LAKE BOURGET.

As early as 1856, while the Mont Cenis and Culoz railway was being constructed, some antiquities were dredged up in the bay of Grésine, in Lake Bourget, which the engineers surmised to be remains of a lake-dwelling. Though this information was formally communicated to the Société Savoissienne, it was not till 1862, in consequence of renewed attention to these discoveries by Baron Despine and M. Desor, that this society took steps to investigate the matter. A preliminary investigation conducted by a committee of seven gentlemen was considered so satisfactory that the committee was renewed, with funds at its disposal for systematic researches among the palafittes. Since then several archæologists have conducted independent researches, among whom may be noted particularly Le Comte Costa de Beauregard, MM. Rabut, Perrin, Revon, Cazalis de Fondouce, and Chantre. (B. 73, 138, 176, 179, 282.)

The combined results of these explorers have now established the fact that there were eight settlements in this lake, all of them belonging to the Bronze Age. The antiquities fished up have been very numerous, but unfortunately they are widely distributed, many indeed being in private collections. The largest proportion is, however, to be found in the Museums of Chambery, Aix-les-Bains, Annecy, and St. Germain, and the private collection of Count de Beauregard in his château on the Lake of Geneva. (For relative position of these stations see Sketch Map of Lake Bourget.)

Conjux.—This station is 200 yards from the shore, opposite the village of the same name. A group of piles only 50 yards from the shore is supposed to have been the ruins of a Roman pottery business, on account of the abundance of characteristic ware found among them. A peculiarity of this station is the number of moulds found on it in proportion to the other objects, no less than 13 being recorded up to 1875, representing all manner of industrial implements, as knives, winged and socketed celts, sickles, hammers, pins, rings, and buttons.

Chatillon.—This settlement occupied a sheltered position about 500 feet from the shore. In one part the piles project out of the mud, and are all inclined towards the east at an angle of 45°, but in the rest of the station they are straight. A vessel of earthenware, like the later productions of the lake-dwellers, was found among these piles with the name Severinus stamped on it in Roman characters. (B. 176, p. 24.) Here were found some 40 or 50 of these very remarkable vessels of black earthenware, ornamented with tin strips forming a combination of pleasing designs (Fig. 193, Nos. 4 and 5); also some fragments of Gallo-Roman pottery, and others of a very early type. Among the relics are eight moulds (one of which is for a lance-head) and about 320 objects of bronze.

Gresine.—The bay of Grésine contains the sites of two settlements—one close to the railway, and the other farther out and of larger extent. The latter appears to have been connected with the Pointe de Grésine, as a gravel bank runs from this part of the shore to the site of the palafitte; and the two stations were connected with a gangway, the remains of which have been traced. The railway just touches the site of the inner station, to which accident the discovery of palafittes in Lake Bourget is due. Although the stations at Grésine have been more frequently searched than any others, owing to their proximity to Aix-les-Bains, they have continued for a long time to be the richest in all kinds of antiquities, carbonised fruits, etc. Among the moulds is one for the handle of a sword on one side, and a buckle on the other. (B. 282, Pl. liv. 2.) No less than five bronze hammers have been found on this station, all of which are socketed and cylindrical in shape. Last summer some remarkable objects were fished up, which I saw in the collection of the finder at the Restaurant Lacustre (Port Puer), some of which are here figured (Fig. 21, Nos. 4, 6, and 12).

Meimart.—The débris of this settlement lies about 100 yards from the shore, under 16 to 20 feet of water, and hence it has been less searched, although it is of considerable extent, and has yielded a few antiquities, notably a bronze sword, moulds, fragments of pottery, and a Roman vase.

Le Saut.—This settlement, like the others, was on a slight elevation some 110 yards from the shore, and at low water its relics have to be fished from a depth of about 10 feet. The station has been well explored, and it has been observed that the ceramic remains indicate greater technical skill the farther out in the lake they are picked up, and where the piles are seen to project higher above the mud. A piece of timber 22 feet long, with numerous mortises at each end, and a bone harpoon with one barb, like those of bronze from Peschiera, are the only objects which distinguish the antiquities of this station, which in general are very similar to those from Grésine.

Les Fiollets.—A small settlement in 15 to 20 feet of water. The few bronze objects found here are covered with calcareous matter. Some of the pins collected on it are of novel forms, but the most interesting object is a small file, which may be seen in the Museum at Chambery. Mortised beams were also fished up from this station.

Charpignat.—Some piles have been observed near the village of Bourget, but the associated industrial remains, if any, have not yet been revealed.

In 1875 Mr. Perrin made a series of elaborate statistics, by which he estimated the entire number of bronze objects from the palafittes in Lake Bourget at a little over 4,000, and tabulated them in various categories according to their uses, indicating the stations on which they were found, and the museums or collections in which they were then located. (B. 282.) Since then so many additional relics have been recovered from the palafittes that Mr. Perrin's tables can offer no approximation to accuracy; but, nevertheless, they have a certain value in showing the relative frequency of the different objects. I have, therefore, taken the liberty of reconstructing from Mr. Perrin's data the following list of the objects found in Lake Bourget, which gives a better general idea of the culture and civilisation of its lake-dwellers than pages of descriptive details:—

General Remarks on the Palafittes of Lake Bourget.—Count Costa de Beauregard, in his excellent article on the "Habitations Lacustres du Lac du Bourget" (B. 176), states that the stakes on which these villages were reared were generally of oak, measuring from six to eight inches in diameter, and that they were placed at a distance of 100 to 200 yards from the shore, in a depth of 4 or 5 yards of water. Their lower extremities almost always bore cutting marks, which could only be made by metal tools. The great differences as regards their state of preservation show that the settlements had been occupied for a long time, necessitating the renewal of the piles at different epochs.

The Count also believes that all the palafittes of Lake Bourget were constructed during the Bronze Age, in regard to which he thus writes:—"Malgré les quelques instruments de silex et les hachettes de pierre rencontrés dans nos fouilles, il est peu probable, comme je l'ai déjà dit, que ces bourgades aient été fondées à l'Epoque de la Pierre. Tout nous porte à croire, au contraire, qu'elles florissaient à l'Epoque du Bronze, période qui a dû être de fort longue durée en Savoie, car il a fallu bien des siècles pour accumuler sur les différents points que nous avons explorés une pareille quantité d'objets et de débris de toute sorte." (Ibid., p. 23.)

Relics.—Weapons.—The swords recovered are few, and of one type (Fig. 20, No. 16). That they are of home manufacture is more than probable from the finding of portion of a mould of the same class of weapon, now deposited in the Museum of Chambery. Only a few tips of scabbards hitherto found (Fig. 21, No. 20). The daggers were both tanged and riveted to their handles. Lance-heads (Nos. 1 to 4) are all socketed, with only one or two exceptions (Nos. 5 and 12), which might be daggers. They are generally unornamented. Arrow-points are formed for the most part of triangular plates of bronze, with two or four holes for fastening them to the stem; but other forms are met with (Fig. 21, Nos. 13, 22 to 26, and 32).

Implements.—Hatchets (Fig. 20, Nos. 9, 10, 11, and 17) are both winged and socketed, and the latter have their sockets round, oval, or rectangular. The chisels and gouges are all socketed (No. 19). Sickles (Nos. 20 and 21) have nearly all a raised button for fixing the handle (in which respect they differ from those of Switzerland), and may be classified under a variety of groups dependent on the degree of curvature and the disposition of their raised ribs. The knives are socketed, tanged, and with a solid handle (Nos. 6, 7, 8, 13, and 14); the former being most, and the latter less, frequent. Razors are of two kinds, with or without a handle (Nos. 22 and 23). Needles have the eye either at the end or middle. Awls and a variety of fish-hooks are abundant; but spears or harpoons are very rare. Rivets, nails, and bits of thin bronze plates, are in some places abundantly met with. Examples of saws and files have been found, but in small numbers.

Ornaments.—Pins with large round heads are very rare, as are also those with wheel heads (the various forms are shown in Fig. 21, Nos. 10, 18, 19, 21, 30, and 31). Bracelets (Nos. 16, 17, and 29), which are numerous, and mostly open, are either solid or hollow (one is of tin); fibulæ and torques rare; finger-rings are of two kinds, plain and spiral (No. 7); portions of girdles, buckles, pendants (No. 5), buttons (No. 33), bronze beads, and small spirals, are abundant; a clasp is like one from Mörigen (Fig. 20, No. 25). Tin appears in ingots, in a bracelet, discs, and thin strips for ornamentation to dishes; also gold in the form of a few portions of twisted wire or leaf. Several bronze vessels. One charming little vase (Fig. 21, No. 14) of cast bronze, figured by Count Costa de Beauregard, and now in his possession, was found at Grésine along with a sword (Fig. 20, No. 16), a knife (No. 8), and about 250 nails supposed to have been used in the manufacture of a shield. Nos. 1 and 2 of Fig. 21 represent two remarkable objects from Grésine, now exhibited in the Museum at Aix-les-Bains. A similar object, but more worn and minus some of its rings, is in the Museum at Chambery; and a fourth is in the Museum Lacustre at the Port (Fig. 195, No. 4). Crescents, spindle-whorls, fragments of cloth, bits of plaited rushes and basket-work, glass in small coloured beads, and amber (Fig. 21, Nos. 27 and 28), also in small beads, are all fully represented.

The articles represented by Nos. 4, 6, and 12, may be the brass ornaments on a set of harness; but as to the two curious vessels of bronze (Nos. 8 and 11), I am unable to assign any use.

The pottery shows great skill in the ceramic art. It is of a grey, black, or red colour (Nos. 34, 35, 36, and 37). Vessels combining the three colours, in various geometrical forms, with linear ornamentation, have been found among them, and others with ornamentation reminding one of the impressions of fern-leaves (Polypodium vulgare).

A few iron spear-heads (Fig. 198) and knives, as well as Roman tiles and pottery, have also been collected from these palafittes.

The domestic and wild animals, so far as they have been identified, are similar to those from the Swiss lake-dwellings.

LAKE ANNECY.

Since 1856 piles have been discovered in several places in Lake Annecy, but owing to the depth of water and the accumulation of mud, their associated relic-beds could not be easily examined. Up to the present time only four stations have been sufficiently investigated to enable us to form some idea of their chronological position with respect to the other remains of lake-dwellings. These are Stations du Port, De Vieugy, Du Chatillon, and Du Roselet.

The first-named (Station du Port) came to light only in the beginning of 1884, when the little harbour at the town of Annecy was being deepened to facilitate the movements of the pleasure-steamers which ply on the lake during the season. In the course of these operations the dredging-machines came into contact with piles, and brought up various kinds of stone implements, etc., in the mud, which, unfortunately, were mostly re-deposited in deep water. The spot where these remains were found lies just at the extremity of the Swan Island; and after the public works were completed, the dredger was put at the disposal of the Société Florimontane, who conducted systematic investigations, both there and at the stations of Roselet and Vieugy. Previous to this time all the stations examined had yielded more or fewer bronze objects, and they were therefore considered to be analogous to those of Lake Bourget, all of which were founded during the Bronze Age. The explorations conducted at the Station du Port upset this view, as from the character of the relics found on it there could be no doubt that its inhabitants lived chiefly during the Stone Age, but the station survived to the Bronze Age. The objects collected from it consist of perforated hammer-axes of serpentine (Fig. 22, Nos. 8 and 9), polished hatchets of serpentine, stone spindle-whorls, beautifully-worked daggers and lance-heads of flint (Nos. 5, 6, and 7), as well as arrow-heads, (one of shale stone No. 10), saws, scrapers, etc., of the same material. Only two metal objects, viz. a bronze hair-pin (No. 2), and a bead (No. 3), probably of copper, like those from Vinelz are recorded; but these are said to have been on the surface of the relic-bed.

Station du Roselet was the first discovered in this lake, but it has yielded only a few relics, among which are fragments of pottery, some fine spindle-whorls, a hatchet of serpentine, and a bronze bracelet.

Chatillon was in a depth of 8 to 13 feet, and among its relics are a socketed knife and a couple of bracelets (No. 4).

The settlement at Vieugy was discovered in 1868, and the most important objects from it are a bronze hatchet of the flat type (No. 1), some stone moulds, and a few rubbers. (B. 315.)

LAKES OF AIGUEBELLETTE AND THUILLE.

M. Troyon reported the existence of piles at two places in the Lake Aiguebellette, and one in Lake Thuille; but of these I find no further records. (B. 31.)

LAKE CLAIRVAUX (FRANCE).

The Lake of Clairvaux is situated on the first rising plateau of the Jura mountains, not far from the town of Lons-le-Saulnier, and covers an area of about 200 acres. At its north-west extremity there is a tongue of land projecting into the lake called La Motte-aux-Magnins, which is believed to have been an island in former times, but is now continuous with a tract of marshy ground which extends between the Motte and the town of Clairvaux.

It has been recorded that at various times prior to 1870 diverse antiquities were found in the course of drainage operations in this marshy ground, such as horn implements, stone axes of flint and jade, boars' tusks, bits of pottery, bronze celts, a fibula, and an armilla; also Gallo-Roman remains, including a Gaulish gold coin and Roman coins. In the lake itself there were no discoveries made, with the exception of a group of five piles known to fishermen. But none of these discoveries had ever suggested to any one the idea of a lake-dwelling, the common and accepted opinion being that they were remains of Druidical times and customs.

On the 27th of June, 1870, when the water was about its lowest, Mr. Le Mire happened to be walking on the shore and accidentally stumbled on the top of a black pile of oak. His attention being thus directed to such a curious object, he looked about and detected many others just protruding from the lake-bottom. He then determined to investigate the matter, and at once employed some labourers to make excavations. The place selected was 100 yards to the west of the Motte-aux-Magnins, and 25 yards to the east of the canal which forms the outlet of the lake. Trenches were dug about 1 yard in width and the same in depth (a greater depth being prevented by the oozing up of water). During these operations piles were abundantly met with, but no relics were found, and it was remarked that there was no change in the stuff thrown up from the trenches, it being the ordinary whitish deposits similar to what is seen on the present surface of the strand. The piles were of oak, fir, yew, poplar, willow, and hazel, and measured from 4 to 6 inches in diameter.

Mr. Le Mire then shifted his operations to the south side of the Motte-aux-Magnins, and after passing through 6 to 8 inches of the whitish surface deposits he came upon a blackish peaty layer containing roots of water-plants and other organic débris, which turned out to be the veritable relic-bed of the lake-dwellers.

Here he continued the excavations for about three weeks with a couple of workmen, and in this way an area of about 120 square yards was examined, which he thinks was not more than a twentieth part of the total site of the lake-dwelling. The piles did not reach the surface, but they were met with abundantly, no less than 150 being counted in the space examined. The breaking-out of the war put a stop to these excavations, and the subsequent return of the water to its ordinary level prevented their renewal.

Mr. Le Mire has published an illustrated report of his investigations (B. 219), from which these notes are taken, but the accompanying illustrations (Fig. 23) are from a selection of objects exhibited in the anthropological section of the Paris International Exposition of 1889. In this collection I note that there are a few relics, such as the two bronze objects, to which Mr. Le Mire does not refer in his report; probably these may have been found since its publication.

Among the relics staghorn implements take a prominent place. Handles and fixers for stone weapons amounted to 49, and the two here illustrated (of five exhibited in Paris) still retain their celts (Nos. 6 and 8). There are several perforated horn hammers, one of which (No. 7) retained portion of its wooden handle when found; another is a foot long, and the perforation is nearer the burr of the horn which forms the hammer-end. A tyne 11 inches long forms the handle to a small stone chisel. Another object (No. 5) is a unicum of its kind. It is a chisel of horn formed from the tyne, with the body of the horn forming a neatly-polished handle.

The polished daggers or pointers are also finely made, and almost remind one of those at Laibach. Twenty-six were exhibited in Paris, three of which are here represented, including the largest and smallest (Nos. 9, 13, and 14).

About a dozen triangular or leaf-shaped arrow-points, and one or two spear-heads of flint. Of the latter, one (No. 2) is remarkable for its size and elegant workmanship.

Wooden dishes formed out of the solid, all having a round base, were collected to the number of 15, and some large globular pieces of wood were supposed to be the primary stage of their manufacture. One dish figured by Le Mire is here reproduced, and shows a neat handle (No. 15). There were also wooden mallets perforated for a handle. Three bits of a bow, one showing the tip with a notch for the string. An axle-tree for a waggon Le Mire considers interesting, as showing a knowledge and use of traction by wheels.

Besides a few stone celts and chisels, most of which were still in their horn handles, there were a few flint knives (No. 1), three sharpening stones, two curious and novel objects of polished stone, one of which is illustrated (No. 10).

The pottery includes 140 fragments of dishes, showing various forms of handles and linear ornamentation.

Of bronze there are just two objects, a small awl or chisel and a much-worn dagger (Nos. 11 and 12).

Animal bones collected to the amount of 150 kilogrammes were not reported on by a skilled person; but, according to Le Mire, they belonged chiefly to the ox, stag, boar and pig; among them was a fine specimen of a bear's skull. Among other organic remains were a few grains of wheat and acorns.

MARSH DWELLING IN THE TOWN OF BORDEAUX.

In 1867 Mr. Delfortrie (B. 136) published a notice of prehistoric antiquities of the Neolithic Age found in the course of excavations for the improvement of the town of Bordeaux, which point to the existence of some kind of marsh dwelling in the very centre of the town. Attention was first directed to the matter by the quantity of bones thrown up from the lower part of the excavations, among which Delfortrie detected some worked ones, and associated with them were various stone and flint implements.

In regard to the osseous remains, he observes that the lower jaws of ruminants, which were relatively in great abundance, had their incisor teeth purposely removed, but the molars were retained. On the other hand, the upper jaws were entirely absent or broken, like all other marrow bones.

At three different points forming an almost equilateral triangle of 200 mètres the side, he procured sections of these street cuttings, in all of which the succession of strata and relics indicated similar conditions. At one point he gives the following details of a section:—

The bones represented the following animals:—Great ox, smaller ox, stag, pig, wild boar, horse (a small kind), goat, sheep, and dog. Mr. Delfortrie thinks the bones of the horse show that the animal was not domesticated. No piles were discovered except in the Gallo-Roman period. The following shells were found in the marine bed C.:—Ostrea edulis, Pecten maximus, Mytilus edulis, Venus decussata, Cardium edule, Mactrea solida, Turbo neritoides, and Trochus cinerarius.

The relics were found chiefly in bed E., among the ashes, a few being from the sandy peat above it. These consist of pointers, needles, polishers, spatulæ, arrow-points, and an implement of bone called a whistle; flint saws, a polished celt also of flint, three small polished stone celts of serpentine or quartzite, and about a dozen flint knives.

The opinion of the narrator is that there was here a marsh dwelling of the nature of the Kökkenmödings of Denmark which in point of time preceded the Swiss lake-dwellings, but was posterior to the Reindeer Period of Central France.

In my opinion, the character and finish of the relics furnish no grounds for supposing that this habitation was prior to the early Swiss lake-dwellings; nor are we warranted, from such limited explorations as could be made in the streets of a town, to exclude the more probable idea that this was an ordinary palafitte, notwithstanding that piles were not observed.

Second Lecture.
SETTLEMENTS IN EASTERN SWITZERLAND, THE DANUBIAN VALLEY, AND CARNIOLA.

The remains of lake-dwellings which I have hitherto described were, with one or two exceptions, situated on the borders of large lakes, and the industrial remains recovered from them were found more or less buried in the lake sediment. But these are not the invariable circumstances in which such antiquities are met with, as has already been noticed in the case of Wauwyl; but their differentiating points I did not then discuss, reserving them for this special occasion.

Every careful observer of natural phenomena must have noticed how, under certain well-defined conditions, the superficial areas of lakes are becoming gradually encroached upon, not only by the accumulation of débris carried into them by streams and rain-wash, but by the growth of peat on their margins. This latter process occurs more frequently in the smaller lakes—so much so that some of them have now almost entirely disappeared owing to the complete filling up of their basins. Though the growth of peat is slow, and almost imperceptible to individual observers, whose lifetime is generally too short to mark its progressive character, it has proved a most formidable antagonist to lake settlements by destroying their lacustrine character, and thus compelling their inhabitants to abandon them altogether. The peat has, in some instances, actually engulfed entire villages, with the accumulated débris of their industrial equipments, thus hermetically sealing up everything in one of the best antidotes to natural decay. Cities and mighty empires have risen, flourished, and disappeared, without transmitting to future ages a single record of their existence, like flowers born to blush unseen. Such, indeed, might have been the fate of many of these pile-villages, notwithstanding the favourable conditions in which their ruins have been sealed up, had it not been for the mere accident of peat cutting, which has disclosed so many of their buried treasures. These remarks are peculiarly applicable to the celebrated settlement at Robenhausen, with which I begin to-day's lecture.

LAKE OF PFÄFFIKON.

The small lake of Pfäffikon, which lies to the east of Lake Zürich, contained two settlements, viz. Robenhausen and Irgenhausen.

Robenhausen.—This well-known station, which has furnished specimens of lake-dwelling remains to most of the European museums, is situated near the middle of an extensive tract of pasture-land on the south side of the lake. Although its site is now several hundred yards from the lake, there can be no doubt that, originally, it was completely surrounded by water; the nearest land, that on the west, being some 2,000 yards distant. On the east side the old lake-shore is 3,000 yards distant, and towards this, notwithstanding its greater distance, there extended a gangway, the remains of which can still be traced. Underneath the grass there is a thick deposit of peat, which has been utilised as fuel according to the needs of the surrounding community; and a mere glance at the locality shows that the whole expanse is but an encroachment of the peat on what was formerly part of the lake. The meadow belongs to peasant proprietors, among whom it is parcelled into small plots. During the winter of 1857-8 Mr. Jacob Messikommer, the owner of one of these plots, discovered the remains of a pile-dwelling on his portion, and to its investigation he has ever since devoted himself. His efforts were greatly encouraged by Dr. Keller and other members of the Antiquarian Society at Zürich, to whose museum many of the principal relics have been sent. A few years after its discovery, the project of deepening and widening the outlet, which, as it so happened, passed through the lake-dwelling, afforded a splendid opportunity to archæologists for investigating its antiquarian remains. Messikommer was appointed superintendent of the proposed excavations. Since then he has on several occasions when the waters were low, as in the years 1864, 1865, 1870, 1875, 1882, 1884, and 1886, made more or less extensive diggings in different parts of the settlement for the purpose of clearing up obscure or disputed points. Altogether he has made good use of his advantages, and to his intelligent and watchful care we are indebted for a careful record of the relics, as well as a series of shrewd observations bearing on the character and duration of this settlement, which has made it one of the most instructive in the whole range of lacustrine research.

The space occupied by the settlement formed an irregular quadrangle, little short of three acres in extent. The piles were made from the round or split stems of trees—oak, beech, and pine being the prevailing kinds. On the supposition that they were placed at uniform distances throughout, Messikommer calculates from the data supplied by the Aabach Canal, which involved an area of about 4,000 square feet, that 100,000 piles were required for the construction of the entire lake-village.

In order to get at the relics, one has to dig through 5 or 6 feet of peat, in which no relics are found, with the exception of the piles, the tops of which nearly reach to the surface. Such pits are soon filled with water, as all the relic-beds are below the level of the lake.

As the excavations progressed, Messikommer made the important observation that the piles could be distinguished into three sets, corresponding with so many relic-beds.

The first set of piles penetrated into the shell marl some 10 or 11 feet below the present surface; and immediately over this marl there was a bed of greasy peat only 4 or 5 inches thick containing a few relics. Then followed a bed of charcoal with carbonised wheat, barley, cloth, etc., the result, according to our investigator, of a general conflagration which destroyed the entire settlement. After this catastrophe a new superstructure was reared, the piles of which were so closely set that, on an average three or four could be counted in each square foot. This new village appears to have flourished for a long time, as its duration is represented by a peaty deposit nearly 3 feet thick containing a variety of relics, as bones, pottery, portions of clay flooring, etc. Then followed a second bed of burnt materials, as corn, fruits, bread, and the usual industrial implements of stone—all of which point to a second conflagration. But, apparently undiscouraged, the lake-dwellers again undertook the task of reconstructing their peculiar dwellings, and Messikommer distinguishes this third series of piles by their not penetrating so deeply as those of the previous habitations. While the piles of the earlier dwellings penetrated into the shell marl, those of the third structure fell short of the former by 2½ feet and terminated in the intervening accumulated débris. On the other hand, however, their tops reached higher in the peat, coming nearly to the present surface. Further, he observed that it was only in the third settlement that the piles were split, those of the two former being round and much more decayed. Also, corresponding to its duration, there was a deposit of peat 3 feet in depth containing various relics, but no evidence of a conflagration, and above this point the peat was entirely destitute of the remains of human industry. It would thus appear that the lake-dwellers voluntarily abandoned their village, either on account of the accumulation of peat or because, in the exigencies of civilisation, they found more congenial conditions of habitation elsewhere.

During the excavations in the Aabach canal the above facts were amply demonstrated, as, indeed, they can be at the present time by any one who chooses to make the necessary excavations, permission for which the proprietor freely gives.

From the peculiar grouping and distribution of the relics over certain areas Mr. Messikommer came to the conclusion that while each cottage had its special appliances, as a hearth, a millstone, sharpening stones, and weaving materials, there were other relics specially localised. Thus there were large quantities of corn in one place, dried fruits in another, flax in a third, etc. He also learned to recognise from the kind of litter used, and the droppings of the animals, where the stalls for cattle, sheep, and goats were located; which, according to him, were in the intervals between the cottages. Bones, scales of fish, dried fruits, water-chestnuts, beech and hazel-nuts, acorns, and other remains of food, were very abundantly met with. The following are some of the more interesting relics from this vast deposit of the industrial remains of many ages:—

Wooden Objects.—A bow of yew, five feet long, still retaining the notch at both ends for the string; another specimen measures only 3½ feet in length. A large tub-like dish, nearly 16 inches in diameter, and a variety of ladles. A yoke for cattle, made of a hazel branch. A large door of wood, so arranged as to turn on a pivot, and measuring 4 feet 9 inches by 2½ feet wide, and 1½ inch thick; a canoe 12 feet long, 2½ feet wide, and 5 inches deep; a large assortment of handles, knives (Fig. 24, No. 15), clubs (No. 26), dishes, suspension hooks, etc.

Horn and Bone.—Haftings for stone axes and chisels (Nos. 8 and 9), daggers, chisels, perforated axe-hammers (No. 12), arrow-points (No. 2), agricultural implements, small cup (No. 7), etc.

Stone.—Axes of nephrite are scarce, but they are abundant of the ordinary materials (No. 23)—some are perforated; flint saws in their handles, hammer-stones (No. 10), pendants (No. 3), a stone disc polished and perforated in the centre with a round hole (No. 13), arrow-points and scrapers of flint (No. 1), two small objects of redstone perforated with a series of holes (Nos. 5 and 6).

Pottery.—Earthenware cups, spoons, and various kinds of vessels (Nos. 14, 16, 17, and 18). Particularly noteworthy is one with a conical base requiring a ring-support (No. 18). Several coarse crucibles with handles (No. 22). When the first of these objects was discovered, it was supposed to be a large spoon, but latterly traces of copper were found in the pores of one, and thus their true nature was recognised. These crucibles were found a few years after the discovery of the lake-dwelling, and although Messikommer was constantly on the look-out for metal objects it was not till 1882 that his search was rewarded. This was a small copper celt of the flat type (No. 4), but as it was in stuff thrown out of the trenches for some time, it was impossible to say to which settlement it belonged. (B. 383, p. 324.) In 1884 Messikommer announced that a crucible which had evidently been used was found in the stratum of débris corresponding with the second settlement at Robenhausen.[21]

On the 4th of October, 1887 (B. 454), Mr. H. Messikommer, while making excavations in an undisturbed part of this lake-dwelling, found near the surface of the peat, and on a level with the tops of the piles, another hatchet of the flat type made of bronze (No. 11). It is clear from these respective finds that the Robenhausen lake-dwelling came to an end before bronze came into general use.

Weaving Materials.—A great many specimens of flax, yarn ropes, balls of thread, bits of ribbon, and variously-woven cloths, fishing and hair nets,[22] plaited borders, fringes, and mats (Fig. 25). Loom-weights (Fig. 24, Nos. 20 and 21) and clay pirns were also met with, but, singularly enough, hardly any spindle-whorls. It is not very clear in what position these recorded relics have been found; but in 1882, when the water happened to be very low, the lowest relic-bed was carefully searched, and similar remains were found in it. In consequence of these finds, Messikommer announced, in 1882 (B. 383a, p. 379), that he was convinced that all manner of weaving was thoroughly known at the very commencement of the Robenhausen lake-dwelling.

The third settlement has yielded very little cloth or thread, probably owing to the fact that no conflagration took place, by the charring of which such remains are preserved from decomposition. On the other hand, jade implements, among which is an arrow-head of nephrite, and some 60 seeds, and fruits, have been collected. Among the latter the water-chestnut (Trapa natans) may be especially noted, as it no longer grows in the locality. (B. 462.)

Irgenhausen.—Only one other settlement has been recognised as a true lake-dwelling in this lake-basin, viz. Irgenhausen, situated about half an hour's walk to the east of Robenhausen. The station ran parallel to the shore for a distance of about 300 feet, with a breadth of only 30 feet. The relics found on it are similar in character to those from Robenhausen, the most remarkable of which are specimens of embroidered cloth and checked muslins. (B. 126, Pl. xvi. Fig. 2 and 2a.) Messikommer believes that only one row of cottages occupied this site. Almost the whole site of this lake-dwelling has disappeared into the depths since 1881, and can no longer be found.[23] Only a yawning deep (eine gähnende Tiefe) is now to be seen where formerly stood the remains of the Pfahlbau. This phenomenon is, however, not singular in the Swiss lakes, as evidence of which we have the recent catastrophe in Lake Zug, which demolished not only the site of a prehistoric lake-dwelling, but also a large part of the town of Zug.