Chapter XVIII: Part 18

"The piles are of the same scantling, also of oak, but somewhat longer than those of London Wall; the points are driven into the gravel; the peat is three to four feet thick; large beams of the same size as the piles have been laid across them horizontally, and Roman pottery is found at all depths in the peat. Judging from the extent over which these piles have been discovered, there can be little doubt that in digging for the foundations of the many large warehouses and other buildings that are now being built within this district the remains of early habitations are constantly turning up and are destroyed without receiving attention."

As to the relics from these London pile-dwellings let me finally observe, that, to a certain extent, both in character and surrounding conditions they correspond with those from the Terp mounds in Holland and North Germany, from which it is probable the earliest Anglo-Saxon invaders hailed.

CRANNOG IN LLANGORSE LAKE, NEAR BRECON, SOUTH WALES.

Only one lake-dwelling has hitherto been recorded in Wales, viz. that of Llangorse. The partial exploration to which it has been subjected was undertaken by the Rev. Mr. Dumbleton, and the results are recorded by him in the _Archæologia Cambrensis_ for 1870 and 1872. (B. 173.) The following extracts from these reports clearly show that the island was entirely artificial and constructed after the manner of the Scottish and Irish crannogs. Its structural features were well seen in the surrounding stockades and log-floorings, while the heaps of charcoal, remains of food-refuse, and other indications point to a prolonged period of human occupancy. Mr. Dumbleton states that until about seven years ago, when the lake was artificially lowered a foot and a half, this island was not half its present size. He then advances various evidences to show that formerly the level of the water was still lower, when, therefore, the island would have been larger than now. This opinion may be, and probably is, correct; but we must remember that another factor has to be taken into account when discussing the invariable submergence of these islands, viz. their own pressure on a yielding lake sediment, together with the decay of the brushwood and other organic materials which generally formed their under strata. It is to be regretted that no relics were found on this island, and I cannot help thinking that, in the circumstances, a more careful search would have furnished some scraps of the handiwork of its occupiers. From the description it is clear that metal tools were used in manipulating the woodwork, but otherwise, and in the absence of any historical notice, we have no means of determining either the age of this singular lacustrine abode or the social condition of its inhabitants.

"Immediately beneath the southern spurs of the Black Mountains, and in the hollow of the great geological fracture which parts that chain from the Brecknockshire Beacons, is situated a sheet of water now called the Lake of Llangorse. Its name was formerly Llyn Savathan, or the lake of the sunken land. The area of water was once far more extensive than it is now; and it has subsequently been, as I think, considerably less than at present. A circuit of five miles will now enclose it. The margin is flat and swampy, except on the north-east, where the mountain descends upon the shore-line somewhat abruptly. The depth, though by vulgar report vast and fearful, Leland has rather overstated in assigning to it thirteen fathoms."

"Within a bow-shot of the flat meadows on the north side there is an island that would appear but little above the water, were it not for some small trees and brushwood that have fastened upon it.

"Sailing by the island one day in 1867, I observed that the stones which stand out on the south and east sides were strangely new looking, and most unlike the water-worn, rounded fragments that on the main shore have been exposed to the action of the waves; neither did there seem to be any original rock-basis at all. It was, in fact, nothing less than a huge heap of stones thrown into water two or three feet in depth. Was this the key, I thought, to the old tradition of a city in the lake? In the summer of last year, my brother, then living in the neighbourhood, first discovered a row of piles or slabs; some standing a few inches above water, for the lake was very low. We have together made some investigations during the past month, the results of which I will detail.

"The island, as now above water, measures 90 yards in circumference, its form being that of a square with the corners rounded off. The highest part is nearly in the centre, and is 5 feet above the water-level. The sides most exposed to weather, where also the water is deepest, are composed of stones sloping into the water, and extending to the distance of fifteen yards from the edge. Under the water, however, they are not nearly so thickly strewn as above. It is remarkable that on the leeward or northern side, about one quarter of the island is almost destitute of stone protection with which the greater part is covered. There is simply a surface of vegetable mould, inclined towards the water. Neither in the water, which is there very shallow, are there more than a score of stones to be found on that side. I must now speak of the piles. These are of two sorts, the most obvious being either at the margin or within a few feet of it. Like the stones, they are most numerous where the action of the storm would be most felt, and upon the shallow side they disappear entirely. They have been disposed in segments of circles, the stones being heaped inside them, and thus saved from being torn away by the waves. These piles (or rather slabs) are of cleft oak, and have been pointed, as it seems, by cuts from a metal adze. We have counted about sixty. They have been driven tightly into the shell-marl, to the depth of four feet. There are also other piles, which are round, generally of soft wood, and are found outside the present edge of the island. Several are in water two feet deep, and are driven into the marl only twelve or eighteen inches. These would have been quite powerless to confine the stones, and were evidently for another purpose.... Is it not likely that the island itself was central common ground? and that the habitations were projected from its edge towards the water and were supported by these thick round piles? Something like a ring of these is found near the oak slabs before mentioned; and traces of a second set are at the distance of twelve or fifteen yards, in water about two feet deep. Between the two, small wood is found abundantly, a few inches in the marl. At about ten yards from the shore, and in two feet of water, there appear to be the actual remains of a sunken platform. Three trunks of soft wood lie nearly parallel to one another. A 6 feet stem of oak, which I cannot account for, was with them. The top of this we sawed off, as it exhibits the marks of some heavy cutting instrument where, in modern days, a saw would have been used.

"I have to add to this subject the discovery of two much more perfect platforms in a perplexing situation, namely, within the oak slabs. They were composed of eight straight trunks, about six inches in diameter, lying side by side. Their direction is from the centre to the water; their ends, towards the shore, are thrust against the slab piles; others are closed in one case by a transverse oak beam....

"The examination of the interior would, of course, unfold the process of the construction. We therefore made several perpendicular openings; and these invariably led us down to the shell-marl, showing first a stratum of large, loose stones, with vegetable mould and sand; next (about eighteen inches above the marl), peat, black and compact; and beneath this, the remains of reeds and small wood. This faggot-like wood presented itself abundantly all round the edges of the island, and in the same relative position, namely, immediately upon the soft marl; the object of it being, of course, to save the stones from sinking.

"On digging through the before-mentioned low portion of the crannog a different order of materials exhibited itself. As I said, the stones are very few; the depth is 3 feet instead of 5; 18 inches of vegetable mould; 6 inches of earth mixed thickly with charcoal; and 1 foot of peat, small wood or reeds. I may here say that this charcoal is found under water, in very frequent small fragments, on this north-eastern side; and is covered, not with marl or stones, but with sand. Bones are found in numbers amongst the stones where the water is quite shallow; every spadeful of marl, in some parts, would, as the water dripped off, show one or more small bone fragments or teeth."

The osseous remains were more or less identified by Professors Owen, Rolleston and Boyd Dawkins as belonging to _Bos longifrons_, horse (small and large variety), red deer, and wild boar.

LAKE-DWELLINGS IN BERKS, ETC.

In 1878, Professor T. Rupert Jones, F.R.S., communicated to _Nature_ a short notice of "English Lake-Dwellings and Pile-Structures," in which, after drawing attention to the previously published articles of General Lane Fox and Sir Charles Bunbury, he writes as follows:--

"Since writing the above I have been informed that Mr. W. M. Wylie, F.S.A., referred to this fact in _Archæologia_, vol. xxxviii., in a note to his excellent memoir on lake-dwellings. I can add, however, that remains of _Cervus elaphus_ (red deer), _C. dama_? (fallow deer), _Ovis_ (sheep), _Bos longifrons_ (small ox), _Sus scrofa_ (hog), and _Canis_ (dog), were found here, according to information given me by the late C. B. Rose, F.G.S., of Swaffham, who also stated in a letter dated August 11th, 1856, that in adjoining meres, or sites of ancient meres, as at Saham, Towey, Carbrook, Old Buckenham, and Hargham, cervine remains have been met with; thus at Saham and Towey, _Cervus elaphus_ (red deer); at Buckenham, _Bos_ (ox) and _Cervus capreolus_ (roebuck); at Hargham, _Cervus tarandus_ (reindeer).

"The occurrence of flint implements and flakes in great numbers on the site of a drained lake between Sandhurst and Frimley, described by Captain C. Cooper King in the _Journal of the Anthropological Institute_, January, 1873, p. 365, etc., points also in all probability to some kind of lake-dwelling, though timbers were not discovered.

"Lastly, the late Dr. S. Palmer, F.S.A., of Newbury, reported to the Wiltshire Archæological Society in 1869 that oaken piles and planks had been dug out of boggy ground on Cold Ash Common, near Faircross Pond, not far from Hermitage, Berks." (B. 312, p. 424.)

The following is Dr. Palmer's notice of the pile-structures at Cold Ash Common above referred to:--

"Recurring to the antiquities of the peat proper, I would refer to the subject of lake-dwellings. I do not despair of finding them in our neighbourhood, for I believe traces of them have been found near Cold Ash, some such structure having been uncovered in digging bog-earth for horticultural purposes. It was circular, measuring 30 feet across, and the planks were 16 to 18 feet in length, roughly hewn, and with beams crossing from side to side, and resting on the piles. There was also a kind of causeway to it. It was on the borders of a morass, the resort of wild fowl within the memory of man. The general appearance of the valley at this place leads me to surmise that it was not long since covered with water; there is still a pond in the centre. The bog-earth had been carted away before I heard of the discovery, so that I had no chance of examining it for animal or other remains."

The editor of the _Transactions of the Newbury District Field Club_ adds the following note to the above extract:--

"Mr. Walter Money, F.S.A., has gathered some information about this interesting relic of the past. It is situated on a part of what was Cold Ash Common ... and has long been known as 'Wild Duck Pond;' it is now an oval piece of water, not much more than 20 feet across, surrounded by arable land.

"About thirty years ago, before the Common was enclosed, the season being dry, the 'Wild Duck Pond' was cleared by Mr. Whiting, of Longlane Gate, who thought the accumulated soil or mud might be useful on the land. After the removal of the top soil, some rough timber framing was met with, lying across the centre of the pit, forming, it would seem, a rude platform. A space was cleared about ten feet deep, where a heavy log of oak was found lying across from side to side. This was not removed. The work was then abandoned; the soil taken out being found to be of no use to the land. About thirteen years ago, the excavation was repeated by Mr. Lancaster, the then tenant of this part of Col. Loyd-Lindsay's property; but the investigation was not pursued far, and the water having flowed into the digging, 'Wild Duck Pond' was again restored nearly to its former condition." (_Trans. of Newbury District Field Club_, vol. ii. p. 148.)

Remains suggestive of a pile-structure were also observed by Mr. Dolby in 1870 in one of the ponds at Fence Wood, near Hermitage. Here in digging they found "a sort of pyramidal dwelling beneath the ground, the roof being covered with clay about a foot thick. This roof was supported by a large piece of timber, some twenty-six feet long, which they had got out. There were causeways there also at a depth of fifteen or sixteen feet. The water had long since rushed in and filled up the excavation, so that nothing further is known of this place." (_Ibid._, vol. i. p. 123.)

LAKE-DWELLINGS IN HOLDERNESS, COUNTY OF YORK.

The discovery of lake-dwellings in Holderness is due to Mr. Thomas Boynton, Bridlington (lately of Ulrome Grange), whose attention was first directed to the subject in the spring of 1880. Previous to the excavation of a great drainage scheme about the beginning of the present century this district appears to have been intersected by a series of sinuous and irregularly shaped lakes, whose surplus waters partly found an outlet, not in the present artificially constructed channels which convey them directly into the German Ocean, but in quite a different direction, along a sluggish watercourse, still extant, which falls into the Humber near Hull. That this latter was in former times the natural drainage course of the entire waters of Holderness is the opinion of Mr. Boynton and other geologists with whom I had the pleasure of discussing the matter. Mr. G. W. Lamplugh believes that the Gypsey Race--a stream which now enters the sea at Bridlington--at some former period continued its course through this chain of lakes and finally debouched by the same route into the Humber. The natural causes which have effected this great change in the hydrographical conditions of Holderness are to be found in the steadily progressing encroachment of the sea on the land, which here goes on at a very rapid rate. When the sea lay many miles farther off, which undoubtedly was the case in former times, it is supposed that the intervening land stood somewhat higher, and that consequently Holderness was a complete water-basin, with its outlet towards the Humber. But as the sea advanced, gradually undermining and washing away the soft glacial deposits which here form its shores, this natural basin became, as it were, tapped in the middle and so allowed the waters of its upper reaches to escape directly into the sea--a process precisely analogous to that by which its final drainage was effected by human agencies.

Nor is this opinion based exclusively on geological considerations, as we have positive historical proofs in the early annals of the country that formerly towns existed whose sites are now far out in the sea. Thus Mr. Poulson ("History of Holderness," p. 467) states that "the writer of the chronicle of the Abbey of Meaux, in lamenting the losses which the abbey had sustained, observes that they received nearly £30 from the town of Hythe, in the parish of Skipsea, chiefly from the tithe of fish; but now, says he, 1396, _the place is totally destroyed_--a proof that it was gone into the sea before the commencement of the fifteenth century." The lake of Withou, which is recorded as having paid tithe for its fish in 1288 (_Ibid._, 468), is not only at present completely drained, but more than half of its bed is washed away, and the sea beach, which runs right across it, presents a most instructive section of its sedimentary deposits and subsequent growth of peat.

From these remarks it will be seen that, in estimating the precise physical conditions that prevailed when the lacustrine abodes I am now about to describe were constructed, we have to deal with problems of a somewhat discursive character, and which, consequently, lie beyond the scope of this work. It is clear, however, that, previous to its artificial drainage, the district was overspread with a succession of shallow lakes and marshes, pre-eminently well adapted for the construction of lake-dwellings. The lakes are now gone and instead of them we have artificial drains winding along the lowest portions of their former beds. It is along the steep banks of these sluggish water-channels that Mr. Boynton has detected, in various places, piles and transverse beams, which he justly considers to be the remains of ancient lake-dwellings. Up to the present time indications of five stations have been observed, which for facility of reference the discoverer names as follows--(1) West Furze, (2) Round Hill, (3) Barmston, (4) Gransmoor, and (5) Little Kelk.

These are situated at considerable intervals from each other, varying from half a mile to two or three miles, and as they are deeply buried their investigation entails a considerable amount of labour and expense. It is only the stations at West Furze and Round Hill that have as yet been subjected to anything like a systematic exploration. A few years ago Mr. Boynton at his own expense carried out a series of excavations at the former station by which its character has been satisfactorily determined, and subsequently he has undertaken to examine the second with a grant from the Society of Antiquaries; but these works are not yet completed, and at present they are entirely suspended owing to the volume of water in the drain.

I may state that I have on several occasions visited the locality and so became practically conversant with the general features of these discoveries. Moreover, for the special object of this work, Mr. Boynton has freely placed all the materials in his possession at my disposal and given me permission to add to my notes the accompanying illustrations of a few of the more interesting objects.

_West Furze._--This was the first discovered, and the circumstances that led to the discovery are thus described by Mr. Boynton (B. 373, p. 300):--

"In the spring of the year 1880 the Commissioners of Beverley and Barmston Drainage found it necessary to deepen one of these drains (the branch called the Skipsea drain).

"A short time after this was done I was walking in one of my fields adjoining, and picked up some perforated bone implements. I shortly afterwards had the earth, which had been excavated at this place, turned over, and found more implements of the same class. Also two made from the antlers of the red-deer, and a small piece of red ochre, with several stones which bear traces of having been utilised.

"In the month of May, 1881, the water in the drain at that time being very low, and having obtained the services of half a dozen men accustomed to similar work, I had the water dammed, and dug through peat to a bed of gravel, 9 feet 6 inches from the surface.

"We found three more perforated bone implements, all in the side of the drain, and at the depth of 7 feet, also several stakes and piles with remains of brushwood. I then determined, when opportunity offered, to excavate in the field, and proceeded to do so in December last (1881). We commenced by digging a trench parallel with the drain and 60 feet in length. This trench and the drain formed two sides of a square, running north and south."

Subsequently Mr. Boynton cleared out the entire enclosure thus marked out by these primary trenches and found the whole of it to be occupied with an artificial structure of wood like the so-called fascines of Switzerland or the crannogs of Scotland and Ireland. The depth of decayed brushwood was very considerable, and it was pierced here and there with upright piles. At the margin these piles were thicker, and in one place, the south-east corner, he states that they met with great "numbers of stakes, with some brushwood, the earth being a peaty marl." Further progress from this point is thus described:--

"When clear of the slope there is a decided layer of brushwood about two feet thick, also studded with stakes, and along the inner side of the south trench we found a number of piles from 5 to 7 inches in diameter, in a line, and mostly upright. One of these we got out quite perfect. It is of oak wood, 4 feet in length, 6 inches in diameter, and has a forked top which has apparently been intended for carrying a horizontal beam or support. The piles are about 4 feet apart. One had given way and had been replaced.

"As the trench is not exactly in a line with the piles, several are now left standing and partially exposed. In this portion of the digging we found several bones of animals, a peculiar grinding-stone of whinstone or granite, almost semicircular in shape, 12 inches long by 7 broad, a flint core, a stone with the centre hollowed, a hammer-stone, and two fragments of rude pottery.

"Hazel-nuts are numerous; several I have picked out appear to have been opened by squirrels."

The drain appears to have intersected the woodwork, and as the excavations were confined to one side, the exact dimensions of the lake-dwelling cannot be stated. Its length was approximately about 70 feet, and its breadth probably one-third less. On my first inspection of the locality after these excavations had been completed I was struck with the narrowness of the lacustrine area in which the structure was reared. From the nature of the adjacent ground it was readily seen that the lake widened very considerably both above and below; but here it was so contracted that the woodwork appeared to occupy the entire breadth of the waterway--a fact which suggested to me the idea of its being a bridge or military stronghold. However, on closer inspection I saw that the accumulation of rain-wash had considerably encroached on the original bed of the lake, and I am satisfied that there would be, in former times, sufficient space for giving to the dwelling a complete insular character.

The following relics, now in the possession of Mr. Boynton, were collected in the course of the investigations:--

_Horn and Bone._--The perforated bone implements (=Fig. 176=_a_, Nos. 1 and 2), of which not less than eighteen were collected, are the most remarkable objects. They all consist of the articulate extremities of the long bones of some large bovine animals, with the exception of two, one of which was the thick end of a scapula and the other a cervical vertebra. The latter was not manipulated, and the reason it is here classified as an implement is that a portion of a wooden handle, which had been inserted into the spinal aperture, still remained. In this manner the vertebra became a formidable weapon, which, when used as a club or skull-cracker, could scarcely be matched by any work of art. I am of opinion that all these perforated bone implements were simply warlike weapons. Three handpicks, made from the horns of the red deer--the brow antler forming the pick and the body of the horn, stripped of its antlers, the handle. Also a club, or broken pick, and several portions of worked tines.

_Stone._--Three hammer-stones of natural pebbles; two anvils, one flat and circular and the other having a slight cavity on one side; six polishers, or rubbers; two flint cores, and about 50 substantial-looking flakes. One flake was a good example of a knife, and showed evidence of having been used; three other flakes were secondarily chipped and converted into neat scrapers and a saw (No. 3).

_Bronze and Jet._--One bronze spear-head (No. 4), and a fragment of a jet arm-band, like those from the Ayrshire crannogs.

_Pottery._--Fragments of a coarse unornamented pottery were found, out of which one vessel has been restored, having the following dimensions:--11 inches wide at mouth; 12 inches in the widest, a little below the mouth; and 7½ at base. Height, 7½ inches.

About thirty yards distant from the lake-dwelling, in a peaty hollow in the field, Mr. Boynton found pottery of a similar character. It was buried about three feet in the peat. The depth of peat over the lake-dwelling was somewhat more, being nowhere less than 4 feet.

_Fauna._--No expert has as yet made a report on the osseous remains, but they are believed to represent the following animals:--_Bos longifrons_ and _primigenius_, horse (a small breed), dog or wolf, beaver, ox, pig, sheep or goat, deer, otter (?), goose, and some small birds.

One well-formed human skull, with portion of an upper jaw.

ROUND HILL.--So far as the excavation of this station has been prosecuted the woodwork appears to have been precisely similar to the former, but the area occupied is of larger dimensions. Mr. Boynton thinks that the piles here belong to different periods of time, and a curious fact which he pointed out to Canon Greenwell and myself seems to support this view. He showed us the point of one pile which had penetrated and terminated in the stump of another, from which he inferred that before the former had been inserted the latter had already been in a state of decay. The decayed brushwood had also a greater thickness than at West Furze. The station has not, however, yielded many relics, the principal objects being a small stone celt, portion of a perforated stone hammer, and the half of a jet bracelet. The latter appears to be unique. It is of a flattish form, and ornamented on its outer side by five prominent ridges, running circularly. The marginal ridges are separated from the three central ones by a wider interval, in which runs a smaller ridge or bead. These ridges were evidently manipulated without the use of a turning machine, as they are not perfectly uniform, though the artists intention was to make them so.

In regard to the other three stations there are only indications of their being of a similar character, such as piles and transverse woodwork along the bottom and sides of the drain. At Barmston, a stone axe, a perforated bone implement, like those from West Furze, and bits of charcoal were found. At Gransmoor a very large quantity of broken bones lay exposed in the bottom of the drain, amidst a profusion of oak piles and beams, but among them no implements have been found.

IV.--GENERAL REMARKS ON THE LAKE-DWELLINGS OF GREAT BRITAIN AND IRELAND.

Having placed before you, with a considerable amount of fullness, certain details of the investigations of ancient lake-dwellings that have been made within the British Isles during the last half century, I proceed now to the discussion of some facts bearing on the ultimate question of their origin and development. As my conclusions are of a somewhat argumentative character, involving the consideration of some collateral phenomena as well as a critical analysis of the special materials derived from archæological research, it will be advisable, in order to secure, as far as possible, precision, at least in methods, to concentrate attention on a few definite problems--convenient foci as it were for grouping my observations. I propose accordingly to deal successively with their structural peculiarities; their range in space and time, and how far this range coincides with ethnography; and, finally, their relation to analogous remains in Europe.

Except in a very few instances, which will be afterwards more specially referred to, all the lake-dwellings hitherto examined in Great Britain and Ireland were constructed on artificial islands made generally of wood, but sometimes of stones and such other materials as might be considered suitable. Although no such instructive examples as those at Lochlee, Buston, etc., have been recorded in Ireland, there can be no doubt that those of the latter country were built on the same general principles. Indeed, few of the writers on Irish crannogs have paid much attention to the structure of the islands, and, beyond the mere statement that they were stockaded, palisaded, or surrounded by one or more circles of piles, they have supplied no explanation of the attachments and proper function of the surrounding piles. But though the purpose of the mortised beams does not appear to have been at first well understood in Ireland, it is of importance to observe that their existence has not been entirely overlooked. Dr. Reeves, writing of a crannog in the county of Antrim, says: "These piles were from 17 to 20 feet long, and from 6 to 8 inches thick, driven into the bed of the lough, and projecting above this bed about 5 or 6 feet. They were bound together at the top by horizontal oak-beams, into which they were mortised, and secured in the mortise by stout wooden pegs." (_Proc. R. I. A._, vol. vii. p. 155.)

Mr. G. H. Kinahan in a paper on the crannogs of Lough Rea thus incidentally alludes to the subject:--"A little north-west of the double row, in the old working, there is a part of a circle of piles; and in another, a row of piles running nearly east and west. Mr. Hemsworth of Danesfort, who spent many of his younger days boating on the lake, and knows every part of it, informs me that on the upper end of some of the upright piles there were the marks of where horizontal beams were mortised on them. These seemed now to have disappeared, as I did not remark them." (_Ibid._, vol. viii. p. 417.)

These are by no means isolated observations on this point, and when we consider how readily the exposed woodwork of an uninhabited crannog would be destroyed, either by the hand of man or the natural processes of decay, we need not wonder that it is only the stumps of the piles and generally submerged portions of these singular structures that remain to the present day.

The construction of a crannog must have been a gigantic operation in those days, requiring in many cases the services of the whole clan. Having fixed on a suitable locality--the topographical requirements of which seemed to be a small mossy lake, with its margin overgrown with weeds and grasses, and secluded amidst the thick meshes of the primæval forests--the next consideration was the selection of the materials for constructing the island. In a lake containing soft and yielding sediment of decomposed vegetable matter, it is manifest that any heavy substances, such as stones and earth, would be totally inadmissible, owing to their weight, so that solid logs of wood, provided there was an abundant supply at hand, would be the best and cheapest material that could be used.

The general plan adopted was to make an island of stems of trees and brushwood laid transversely, with which stones and earth were mingled. This mass was pinned together, and surrounded by a series of stockades, which were firmly united by intertwining branches, or, in the more artistically constructed crannogs, by horizontal beams with mortised holes to receive the uprights. These horizontal beams were arranged in two ways. One set ran along the circumference and bound together all the uprights in the same circle, while others took a radial direction and connected each circle together. Sometimes the latter were long enough to embrace three circles. The external ends of these radial beams were occasionally observed to be continuous with additional strengthening materials, such as wooden props and large stones, which, in some cases, appeared also to have acted as a breakwater. The mechanical skill displayed in their structure was specially directed to give stability to the island and to prevent superincumbent pressure from causing the general mass to bulge outwards.

South of the Scottish border the remains of lake-dwellings are too much decayed or imperfectly observed to furnish many reliable data bearing on this subject. So far, however, as the evidence goes it would appear that the artificial island in Llangorse and the lacustrine dwellings in Holderness were true fascines; the former, indeed, having all the appurtenances of the typical crannog.

The crannogs were made accessible by various means. Some had moles or stone causeways, the existence of which, in some instances, became known only upon the drainage of the lake. Hence it is conjectured that these approaches might have been always submerged, and so supplied, on emergencies, a secret means of communication with the shore. This idea was suggested by the tortuous direction which many of them assumed, as for example the causeway discovered in the Loch of Sanquhar which had a zig-zag direction and so could only be waded by persons intimately acquainted with its windings. Others were approached by a wooden gangway, the evidence of which now consists only of the stumps of a double row of piles. Others again were completely insulated and accessible only by boats. One feature regarding some of the wooden gangways deserves particular attention. Both at Lochlee and Lochspouts the piles were found to be tightly embraced at their lower extremities by a curiously constructed network of transverse beams. As the surface of these elaborate structures was buried from 3 to 7 feet beneath the lake-bed, my first impression was that they might have been used, like the submerged stone causeways, as a concealed means of communicating with the shore. To test this suggestion I had a special excavation made along the line of a gangway at the Miller's Cairn in Loch Dowalton. (B. 426, p. 102.) After digging through 3 feet of the consolidated and hardened mud, we came upon a stratum of fine blue clay, extremely tenacious, and little liable to displacement. The pointed stakes of the gangway, which penetrated into this clay only a few inches, here met with a firm resistance. It then occurred to me that the ingeniously arranged wooden beams at Lochlee and Lochspouts served merely the same end as the blue clay at the Millers Cairn, and that they were to be found only in localities where there was a great depth of mud incapable of affording a sufficient basis of resistance to the piles. Such difficulties have been encountered by the constructors of pile-dwellings in all countries; and it is curious to note the variety of methods by which they were overcome. The Swiss lake-dwellers sometimes surrounded the piles with heaps of stones which now go under the name of _steinbergs_; at other times split planks were laid on the soft mud into which the piles were mortised. The former plan was adopted on rocky shores too hard for piles to be driven in, and the latter where there was a great depth of soft mud, as at Wollishofen and other stations adjacent to the town of Zürich. In North Germany, as Persanzig, Aryssee, and other localities, the log-house principle, which greatly economised the materials, was adopted in the construction of the subaqueous foundations. It appears to me that this was the principle adopted in the structure of the great Irish crannog of Lagore, as Sir W. Wilde distinctly states that it was "divided into separate compartments by septa or divisions that intersected one another in different directions." It was in these compartments, which were filled with bones and black mud, that the antiquities were found; so that the crannog-dwellers must have used them as kitchen-middens. Originally they contained only water, but in the course of time they became filled with food refuse and other _débris_. House-cleaning was thus reduced to a minimum, while the laws of sanitation were not more violated than in the underground cess-pools of many of our modern dwellings. A curious statement by Wilde in regard to the disposal of bones at Lagore is that "the remains of each species of animal were placed in separate divisions, with but little intermixture with any others."

It may be also mentioned that the log-house structures described by Pigorini as lining the inside of the surrounding dyke in the terramara of Castione were perfectly analogous, only in this case the compartments were filled with clay and rubbish, so as to act better as _contraforte_ to the clay wall.

Canoes are so invariably found associated with crannogs that their discovery in lakes and bogs has been considered by Dr. Stuart as an indication of the existence of the latter. This may be true in some cases; but in others, such as Closeburn, Lochwinnoch, and Loch Doon, three of the examples cited by him, it is more probable that the canoes were used by the occupiers of the mediæval castles in the vicinity of which they were found. From these and other instances that have come under my notice I have come to the conclusion that dug-out canoes do not indicate such great antiquity as is commonly attributed to them, nor do they therefore necessarily carry us back to prehistoric times.

There is no peculiarity in the structure or form of these dug-outs which distinguishes their age or nationality. There is a good collection of them in the Museum of the Royal Irish Academy. Some have pointed prows and square-cut sterns; others have both ends pointed; some have cross bands, like ribs, left in the solid oak at regular intervals, as if to strengthen the vessel; while others are uniformly scooped out without any raised ridges. They vary much in size and shape. The largest is thus referred to in the small handbook to the Museum:--"Down the centre of the room extends the largest known canoe, formed of a single tree. The remains measure 42 feet in length, and the canoe was probably 45 feet long, by 4 to 5 feet wide, in its original state. It was recovered from the bottom of Loch Owel, in West Meath, and cut into eight sections for purposes of transport. There is a curious arrangement of apertures in the bottom, apparently to receive the ends of uprights supporting an elevating deck."

One of the canoes found at Lochlee, the remains of which are still preserved in the Burns' Museum at Kilmarnock, measured when disinterred 10 feet long, 2½ broad, and 1¾ deep. There were nine apertures in its bottom, arranged in two rows, four on each side, with the odd one at the apex. These holes were perfectly round, and exactly one inch in diameter; but when the boat was found they were quite unobserved, being all tightly plugged up, and it was only long afterwards that the plugs, upon drying, dropped out and so revealed their existence.

During the summer of 1874 a canoe (=Fig. 177=) was discovered in Loch Arthur, or Lotus Loch, in the stewartry of Kirkcudbright, in the vicinity of a small artificial island, which is thus described by Rev. James Gillespie:--

"When fully exposed to view by the trench which was dug around it, the canoe was seen to be of great size, ornately finished, and in a fair state of preservation. It had been hollowed out of the trunk of an oak, which must have been a patriarch of the forest, the extreme length of the canoe bring 45 feet and the breadth at the stern 5 feet. The boat gradually tapers from the stern to the prow, which ends in a remarkable prolongation resembling the outstretched neck and head of an animal. When excavated this portion of the canoe was entire. At the neck of the figurehead there is a circular hole, about 5 inches in diameter, from side to side. At the prow a small flight of steps has been carved in the solid oak from the top to the bottom of the canoe. The stern is square, and formed of a separate piece of wood, inserted in a groove about an inch and a half from the extremity of the canoe.

"Along the starboard side (which when found was in good preservation, except near the stern) there could be traced seven holes about three inches in diameter. The three front holes were nearly perfect, but at the stern the side was so broken that only the lower parts of the holes could be observed. They are about five feet apart, and the front hole is about that distance from the prow--the last being about seven feet from the stern. There are three holes pierced through the bottom at irregular intervals." (_Proc. Soc. Antiq. Scot._, vol. xi. p. 21.)

A curious feature presented by some of these canoes was that accidental defects had been repaired, and the method adopted in its execution is worth noticing. The canoe found close to the Buston crannog already described (page 428), showed this peculiarity in a marked degree. Another from the Loch of Canmor is thus described by the Rev. James Wattie:--

"On the 16th June, 1859, there was fished up from the bottom of the loch, near the north shore, opposite to the Prison Island, a canoe (=Fig. 178=) hollowed out of a single oak-tree, 22½ feet long, 3 feet 2 inches wide over the top at the stern, 2 feet 10 inches in the middle, and 2 feet 9 inches at 6 feet from the bow, which ended nearly in a point. The edges are thin and sharp, the depth irregular--in one place 5 inches, the greatest 9 inches. There are no seats, nor rollocks or places for oars; but there may have been seats along the sides, secured by pins through holes still in the bottom. There are two rents in the bottom, alongside of each other, about eighteen feet long each; to remedy these, five bars across had been mortised into the bottom outside, from 22 to 27 inches long and 3 inches broad, except at the ends, where they were a kind of dovetailed, and 4 inches broad. One of these bars still remains, and is of very neat workmanship, and neatly mortised in. The other bars are lost, but their places are quite distinct. They have been fastened with pins, for which there are five pairs of holes through the bottom of the canoe, at the opposite side, at a distance of from 18 to 20 inches, the bottom being flattish. There are also five pairs of larger holes through the bottom, etc." (B. 94, p. 167.)

Exact parallels to all these have been found in the Continental lake-dwellings. Of two found at Vingelz, Lake of Bienne, the largest was 43½ feet long, 4 feet 4 inches wide, and had 4 ribs left in the solid. It had iron cramps also, apparently to strengthen it, and belonged to the pre-Roman Iron Age. One at Cudrefin had also these solid cross ribs. One of the best preserved was found a few years ago at Vingrave (Lake of Bienne) covered with 2½ feet of mud, and is now deposited in the Museum of Neuveville. It is roughly made, having thick sides and a square-cut stern, with a groove for a movable stern-piece. From measurements lately taken by myself I found it to be 30½ feet long, rather less than 3 feet wide, and its greatest depth 1 foot. Its sides had four or five cuts along their margin, apparently for the use of oars. (B. 392, p. 20.)

That the crannogs in Scotland and Ireland lingered on sufficiently long to come within the borderland of history requires no great amplification here. The references to crannogs in the Irish annals are very numerous, extending over a period from the middle of the ninth to the seventeenth century.

In 1870 there was published in the _Journal of the Royal Historical and Archæological Association of Ireland_ (B. 171_a_) an account of an unsuccessful attack on a crannog near Omagh, in the year 1566, by an English army under the command of Deputy Lord Sydney. This document, which was copied by Dr. Caulfield from despatches in the Public Record Office, London, gives a vivid description of the methods adopted in the attack and defence. A kind of pontoon was constructed on "floating barrels," which conveyed the attacking party to the island; but they found it "so bearded with stakes and other sharp wood, as it was not without extreme difficulty scaleable, and so ramparted as if the hedge had been burned--for doing whereof the fireworks failed--without a long time it was not to be digged down. Yet some scaled to the top, whereof Edward Vaughan was one, who, being pushed with a pike from the same, fell between the hedge and the bridge, and being heavily armed--albeit he could swim perfect well--was drowned, and two others hurt upon the rampart and drowned," etc.

That these island forts, however impregnable they might be considered in previous ages, had ultimately to succumb before the more modern resources of warfare, is shown by the following narrative taken from the Calendar of State Papers of Ireland, vol. 156, p. 374:--

"There was one Dualtagh O'Conner, a notorious traitor, that of all the rest continued longest as an outlaw, of power to do mischief. He had fortified himself very strongly after their manner in an island or crannoge within Lough Lane, standing within the county of Roscommon and on the borders of that country called Costelloghe. A few days ago, as opportunity and time served me, I drew a force on the sudden one night and laid siege to the island before day, and so continued seven days, restraining them from sending any forth or receiving any in, and in the meantime I had caused divers boats from Athlone and a couple of great iron pieces to be brought against the island, and on the seventh day we took the island, without hurt to any on our side, save my brother John, who got a bullet-wound in the back. When our men entered the island there was found within it 26 persons, whereof 7 were Dualtagh's sons and daughters; but himself and 18 others, seeking to save themselves by swimming, and in their cot to recover the wood next the shore, were for the most part drowned. Some report that Dualtagh was drowned, but the truth is not known. It was scarce daylight, and the weather was foggy when they betook themselves to flight. The Irishry held that place as a thing invincible."--Sir R. Bingham to Burghley, Dec. 16th, 1590.

In addition to the historical evidence we have that of the relics found on many of these crannogs, which includes iron pots, guns, leaden bullets, coins, etc. Thus associated with two crannogs in Lough Annagh were an iron cuirass, matchlock guns, pistols, antique keys, spurs, various implements of iron, a bronze ladle, bronze spear-head, etc. (B. 149, p. 156.)

To the literary researches of the late Dr. J. Robertson we are indebted for equally explicit historical notices regarding the Scottish crannogs:--"Among the more remarkable of the Scottish crannogs is that in the Loch of Forfar, which bears the name of St. Margaret, the Queen of King Malcolm Canmore, who died in 1097. It is chiefly natural, but has been strengthened by piles and stones, and the care taken to preserve this artificial barrier is attested by a record of the year 1508. Another crannoge--that of Lochindorb, in Moray--was visited by King Edward I. of England in 1303, about which time it was fortified by a castle of such mark that, in 1336, King Edward III. of England led an army to its relief through the mountain passes of Athol and Badenoch. A third crannoge--that of Loch Cannor or Kinord, in Aberdeenshire--appears in history in 1335, had King James IV. for its guest in 1506, and continued to be a place of strength until 1648, when the Estates of Parliament ordered its fortifications to be destroyed. It has an area of about an acre, and owes little or nothing to art beyond a rampart of stones and a row of piles. In the same lake there is another and much smaller crannoge, which is wholly artificial. Forty years after the dismantling of the crannoge of Loch Cannor, the crannoge of Loch-an-Eilan, in Strathspey, is spoken of as 'useful to the country in times of troubles or wars, for the people put in their goods and children here, and it is easily defended.' Canoes hollowed out of the trunks of oaks have been found, as well beside the Scotch as beside the Irish crannoges. Bronze (brass) vessels, apparently for kitchen purposes (=Fig. 179=), are also of frequent occurrence, but do not seem to be of a very ancient type. Deers' horns, boars' tusks, and the bones of domestic animals, have been discovered; and in one instance a stone-hammer, and in another what seem to be pieces for some such game as draughts or backgammon, have been dug up" (=Fig. 180=).

"Before the recent drainage of the Loch of Leys--or the Loch of Banchory, as it was called of old--the loch covered about 140 acres, but, at some earlier date, had been four or five times as large. It had one small island, long known to be artificial, oval in shape, measuring nearly 200 feet in length by about 100 in breadth, elevated about 10 feet above the bottom of the loch, and distant about 100 yards from the nearest point of the mainland. What was discovered as to the structure of this islet will be best given in the words of the gentleman, of whose estate it is a part, Sir James Horn Burnett, of Crathes. 'Digging at the Loch of Leys renewed. Took out two oak trees laid along the bottom of the lake, one 5 feet in circumference and 9 feet long; the other shorter. It is plain that the foundation of the island has been of oak and birch trees laid alternately, and filled up with earth and stones. The bark was quite fresh on the trees. The island is surrounded by oak piles which now project 2 or 3 feet above ground. They have evidently been driven in to protect the island from the action of water.' Below the surface were found the bones and antlers of a red deer of great size, kitchen vessels of bronze (brass) (=Figs. 181= and =182=), a millstone (taking the place of the quern in the Irish crannogs), a small canoe, and a rude, flat-bottomed boat about 9 feet long, made, as in Ireland and Switzerland, from one piece of oak. The surface of the crannog was occupied by a strong substantial building (=Fig. 183=). This has latterly been known by the name of the Castle of Leys, and tradition, or conjecture, speaks of it as a fortalice, from which the Wauchopes were driven during the Bruces' wars, adding that it was the seat of the Burnetts until the middle of the sixteenth century, when they built the present castle of Crathes. A grant of King Robert I. to the ancestors of the Burnetts includes _lacum de Banchory cum insula ejusdem_. The island again appears in record in the years 1619 and 1654 and 1664, under the name of 'The Isle of the Loch of Banchory.'"

That Scottish lake-dwellings were known by the same name, _crannog_, as the Irish, Dr. Robertson adduces the following extract from the Register of the Privy Council to show:--

"Instructions to Andro bischop of the Yllis, Andro lord Steuart of Vchiltrie, and James lord of Bewlie, comptroller, etc.... That the haill houssis of defence, strongholdis and _cranokis_ in the Yllis perteining to thame and their foirsaidis sal be delyverit to his Maiestie and sic as his Heynes sall appoint to ressave the same to be vsit at his Maiesty's pleasour, etc., 14 Aprilis, 1608."