and on the banks of the Amazon they are also to be seen perched on piles driven into the muddy bottom. The delta of the Parana, which is scarcely above high-water mark, is called the Venetia of South America. Here the houses are built on piles in order to keep the flooring free from the sudden rising of floods. Near Rosario, in Buenos-Ayres, gauchos, who have given up horses and horse stealing and taken to canoes and petty larceny, have constructed rude abodes in the swamps along the banks of one of the tributaries of the La Plata.[29] Captain Burton mentions an African tribe who had erected their dwellings in the same way at a great distance from the shore.[30] Lake Moyhrga or Realmah, in Central Africa, presents at the present day the characteristic features pertaining to the descriptions of an ancient Irish lacustrine site. Cameron[31] describes this lake as lying in a small basin embosomed in low-wooded hills, interspersed with patches of cultivated ground. The lake is small—only two miles in length by one in width—the margin girt with floating vegetation corresponding to the reeds, flags, and aqueous growth of more temperate climes. Three villages, besides detached huts, were scattered irregularly over its surface: the huts were built on platforms raised about six feet above the surface of the water, supported on stout piles driven into the bed of the lake; underneath the platform canoes were moored and nets hung to dry. Some huts were oblong in shape, others round, with occasionally, a projecting roof over the door, and they were constructed in a manner precisely similar to those on terra firma. The inmates live in these dwellings with their fowl and goats, only going ashore in their “dug-outs” to cultivate their small patches of tillage, and to allow their goats to graze. These lake-dwellers were very unapproachable when on land, always running to their canoes and paddling away on the appearance of a stranger. The paddles are described as resembling large circular shallow spoons with long straight handles, the canoes about 25 feet in length. In Eastern Africa also pile dwellings are to be found, as at Mayunga, in Madagascar. At Mirambaus, in Eastern Africa, the score or so of houses composing the village are constructed on posts raised some 8 feet above the ground,[32] whilst in the neighbouring village of Mujurumbe huts formed of bamboo and cocoa-nut leaf are erected on piles. Thus we see that a habitat on the water has been, or is still, used by races of men not having the slightest affinity to each other, as the boat-living population of China, the lake-dwellers of Central Africa, Europe, New Guinea, Great Britain, Ireland, &c.

Although in 1810 a lake-dwelling in Ireland had been discovered by Mr. William Trench near Roscrea, Co. Tipperary, its original use was not known, no interest was excited, nor was the importance of the event appreciated; and it was not till 1839[33] that the archæological value of the investigation of lacustrine sites was perceived. The following year Mr., afterwards Sir William, Wilde published an account of the discovery at Lagore, Co. Meath, of the first great lacustrine retreat noticed in Ireland in modern times. It had been anciently the residence of the chief of a small territory in Meath. The formation or eruption of the lake is chronicled in the Irish Annals as having occurred A.M. 3581. Under dates A.D. 848, 933, 967 mention is made of the destruction of the Lagore crannog, and from its being situated near the village of Dunshaughlin, either name is indiscriminately applied to it in the present day. W. F. Wakeman, the only archæologist probably now living who can speak from personal knowledge of the circumstances connected with the original discovery, relates that, when he was with Petrie in the office of the Topographical Department of the Ordnance Survey of Ireland, he recollects how a dealer in Irish antiquities, named Underwood, made his appearance day after day, bringing for sale objects of more or less archæological interest, which he stated had been found near Dunshaughlin. Struck by the frequency of the dealer’s calls, Petrie decided to unravel the mystery by visiting the place; accordingly, accompanied by Wilde, he reached Lagore House, and there, carelessly thrown on the floor of a barn, he saw a large and miscellaneous collection of antiquities, consisting chiefly of weapons composed of iron—swords, daggers, spear and axe-heads, saws, chains, shears, small culinary vessels, &c. &c; pins and brooches of bronze, articles of bone, and even of wood, all of which had been found by labourers engaged in making a drain through the ancient lake bed. Petrie endeavoured to procure some of these for his collection, but the steward, or caretaker, in the absence of his employer, would not part with any; and when the dealer returned after a short lapse of time the collection was already greatly scattered: many objects had been presented to curiosity-seekers in the neighbourhood, and many, it was supposed, went to England. A few, however, found their way to the Royal Irish Academy, to Dean Dawson, Lord Talbot de Malahide, and Petrie.

For some years after the drainage operations, the soil of Lagore remained unturned by the spade; but in 1846, 1847 and 1848, the site of the crannog was reopened by men engaged in the process of turf-cutting, and, as on the previous occasion, quantities of bones were exhumed, and with them a surprising number of antiquities, together with remains of the ancient stockading, and the ruins of several structures evidently used as huts; one of them is thus described by W. F. Wakeman: “Let the reader imagine a foundation formed of four roughly-squared planks of oak, each about twelve feet in length (so arranged as to enclose a quadrangle), the ends of which were carefully fitted together. From the angles of this square rose four posts, also of oak, to the height of about nine feet. In these grooves were cut, into which roughly-split planks of oak had been slipped so as to form the sides of the house; the irregularities between the boards were tightly caulked with moss; a low and narrow opening in one of the sides had evidently served as an entrance. There were no traces of window or chimney.”[34]

Lagore crannog differed from others brought to light at a later date, in not being either submerged or surrounded by water at the time of its discovery. Wilde describes it as consisting of a circular mound of about 520 feet in circumference, slightly raised above the surrounding bog or marshy ground, which forms a basin of about a mile and a-half in circuit, and is bounded by elevated lands. “The circumference of the circle was formed by upright posts of black oak, measuring from six to eight feet in height; these were mortised into beams of similar material laid flat upon the marl and sand beneath the bog, and nearly sixteen feet below the present surface. The upright posts were held together by connecting cross-beams, and (said to be) fastened by large iron (?) nails: parts of a second upper tier of posts were likewise found resting on the lower ones. The space thus enclosed was divided into separate compartments by septa or divisions that intersected one another in different directions; these were also formed of oaken beams in a state of great preservation, joined together with greater accuracy than the former, in some cases having their sides grooved and rabbeted to admit large panels driven down between them.”

It may be inferred that fire was the final agent of destruction, as almost everywhere amongst the timbers lay half-consumed logs and numerous pieces of charcoal. Unfortunately, investigations on the site do not seem to have produced results such as might have been expected, owing principally, as was supposed, to want of friendly co-operation on the part of those engaged in searches. The “find” is now widely dispersed, and can never be re-collected; but the Petrie Museum deposited in the Royal Irish Academy, as well as some private cabinets, would still furnish material for scores of illustrations.

At about the same date as the Lagore crannog, that of Lough Gur, Co. Limerick, became visible when the waters of the lake were lowered. The existence of this artificial island in ancient times was traditionally known to the neighbouring peasantry: the stones by which it was overlaid were in places distinctly visible to a person passing over them in a boat when the lake level was low and the water calm. A large tripod composed of iron was at times plainly to be seen, and had attracted Mr. Wakeman’s attention; but, from superstitious feelings on the part of the boatmen by whom he was accompanied, no attempt at its recovery was then made; it is said to be now in the collection of the Royal Irish Academy. A number of antiquities have been from time to time obtained from the site, and there was a vast collection of bones; but, according to the evidence of a clergyman then living in the neighbourhood of Bruff, there does not appear to have been any vestige of staking on or around this crannog. Other remains of the same class were brought to light in rapid succession, so that in 1857 no less than forty-seven of them had been recorded: since that time the number has increased to about 200, and every succeeding year adds to the total.

In 1853-4, owing principally to the writings of Keller, the entire attention of archæologists was attracted to the lacustrine remains of Switzerland, and it was the more easily diverted from the Irish sites by the pronouncement of Sir William Wilde, that no stone, and but few bronze weapons, had been found in them: he also suggested, that from the ninth to the seventeenth century might be considered the probable date of the lacustrine period in Ireland. These theories have since been abundantly confuted; but they served at the time to check investigation, because detracting from the impression of antiquity.

Crannog, derivation of the word.—In the Irish Annals, lake-dwellings are called crannogs, derived from the term crann, which signifies a tree. That word, always spelled with a double n, is in Irish generally applied to a tree with foliage and branches, as crann na coille, trees of the wood, but in its primary meaning it refers to some massive object of timber standing erect—as stems of the forest, or a ship’s mast, which is also called crann. The last syllable in crannog, i.e. og, is the terminal form of numerous Irish words; “sometimes it carries the force of a diminutive, but more frequently not. Crannog is no doubt formed in the same manner as cuaróg, the name usually applied to a wild bee’s nest, and derived from the noun cuar, any conical hollow with the diminutive ending og attached thereto. In some country places, the old kind of pulpit or form was called crannóg, and in others crannghail or crannghaoil, a word of somewhat similar meaning. This latter word, too, was the Irish term for a hurdle, and was commonly used for those wicker-chimneys so common formerly in country cottages.”[35]

It is doubtful whether the term crannog was originally applied to the timber framework of which the island was constructed, or to the wooden huts erected on it; though now-a-days it is generally understood to include the whole structure, both island and dwelling. In its topographical sense, the word is applied to wooden lake dwellings, but in another sense the Anglo-Irish employed it to designate a basket, hamper, or measure of a certain size for measuring or gauging grain or corn.[36] G. H. Kinahan states[37] that, although “crannog” is now the generally-accepted appellation for the ancient lake dwellings of Ireland, it is, nevertheless, “a modern term introduced to supply the place of the ancient one, which is unknown or unrecognized.”

Lake dwellings bore in Scotland the same designation. In a document dated 14th April, 1608, directed to State officials, concerning the surrender of some rebellious clans, it is ordered, that “the haill houssis of defence, strongholdis and cranokis in the yllis perteining to thame and their foirsaidis sal be delyveret to his Maiestie,” &c. Although the term crannog is to be found in the Irish Annals, yet in the earliest entries these dwellings are designated simply as Inish, i. e. island: for instance, in an old Irish MS., “The wars of the Gaedhiel with the Gaill,” it is recorded that, in the year 1013, Brian Boru repaired inir locha Gair, i. e. the island of Lough Gur, county Limerick, which is one of the most important as also apparently one of the most ancient Irish lacustrine sites; for, as before stated, remains of the reindeer, the Megaceros Hibernicus, and bear, were found in the lake bed.

Crannog, a common Townland Name.—There are numerous localities throughout Ireland in which the term “crannog” is embodied in the name, and where, consequently, must have been formerly a lake or swamp, with its accompanying artificial island, although in some cases the lake has now disappeared, and the swamp has been drained. In most of the districts in which these islands were found several small lakes are clustered together. In Connaught, near the demesne of Longford, county Sligo, the residence of Sir Malby Crofton, Bart., in a small pond, almost dry in summer, there is an islet, still called by the country people “crannog:” it has bequeathed its name to the townland in which it is situated, i.e. “Lochanacrannog,” signifying the little lake of the crannog.[38] In the parish of Cloonclare, county Leitrim, is a locality called Crannog Island. In Connemara, county Galway, there is an island in Ballinahinch lake which has given name to a parish and to a barony, the word Ballinahinch signifying village of the island. In Ulster there is a townland in county Tyrone called Crannog. In county Donegal, Crannog-boy, or the yellow lake dwelling, was the head quarters of The O’Boyle, a steady supporter of the O’Donnels. In county Antrim, Lough Crannagh, situated over Fair Head, probably derives its name from a small artificial island which lies in the centre of the lake.[39] In Loughmacnean, county Fermanagh, are several artificial islets, one of which is called “Crannog Island.” In Leinster there is a townland in county Kilkenny called Cronoge. In Munster there is a place in the parish of Ardagh, county Limerick, called Coolcranoge, the angle of the lake dwelling. “Crannog Island,” “sunken island,” “drowned island,” “Inish,” or “the island,” are common designations for diminutive artificial islets, showing themselves above water only during summer drought in many lakes throughout the kingdom. There is hardly a large sheet of water to which is not attached a tradition of a frightful outbreak of flood, covering what was formerly “a town;” or which does not possess its legend of an enchanted well, which, consequent upon some affront offered to its guardian spirit, covered the valley, its inhabitants, and houses. May not these traditions be traceable to lingering remembrance of former lacustrine habitations, for Giraldus Cambrensis, writing in the twelfth century, described the tradition that then prevailed in the north of Ireland, of waters having overwhelmed the plains now occupied by Lough Neagh, a locality the most thickly studded with these remains; and the legend has been immortalized by Moore, who thus alludes to it:—

Similar legends of submerged towns are prevalent in the south of Scotland, on the littoral facing the Irish coast. In the neighbourhood of Carlingwark Loch, Kircudbrightshire, in which are sites of crannogs, there has been from time immemorial a tradition that “there had been a town in the loch which sunk or was drowned;”[40] whilst a Scotch writer states that, “amongst other points of coincidence between Scotch and Irish crannogs may be noticed a tradition connected with some of them, common to both countries, which seems to have arisen from the submersion of the island homes by the rising of the waters in the loughs.”[41] The same idea is held by the natives of Central India: “from this hill is heard the sound of fairy drums; in that lake are seen reflected the ruins of a buried city.”[42]

The tradition of towns buried beneath the waters is not confined merely to the lakes of Ireland: there is the beautiful fable of the City of Gold, hid beneath the angry ocean, sometimes seen, but ever in different localities:

Submarine Crannogs.—After a very high tide at Ardmore, near Youghal, the waters retired more than customary, disclosing the fact that this particular portion of the sea-shore had been the site of a forest, as remains of trees were found in various parts of the submarine deposit. This submerged tract extends to between the four and five fathom line, but it has not been ascertained to what further distance it may stretch seaward. A bank of shingle having shifted by a change of current, laid bare the substructure of an undoubted crannog: at high water it was covered by the tide to a considerable depth. Either this dwelling had been erected when Ireland was joined to Great Britain, or it was existent when Ireland was at a greater elevation above the sea, and therefore of greater extent than at present. The theory is enunciated by eminent geologists that many of our present harbours had been inland lakes in prehistoric times, and thus the Bay of Ardmore may have formerly presented a similar appearance to the Gulf or Lake of Maricaibo in Venezuela at the present day.

In the Irish Records, lake dwellings are likewise designated “Stockaded Islands,” and in the Ulster Inquisitions of 1605, the term applied to them is insula fortificata. Although antiquarians have differed in opinion respecting the age of these remains, yet after patient analysis of the characteristic features of the numerous excavations made in recent years, the weight of evidence seems to indicate that these constructions were of all ages, some being very ancient; it is quite apparent that they have been built and rebuilt, and in them have been found implements of stone, bronze, and iron in their respective strata. It would seem, indeed, that the “age of bronze” had in many instances largely overlapped that of iron. The chronology of this period is a subject that has not as yet been determined, and the time when it commenced and when it ended is still unsettled.

Favourite Sites for Crannogs.—Marshes, small loughs surrounded by woods, and large sheets of water, were alike suitable for the home of the Irish lake-dweller, his great and primary need being protection; he was bound by no conventional engineering rule, he did not exclusively employ wood, but appears to have been guided by surrounding circumstances. On peaty or muddy sites a wooden substructure was essential; on hard bottoms, stone, gravel, or earth were, if convenient, employed. As providing good fishing grounds, the entrance or exit of stream from lake was a favourite site, and natural shoals thus placed were eagerly selected.

Mode of Construction of.—Having decided on the position, the crannog builder set to work by driving stakes into the bottom of the lake in a circle of from sixty to eighty feet in diameter, a considerable length of the stake sometimes projecting above the water; these were in many instances joined together by horizontal beams, the interior filled up by branches of trees, stones, gravel, earth, and bracken. Often an inner row (or more than one) of piling is found about five feet distant from the outer, and piles are driven in various parts of the interior, either to consolidate the mass or to act as stays for the walls of the dwelling. Next were placed one or two layers of round logs, cut into lengths of about six feet, generally mortised into the upright piles, kept in position by layers of stone, clay, and gravel. In some cases, where the foundation was soft, the superincumbent layers of timber were of great depth; in other cases, where the bottom of the lake was firm, the platform of timber was confined to a portion of the island. The side most affected by the action of the water was frequently strengthened by rows of piles, sometimes five or six deep, as well as by a breakwater of stones.[43] On the foundation, when raised sufficiently above the water, the dwelling was erected; the hearth was in the centre of the island, for in almost every case a collection of flag-stones has been discovered in the interior of the enclosure, bearing on them marks of fire. At times several hearths occur. In some instances there are indications of these structures having had additions made to their height at various times, either to keep pace with the chronic rising of the level of the lough; or, taking into consideration the compressible nature of the component parts of the foundation, the island may have required increased elevation, owing to the effect of natural subsidence. The enormous amount of wood employed in the formation of a crannog, despite the mass of stones and other material with which it was laden, must have had at first a buoyant tendency, but according as the wood became completely saturated with water, consolidation of a more or less pronounced character would, in the course of time, become observable. The stones deposited over the lower strata of fern, bracken, and branches, were indubitably used to compress and solidify the substructure, as likewise were the stones on the upper strata used for a similar purpose. Considerable ingenuity was displayed in the formation of these island homes, which were frequently constructed in a depth of twelve to fourteen feet of water; and, apart from having served in their day as secure retreats for large numbers of persons, they have proved their durability by resisting successfully the ravages of time, which may be reckoned by centuries.

The Scottish lake dwellings were formed in almost identically the same manner; in fact, the structural details are so completely analogous as almost to necessitate the belief of their having been erected by the same race. Up to the present time, out of the fifty-three lake dwellings constructed of timber, discovered in Scotland, thirty-three have been found in Wightown,[44] Kirkcudbright,[45] Dumfries,[46] Ayr,[47] and Bute,[48] in south-western Scotland, whilst the greater number of those found in Ireland are situate on the opposite coast, i.e. in Ulster.

G. H. Kinahan, who explored crannogs in four localities,[49] states that in all of them floors of wicker- or basket-work were found. The first discovered by him was in the large crannog of Loughrea. In one of the excavations there became apparent a perpendicular single wicker-work wall or partition that went down to the level of the basket flooring; from it, for eleven feet towards the north there was a rough pavement, on which was a thin layer of gravel. The surface of the pavement was on a level with the basket flooring. The rods of which it was composed were soft and rotten—in fact reduced to mould, gave no resistance to the spade, and might easily escape ordinary observation. The heads of the piles forming the partition walls may, according to this authority, be seen in an unexcavated crannog near Strokestown, county Roscommon. There seems, therefore, reason to suppose that all piles situated in the interior of crannogs point out the direction of the partition wall or walls of the habitations, as, whatever may otherwise have been the mode of construction—whether of wicker-work or rods—they were supported at intervals by piles. May not the so-called basket flooring have been a portion of the wattle walls which had fallen down? Other experts on the subject have been unable in their explorations to detect the presence of wicker-work floors; but in the year 1858, in a small island about 200 yards from the shore, in the lake formed by the Shannon at Castleforbes, county Longford, the Earl of Granard discovered traces of stockading and piling, together with a coarse sort of wicker-work.[50] This, however, was in the encircling palisades, and W. F. Wakeman has recently noticed similar remains around the crannog of Lisnacroghera.

Stone Lake Dwellings.—On some of these artificial islands the last structures seemed to have been formed of stone: for example, in Loughtamand, county Antrim, a stone house or castle, the stronghold of the MacQuillans, replaced a circular structure composed of wood; a similar change seems to have been made in one of the crannogs of Loughrea. The transition from buildings of wood to those of stone has also been observed in Scotland. Mention may be made of a few stone-built island fortresses; and although, strictly speaking, the term “crannog” is scarcely applicable, yet many of them that present a modern appearance are structures erected on ancient foundations. In parts of Galway and Mayo, where timber was either scarce or of stunted growth, buildings of stone seem to be most numerous, whilst in Ulster they occur in the rocky districts of Antrim and Donegal.

Goromna Island, in Lough Hilbert, county Galway, is a peculiar structure, which, though not formed of wood, is somewhat allied to a crannog, being wholly or in part an artificial island. A large and good specimen, Caislen-na-Caillighe, or the Hag’s Castle, stands in Lough Mask. It is one of the oldest fortresses mentioned in the Irish Annals, being noticed at the date A.D. 1195. In 1233 the Anglo-Norman castle erected on its site was demolished by Felim O’Conor, chief of Connaught, and so late as 1586 it was with difficulty captured by Sir Richard Bingham. This great circular enclosure, ninety feet in diameter and thirty feet in height, occupies almost the entire island. The walls, which are still eight feet in depth at top, exhibit the characteristic inward slope peculiar to the cashel; its situation near the mouth of a river (the Robe) is characteristic of crannogs. In Lough Bola there is a curious cashel, or stone lake dwelling; the accompanying sketch gives an idea of its present appearance. A primitive habitation lies in Lough Cam, north of Roundstone, and two miles west of Toombeola. All these islands have a crannog-like aspect. To the south of Ballinahinch lake lies that of Ballinafad, and in its northern portion there becomes visible, when the water is low, a circle of stones and a small island, evidently the remains of some artificial structure. O’Flahertie, in his History of Iar-Connaught, mentions that the ancient castle of the O’Flaherties of Bunowen, in the lake of Ballinahinch, was built on an artificial island: this had been constructed by one of the original septs long prior to the occupation of the country by the O’Flaherties, who in their turn were driven out by the Martins.[51]

Theory of Crannogs being only Temporary Refuges.—The opinion has been frequently advanced that crannogs were merely refuges used in case of hostile incursion—not a general or habitual residence; and if a rath chance to be in the immediate vicinity, it is pointed out as the abode of the chief on shore. The great mass of bones, however, and remains of household gear, found upon and around some of the sites, point conclusively to their having been places of residence for long and various periods, and many of them bear signs which denote their having been often demolished, burnt, and rebuilt, the Celt clinging to his watery home with as much pertinacity as in latter days he clings to his cottage on terra firma. Keller was of opinion that the great distinction between the continental pile dwellings (pfahlbauten), and the crannogs of Ireland and kindred lake dwellings of Scotland, is that the latter served merely as places of refuge for chieftains, their families, and property, i. e. strongholds belonging to individuals, whilst the Swiss lacustrine dwellings were inhabited century after century by groups of families; here they fabricated their pottery, their utensils, their wearing apparel, their fishing implements, their weapons of warfare; and hence in Swiss waters rows of huts are found, each furnished with its hearth and its weaving loom. Although Irish and Scottish lacustrine dwellings may have been the abode of chiefs, yet the clan clustered around them, for in many Irish lakes a large crannog is observable, whilst in close proximity—sometimes connected with it—are others of smaller size and meaner construction; might we not fairly surmise that the larger one was the abode of the chief, the others those of his followers? The theory that these dwellings were places of merely temporary refuge can be refuted on many grounds. The evidences of repeated renovations and re-erection of crannogs after having been burnt,[52] also the traces of cattle, which appear to have been kept in them, the enormous quantity of bones contained in the kitchen middens, the hearths, the remains of fire, of cooking utensils, of spindle whorls, of household gear, the domestic arrangements made in each hut for the accommodation of a family, the various articles for personal adornment and for amusement, all tend to denote continuous occupancy; finally, we must bear in mind the enormous expenditure of time and labour that would be required for erection of what was, on such a theory, to serve only as a place of temporary shelter.

Palisades and Dwellings.—Palisaded fortifications are unquestionably very primitive. Around the cyclopean wall of Dun Ængus and other prehistoric forts in the Isles of Arran, county Galway, are found palisades, or chevaux-de-frise, of sharp-pointed stones, and such defences have in all ages proved effective under skilful management. Even in the present advanced state of military science, the various wars in New Zealand have demonstrated the formidable nature of the Maori pah or fort. The outer range of piles around crannogs rose considerably above the water, and thus formed a stockade or breastwork for repelling an attack from enemies. The Lord Deputy Sidney, describing to Elizabeth a repulse of her troops from one of these retreats, says that at a distance it did not appear formidable, as it was simply encircled with a stout palisade bristling with rows of sharpened stakes; but when assaulted, the soldiers found it impossible either to scale or undermine it. This style of defensive work was of ancient origin in Ireland, for the Annals state that, in the year 990, the island as well as “the dreach and rampart of Lough Cimbe” (now Lough Hackett, county Galway) were swept away by a violent storm.

Within the area enclosed stood the hut or huts in which the families lived; the stockade served equally for shelter and defence, fulfilling the same purpose as did the circumvallation of the rath or doon on terra firma. Edifices constructed of logs, of wattling, and hurdles daubed over with clay, and thatched with reeds, were in early times considered characteristic of the Irish; even public buildings were constructed of these materials. Bede states that an Irish monk, who had been elected Bishop of Lindisfarn, built a church for his see “altogether of sawn oak, and covered it with reeds, after the manner of the Scots” (i. e. the Irish). We need not, however, refer to history, or depend upon conjecture, in order to reconstruct these island dwellings; for the foundations, and even some of the log walls, have been exposed to view. Good examples are presented by the flooring of an oblong house at Drumaleague, county Leitrim, and at Cargaghoge, county Monaghan; remains of a wooden hut at Kilnock, county Antrim; and the following in county Fermanagh—mortised beams at Loughavilly, a house at Gortalough, angle posts at “The Miracles,” and lower framework of a house at Ballydoolough. At Kilnamaddo, in the same county, log huts were found buried under seventeen feet of peat; these were practically almost perfect, wanting nothing save the roof; they were very low, the side walls scarcely four feet in height, and they might be looked upon rather as lairs for sleeping in, than dwellings in the modern sense of the word; indeed, the primitive races of Ireland, whether building in stone or wood, made use of low roofs and consequently low doors. The openings left for egress and ingress were probably closed by hurdles of wicker-work, as demonstrated by the old Irish proverb, Fuair se air suibhal eadar cliath a’s ursainn (he got off ’twixt hurdle and door-post), equivalent to the saying, “he escaped by the skin of his teeth.”

A considerable portion of the townland of Kilnamaddo (the wood of the dog) had apparently, in olden times, formed the basin of a sheet of water, and upon one of its shoals some primitive tribe had erected a habitation. The piling can still be traced, but the chief antiquarian interest attached to the discovery consisted in the remains of the huts already referred to, and which were constructed of oak beams. They were two in number, stood about fifty feet apart, and somewhat resembled the hut exposed to view by Captain Mudge in Donegal, but differed from it in not being divided into an upper and lower story. The Kilnamaddo huts were quadrangular, and the larger and more perfect specimen measured on the outside eleven feet six inches by ten feet. To make the structure, four massive posts of oak, averaging seven feet in length and seven feet in circumference, were set in the ground. These timbers, near their upper ends, have mortised holes averaging eleven inches in height by eight and a-half in breadth, through which passed the ends of beams to which slabs of oak were attached, and the floor was also composed of oaken planks. The roof, as well as a great portion of the sides, did not remain in situ at the time of the discovery, but a number of timbers found immediately adjoining each of the structures were admirably suited for the purpose of forming side walls and roofing. The lower frame of the work appears to have been very similar to that upon which the roof had rested. In the lower portions of the four upright posts were very curious mortise holes, evidently intended for the reception of beam ends. The sides of this edifice were retained in position and pressed to the frame below by a row of small, well-sharpened piles; no sign of pinning or grooving in the upper portion of the frame could be discovered, therefore the conclusion may be drawn that the huts were originally built within an excavation, the peat was then heaped against the sides, probably even over the roof, and a small passage left as an entrance. In short, it was a structure of wood, but otherwise analogous to the stone-lined chambers and passages so generally found in raths. Almost in the immediate neighbourhood, may still be observed a subterraneous work of uncemented stones, styled “St. Lasser’s Cell.” In it are three very small, low-roofed chambers, connected by passages equally low; but there is no reason to suppose they had communication with any other structure, and, although named after a “saint,” they rather convey the impression of pre-Christian origin. Unlike the remains of most crannogs, that of Kilnamaddo presented no indication of having been destroyed by fire; had such been the case, the posts would have shown undoubted signs of charring. Evident traces of a kitchen midden were discovered in the bog close to the hut, but at a level slightly above it. The height of the interior of the most perfect of the huts, without making any allowance for a sloping roof, being only four feet, they would thus correspond to the souterrains of a rath. It has been suggested that they were used by the crannog occupants only for the preservation of perishable commodities; and in seeming corroboration of this theory, large lumps of the substance called “bog-butter” were discovered in them carefully rolled up in cowhide.

It is difficult to arrive at even an approximate estimate of the age of these curious structures: the mortise-holes, pierced clean through timbers more than seven feet in circumference, could hardly have been formed by the aid of stone hatchets or chisels: several of the piles and other portions of the work show traces of long, clean cuts, such as are given by sharp metallic tools; yet the only implements found here were of stone, and by the aid of such it is possible that some of the timbers were more or less worked. The roots and part of the trunk of a yew were found in situ in the bog, on a higher level than the roof of the more perfect hut. It has been computed, from the girth of this yew, that it would take at least one thousand years to attain the bulk it showed when first laid bare, and the eleven feet of bog by which it was covered would take at least another thousand years to grow.[53]

Plate I., fig. 4, represents the most perfect of the Kilnamaddo huts restored. When the drawing was taken, all the timbers in the illustration were on the spot, and some of these retained their original position. Fig. 3 is a slab of oak measuring six feet by two feet nine; it is not so thick as fig. 2, which represents a slab of oak five feet ten inches in length by one foot six inches in breadth, the depressions at its broader end being five inches by six, and it is one foot two inches in thickness; this and fig. 3 were found lying by the side of the hut, and it is conjectured that they formed a portion of the roof. Fig. 1 represents an angle-post five feet six inches in length, and four feet ten inches in circumference. This style of construction appears, as far as is yet known, to have been confined to the north-west portion of Ireland, the two somewhat similar huts being those discovered by Captain Mudge in Donegal, and by Mr. Morant in Monaghan.

A representation is given (plate II.) of the wooden hut discovered in 1833, by Captain W. Mudge, R.N., in the bog of Drumkelin, parish of Inver, county Donegal, it being the most perfectly preserved primitive dwelling of that material yet brought to light in Ireland. It was surrounded with a staked enclosure; portions of the gates also were discovered. The flooring of the house (plate III., fig. 3) rested on hazel branches covered with a layer of fine sand; a paved causeway (plate III., fig. 4), over a foundation of hazel branches and logs, led from the door of the house to a fireplace, on and around which lay ashes, charred wood, and half-consumed turf. This unique structure was nearly square, twelve feet wide, and nine feet high, formed of rough logs and planks of oak, apparently split by wedges, the interstices filled with a compound of grease and fine sea sand. One side of the hut, supposed to be the front, was left entirely open. The framework[54] consisted of upright posts and horizontal sleepers, mortised at the angles, the end of each post being inserted into the lower sleeper of the frame, and fastened with a large block of wood. The discoverer states that the mortises were very rough, as if made with a kind of blunt instrument, the wood being bruised rather than cut, and it may be inferred that a stone celt found lying upon the floor of the house was the identical tool with which the mortises had been formed (plate III. figs. 6 and 7). By comparing the chisel with the marks of the tool used in making the mortises and grooves, it was found to correspond exactly with them, even to the slight curved surface of the chisel; but the logs had evidently been hewn with a larger instrument in the shape of an axe, undoubtedly of stone, as the marks, though larger than those the chisel would have made, are of the same character, being somewhat hollow and small cuts, not presenting the smooth surface produced by a common iron axe. A second but larger stone was also found on the floor, which, being ground at one end to an edge, was probably used as a wedge for splitting timber (plate III. fig. 5). There was also a flat freestone slab, measuring three feet by one, and two inches thick, having a hollow in the middle about three-fourths of an inch deep; large quantities of whole and broken nuts were found near it, together with several round shingle stones. The hollow in the flag is supposed to have been a receptacle for the nuts, the small stones being used as hammers to break them.