20th, 1901.

Note 1. An accurate plan of this circle as it stood in 1872, from which the one accompanying this paper is reproduced, illustrates a short descriptive account of it, which will be found in these Transactions, vol. v. (1880), pp. 47-50; and in the Journal of the British Archaeological Association, vol. xxxiv. (1878), p. 33. Reference may be made to those papers (especially to the former, as by much the fuller) for details not immediately connected with the present subject.

Site.-The site is an old meadow, not known to have been ploughed, except as to a small portion of it. It dips very gently from west to east; but, immediately to the north-west of the ring, the ground rises in one of those low knolls which, in that district, not unfrequently diversify the surface of areas otherwise smooth. Hard by, on the south, flowing eastward, is a runnel fulfilling the purpose of a land drain. It is probable that, anciently, this did not form a distinct watercourse.

Lithology.—The stones of the circle are such as are locally known as "grey cobbles." They are, however, not of uniform constitution. The majority are of a metamorphic slate, exhibiting sharp angular fracture, and variously tinted—nearly black, dark grey, red, reddish yellow, and whitish grey. These characteristics are well seen in two of the taller stones in the northern quarter of the circle, and in the newly-built fence walls near it.

Excavations.—Trenches, 18 inches in width, and of the total length of 360 feet, as recorded on the accompanying plan, with their depths at several points (measured from the surface), were laid out to explore the ground most likely to yield profitable results—as much by way of scientific induction as in the form of "finds." The strata proved to be:—(1) grass-sod (hereinafter called "sod"), 4 to 5 inches; (2) a bed of gravelly soil (hereinafter called "soil "), containing comminuted slate, small stones, with a few cobbles, and varying in thickness from 1 inch to 25 inches; its colour usually dull brown, but here and there whitish brown, reddish brown, and bright ochreous yellow; (3) sub-soil of firm, gravelly, yellowish marl, locally known as "pinnel." The depth at which this was struck varied from 6 inches to 29 inches; the variation being due to a wavy irregularity in the top of the "pinnel", to which the almost even grassy surface does not conform. In several places the depth was 16, 18 and 20 inches. The maxima occurred in two spots—one at the middle of a flat hollow, about 6 feet in diameter, which was met with in the north-eastern trench (where a short lateral trench was cut, 12 feet from the intersection at the centre of the circle), where the depth is 28 inches; the other in the middle of the low mound by the gateway, where it was found to be 29 inches. This spot was reported to have been disturbed by some casual digger in the time of the father of the present tenant. The shallowest cuttings were at the western edge of the said mound, and in the radial trenches across and outside the line of stones. The small stone, or stones, split and sunken across the inner end of the gateway, and commonly called "the altar ", appeared to have been dug around by some previous explorer. The results of the work, so far as it was carried, convinced us that to dig into the sub-soil itself would be but to waste time.

Objects found.—The only objects found in the trenches were the following:—a piece of charcoal of the size of a large filbert, about 15 inches below the surface, at a point 12 inches north-east from the centre of the circle; also a few very small fragments of the same material, about 12 inches deep in the gateway-trench; a minute splinter of decayed bone, within a few inches of the first-named piece of charcoal; two small bits of red stone, having a good deal of iron in them, which made a mark when rubbed, and which Mr. Barlow-Massicks thought to be, not of local origin, but "drift", in the south-eastern trench, not far from the centre of the circle; two small shards of modern glazed pottery; and, finally, a "Lancaster halfpenny", in the sod near the southern end of the peripheral trench along the eastern gap. On the obverse is the head of John of Gaunt, coroneted, and an incised inscription:— JOHN OF GAUNT DUKE OF LANCASTER.

On the reverse is a faint image of a shield; and around the rim, in raised letters: PAYABLE . AT . THE . WAREHOUSE . OF . THOS . WORSWICK . X . SONS.

Mr. W. O. Roper, F.S.A., says that the dates of these halfpennies range from 1789 to 1794; and that the commonest of them resembles this, with the arms of Lancaster and the legend:- LANCASTER HALFPENNY 1791 on the reverse.

Observations and Conclusions.—The following statement of facts, and of particular conclusions to which we have been led, is arranged in sections lettered to correspond with those of the preceding programme:-

(a) The places where there are hollows, more or less distinctly marking the positions of stones which have disappeared, are indicated on the plan by crosses. Where there are no crosses, no hollows could be seen.

In a case like this, with the members of a peristalith in close order, as though to form a kind of continuous fence, it seems very improbable that the number of stones was a matter of any moment. Hence, the small value that may attach to the observations made under this head.

(b) and (c) The stones were founded on the top of the "pinnel", and were steadied by a packing of small cobbles—some larger ones being wedged in (especially on the inner side), so as to trig up the standing-stones.

As noted in my earlier account of this circle, all the stones, but one fragment resting on the ground, have fallen inward. There can be little doubt that, in most cases, if not in all, this may be attributed to the usual cause—the sheltering of sheep under the lee of the larger stones. In course of time this wears hollows, in which water collects and softens the ground; so that the blocks, partially undermined, begin to decline, and, sometimes, finally fall.

One of the leaning stones, at the W.N.W. point, which remains as it was at the date of my survey, 29 years ago, has now been protected by a bank of cobbles, which will at least prevent the sheep from continuing to worry the ground close to the stone.

The rowan-tree, which had sprung up in, and riven apart one of the standing-stones in the north quarter, has been dead for several years.

(d)No trace of either ditch or pavement was found; and the gateway was not paved; though the soil was full of stones—probably a portion of the packing for the jambs of the entrance.

(f) The discrepancy between the irregular top of the "pinnel" and the fairly even upper surface may indicate an endeavour to flatten the area of the circle—perhaps still farther improved by years of annual mowing. Had there been an intention to make a level platform (such as was rarely, if ever, attempted in these situations), the soil on the eastern and lower side would have been deeper than that on the western and higher—which is not the case.

The uniform depth of the sod, the homogeneous nature of the lighter-coloured soil below it, and the very slight exposure of the stone-packing¹, sufficiently show that the height of the present surface is practically identical with that of the ancient one.

Note 1. Gough says that some part of this "rude pavement of smaller stones" was raised about half-a-yard above the surface of the earth." This must have been an exaggeration.

(g) There were no indications that the place had been used for cremation, or for funeral-feasts. The various whitish, reddish, and yellow colouring of the soil in four or five spots was, doubtless, derived from the rocks; and the surface of the sub-soil throughout was found to be uniformly clean and tinted somewhat like yellow clay. The occurrence of the charcoal and bone was plainly accidental—probably resulting from top-dressing—and the incidents were much too insignificant to support any theory involving the practices of which evidence was sought.

No traces of interments were anywhere met with.

(h) A thorough and repeated scrutiny has failed to detect any archaic markings upon the stones.

Summary of Results.—From the above it will be seen that the results of this exploration were almost wholly of a negative character. This, however, is not equivalent to saying that, to the practised archæologist, the inquiry has been a barren one. It indicates that, whatever may have been the purpose for which some other structures of this class were set up, "Sunken Kirk" must have been destined to be a locus consecratus, not necessarily connected with sepulture, but devoted to certain ceremonial observances suggested by the provision of the gateway—otherwise meaningless.

Comparison of Circles.—It would extend this paper to an unreasonable length were it to embrace an adequate account of all the different forms given to these monuments, even in the closely connected regions of North and West Scotland, Cumbria, Man, the Midlands, South and West of England, Wales, East and West Cornwall, and Brittany—implying almost as great a variety of forms of use; and suggesting that they were erected by tribes with rites and customs which, while they all had certain characteristics in common, were, nevertheless, in some particulars, distinctly differentiated. But, to discover the place and affinities of "Sunken Kirk" in such a collection of examples, it may be useful very briefly to indicate the main features of a convenient classification; to the end that its points of resemblance to, and of difference from other monuments of its species may be clearly seen.

The principal local examples, referred to the foregoing numbers, are these: - Eskdale-moor [Map], 1; Oddendale [Map], 2; Gunnerkeld [Map], 2, 3, 9; Long Meg, 3, 6¹, 9; Castle-rigg, 3, 6¹, 9; Swinside, 3, 5², 9; Shap, 6, 1 i [?] .

Note 1. No signs of sepulture have yet been discovered in either of these two cases. The statement of Camden with regard to Long Meg (Britannia, Gibson's ed., 1695, p. 831) that "within the circle are two heaps of stones, under which they say there are dead bodies bury'd," though supported by a report made to Aubrey, is discounted by the positive assertion of Gibson that "the heaps of stones in the middle of this monument, are no part of it; but have been gather'd off the plough'd-lands adjoyning, and (as in many other parts of the County) have been thrown up here together in a waste corner of the field." There remained no mark of their former existence at the time of Hutchinson's visit in 1773.

Note 2. Probably the "stones still to be seen, though sunk deep in the earth," in the middle of this circle, which Gough supposed to be remains of an altar, if not those by the gateway, were nothing more than such cobbles as were found in digging the trenches.

To sum up the distinctive characteristics of "Sunken Kirk [Map]":—it is a circle with stones in close order, each set in a bed of small stones, having a gateway, and yielding no evidence of having been sepulchral or connected with funeral rites.

Finally, it may be instructive to note the points of resemblance and of difference between "Sunken Kirk [Map]" and Gunnerkeld [Map]. Both are furnished with gateways; the former is not sepulchral, the latter is. The general affinities of "Sunken Kirk [Map]" are with Long Meg [Map] and Castle-rigg [Map]; those of Gunnerkeld with the double ring at Oddendale [Map].

The following is a copy of Housman's account of the circle at Swinside, to be found on pp. 471-2 of his Topographical Description, before referred to. As an early authority, it may be well to quote it here, to supplement the series of extracts which were published in the fifth volume of these Transactions, pp. 47-50:-

About three miles north-west from Duddon-bridge, in a wild country, among dreary mountains, there is a druidical temple, called Sunken-kirk [Map], which consists of a circle of large stones standing erect in the ground, and pretty close together, some of which are more than eight feet above the surface, and eleven feet in circumference. The larger stones are 50 in number, with an infinite number of smaller ones scattered about their bases. The form is nearly a circle, being 87 feet by 84 in diameter. It is seated in the level part of a high but swampy meadow, and near a solitary farmhouse. At the entrance there are four large stones, two placed on each side, at the distance of six feet."

SUPPLEMENTARY EXCAVATIONS.

Though not specified in our commission, we incidentally more or less examined three mounds which had a cairn like appearance. One of these was a small heap in a part of the moor on the north-western flank of Knott-hill. Nothing but stones and earth was found in it. A similar heap on the edge of a rushy flat in the lower part of Broadgate New-close, near Crag-hall, yielded the same results—the earth being very red. The third was a barrow-like mound in the upper part of the same close, 75 feet from the wall on its western side, and a few feet north of the fell-gate. The heap, on sloping ground, is 24 feet in length and about 12 feet in width. Three large cobbles rest, or have been set up, in a line crossing its crest; and a group of similar stones lie a little below its foot, as though rolled down from the mound. A few feet of trench were cut, but nothing was found, except some cobbles. The examination could not then be continued; but it is by no means certain that a farther search might not be rewarded by interesting discoveries.

It may be well to mention here that Swinside is the centre of a district full of remains and relics of ancient times, many of them, presumably, of neolithic age:- e.g., (1) megalithic monuments at Lacra and Kirksanton:¹ (2) very many cairns and hut-circles, from Thwaites-fell, a mile and a half north of Swinside, across to Bootle, and thence, northward, to Barnscar; none of which have yet been explored: (3) many stone implements have been found; as two good axes found at Fenwick, close to Swinside, an axe at Lowscales (now at Broadgate), a stone celt and stone hammer at Silecroft, a stone hammer and arrow-heads at Bootle, a flint arrow-head at Corney, a "British" urn at Beck, near Millom (now at Broadgate).

Note 1. Stone circles formerly existed at Annaside, Gutterby, and Hall-foss; but they have been destroyed.

The only sufficiently level sites which we noted between Swinside and Broadgate were the following:- On the east side of Knott-hill—(1) a flat, overlooking Blackbeck, in Whinny-hill, 141 O.M.; (2) a mound (much too small) in Hog-house meadow, 281 O.M., 12 chains N.N.E. from Crag-hall; (3) a flat hollow in High-crag, 296 O.M., just N. from Crag-hall; (4); a flat in the centre of High-leys, where there is a spring, 301 O.M.; (5) a flat near the centre of Broadgate New-close, 291 O.M., on the north side of Knott-hill plantation; (6) a rushy flat in the lower part of the same close, about 100 yards from the road. On the west side of Knott-hill—(7) a flat, 122 O.M.; and (8) another, 127 O.M., S.W. from Knott-hill. From every one,of these sites some part of Duddon-sands may be seen, in directions varying from S. to S.E. If one of them should be the spot to which Housman refers, his "south-west" (as Mr. Gaythorpe has pointed out) must be a misprint for south-east; unless his bearings were magnetic, which might make his text approximately correct for some of the positions.

While on the spot, we availed ourselves of an opportunity, kindly afforded to us by Mr. Lewthwaite, of inspecting his ordnance and estate-maps, to see whether they had any note of the missing monument. No memorandum was found upon the older edition (1862) of the 25-inch ordnance-map. Next, in order of date, was a map of the whole estate, dated 1835, showing the ring at Swinside, but not the wall between it and the farm-road, which had been built before the date of the ordnance map. It contained no indication of a second circle. Going back, was an undated map of a portion of the estate, not including Swinside, which, from the watermark on the paper, was discovered to have been made soon after 1798. Delineated thereon, with a fair degree of accuracy, were the rocks, blocks and thorn-trees in Broadgate New-close, which then, as now, appears to have been distinguished from all the other fields in the neighbourhood by these rough features. But if the lost circle (then reported to be in existence) was placed there, it is very singular that no record was made of so noteworthy an object, when much ingenuity was exercised in planning the positions of the natural ones. Nor was any reference to it found in any other part of the map. The same can be said of a still older, undated map of a portion of the estate not including Swinside, with watermark showing that it was made after 1770.

One other testimony remains to be examined,—that of the Rev. T. N. Postlethwaite, of Ulverston, who is nearly connected by family ties with the locality. In an account which he has kindly sent, he says that his father, who was familiar with the neighbourhood, averred that "a second circle existed; and, in the absence of the then owner of Broadgate (possibly [the present] Mr. Lewthwaite's great grandfather) in London, the farmer at Swinside—or Crag-hall—wanted to build a barn, and, seeing the stones handy, carted away one circle for the purpose, and was continuing operations on the second and existing circle, when Mr. Lewthwaite returned and rescued it. I have heard two versions—one referring to a distinct circle; one to a circle within—or without—the present one." Mr. Gaythorpe says that "the barn at Crag-hall was built about 1829 or 30, but neither Thomas or Margaret Garnett who live there (the latter being born in 1819) recollect anything of this removal." It is just possible that some of the Swinside stones were removed at about that time, to furnish the foundation of the neighbouring wall; but the Garnett's testimony throws great doubt on the statement that the barn at Crag-hall was partly built with materials taken from the second circle; and the story is made still more dubious by the doubtful way in which Mr. Postlethwaite writes of "Swinside—or Crag-hall", and by the implication that the same man was tenant of both farms, which are three-quarters of a mile apart. On the other hand, the statement is true of another place, of which a separate account is hereunder given.

ANCIENT REMAINS ON GORNAL-GROUND.

Taking hint from a memorandum at the end of Mr. Gaythorpe's newspaper-article, mentioning a curious place, of which he heard only after his descriptive notes had been written, I went a-prospecting to the spot, with small expectation of being rewarded for the trouble. Mr. Gaythorpe was told that "there is a circular place in a field at Gornal-ground adjoining the old 'Roman' road, which is thought to have formerly been a Druidical resort, and there are a number of huge stones lying about the place." Gornal-ground is a homestead, unnamed on the one-inch ordnance-map, three furlongs east from Broadgate, and nearly a quarter of a mile west from Hazel-mount. An intelligent lad at the farm pointed out the way to the place of which I was in quest. It is so secluded that the stranger hardly sees it until he is close upon it. The site is in a small sheltered bottom, one-fifth of a mile south-east from the house. It is oval; has been slightly excavated and levelled; and is crossed by a small beck. From it the ground rises gently north-westward, up-stream, steeply, on one side of the valley, westward, and less steeply on the other side, north-eastward; while, south-eastward, down-stream, over the brow of a gentle bank which completes the basin-like form of the little amphitheatre, there is a considerable fall toward the outlet of the valley. The inclosed space is, in fact, a shelf, almost entirely formed by nature, and sheltered on every side but the last-named; in which direction it commands a view of Duddon-sands ranging over 40° of the compass—viz., from S. 20° E. to S. 60° E.

The cleared area (see the accompanying plan), with inner diameters of 184 feet and 150 feet, has been cinctured by a rampart built partly with large boulders, partly with rude walling, and partly with slabs set up on edge in the ground. A rough wall, of which some of the lower courses may be old, fences it on the eastern side; while the western half of the circumference, from which most of the materials of the walling have been removed, is now unprotected. Along this ragged open portion of the circuit are bedded stones of various sizes,—remnants of the original rampart,—the finest being those in the northern quarter between the stream and the wall. A few large blocks lie in the south-eastern part of the area,— one of them, 9 feet in length and about 5 feet in breadth, having fallen flat and become imbedded in the soil. There is no distinctly-marked gateway; but a cart-track passes through a gap in the western side which probably was the position of the original entrance. A low mound, rising in the midst of the work, is a peculiar and noteworthy feature.

Skirting the inclosure for the distance of iio feet on its north-eastern side is a narrow and sinuous pack-horse way, half a mile in length, which connects Gornal-ground with the road near Lady-hall, and has been mistaken for, and miscalled, a "Roman road." The lower portion has been modernized into a good cart-road. The middle portion appears to retain its original character; the several ascents, for a total length of 155 yards, being neatly pitched and paved;—the former style prevailing on the steeper slopes, and, here and there, approximating to.a rude stair-way; the latter prevailing on the easier gradients. The upper portion—now all but impassable by reason of the overgrowth—is little better than a stony water-course. There appears to have been no direct access from this lane to the adjoining inclosure; but, about 80 yards above the point where it leaves it, there is a gateway, now walled-up, which opened opposite to the head of another such lane on the western side of the beck, at a place where the two roads and the beck are close together. This western way consists of three sections:—(1) a narrow stony lane, like the upper portion of the principal pack-horse way, walled for a short distance in its upper part; this leads to (2) a hollow-way carried along a rather steep slope (see plan and section on the plate); which in turn is succeeded by (3) a green ramp, leading down nearly to the present entrance to the ring. There are some slight indications that the hollow-way may have pursued a direct south-ward course; but they are too faint to justify a stronger statement.

It needed but a glance to determine that these are the remains, not of a megalithic circle, but of an occupied spot, with which the adjoining way was intimately connected: the approach therefrom being apparently along the route provided by the western branch-way.

As the uncommon incidents of the traversing stream and the single central mound invested the place with peculiar interest; and as the latter seemed not unlikely to furnish the key to the purpose for which the location was established; I asked Mr. Collingwood to accompany me on a second visit to the spot, in order to make a good plan of it, and to open the mound. It was also desirable that, by a personal inspection, he should be in a position to form an independent opinion as to the value of a conjecture which I had ventured to hazard, that this was the site of the long-lost circle. So, on the 30th of April, we carried out the proposal; and, while the survey was in progress, Mr. Collingwood dug into the mound. His observations thereon are given in the following paragraph.

To set against this, we have only, (1) the alleged proximity of the lost circle to Crag-hall; which, however, after all, appears to be extremely doubtful; and (2) the curious fact that, if this was the place, neither of the early authors (or their informants) mention such note-worthy matters as the passage of the stream through the circle, or the contiguous ancient way.

But a reasonable inference in favour of our provisional conclusion may be drawn from a statement made by Mr. Hartley, the owner of Gornal-ground, to Mr. Collingwood, on a subsequent visit, that very many loads of stones were taken by his father from the ruins of the ring to furnish materials for some of the farm buildings.

Perhaps we may not be far wrong in thinking that this may have originated the story, which, by mistake, became associated with Crag-hall; and may do away with the principal reason for hesitating to believe that this is the site of which we were in search.

So far as we know at present, this place (if its use has been rightly divined) is an unique example of a mediæval station of a peculiar type, the singularity of which may not improbably be due to its position and environment. Should farther research result in the discovery of other not very dissimilar sites on ancient lines of communication through the district, it will tend strongly to support this conjecture. Meanwhile, it will be hard to imagine any other purpose for which the spot could have been inclosed.

ADDITIONAL OBSERVATIONS ON ANCIENT WAYS.

In connection with the investigations here recorded, the opportunity was seized of going a little a-field on the traces of ancient ways.

Continuation of the Gornal-ground Road.—It has been stated that this road can be traced to Crag-hall, and so on to Swinside. We carefully examined the ground as far as Crag-hall; but saw no marks of such a continuation.

Mr. Collingwood supplies the following notes of his own earlier observations on this head:—"There are old massive stepping-stones across the Black Beck, a little above the mill, and a continuous footpath going up to the road from Broadgate to Swinside. Past Swinside there is a track, in which one bit seems to be paved, the rest being on the stony pinnel, straight to the peat house on the moor, where it crosses the beck, and goes straight to Stoneside; thence by old lanes in a direct line to Waberthwaite. This seems to be a section of an old road which goes in a direct line from the old Leven ford through Ulverston, by Horrace to Sandside at Kirkby, and there skirts the sand to Foxfield. The crossing of Duddon is easy (experto crede) to the lane leading up to Ladyhall and joining our Gornal-ground road, which seems to be a section of an ancient packhorse track leading from Furness by the shortest and easiest route to the old harbour of Ravenglass. There are many old paved packhorse tracks in this part of Cumberland, none of them of course rightly described as Roman."