FOOTNOTES:

[51] Hibbert's Shetland, p. 452.

[52] Account of the Islands of Orkney, by James Wallace, M.D., 1700, p. 58.

[53] Notices of remains found in tumuli and cists, of gigantic stature, frequently occur in the Statistical Accounts and other local records, but the statements are generally too vague to be of any value. Erroneous opinions, I believe, most frequently arise from comparing the femur or thigh-bone with the apparent length of the thigh, by persons ignorant of anatomy. Nothing, however, more readily secures distinction among a rude warlike people than the personal strength accompanying superior stature, if combined with corresponding courage; it need not therefore excite surprise if the larger tumuli should occasionally be found to cover the remains of some primitive chief of gigantic stature.

[54] The account which Tacitus gives of the simpler rites of the ancient Germans probably more nearly accords with those of the primitive Britons: "Funerum nulla ambitio; id solum observatur, ut corpora clarorum virorum certis lignis crementur. Struem rogi nec vestibus, nec odoribus cumulant; sua cuique arma, quorundam igni et equus adjicitur."—Tacit. de Morib. Germ. cap. 27.

[55] Ipsum cremare apud Romanos non fuit veteris instituti, terra condebantur.—Hist. Nat. lib. vii. c. 54.

[56] Cases occur where the original tumulus has been adopted as a place of sepulture long subsequent to its original construction. Care is therefore required to discriminate between superficial interments of late date, and the original cist or urns; but it is rarely difficult to detect the evidences of intrusion. The slight depth at which they are generally interred affords in itself a striking contrast to the labour exercised by the constructors of the sepulchral mound. It is also to be borne in remembrance, that all the urns found in tumuli are not sepulchral, or proofs of cremation.

[57] De Bell. Gall. lib. vi. chap. 19.

[58] Dr. Hodgkin read a paper at the meeting of the British Association, held at York in 1844, on the dog as the associate of man, chiefly with a view to shew how much the study of the inferior animals which, by accident or design, have accompanied man in his diffusion over the globe, is calculated to throw light on the affinities of races.

[59] Ure's History of Rutherglen, p. 124.

[60] Ure's History of Kilbride, pp. 216-219.

[61] By N. K. Sjöborg. Two vols. quarto. Stockholm, 1822.

[62] Graham's Antiquities of Iona, Plate III.

[63] Worsaae's Primeval Antiquities, p. 109.

[64] Sinclair's Statist. Acc. vol. xix. p. 441.

[65] New Statist. Acc. vol. x. p. 717.

[66] Petrie's Eccles. Architect. of Ireland, pp. 103-5.

[67] Add. to Camd. Brit. in Radnorshire.

[68] New Statist. Acc. vol. iv., Kirkcudbright, pp. 132, 133.

[69] Pennant's Tour, vol. i. p. 156.

[70] New Statist. Acc., Dumfriesshire, vol. iv. p. 475.

[71] Itiner. Septen. Append. pp. 171-177.

[72] Dumfries Journal, June 24, 1828. MS. Communication, Soc. Antiq. Scot., Andrew Brown, Esq., read March 9, 1829.

[73] Sinclair's Statistical Account, vol. i. p. 292.

[74] Ure's Kilbride, p. 213.

[75] Itin. Septen. p. 42.

[76] Sinclair's Statistical Account, vol. viii. p. 497.

[77] This point has been conclusively established in the valuable communications of Mr. F. C. Lukis to the Archæological Journal, on the Primeval Antiquities of the Channel Islands, vol. i. pp. 142, 222. The original merit, however, of showing that cromlechs are "sepulchral chambers," and not "Druidical altars," is, I believe, due to a well-known and zealous antiquary, Mr. John Bell, of Dungannon, who published his views in the Newry Magazine, 1816, vol. ii. p. 234, from whence they were copied into various other journals.

[78] E. B. Barrett.

[79] Martin's Western Isles, p. 220.

[80] Archæol. Scot. vol. iii. p. 43.

[81] Journal of Brit. Archæol. Association, vol. iii. p. 342.

[82] While this sheet is passing through the press, I have had an opportunity of exploring this cromlech. The natural rock was laid bare at a very little depth without meeting with the slightest traces of sepulchral remains, and were it not for the remarkable line of perforations along the centre of the capstone, the whole might have been ascribed to a natural origin. It was found impossible, however, to get directly under the great stone, without the risk of overthrowing the whole.

[83] MS. Letter, George Petrie, Kirkwall.

[84] Archæol. Scot. vol. iii. p. 42.

[85] Archæol. Scot. vol. ii. pp. 77, 100.

[86] Archæol. Scot. vol. iii. p. 48.

[87] Archæol. Scot. vol. iii. p. 49.

[88] Ure's Kilbride, p. 213.

[89] MS. Letter, G. W. Knight, Libr. Soc. Antiq. Scot. 1829.

[90] Ure's Rutherglen, p. 223.

CHAPTER IV.
DWELLINGS.

Before proceeding to examine in detail the varied contents of the Scottish tumuli, it may be well to glance at the evidence we possess of the nature of the habitations reared and occupied by the constructors of such enduring memorials of their dead as have been described in the preceding chapter. Scattered over the uncultivated downs both of England and Scotland, there still remain numerous relics of the dwellings of our barbarian ancestry, which have escaped the wasting tooth of centuries, or the more destructive inroads of modern cultivation. Sir Richard Colt Hoare remarks, in his "Ancient Wiltshire,"—"We have undoubted proofs from history, and from existing remains, that the earlier habitations were pits, or slight excavations in the ground, covered and protected from the inclemency of the weather by boughs of trees and sods of turf." Of these primitive pit-dwellings numerous traces are discernible on Leuchar Moss, in the parish of Skene, and in other localities of Aberdeenshire; on the banks of Loch Fine, Argyleshire; in the counties of Inverness and Caithness; and in various other districts of Scotland still uninvaded by the plough. They are almost invariably found in groups, affording evidence of the gregarious and social habits of man in the simplest state of society. The rudest of them consist simply of shallow excavations in the soil, of a circular or oblong form, and rarely exceeding seven or eight feet in diameter. Considerable numbers of these may be observed in several districts both of Aberdeenshire and Inverness-shire, each surrounded with a raised rim of earth, in which a slight break generally indicates the door, and not improbably also the window and chimney of the aboriginal dwelling. To this class belong the "pond barrow," already referred to as erroneously ranked among sepulchral constructions. Within a few miles of Aberdeen are still visible what seem to be the remains of a large group, or township, of such rude relics of domestic architecture. These, Professor Stuart suggests, may mark the site of the capital of the Taixali, when the Roman legions passed the river Dee in the second century.[91] They consist of some hundreds of circular walls scattered over more than a mile in extent, of two or three feet high, and from twelve to twenty feet in diameter. Their varying sizes may be presumed to indicate the gradations of rank which, we know, were established among the northern Britons, who were undoubtedly, at the period of the Roman invasion, a race far in advance of the first constructors of the rude pit-dwelling or "pond barrow" previously referred to. Nothing, however, has yet been discovered on this site to indicate any traces of Roman influence. On digging within the area of the pit-dwellings, a mass of charred wood or ashes, mingled with fragments of decayed bones and vegetable matter, are generally found; and their site is frequently discernible on the brown heath, or the grey slope of the hill-side, from the richer growth and brighter green of the grass, within the circle sacred of old to the hospitable rites of our barbarian ancestry, and where the accumulated refuse of their culinary operations have thus sufficed to enrich the soil.

The first indication of a slight advancement in the constructive skill of the primitive architect is discernible in the strengthening of his domestic inclosure with stone. This is not infrequently accompanied with small circular or oblong field inclosures, as if indicating the dawn of civilisation, manifested in the protection of personal property, and the rudiments of a pastoral life, in the folding of sheep and cattle. Still greater social progress would seem to be indicated in those examples, also occasionally to be met with in various districts, where a commanding site appears to have been chosen for the location; and traces still remain of an earthen rampart inclosing the whole, as on the Kaimes Hill, in the parish of Ratho, Mid-Lothian. Such, perhaps, may be the remains of a British camp, or of a temporary retreat in time of war.

With this class also may be grouped the "Picts' kilns," on which Chalmers, Train, Sir Walter Scott, and other antiquaries, have expended much conjecture and useless learning. These are of frequent occurrence in Wigton and Kircudbright shires, as well as in parts of the neighbouring counties. They consist of elliptical or pear-shaped inclosures, measuring generally about sixteen feet in length and seven or eight feet in breadth. Externally the walls appear to be of earth, sometimes standing nearly three feet high. On removing the surface they are found to be constructed internally of small stones, frequently bearing marks of fire. They are popularly believed to be ancient breweries reared by the Picts for the manufacture of a mysterious beverage called heather ale. Sir Walter Scott suggests, with not much greater probability, that they are primitive lime kilns. They are said by Mr. Train to be invariably constructed on the south side of a hill, close to the margin of a brook, and with the door or narrow passage facing the stream. Future excavations on their sites may perhaps furnish more conclusive evidence of their original purpose.

Greater art is apparent in the relics of another class of ancient Scottish dwellings occasionally met with in different parts of the country. In the Black Moss, already referred to, on the banks of Etive, Argyleshire, at various points where some advance has been made in recovering the waste for agricultural purposes, the progress of cultivation has uncovered rough oval pavings of stone, bearing marks of fire, and frequently covered with charred ashes. These are generally found to measure about six feet in greatest diameter, and are sometimes surrounded with the remains of pointed hazel stakes or posts, the relics, doubtless, of the upright beams with which the walls of the ancient fabric was framed. Julius Cæsar describes the dwellings of the Britons as similar to those of the Gauls;[92] and these we learn, from the accounts both of Strabo and Diodorus Siculus, were constructed of wood, of a circular form, and with lofty tapering roofs of straw. Such apparently were the structures, the remains of which are now brought to light within the limits of the Dalriadic possessions. These ancient Caledonian hearths, now quenched for so many centuries, are discovered beneath an accumulation of from eight to ten feet of moss, under which lies a stratum of vegetable mould about a foot deep, resting upon an alluvial bed of gravel and sand; the original soil upon which the large sepulchral cairns of the same district have been reared.

A discovery made at Dalgenross, near Comrie, in 1823, though described in a communication to the Society of Antiquaries of Scotland as an ancient tomb, manifestly furnishes another example of the same class of primitive dwellings. At Comrie, as in twenty other equally probable localities, antiquaries of the district have pronounced some imperfect and half-obliterated earth-works to be the remains of a Roman camp, and the scene of Agricola's famous victory of Mons Grampius! The writer, learning that workmen were trenching this interesting spot, remarks,—"I hastened to where the men were at work, and found that they had come upon a circle of flat stones set on edge, and at the bottom a paving of large flag-stones. The cavity was filled with a kind of black earth, pieces of charred wood, and also some fragments of iron, but so completely defaced with rust that it is impossible to say to what purpose they have been applied." On pursuing the investigation further, pieces of charcoal and burned wood were found, along with charred wheat, which might possibly suggest its having been a granary; but its general characteristics much more nearly assimilate it to a native dwelling, to which, it may be, the torch of the Roman legionary applied the brand that reduced it to a blackened ruin.

Among the relics of primitive domestic architecture brought to light in later times, no class is more remarkable than the weems, or subterranean dwellings which have been discovered in different parts of Scotland. Of this class are two structures discovered under ground in the parish of Tealing, Forfarshire. One of them consisted of several apartments formed with large flat stones without any cement; and in it were found wood-ashes, several fragments of large earthen vessels, and an ancient stone hand-mill, or querne. The other was a single vault constructed in the same manner, measuring internally about four feet both in height and width, and in which were found a broad earthen vessel, and a stone celt or hatchet.[93] In another opened in the parish of Monzie, Perthshire, a stone celt and bronze sword were found, both of which are preserved at Monzie House. Chalmers supplies a curious list of similar subterranean dwellings discovered at various times in Forfar, Perth, Aberdeen, Ross, Sutherland, and Inverness shires, and in the Orkney Islands.[94] The like structures are noted by Martin, among the antiquities of the islands of Walay, Erisca, and Skye;[95] and by Pennant also in the latter island. They are described by Martin as "little stone-houses, built under ground, called earth houses, which served to hide a few people and their goods in time of war."

The general name applied in Scotland to these subterranean habitations is Weems, from the Gaelic word uamha, a cave; and as this name is in use in the low countries, where nearly all traces of the Celtic dialect have been lost as a living language, probably since the era of the "Saxon Conquest," it may be accepted as no insignificant evidence of their Celtic origin or use. In Aberdeenshire, where they have been found in greater number than in any other single district, they are more generally known, as in the Hebrides, by the name of eirde (i.e., earth) houses.

An interesting account of a large group of weems discovered in Aberdeenshire, is given by Professor Stuart in the Archæologia Scotica,[96] and since then many more have been brought to light in the same district. Several of these opened of late years in Strathdon are described with great minuteness in the Statistical Account of that parish.[97] On a bleak moor in the adjoining parish, not far from the old castle of Kildrummie—which, from many large fossil trees dug up in it, appears to have once been an extensive forest—the largest assemblage of these singular habitations occurs which has yet been discovered in Scotland. Others have been found about six miles further up the country, at Glenkindrie, at Buchan, and near the source of the Don, one of the wildest districts of the Highlands. They are indeed scarcely less common than the sepulchral cairn. My object, however, is not so much to accumulate numerous examples, as to select a few characteristic types of each class of Scottish antiquities; though these weems appear to possess peculiar claims to minute description, from their very frequent occurrence. In general, no external indication affords the slightest clue to their discovery. To the common observer, the dry level heath or moor under which they lie presents no appearance of having ever been disturbed by the hand of man; and he may traverse the waste until every natural feature has become familiar to his eye, without suspecting that underneath his very feet lie the dwellings and domestic utensils of remote antiquity.

The Aberdeenshire weems are constructed of huge masses of granite, frequently above six feet in length, and though by no means uniform either in internal shape or dimensions, a general style of construction prevails throughout the whole. Some of them have been found upwards of thirty feet long, and from eight to nine feet wide. The walls are made to converge towards the top, and the whole is roofed in by means of the primitive substitute for the arch which characterizes the cyclopean structures of infant Greece, and the vast temples and palaces of Mexico and Yucatan. The huge stones overlap each other in succession, until the intervening space is sufficiently reduced to admit of the vault being completed by a single block extending from side to side. They have not infrequently smaller chambers attached to them, generally approached by passages not above three feet in height; and it affords a curious evidence of the want of efficient tools in the builders of these subterranean structures, that where these side apartments are only separated from the main chamber by the thickness of the wall, the stones, though placed flush with the walls of the latter, project irregularly into the small cells, giving them a singularly unshapely and ragged appearance. Similar structures, but of smaller dimensions, have been discovered in Lanarkshire, at Cartland Craigs, in the neighbourhood of Stonebyres, and at a place called Cairney Castle. In these last were found quernes, deers' horns, and bones. In one uncovered in the parish of Auchterhouse, Forfarshire, a brass ring was discovered; and both there, and in another in the same parish, were ashes, bones, and quernes.[98] The Rev. Thomas Constable furnishes a very interesting description of one near Lundie House, in the latter county, which was minutely surveyed by the eminent antiquary, Lord Hailes. Its contents were of the usual description, including several quernes about fourteen inches in diameter.[99] So also, in a minute account of similar structures in Caithness and Sutherland, furnished to Pennant by the parish minister of Reay, the writer remarks:—"We found in them nothing but hand-mills, or what the Highlanders call quernes, which were only eighteen inches in diameter, and great heaps of deers' bones and horns, as they (the Picts) lived much more by hunting than any other means."[100] The discovery, indeed, of the primitive hand-mill in these ancient dwellings is so frequent as to be worthy of special notice, and might seem to indicate that their original destination had been for store-houses or granaries, did not the constant occurrence of the bones of domestic animals, or of those most prized in the chase, and frequently in considerable quantities, leave no room for doubt that they must have been occupied as places of habitation. They agree very nearly with the description furnished by Tacitus of the winter dwellings of the Germans, whom he represents as digging caves in the earth, in which they lay up their grain, and whither they retire in the winter, or on the advance of an enemy to plunder the open country.[101] The entrance to such of these subterranean dwellings as have been found sufficiently perfect to afford indications of their original character, appears to have generally been by a slanting doorway between two long, upright stones, through which the occupant must have slid into his dark abode. Occasionally a small aperture has been found at the further end, apparently to give vent to the fire, the charcoal ashes of which lie extinguished on the long-deserted floor. In some a passage of considerable length has formed the vestibule; but so far as now appears, a solitary aperture served most frequently alike for doorway, chimney, ventilator, and even window, in so far as any gleam of daylight could penetrate into the darkened vault. One is forcibly reminded, while groping in these aboriginal retreats, of Elia's realisations of the strange social state to which they pertain, in his quaint rhapsody on Candle-light, "our peculiar and household planet! Wanting it, what savage unsocial nights must our ancestors have spent, wintering in caves and unilluminated fastnesses! They must have lain about and grumbled at one another in the dark. What repartees could have passed, when you must have felt about for a smile, and handled a neighbour's cheek to be sure that he understood it! This accounts for the seriousness of the elder poetry. It has a sombre cast, derived from the tradition of these un-lanterned nights!" The grave humorist goes on to picture a supper scene in these unlighted halls, rich with truthful imaginings, mingled with his curious but thoughtful jests:—

"Things that were born, when none but the still night,
And his dumb candle, saw his pinching throes."

In truth, these dwellings, constructed with such laborious ingenuity in every district of Scotland, seem to throw a strange light upon that dim and remote era to which they belong, giving us some insight into the domestic habits and social comforts of a period heretofore dark as their own unillumined vaults.

Adjoining many of the weems small earthen inclosures are discernible; some of which are square, measuring about fifteen paces each way, with the area somewhat below the surrounding soil, and have probably been constructed for folding sheep or cattle. Others are circular, and so small as to leave little doubt that there must have stood the slight huts, constructed of turf and branches of trees, in which the architect of the cyclopean structure dwelt during the brief warmth of summer, while he sought refuge from the frosts and snows of our northern winter in the neighbouring subterranean retreat. The number of weems frequently found together appears altogether inconsistent with the idea of their construction as mere places of concealment. They are manifestly the congregated dwellings of a social community, though strangely differing from any that have dwelt in the land within the era of authentic history. When we compare these dwellings with the clay huts still common in many a Highland district, or with such humble Lowland biggings as those which have won a new sacredness as the birthplaces of Hogg or Burns, it is impossible to overlook the remarkable differences presented by the two states of society, separated not more widely by time than by variance of habits and ideas. How striking is the contrast between the artlessness of the Ayrshire cottage, that sufficed, with its straw roof, to satisfy the wants of one among the great master-spirits of all times, and the labour and ingenuity expended in producing these retreats of the Scottish aborigines. In rudeness of result perhaps both are on a par. The ingenious and methodic skill, however, entirely belongs to the old builders. Their mode of constructing with huge unhewn stones, frequently brought from a considerable distance, seems to point them out as the architects of that same remote era in which the rude monumental standing stones and circular groups of monoliths were reared, which still abound in so many districts of the Scottish mainland and surrounding isles.

Similar subterranean structures have been discovered at different times in Orkney, some of them of considerable extent, and including various recesses and chambers branching off from the chief central apartment. An unusually minute and interesting account of one in the parish of Shapinshay is given in the Old Statistical Accounts,[102] by the Rev. Dr. George Barry, the historian of Orkney, in which was found a beautiful torquated ring, evidently of primitive workmanship.

Structures of the same character, on the mainland of Orkney, were explored by Lieutenant F. W. L. Thomas, R.N., while engaged in the Admiralty Survey in 1848. In the course of his investigation of one of these at Savrock, about a mile to the westward of Kirkwall, and close to the sea-shore, some curious evidence was disclosed, showing the primitive arts of its builders, and their inability to overcome an obstacle requiring unusual skill or effective tools. In excavating the site for this subterranean dwelling they appear to have cleared away the soil till they reached the natural rock, which forms the floor of the vault. Pillars constructed at irregular intervals admit of the whole being covered by immense slabs resting on them, where the width is too great to be overarched at so slight an elevation by converging walls. A long passage leads from this chamber, floored, like it, with the natural rock. In one place, however, an irregular elevation of the strata occurs. Such an obstacle was either beyond the skill of the laborious architects, or at least demanded more exertion than they cared to expend on its removal; and the roof has accordingly been elevated so as to admit of free passage by ascending and descending over the superimposed rock. The passages, as in nearly all the structures of this class which have been carefully explored, are extremely straitened. Unfortunately this primitive dwelling supplied materials for building a neighbouring farm-house and offices before Lieutenant Thomas had an opportunity of exploring it; so that what remained was in a very imperfect and dilapidated state. Portions of the roof still entire, constructed of huge masses of unhewn stone—one of them measuring about five feet long—afforded abundant evidence that no amount of mere physical labour was grudged in the completion of the edifice, and seem to justify the probable assignment of it to a period prior to the introduction of metallic tools. In another of these subterranean buildings, however, situated on the Holm of Papey, Lieutenant Thomas observed some doubtful indications of the use of tools. "On the side wall, near the entrance," he remarks, "and about six feet from the floor, there is a neatly engraved circle, about four inches in diameter; there is also another stone, with the appearance of two small circles touching each other, cut upon it; but it is so common to find geometrical figures upon the Orkney flags, arising from a semi-crystallization of the pyrites which they contain, that I am unable to decide whether these are natural or not." The height of the passage where it remains perfect is only two feet seven inches; but nearly one-half of it is unroofed, and heaps of large stones lying scattered about afford evidence of the great extent of the building when complete. Within and around the area of this ancient structure abundant indications were discovered of its having been used as a dwelling-place. A large accumulation of wood or peat-ashes shewed that it must have been occupied for a lengthened period; and this was further proved by the great quantity of the bones of domestic animals scattered about the place. Those of sheep, apparently of the small northern breed still found in Orkney, were the most numerous; but besides these, there were skulls and bones of horses and oxen, the skull and portions of the horns of a deer, and a large bone of a whale. A thick layer of the shells of the periwinkle, L. Littoralis, covered the building and the adjacent ground, mixed sparingly with the oyster, the escallop, the common whelk, and other edible mollusca, which had evidently been consumed in great quantities on the spot. Along with these were also found a few extremely rude implements, the relics of the primitive arts of the builders, besides an antler of a deer artificially severed from the complete tyne. These objects were roughly fashioned from the thigh-bone of an ox, and designed apparently as handles for some weapon or cutting implement, most probably of shell or flint. Other Orkney relics of the same interesting class, but exhibiting more completeness of design, and accompanied with attempts at ornament, are described and figured in a subsequent chapter.

This large, though very imperfect example of the dwellings of primitive communities of the ancient population of the Orkneys, may be properly classed with the weems of the Scottish mainland, though it is not entirely subterranean. The floor is nine feet below the natural surface of the ground; and from the mode by which the whole appears to have been in-roofed with immense overlapping stones, it must have projected somewhat above the surface, and was probably covered over with a raised mound of earth. In this respect it approaches, in some degree, to another class of buildings, which appear to be peculiar to Orkney and the neighbouring districts of Caithness and Sutherland, though it is possible enough that they may have been at one time no less common on the whole Scottish mainland. These structures, for which it may be convenient to retain the popular name of Picts' houses, are not strictly speaking, subterranean, but erected generally on the level ground, or, at furthest, excavated in part out of the side of a hill, so as to admit of a level entrance. Externally they are scarcely distinguishable from the larger tumuli, but on digging into the green mound it is found to cover a series of large chambers, built generally with stones of considerable size, and converging towards the centre, where an opening appears to have been left for light and ventilation. These differ very little from many of the subterranean weems, excepting that they are erected on the natural surface of the soil, and have been buried by means of an artificial mound heaped over them. Barry has minutely described one, which he calls an "ancient Pick house," opened at Quanterness, near Kirkwall.[103] Another relic of the same class was explored during the past year by Mr. George Petrie of Kirkwall, to whose valuable communications on kindred subjects I have already had occasion to refer. Through his kindness I have been favoured with a minute account of the result of his labours, as well as with the plans engraved, drawn from careful measurements taken at the time.

[Drawn by Lieut. F. W. L. Thomas R.N. from Plans by George Petrie, Esq.
PLAN & SECTIONS OF PICT'S HOUSE, WIDEFORD HILL, ORKNEY, Explored in 1849.
Published by Sutherland & Knox, Edinburgh.]

In the month of October 1849, Mr. Petrie's attention was directed to a large tumulus or green knoll, which stands about half-way up the western declivity of Wideford-hill, overlooking the beautiful bay of Frith on the mainland of Orkney, and within a short distance of the Pict's house of Quanterness, described in Barry's History of Orkney. Being on a steep and unfrequented part of the hill, it appears to have almost entirely escaped observation. An opening, however, had been attempted at some former period, but abandoned after an excavation of about a couple of feet in depth had been effected. Mr. Petrie employed men to make a section into the mound, and himself superintended and assisted in the operation, which proved one of considerable time and labour, from the large stones and the quantity of clay used in completing the external mound, as well as in the masonry of the structure found underneath. The building appears to have been constructed in the following manner:—A place for the site having been scooped out of the side of the hill, the cells or apartments were built of large unhewn stones, the walls being made gradually to converge as they rose in height, until they approached to within a foot at top. Externally the work was bounded by a wall of about two feet high. The entire structure was then brought to a conical shape with stones and clay; the stones being disposed with considerable regularity, and over all a thick layer of turf or peat had been laid. The mound which encloses the whole is about one hundred and forty feet in greatest circumference, and forty-five feet in diameter. The work of exploration was commenced by making a cut, six feet in breadth, upon the north side, and clearing away the stones and clay in the direction of the highest part of the mound. On penetrating towards the centre, at about six feet from the top, a stone was exposed placed on edge, about eighteen inches long and nine inches thick, underneath which lay another, which was found to cover a hole of about a foot square, at the top of the chamber marked D in the plan. (Plate I.) On obtaining entrance to this chamber or cell, it proved, like those subsequently opened, to be constructed with walls gradually converging on all sides towards the top, and to measure five feet nine inches in length from north to south, four feet eight inches in breadth, and five feet six inches in height. On the west side of the chamber, the small passage, marked h, was observed appearing to communicate with another apartment, but it was so blocked up with stones and rubbish, that excavation had to be resumed from the exterior. After working for upwards of an hour, the large stone, marked m, was reached, and on removing it an entrance was effected into the central chamber A. This was about three-fourths filled with stones and rubbish, heaped up under the opening marked i, on digging into which bones and teeth of the horse, cow, sheep, boar, &c., were discovered mixed with the rubbish, and also some which were supposed to be those of deer, but not a vestige of human bones.

The general plan will convey the best idea of the form and arrangement of the chambers. The central apartment, A, is an irregular oblong vault, ten feet long, five feet in greatest width, and 7½ feet in height from the bottom to the lower edge of the stones marked o o. Above this extends the opening i, which had no other covering than the outer layer of turf. Mr. Petrie came to the conclusion, after a thorough examination of the whole, that the rubbish found in this large chamber was the debris of some later building erected above the mound, the materials of which must have been precipitated through the narrow opening, as no part of the subterranean structure was found imperfect with the exception of the passage g. From the floor of the chamber to the extreme height of the mound is twelve feet. At the north end of the central chamber the passage e leads to the cell C, measuring five feet seven inches long, four feet wide, and six feet high. From the east side of this a passage extends a considerable way, until it is abruptly terminated by the native rock. The chamber D, which was first entered, communicates with the central apartment by a short passage h, directly opposite to which is the long gallery b, which formed the entrance to the building from the western side of the mound. A third passage, a, proceeds in an oblique direction from the central chamber to a cell B, the proportions of which are six feet in length, three feet seven inches in width, and 6½ feet in height. Nothing found in this or any of the previously explored "Picts' houses" gives the slightest countenance to the idea that they were designed as places of sepulture. The most remarkable feature about them, however, and the one least compatible with their use as continuous dwelling-places, is the extremely circumscribed dimensions of the passages. The whole of them measure about fifteen inches in height by twenty-two inches in breadth, so that entrance could only be obtained by crawling on the ground. The arrangement affords a very striking confirmation of the barbarous state of the people, who were yet capable of displaying so much skill and ingenuity in the erection of these cyclopean structures. It is curious indeed that as civilisation progressed, primitive architecture became not only simpler but meaner, the ingenious builder learning to supply his wants by easier methods; while also the gregarious social ties which such laborious and extensive structures indicate were exchanged for the more refined separation into families, with, as we may assume, the gradual development of those virtues and affections which flourish only around the domestic hearth.

The first step in the descending scale indicative of the abandonment of the cyclopean architecture for simpler and less durable modes of construction, appears in a class of dwellings of similar character to the "Picts' houses," but inferior in their masonry, and generally smaller in size and less complete in design. Examples of this class have also been found in various parts of Scotland. They are generally more entirely subterranean than the "Picts' houses," partaking in this respect more of the character of the weems. They occupy, however, an intermediate position, being excavated for the most part in the side of a hill, so as to admit of an entrance level with the ground. They are also found more frequently in groups, and have probably been each the dwelling-place of a single family.

In these, oaken rafters appear to have supplied the place of the more ancient cyclopean arch, and the walls are generally built of smaller stones. Weems of this more fragile character have been discovered at Prieston, near Glammis, in Forfarshire; at Alyth and Bendochy, Perthshire; and at Pennycuick, Mid-Lothian, as well as in other districts. One in particular, found at Alvie, Inverness-shire, measured sixty feet in length. These may be regarded as works of a later age than the more massive and enduring structures previously described, when the domestic habits of the old builders had survived their laborious arts and monolithic taste. One of the most singular groups of this latter class is a series of contiguous excavations, on the ridge of a hill immediately to the north of Inchtuthill, Inverness-shire, known in the district by the name of "the steed's stalls." Seven circular chambers are cut in the side of a steep bank, separated by partitions of about twelve feet thick. The floors are sunk about twenty feet, and each chamber measures fifteen feet in diameter. A long passage of about four feet wide has formed the original way of ingress, but the rafters, which most probably formed the roof, have long since disappeared, and only a very partial estimate can now be formed of the appearance presented by these singular chambers when complete.

With the same class also may be reckoned certain structures described by Pennant as the repositories of the ashes of sacrifices. One of these, within a few miles of Edinburgh, in the neighbourhood of Borthwick Castle, was brought to light by the plough coming in contact with its rough masonry, at a depth of only a foot below the surface. It may be described as pear-shaped, and with a passage continuing from the narrow end, measuring fifteen feet in length by two and a half in breadth. The masonry was of the rudest description, and nearly the whole space between the walls was filled with a rich black mould, irregularly interspersed with charcoal, fragments of bone, and the teeth of sheep and oxen.[104] A similar building was discovered about the same time in the east of Fife, and one closely corresponding to it has recently been disclosed by railway operations at Newstead, in the neighbourhood of Melrose. In this, as in the example above referred to, the narrow passage pointed nearly north-west; and its masonry was equally rude; but among its contents were various carved stones, apparently corresponding with Roman remains frequently found in that neighbourhood, and one in particular with the cable-pattern, or woollen fillet, so commonly employed by the Anglo-Roman sculptors.

Akin to such subterranean dwellings are the natural and artificial caves which, in Scotland, as in most other countries, have supplied hiding-places, retreats for anchorites, and even permanent native dwellings, and may be described along with this class, though belonging to many different periods. Such caves abound in Scotland, and especially along the coast, but in general their interest arises rather from the associations of popular traditions, than from any intrinsic peculiarity of character pertaining to them. Few such retreats are more remarkable, either for constructive art, or historic associations, than the well-known caves beneath the old tower of Hawthornden, near Edinburgh. They have been hewn, with great labour and ingenuity, in the rocky cliff which overhangs the river Esk. No tradition preserves the history or date of their execution, but concealment was evidently the chief design of the excavators. The original entrance is most ingeniously made in the shaft of a very deep draw-well, sunk in the court-yard of the castle, and from its manifest utility as the ordinary and indispensable appendage of the fortress, it most effectually conceals its adaptation as a means of ingress and communication with the rock chambers beneath. These are of various forms and sizes, and one in particular is pierced with a series of square recesses, somewhat resembling the columbaria of a Roman tomb, but assigned by popular tradition as the library of its later owner, Drummond the Scottish poet. Whatever was the purpose for which these were thus laboriously cut, the example is not singular. A large cave in Roxburghshire, hewn out in the lofty cliff which overhangs the Teviot, has in its sides similar recesses, and from their supposed resemblance to the interior of a pigeon-house, the cavern has received the name of the Doo-cave. Authentic notices of the Hawthornden caves occur so early as the reign of David II., when a daring band of Scottish adventurers made good their head-quarters there, while Edward held the newly fortified castle of Edinburgh, and the whole surrounding district. In the glen of the little river Ale, which falls into the Teviot at Ancrum, extensive groups of caves occur, all indicating, more or less, artificial adaptation, as human dwellings; and in many other districts similar evidences may be seen of temporary or permanent habitation, at some remote period, in these rude recesses. Along the coast of Arran there are several caves of various dimensions, one of which, at Drumandruin, or Drumidoon, is noted in the older traditions of the island as the lodging of Fin M'Coul, the Fingal of Ossian, during his residence in Arran. Though low in the roof, it is sufficiently capacious for a hundred men to sit or lie in it. In this, as in the previous example, we find evidences of artificial operations, proving its connexion with races long posterior to those with whose works we have chiefly to do in this section of archæological inquiry. In the further end a large detached column of rock has a two-handed sword engraved on it, surmounted by a deer, and on the southern side of the cave a lunar figure is cut, similar in character to those frequently found on the sculptured pillars and crosses which abound in Scotland. It is now more frequently styled the king's cave, and described as the retreat of Robert the Bruce, while he lurked as a fugitive in the Western Isles; but like many other traditions of the Bruce this seems to be of very recent origin. Other caves in the same island are also of large dimensions, and variously associated with popular traditions, as, indeed, is generally the case wherever subterranean retreats of any considerable extent occur. Some are the supposed dwellings of old mythic chiefs, whose names still live in the traditional songs of the Gael. Others are the retreats which the primitive confessors of Scotland excavated or enlarged for their oratories or cells. Of the latter class are the caves of St. Molio, on the little island of Lamlash, or the Holy Isle, on the east coast of Arran; of St. Columba and St. Cormac, on the Argyleshire coast; of St. Ninian, in Wigtonshire; of St. Serf, at Dysart, on the Fifeshire coast; and the celebrated "ocean cave" of St. Rule, in St. Andrew's Bay. This last oratory consists of two chambers hewn out of the sandstone cliffs of that exposed coast. The inner apartment is a plain cell, entered from the supposed oratory of the Greek saint. The latter is nearly circular, measuring about ten feet in diameter, and has a stone altar left hewn in the solid rock on its eastern side. Possibly the singular dwarfie stone of Hoy, in Orkney, owes its origin to a similar source. A huge mass of square sandstone rock, which appears to have tumbled from a neighbouring cliff, has been hollowed out into three apartments, with a fire-place, vent, stone-bed, pillow, &c. The traditions of the island preserve strange tales of a giant and his wife who dwelt in this abode, and the "Descriptio Insularum Orchadium," written by Jo. Ben., (John the Benedictine,) in 1592, adds to the account of its internal accommodation the following somewhat whimsical provision for the comfort of the latter,—"Tempore camerationis fœmina gravida fuit, ut lectus testatur; nam ea pars lecti in qua uxor cubuit effigiem habet ventri gravidi." Others of the Scottish caves and oratories are less artificial in their character. They are especially abundant in the Western Isles, and on the neighbouring coast, where the waves of the Atlantic have wrought out caverns far surpassing in extent and magnificence the largest in the interior of the country. Few of these, however, possess such marked features as to distinguish them from similar relics pertaining to no definite period, which are to be met with on every rocky coast exposed to the rude buffets of the ocean waves. One exception, indeed, may well claim to be singled out as unmatched by any other work of nature or art, though belonging to an older system than the primeval period of the archæologist. Amid scenery unsurpassed in the interest of its historic associations, or its venerable relics of medieval skill, stands the wondrous natural cave which popular tradition has associated with the favourite name of Fingal.

"Nor doth its entrance front in vain
To old Iona's holy fane,
That nature's voice might seem to say—
Well hast thou done, frail child of clay!
Thy humble powers that stately shrine
Tasked high and hard—but witness mine!"[105]

To those who are curious in investigating such ancient relics, Chalmers furnishes a very ample list of "Natural Caves in every part of North Britain, which have been improved into hiding-places by artificial means."[106] The associations with many of these retreats are of the most varied and romantic character; and few districts of the country are without some wild or thrilling legend or historic tradition relating to such caverned shelters of the patriot, the recluse, or the persecuted devotee.