During the autumn of 1877 the site of a third pile-dwelling was come upon, about 300 yards from the last mentioned, and on the other side of the Isca (see Sketch, p. 171), under precisely similar conditions as the two former, but owing to want of funds the excavations were discontinued before the entire area was searched. Dr. Deschmann is of opinion that these are by no means the only portions of the moor in which lake-dwellings existed, as indications of them were found in several other places along the bed of the Isca. In further support of this opinion I may mention that in 1857, before lake-dwellings had attracted attention in this quarter, some objects were found at Moosthal, in quite a different part of the moor, which point to its being the site of a lake-dwelling. Here the peat was 10 feet deep, and at this depth, and lying immediately over the lake-silt, were found three perforated staghorn hammers, two canoes, and some other objects of human industry, which, however, were dispersed before Dr. Deschmann became aware of the discovery.

The relics of human industry collected during these systematic explorations, with the exception of a few in the Museum of Vienna, are deposited in the handsome new Museum at Laibach, where they form a remarkably complete and interesting demonstration of the culture and civilisation of the lake-dwellers. Some of these are illustrated on Figs. 42, 43, and 44.

Pottery.—Vessels for household use are extremely abundant, and varied in form and ornamentation. They are all hand-made, and the quality of the paste appears to have been good—that for the larger dishes was mixed with rough sand. All the pottery has a darkish appearance, but most of the smaller vessels had been smeared over with some black composition. Not a few of these dishes were quite whole, so that their varied forms and uses may be readily distinguished. They may be classified as jars, vases, cups, plates, jugs, bowls, flasks, spoons, etc.; and ordinary handles, perforated knobs, tubular borings (ansa canalicularis) appear to have been indiscriminately used.

Some of the smaller dishes have four or five stud-like processes or rudimentary feet, and others have a pedestal-like base, slightly expanding at the lowest point, on the underside of which there is often impressed the shape of a broad cross (Fig. 42, No. 20). The ornamentation, when reduced into its simple elements, may be thus stated:—

(1) Straight or wavy ridges, sometimes notched across, and running upwards or across the body of the vessel; (2) finger or nail marks; (3) checks made with groups of incised lines crossing each other; (4) lozenge-shaped spaces alternately plain and lined; (5) herring-bone pattern; (6) triangles, crosses, wheels, rhombs, and other simple geometrical figures, sometimes with inscribed figures or lines; (7) impressions of strings, points, etc. The style of the more highly ornamented vessels is, though complicated, artistic, and when the incised lines were filled with a white material, as is supposed to have been the case with some of them, these patterns on a dark or black ground must have been very effective. Ornamentation is not always confined to the outside of the vessel, as may be seen from Fig. 43, No. 8. Dr. Deschmann sees a striking resemblance between the Laibach pottery, both in manufacture and ornamentation, to that represented in Dr. Schliemann's works on Troy.

Besides the ordinary dishes, there fall to be enumerated under this heading some two or three hundred spindle-whorls, one or two cylindrical weights, perforated cones (Fig. 43, No. 5), a few crucibles of superior workmanship (Fig. 45, No. 14), a mould for an axe-head (Fig. 42, No. 22), and some other small objects, apparently toys (Fig. 42, No. 21). Among the most remarkable and mysterious objects are some ornamented images, more or less fragmentary, of animals and human beings with fantastically-formed heads (Fig. 42, Nos. 11, 23, and 24; and Fig. 195, Nos. 5 to 8).

Stone.—The stone implements, with the exception of rubbers, hammers, and sharpening-stones, are comparatively rare. The sharpening-stones are well represented by a variety of implements—from the small portable hone with a string perforation, to a large hollow block weighing 220 pounds. Of simple stone axes and chisels there are only about a dozen good examples, but amongst them are two little gems—one a hatchet of nephrite (Fig. 42, No. 12), and the other a miniature chisel of greenstone (No. 9). Perforated axe-hammers number about two dozen (Fig. 43, No. 10); they are mostly of serpentine and well shaped, and the boring is neatly executed. The flint objects, which amount to about four dozen, consist chiefly of neatly-formed lance-heads; but amongst them are a few scrapers and flakes, but hardly one that could be called an arrow-point (Fig. 42, Nos. 1 to 5). Almost unique are two conical anvils, one of which (No. 18) has metallic particles of copper or bronze on its flat surface. There is also a polished stone disc showing the commencement of a perforation near its centre with the core still remaining.

Bone and Horn.—A characteristic feature of the Laibach settlements is the abundance of implements of bone and horn which they have yielded, and which may be thus classified:—

1. Perforated hammer-axes of staghorn, numbering between 300 and 400, in all stages of manufacture. The most typical forms of these implements are sketched on Fig. 44.

2. Polished daggers, pointers, chisels, etc., varying in length from 4 to 10 inches, amount to many hundreds. The smaller pointers, awls, and pins, were made of bone splinters and ground to fine points. The finer daggers were invariably made of the leg-bones of a deer or other animal. It appears that they were manufactured by sawing or cutting the bone lengthways and slightly diagonally, so as to have two weapons out of the one bone, leaving each with a joint for its hilt. One or two bones were found showing this operation in an uncompleted stage. Some of these daggers had perforations near the extremity of the handle-end for suspension.

3. The tynes of deer-horns were converted into coarse needles (Fig. 42, No. 13), and used probably in the manufacture of nets. Over a score of these implements have been collected.

4. A few finely-polished objects like hooks are supposed to have been used as dress-fasteners or buckles (Nos. 6 and 16).

5. Another set of curious objects (No. 7), of which about a score have been collected, is supposed by Dr. Deschmann to have been used as artificial bait to catch large fish, just as we at the present day use an imitation minnow. They are made of the tynes of deer-horns, and vary in length from 2 to 5 inches.

6. About a dozen or so of very sharp and finely-polished needles made from the superficial lamina of a rib. The eye, which is at one end, is either round or elongated (No. 15.)

7. Several hollow bones (some of the wild swan), open at both ends, and varying in length from 5 to 10 inches, have marks inside, as if made by the friction of running threads. They are supposed to have been used in the preparation of thread, and hence go under the name of "Zwirndreher."

8. Some flat portions of the horns of the elk and the underjaws of oxen, minus their teeth, were used as polishers.

Metal Objects.—(Fig. 45.) The total number of metal objects now in the Museum at Laibach, and tabulated as coming from the lake-dwellings, is 24. They are all either of bronze or copper, as hitherto not a trace of iron has anywhere been met with. The following is a list of them:—

1. Two flat-handled bronze swords 21 and 14½ inches long (Nos. 3 and 4).

2. Three bronze daggers, 11½, 8, and 7½, inches long (Nos. 2, 1, and 7). The larger has four rivets for fastening a handle; the next has six rivets, and the blade is beautifully ornamented; and the third has two rivet-holes, arranged differently from those in the other two.

3. A winged bronze celt (No. 5); and one of the flat type (No. 9), said to be of copper.

4. Portions of three bronze pins (Nos. 12 and 13).

5. Two thin bracelets of bronze, much worn.

6. Five peculiar objects of copper, like awls (Nos. 6 and 8).

7. Seven objects like daggers, lance-heads, or knives, rudely hammered, are also supposed to be of pure copper (Nos. 10 and 11).

The winged celt and the larger of the two swords are not noticed in either of the reports of the various investigations, but I am assured they form part of the same find; and, in corroboration of this, I find they are included in a photograph issued by the authorities of the Museum, purporting to be a representation of all the metal objects from the Pfahlbauten.

Objects of Wood.—A canoe 15½ feet long and 2½ feet wide was pointed at both ends. Also a toy canoe. Fragments of a few dishes, such as a large plate, a spoon of yew wood, and some bowls—one of which is scooped out of a large round natural protuberance of a tree. A few elongated pebbles rolled in birch bark. Portions of bast ropes, and some coils of very fine carbonised linen threads.

Two remarkable machines ("Biberfälle") (Fig. 46), each constructed out of one solid piece of wood, and having two movable valves in the centre worked by projecting pivots resting loosely in corresponding holes in the machine. These valves are freely movable when pushed upwards, but this motion is arrested just a little short of the perpendicular by the slanting shape of their posterior edges, so that, when left to themselves, they always fall together, and never backwards. The one here represented is in a very perfect state of preservation; and the other, though now in a fragmentary condition, clearly shows that in its structure it was precisely similar to the former. These peculiar implements, though found at a little distance, are considered of contemporary date with the lake-dwelling remains, as they were in the same archæological stratum, and about the same depth in the peat. The one here figured is made of oak, and measures 32 inches long, 12 inches broad, and 4 inches deep. The aperture, when the valves are open, measures 9 by 5 inches. The most recent opinion as to the use of these machines is that they were beaver traps—an opinion that derives much probability from the extraordinary number of the skeletons of this animal which have been found among the food-refuse of the inhabitants of this lake-dwelling.

Such machines are not absolutely new to archæology, and the little that is known about them rather strengthens the opinion above given as to their use. The first discovered to which attention was directed in archæological journals was figured and described in 1873[32] by Dr. Hildebrandt, of Tribsees, Neu-Vorpommern. It measures 29½ inches long, and 6 inches broad at the ends, and has two movable valves in the centre. It was found in a peat bog at a depth of 5 to 6 feet below the surface, and is now preserved in the Museum at Greifswald. Dr. Hildebrandt conjectured that it was some kind of trap for catching fish.

In reply to Dr. Hildebrandt's notice of the machine found at Tribsees, Professor F. Merkel, of Rostock, wrote to say[33] that two similar ones were found in different parts of North Germany, which he considered to be otter traps rather than fish traps. One of them was found in the moor of Samow, near Gnoien, at a depth of 6 or 7 feet, and is now in the Museum at Rostock.

It is remarkably like the one from Laibach, and differs only in being 4 inches longer, and having three holes in the valves instead of two (Fig. 47). A third[34] was found in a moor at Friedrichsbruch, near Flatow, in the province of West Preussen, which was sent to the Märkisches Museum. At no time within historical times were such machines known to be in use, so that their function still remains conjectural, unless the circumstantial evidence derived from the Pfahlbau at Laibach decides them to be "Biberfälle."

While the proofs of the above remarks were still in my hands, I received from Dr. Luigi Meschinelli, of the Geological Museum of the Royal University of Naples, a copy of an article by him, entitled "Studio Sugli Avanzi Preistorici della Valle di Fontega."[35] The objects described in this memoir were found, in the course of excavating peat, in a small valley which opens into Lake Fimon in the vicinity of Vicenza. Among numerous industrial remains of man, consisting of fragments of pottery, various implements of stone and flint, a bronze celt, and a Roman coin of the time of the Emperor Adrian, were three curious and novel objects of wood shaped like small canoes. One of these machines—the best preserved, though not the largest—is carefully described and figured by Dr. Meschinelli, and from his minute description there can be no doubt it is another example of the same apparatus which I have just described as having been found in North Germany and Laibach Moor.

The body of the Italian machine was constructed out of one piece of oak, and measured 28 inches long, 6¾ inches broad, and 2¾ inches thick (Fig. 47a). The opening in the centre, which was closed by two valves revolving on projecting pivots, and resting along their axis in a deep groove cut on each side of the machine, measured 6½ inches by 3½ inches on the under side, so that this would be the actual size of the aperture when the valves were open. Associated with the machine, as will be seen from the illustration, were several worked portions of sticks, evidently the débris of some kind of mechanism attached to it. Similar sticks were found along with the Laibach examples. It will be observed that the dimensions of the Italian one are a little less than those of the previously described machines, but that in all other respects they are identical. The other two found at Fontega were, according to Dr. Meschinelli, precisely similar to the one he describes.

Among the organic remains from these peat excavations I find no mention made of the osseous remains of the beaver, neither is this animal included by Lioy among the fauna of the lake-dwellings at Fimon. So far, therefore, there is no presumptive evidence that the machines described by Dr. Meschinelli were beaver-traps. That, however, the beaver frequented the Po valley during prehistoric times we have positive evidence in the discovery of its bones in several localities—as, for example, the terremare of Castellaccio (B. 457) and Cogozzo (B. 389a).

Puzzled to account for these curious machines which so fortunately attracted the attention of Dr. Meschinelli, he concludes his notice of them thus:—

"A che cosa poteva servire questo oggetto? Era forse un modello per costruire poi una piroga di dimensioni maggiori per utilità pratica? Portata a queste dimensioni, serviva essa al trasporto, o meglio quei congegni erano stati inventati a facilitare la pesca? Volle invece l'artefice sbizzarrire il suo genio inventivo nel costruire un trastullo cosi grazioso? E perchè allora costruirne due di eguali?"

It may be interesting to add that in 1859 a wooden machine, which evidently comes under the same category as the above, was found in a bog in the townland of Coolnaman, parish of Aghadowey, county Derry, Ireland. It is figured in The Ulster Journal of Archæology (vol. vii. p. 165), as an "antique wooden implement," which is thus described by the editor:—

"It was discovered embedded in a solid bank of turf, at a depth of 4 feet from the surface, the bog extending to a great depth underneath. No other article was found near it. It is entirely of wood, and measures as follows:—Extreme length, 3 feet 5 inches; breadth across the centre, 7½ inches; depth, 2½ inches; lid, 14 inches long and 3½ inches broad; under hole, 12¾ inches long and 3½ inches broad. The upper edges have evidently been higher on all sides, when perfect—probably on a level with the lid or small door—or even extending still higher, so as to form a kind of trough. The lid is now somewhat narrower than the opening which it is intended to close, but, no doubt, was made to fit accurately when in use. It moves up and down on a hinge formed by two projections which lie in corresponding hollows, and seems to have been opened and shut by means of a handle inserted into a hole in its centre. These hinges have, no doubt, been kept in their place by some part of the wood above them which is now lost. From each end of the lid and on a level with its upper surface there runs a hollow groove, sloping regularly downwards to the end of the implement, and terminating in a hole which perforates the bottom, seemingly for the discharge of a liquid. Towards each end are two lateral holes placed opposite to each other, one in each lip of the groove, apparently to receive a rope passed through them to serve as a handle for removing the article from place to place. The under side of the implement is flat, having in its centre an oblong hole (the bottom opening of the cavity covered by the lid), which has all its four edges sloped or bevelled.... Coolnaman, which gives name to the townland, is a considerable hill, entirely cultivated, but surrounded at its base by a bog of unknown depth, which evidently occupies the site of an ancient lake. On the side of the hill where the implement was discovered the turf has become quite solidified, and forms a dense black mass up to the surface."

In looking at Fig. 47b, which shows the upper and under sides of this implement, it will be at once seen that it differs from the Continental examples only by having one valve or lid closing the central aperture instead of two. Neither the editor nor any of the parties who had examined this curious machine at the time had ever seen anything of the kind before, and no rational explanation of its use has ever since been offered. One thought it was a fishtrap intended to be placed in a river; another, that it was a kind of pump; a third, that it was a machine for making peats; and a fourth, that it was a cheese-press (Ibid., p. 289).

To find so many of these machines, of unknown use and so remarkably similar in structure, in such widely separate districts as Ireland, North Germany, Styria, and Italy, must be a matter of interest to archæologists, and no one can say that the correct explanation of their use is to be found in any of the suggestions hitherto offered on this point. I may mention one element which may help in the solution of this problem, viz. that all the examples from Italy, Laibach, and Ireland were found in bogs that were formerly lakes. Perhaps this is true in regard to those from North Germany, but the point is not referred to in the short notices which have appeared of them. If these machines are really traps they could only be used in water where the animal could insert its head from below, and among amphibious animals the otter and beaver are the only ones to which all the conditions involved in the trap theory could apply.[36]

Organic Remains.—In the report of the investigations for the year 1877 Dr. Deschmann gives the following analysis of the osseous remains collected, which shows the relative frequency of the respective animals:—

Some of these bones contained crystals of vivianite, resembling in this respect the osseous remains found on some of the Scottish crannogs, especially that at Lochlee. (B. 373, p. 88.)

The Bos primigenius is also represented by a portion of horn 21 inches long. The presence of hazel nuts with gnawed holes also points to the existence of some small rodents, probably the dormouse. One or two bones (metacarpal bone of a deer and an ulna of the bear) are covered all over with groups of peculiarly-striated markings, as if made with a file; but for what purpose, or whether the work of man or of some rodent animal, remains a mystery.

There is also a considerable quantity of the bones of birds, the spine bones of fish, jawbones of large pike, carp, etc., and a portion of the shell of a tortoise (Emis lutaria).

Of human remains there are two skulls of adults, minus the facial part, another of a child, a lower jaw, and a few bones of the extremities.

Notwithstanding a minute search, no traces of any kind of corn have hitherto come to light; but we must not therefore conclude that the lake-dwellers were ignorant of agriculture and the ordinary cereals, as grain is so apt to decompose unless it happens to be in a carbonised state. It is, however, probable that the cultivation of grain was not the chief industry of the colony, and that the mealing-stones which were in such abundance must have been used for grinding some other kind of food as well as grain, such as the kernels of hazel-nuts and water-chestnuts. The water-chestnut (Trapa natans), according to Deschmann, does not grow at the present time in Carniola; nor has it ever, since the earliest botanical examination of the country by Scopoli, been considered a native plant in the Flora Carniolica. In the last century the monks of the Cistercian order, at Sittich, cultivated it in their ponds. Pliny, however, distinctly states that in ancient times it was used as a food. "Thraces qui ad Strymona habitant foliis tribuli equos saginant, ipsi nucleo vivunt, panem facientes prædulcem, et qui contrabit ventrem." (H. Nat., xxii. 10-12.)

Among the vast quantity of osseous remains there is not a single fragment of the skeleton of the horse. On the other hand, it is calculated that the deer is represented by no less than 500 individuals, and the beaver by at least 140. For the latter this is a colossal figure, seeing that the richest station in beaver remains among the Swiss lake-dwellings, viz. Moosseedorfsee, numbers only eight individuals. The animal is now extinct in the country, nor has it ever been mentioned in any of the historical annals of Carniola.

Third Lecture.
LAKE-DWELLINGS AND PILE-STRUCTURES IN ITALY.

On the 20th of July, 1860, M. G. de Mortillet wrote a letter to Sig. Cornalia, president of the Italian Society of the Natural Sciences, at Milan,[37] in which, while mentioning the discoveries made in Switzerland, he suggested that similar antiquities might be found in the lakes of Lombardy. The reading of this letter led to a discussion which at once elicited one or two statements of archæological importance. The vice-president, Sig. Antonio Villa, recalled the fact that a bronze axe-head and some flint arrow-heads were found in the turf-bog of Bosisio, at a depth of 10 feet, which were described and figured in a Milan journal, Il Fotografo, 2nd August, 1856. The president also mentioned that he possessed weapons of a similar character, which were found, along with some human bones, in the peat-beds of Brenna. Shortly afterwards the celebrated naturalist Gastaldi, in an article in Il Nuovo Cimento, directed attention to certain antiquities which the turf-cutters were in the habit of finding in the "torbiera di Mercurago." (B. 37.) Subsequently Gastaldi visited this locality, and along with Professor Moro, of Arona (who first recognised the importance of the objects in question), made further researches in the peat at Mercurago, the result of which was to leave no doubt that they had here to deal with the remains of a true palafitte analogous to the pile-dwellings in the Swiss lakes. During the next two years Gastaldi's report was considerably enlarged by further finds at Mercurago. (B. 43 and 52.)

About the same time that these discoveries at Mercurago were being made the existence of a palafitte in Lake Garda was surmised from the finding, at various times, of bronze implements and weapons in the harbour at Peschiera; but nothing further of a very definite character occurred till the summer of 1863, when Professors Desor and De Mortillet visited Lombardy in search of lake-dwellings. These eminent archæologists were joined by Professor Stoppani, and the immediate result of their investigations was the discovery of several settlements in the Lake of Varese and elsewhere. (B. 67.) Since then the lacustrine stations south of the Alps have greatly increased in number, there being now scarcely any of the smaller lakes and turbaries of North Italy that have not yielded more or fewer remains of this character.

In addition to these ordinary lake-dwellings, whether in water or in peat, there are, in the valley of the Po, other ancient remains known as "Terremare," which are now shown to be so closely analogous to the former that they fall to be described as land palafittes. They are found only in the eastern part of the valley, and as some of their relics, in common with those of the adjacent palafittes, present some characteristics which are not found in western Lombardy, I fasten on this distinction as a convenient principle of classifying the lake and peat dwellings into a western and an eastern group, reserving the "terremare" for separate treatment. Accordingly we begin with Lake Varese, whose settlements appear to have been the most important and the most central in the western group.

LAKE OF VARESE.

Lake Varese is irregularly shaped, about 5½ miles in length, and less than half that in breadth. It occupies a somewhat shallow basin, its greatest depth being 85 feet, and, although bounded on the north by high hills, its banks are generally flat or shelving. Its surface is 770 feet above sea-level, and 134 feet above that of Lake Maggiore, to which its surplus water is carried by the Bardello, a stream which has its outlet at the north end of the lake. The district around is rich and well cultivated, except on the south side, where the lake becomes contiguous with an extensive peat-bog called "torbiera della Brabbia." When Stoppani and his illustrious friends, along with Desor's experienced fisher, Benz, commenced their lacustrine explorations in Lombardy, they selected Lake Varese to start with, on account of the suitability of its shores for such structures. On the very first day (21st April, 1863) their labours were rewarded by the discovery of the sites of two settlements—one at the south-east side of the little island then called Isolino, or Isola Camilla, but now I. Virginia,[38] and the other opposite the village of Bodio.

Professor Stoppani continued his researches after the departure of his friends, and made further discoveries, not only in Lake Varese, but in some of the other lakes of Lombardy. In November of the same year Captain Angelo Angelucci, of Turin, was attracted to the scene of these discoveries in Lake Varese, and henceforth took an active part in the investigation of its palafittes. (B. 63.) Nor must I omit to mention the Abate Ranchet among the list of the early explorers. He discovered in the same year not only a new station on the south side of the outlet, but also, in the following year, two others in the adjoining lake of Monate. (B. 85.) At the end of the first year's explorations we find, from the reports of Stoppani and Angelucci, that no less than six stations were determined in Lake Varese, all situated on its south-western shore. In 1868, when Dr. Camillo Marinoni published a report on "Le abitazioni lacustri e gli avanzi di umana industria in Lombardia" (B. 159), the number had increased to seven. Although no addition has since been made to their number, much attention has been given, especially in these later years, to their investigation. The Sketch Map of Lake Varese (page 189) shows the names and the respective positions of these settlements, which I shall now shortly describe.

Isola Virginia.—This is a small egg-shaped island lying along the west shore, from which it is distant about 80 yards. It is 240 yards long, with a maximum breadth of 100 yards, and contains besides some fine trees, a house with two storeys, the upper of which is converted into an archæological museum, and at a little distance there is a café for the convenience of the numerous visitors that frequent the locality. Its area is nearly 3 acres, and its highest point is barely 8 feet above the average level of the lake.

Piles were discovered in the lake at the south-east side of the island, in a space extending along its margin for about 100 yards, and about half that distance in breadth. Two years ago, when I visited the locality, the heads of piles were readily seen through the water, just cropping above the sandy bottom. In some cases it was difficult to distinguish them from stones; but a poke with the oar or a long stick at once determined which they were. Professor Stoppani, in his first report (B. 67), describes this as a steinberg, but the idea of the whole island being artificial—an idea first suggested by Desor, who found analogous instances in the Rosen Insel, Lake Starnberg, in the little island at Inkwyl, and in the Irish Crannogs—gained strength by the discovery of similar stumps of piles on its north-west side. Although the local antiquaries—Ranchet, Regazzoni, Quaglia, Castelfranco, and others—occasionally visited these lacustrine stations and made considerable investigations, with the result of adding to their private collections, it was not till 1878 that any systematic researches were made with the view of testing Desor's suggestion that the island was a gigantic crannog. This was first attempted by an Englishman, Mr. W. K. Foster, of London, who happened to be residing in the neighbourhood. In carrying out the necessary excavations he had the assistance of Ranchet and Regazzoni, both experienced investigators of lacustrine antiquities. Five trenches, covering on the aggregate about 80 square yards, were dug in different parts of the island, and in all these, piles, fragments of pottery (one of which had the impression of plaited reed-work), and various other relics of human industry, were encountered. In the sections presented by these trenches the following strata were successively passed through:—

The most noteworthy objects collected in these operations were as follows:—In the first layer a Roman coin of Marcus Aurelius, and a portion of a mould for a socketed lance-head (Fig. 48, No. 19). In the second, two fragments of bronze. In the third, two polished stone celts, with a portion of a third, and two clay weights. In the fourth, a flint saw with a wooden handle, two bone pins, and some sharpening-stones.

The piles were evidently in their natural position, and the conclusion that the entire island had been a pile-dwelling was irresistible; but the questions when and by what means was the transformation accomplished, were as obscure as before. Mainly for the purpose of clearing this matter, Sig. Ettore Ponti, in September of the following year, gave instructions to have further excavations made in different parts of the island. On this occasion 12 trenches were dug, covering an area of about 230 square yards with an average depth of 3 to 4 feet.

The stratification and composition of the stuff were very similar to those experienced in the former excavations. In this space 440 piles were counted, and Regazzoni calculates that at this rate the original number of piles requisite for the construction of the entire lake-village would be from 35,000 to 40,000. Some horizontal beams were also found among the débris. Among the relics the following are noteworthy:—A tyne of deer's horn, with a flint implement inserted into the end of it (No. 3); a small clay weight shaped like a pear; several objects of worked bone, as needles (Nos. 7 to 9), pointers, chisels (No. 25), handles, etc. A knife (No. 5) and a dagger of bronze (No. 6), and two oblong beads of coloured glass with transverse grooves, were found in the stratum immediately below the surface soil.

As a rule, the tops of the piles in these trenches were on a level with the surface of the water, while those in the lake were several feet lower—more or less, according to the depth of water. The cause of this was no doubt the protection given to the former by the accumulation of débris around them. It was observed that the uppermost layer alone had yielded Roman coins, but along with them were objects of both the Stone and Bronze Ages—a juxtaposition which might be accounted for by agricultural and other operations to which the island has been subjected in historical times. The fourth, or that which lay immediately over the ancient lake-sediment, was alone exclusively deposited under water, as it contained some entire dishes, and the associated débris were just the usual contents of lake-dwelling relic-beds, viz. the shells of hazel-nuts, acorns, charred bits of wood, bones of various animals (among others the skull of an enormous wild boar), as well as implements of bone, horn, and flint, pottery, etc. The second and third layers were composed of much the same materials as the fourth, but they appeared to have been the contents of a previously-deposited relic-bed artificially heaped up, as they contained portions of wooden beams which had no definite purpose, but lay in the soil in all directions.

The relics of humanity collected on the Isola Virginia in the course of these various excavations are so numerous that one of the two rooms set apart by Sig. Ponti as an archæological museum for the lacustrine remains of Lake Varese is entirely devoted to their exhibition and preservation, where they have been carefully and neatly arranged under the skilful care of Professor Regazzoni.

Pottery.—As in the other lacustrine stations in this lake, there are two kinds of pottery—one black, and made of fine paste, of which most of the smaller vessels were made; the other is of a greyish colour, but sometimes it has a reddish tinge, and contains a mixture of fine gravel or coarse sand, which gives it a rough appearance. The fragments and entire dishes in the Ponti Museum decidedly testify to considerable skill in the ceramic art. Besides perforated knobs and tubular borings for the insertion of cords (No. 17), there are various forms of handles, as in Nos. 14 and 16, the latter of which is interesting, as it suggests the primary stage of the ansa lunata which is such a prominent characteristic of the pottery in the eastern portion of the Po valley.

The diversity of ornamentation is also worthy of notice—raised dots, nail-marks, perforated rims, lines, corrugated grooves, and cord-markings, forming a variety of combinations (Nos. 13, 15, 22, 23, 26, and 27). One bit shows the impression of plaited reed-work (No. 29). Another, an entire dish made of fine black paste, is a curiosity in its way; it consists of three cups united, and having a communication with each other by a small hole in the dividing septa (No. 24). The coarse pottery indicates vessels of large dimensions. There are also loom-weights, spindle-whorls (No. 21), some conical objects pierced vertically (No. 10), and casts of wicker-work, supposed to be the remains of the cottage walls.

Bone and Horn.—Objects of this class are numerous, as polished daggers, pointers, chisels (No. 25), needles (Nos. 7 to 9); also a few perforated teeth.

Stone.—Celts and chisels are fairly abundant, and among them are one or two of jade. Though I noted only one fragment of a perforated axe-head, the art of boring stone was known and skilfully practised, as there are several spindle-whorls and other implements with neat perforations (No. 18). There are also hammer-stones (some with finger-marks), corn-grinders, and polishers. Among the latter are large flat polishing slabs, and a few hand-polishers made like a stone celt (No. 11), which are peculiar to North Italy, if not, indeed, to the Varese lake-dwellers, as I have seen only one other out of the district, viz. at Viadana.

Among the flint objects are knives, scrapers, saws, arrow-points, chisels (like those in Fig. 68, Nos. 8, 14, and 15), cores, and a large quantity of flakes (Nos. 1 to 3). For small cutting implements flint was not the only substance used by these lake-dwellers, as there are 36 fine flakes of obsidian (No. 4), and some arrow-heads of rock crystal.

Bronze.—The bronze objects in the museum, including fragments, amount only to 15, and represent knives, fish-hooks, etc. (Nos. 5, 6, and 12).

Amber.—There is also a bit of amber which appears to have been an ornament.

Small square or oblong pieces of wood perforated (No. 20) are supposed to have been floats for nets.

The organic remains collected in the fourth stratum, which was considered to be the true relic-bed of the palafitte, were submitted to Professor Sordelli, who recognised, among other seeds and fruits, the following:—Millet (Panicum miliaceum), wheat (Trit. vulgare), bramble (R. fruticosus), and the vine (Vitis vinifera).

Among the bones of animals identified were those of the bear, wolf, badger, beaver, wild boar, stag, roe, etc. The ordinary domestic animals were also represented, and in addition to them I have to mention portions of two human jaws which were found a few inches below the tops of the piles. (B. 324, 326, 341, 343a, 359, and 437.)

Professor Castelfranco (B. 456), who has carefully studied the phenomena presented by these repeated excavations, formulates the following theory as to the succession of events which have brought about the evolution, so to speak, of the Isola Virginia:—

(1) The original palafitte had been destroyed by a conflagration towards the close of the Bronze Age or the beginning of the Iron Age.

(2) Its inhabitants were hunters, fishers, rearers of domestic animals, and agriculturists.

(3) Shortly after the destruction of the pile-village, its subsequent occupiers converted the larger portion of its site—which had already, in parts at least, reached the surface by the gradual accumulation of débris—into a veritable island, by heaping over it stuff dug from the margin and especially from the landward side, where there is now a channel separating the island from the mainland. Thus the upper layers contain the débris of the earlier people, mixed with sand, gravel, and mud. This view is rendered probable by the fact that in one place, towards the north of the island, the second layer was displaced by an artificially constructed bed of large pebbles.

(4) The newcomers, to whom Castelfranco assigns the transformation of the palafitte into an island, were the Ligurians, whose "sépultures à cineration" are so numerously found in the neighbourhood.

Bodio.—The bay opposite this village contains the remains of three stations, the most southerly of which is known as "Keller" or "Del Gaggio," the next as "Bodio Centrale" or "Delle Monete," and the third as "Desor" or "Del Moresco." All these are comparatively near the shore, being only about 30 yards distant, and the central one is about equidistant—some 800 yards—from the other two. (B. 327, p. 47.) The central station appears to have been a true steinberg, as its area was covered with stones; regarding which Stoppani remarks that formerly they were more numerous, because within recent times some were known to have been removed for building purposes. At first more bronze objects were found on Keller, and more pottery on Desor, while the Centrale was characterised by the discovery on it of a hoard of Roman coins. Subsequent investigations have not borne out these early distinctions based on the character of their relics, and they are now generally acknowledged to belong to the same age.

The coins found on the Centrale were mostly small silver pieces, much decomposed, belonging to the last half-century of the Republic. Stoppani collected about 70, and Angelucci, who explored shortly after him, no less than 128. One found by Regazzoni in 1876 (B. 327, p. 52) has on it, along with the head of Mark Antony, the following legend:—M. ANT. IMP. AUG. III. VIR. R.P.C. M. BARBAT. Q.P., etc., which would make the date about 40 B.C. The hoard is supposed to have been lost or deposited here long after the lake-dwelling ceased to be inhabited—a supposition that is borne out by the fact that the coins were confined to one limited spot, only a couple of yards square. In 1876-7 Sig. Ponti made researches on Desor which greatly enriched his museum both in stone and bronze objects. (B. 327.) A selection of objects from these stations is given on Fig. 49.

Cazzago-Brabbia.—This station is situated opposite the village of the same name, and at first it gave such poor results that Stoppani called it a trial station, or an attempt to found a settlement. From the researches made in 1877 it was found to be rich in remains, and exactly similar to those at Bodio. It was, however, farther from the shore, and extended parallel to it for about 150 yards. Its breadth was somewhat irregular, and, judging from the disposition of its piles, it would appear to have been two quadrangularly-shaped stations nearly in contact with each other. Among the bronze objects collected on this station are four lance-heads, a chisel, an awl, 10 fish-hooks, four hair-pins, a fibula, etc. (B. 456.) There are also some fine arrow-points of flint.