Combativeness, to use the language of phrenology, is one of the most lively instincts of humanity. The Bible tells us of the struggle between the sons of Adam, and shows us might making right ever since the days of primeval man. History is but one long account of wars and conquests, victories or defeats, and progress is chiefly marked in inventions which made battles more sanguinary and added to the number of victims slaughtered. At the very dawn of humanity man learned to make weapons; very soon, however, weapons ceased to appear sufficient. The first fortification was doubtless the cave, which its owner strengthened by closing the entrance with blocks of stone and piles of broken rock, or by digging deep trenches about it.
Population rapidly increased and war was declared between tribe and tribe, nation and nation, race and race. Terrible must have been the struggles between invaders and the original possessors of the soil. Means of defence were multiplied to keep pace with new modes of attack, and our ancestors of the Stone age were intelligent enough to make places of refuge in which on necessity they could shelter their wives and children, and later, when they became sedentary, their flocks and their stores of grain. In many different localities we find the remains of camps and fortifications, which, to avoid using a more ambitious term, we may characterize generally as enclosures.1
These primitive enclosures, says Bertrand in his “Archéologie Celtiquc et Gauloise,” may have been very much more numerous than is supposed, if we include amongst them, as it appears we ought, many ruins long thought to date from the Roman era.
There is no doubt as to the purpose served by the camps, but we are not prepared to speak as positively as does Bertrand as to their origin, and the difficulty of deciding is very greatly increased on account of these camps having been successively occupied at different epochs by different peoples. Bearing in mind this reservation, we will now sum up to the best of our ability all that is so far known about the most important remains hitherto examined.
The residence of prehistoric man in the rich districts between the Sambre and the Meuse is proved by worked flints, fragments of pottery, and human bones dating from most remote times. The stations successively occupied were situated near watercourses or copious springs, and, where possible, on isolated escarped plateaux surrounded by ravines. Hastedon, about a mile and a quarter from Namur, is one of the best examples we can quote.2 The camp, first made out in 1865, formed a long square, covering some thirteen hectares, or about thirty-two acres. It is situated on an isolated mound connected with the main plateau by an isthmus 227 feet long, and is protected on the south and west by a deep ravine: To these natural defences men had added important works to those parts that were accessible. The cutting of trenches a few years ago brought to light walls of a mean thickness of more than nine feet, formed of masses of rock and sand and round pieces of wood parallel with a revêtement of dry stones surmounted by a palisade consisting of three pieces of wood parallel with the walls, and seven perpendicular traverses. All the wood was charred; the besieged had evidently been driven out by fire. Excavations led to the finding of Roman coins; this and the resemblance of the palisades to those described by Cæsar,3 the very name of Hastedon, and the tradition everywhere prevalent in the district, that this bad been the site of a Gallic Roman camp, led to the general adoption of that opinion. In fact, Napoleon III. actually ordered excavations to be made in the hope of finding traces of the Atuatuques, one of the roost warlike of the tribes of northern Gaul; but side by side with historic relics were no less than ten thousand flints. These are chiefly merely chips or nuclei which had served as hammers, or long thin slices, with some few arrow- and lance-heads often skilfully cut, some polished hatchets, and saws with fine teeth. Nearly all are notched and worn with use, which does away with the idea that the place where they were found was the site of a workshop such as I have already described. With these worked flints were found some fragments of coarse pottery, which could not possibly be confounded with Roman or Gallic work. The flints and pottery, and the walls put together without cement, point to the conclusion that if the camp of Hastedon was occupied by the Roman legions, it was long previous to their day inhabited by some Neolithic race, ignorant of the use of any but stone weapons and implements.
The camp of Pont-de-Bonn in the commune of Modave (Namur) very much resembles in its arrangement that of Hastedon.4 A mound stands out upon the plain protected on the north and west by rocks difficult of access and connected with the main plateau by a very narrow tongue of land. Outside we can make out regular trenches parallel with each other, and connected by a wall of masonry, at the foot of which wall were picked up a good many iron nails. Inside the enceinte itself worked flints were associated with Roman coins. Are not these proofs in the first place of a long Neolithic occupation, then of the residence of Gallic Romans, and yet later of even more modern people of whom the masonry walls and iron nails are relics?
Limburg also contains some defensive works, many centuries old, which are as yet but little known. We may mention amongst them the so-called dyke of Zeedyck, near Tongres, a formidable intrenchment some 2,186 yards long by more than 325 feet wide at the base, and of a height varying from 49 to 65 feet; the earthen ramparts of Willem on the Geule, the not less important ones of Houlem, with many others far away from the great highways of communication, but within the limits of the two provinces of Liège and Limburg.5
A few years ago Bertrand said that there are in France some four hundred earthen enceintes, only sixty of which contain relics connecting them with the Gallic Romans. Since Bertrand’s announcement this number has been greatly increased, thanks to eagerly prosecuted local researches. De Pulligny mentions a hundred in Upper Normandy6; Martinet says they are very numerous in Berry; one of the most remarkable, the quadrilateral of Haute-Brenne, covered an area of nearly three thousand acres.7 Amongst the forests on the Vosges Mountains were discovered long single and double walls, the course of which follows the crest of the ramparts overlooking the valley of the Zorn, between Lutzelbourg and Saverne.8 At Rosmeur, on Penmarch Point (Finistère), Du Chatellier excavated two tumuli which appear to have been connected with a series of defensive works encircling the whole promontory.9 It would be merely fastidious to multiply instances, we will content ourselves with describing a few of the most interesting of these antique fortifications.10
The camp of Chassey (Saône-et-Loire) may be compared with those of Belgium. It is situated on a plateau 2,440 feet long by a width varying from 360 to 672 feet. A huge natural rocky barrier rises on the south and east, whilst on the northeast and southwest we find two important intrenchments made of huge blocks of stone with a revêtement of earth. One of these intrenchments is 45, the other only 29 feet high. There is no trace inside of springs, and the inhabitants must always have had to obtain their water-supply by artificial means. The cisterns now in this camp appear to have been dug out with iron implements, and are certainly of later date than the first occupation of the plateau. Numerous objects picked up in the Chassey Camp belong to Neolithic times, but the people who have occupied it since those remote days, the men of the Bronze and Iron ages, the Gauls, the Romans, and the Merovingians, have so turned over the ground that products of industries, completely strange to each other, are everywhere mixed together in inextricable confusion.11
There were originally a good many hearths about the camp, and it was near to one of them that the spoon was found, figured in an earlier chapter of this book (Fig. 25). With it were picked up polished fibrolite, basalt, chloromelanite, serpentine, and diorite hatchets; evidently made in the neighborhood, as is proved beyond a doubt by the numerous chips and partly worked pieces lying about, as well as the discovery of no less than thirty polishers, many of them showing signs of long service. Bone implements of all kinds and whistles made of the phalanges of oxen are also constantly found. Even if the presence of these objects does not enable us to come to any final conclusion, they are at least most useful and interesting in enabling us to put together little by little a picture of the life of the most ancient inhabitants of France.
The camp of Catenoy, Dear Liancourt (Oise) is arranged very much in the same manner as that of Chassey.12 Cæsar’s Camp, as it is called by the people of the neighborhood, forms a long triangle, the apex of which rests on the eastern extremity of the plateau. Excavations have yielded a number of Gallic-Roman objects, with some polished hatchets, some broken, others intact, with stone and bone weapons, resembling but for a few slight differences those we have described so often. Numerous fragments of pottery were also picked up, which pottery, hand-made and mixed with crushed shells, seldom has either handles or any attempt at ornamentation. Weapons, implements, and pottery are all alike totally different from any Roman or Gallic work known. It is impossible to study the relics at Catenoy without coming to the conclusion that the camp was occupied at periods prior to Gallic and Roman times, and that there, as in many other districts, the Latin conquerors had succeeded an unknown vanquished race.
De Quatrefages has accurately made out a series of works extending along the left bank of the Nive, as far as Itsassou, and of which the Pas-de-Roland marks the extreme limit. A merely superficial examination is enough to show that these defences existed only on the side to which access would otherwise have been easy, while the height overlooking the river on the other side, which is impregnable by nature, has been left untouched. Here too we find the name Cæsar’s Camp given to the relics, a fact of common occurrence all over France, where the great captain was long held in honor. Quatrefages is, however, of opinion that the works are neither Roman, Gallic nor Celtic, and he even arrives by a process of elimination at the conclusion that they were erected by the Iberians, who preceded the Aryans, and have left so deep an impress on all the countries they successively occupied. We do not feel able to accept entirely this hypothesis; but no suggestion of the eminent professor must be overlooked by those who earnestly seek with unbiassed minds to ascertain the truth.
Gregory of Tours relates that at the time of the invasion of the Vandals, the Gabali took refuge with their families in the Castrum Gredonense, and there, for two years, energetically resisted the invaders.13 Grèze, now a little market town of the department of Lozère, is the castrum of which the old French chronicler speaks, and Dr. Prunières there collected forty stone hatchets, differing in no material respect from others found in such numbers elsewhere, with flint knives and scrapers, bone stilettos, and millstones, doubtless used for grinding grain, all of which are to the learned French professor proofs of the existence there of a Neolithic station before the historic period.
In the department of Alpes-Maritimes a series of defensive works crown the circle of mountains which rise from the shores of the Mediterranean. These intrenchments certainly date from a remote period, though we cannot assign them to any definite time, and the fact that they have been repaired at different epochs proves that they were successively occupied.14 They consist principally of circular or elliptical enceintes surrounded by walls of stones without mortar, and they vary in diameter from some 39 to 328 feet. One of the largest is that on the Colline des Mulets, above Monte Carlo.
Prehistoric spoon and button found in a lake station at Sutz (Switzerland).
Although the pile-dwellings of Switzerland and of the terremares of Italy would appear to have been in themselves protection enough, their inhabitants did not neglect other means of defence, from which we may gather that they were engaged in constant and terrible struggles. The terremares were generally surrounded by a talus or rampart of earth, with an external fosse which protected the approaches to the dwellings. The rampart of Castione (Parma), which dates from the Bronze age, was even strengthened inside with large timber caissons.15 In Switzerland, some works recently undertaken to deflect the course of the Aar, on its exit from Lake Bienne, have led to the discovery of a village of the Stone age, with the bridges leading to it and the little forts intended to protect it.16 As have the neighboring settlements, this station has yielded a great many arrows, hatchets, scrapers, and harpoons. We give an illustration of a curious marrow spoon, and of a round object which seems to have been a button (Fig. 84), as they mark the progress made.
Great Britain is intersected by lines of fortifications of unknown origin, but certainly of extreme antiquity. We may mention Dane’s Dyke, Wandyke, the Devil’s Dyke at Newmarket, and Offa’s Dyke, running from the Bristol Channel to the Dee, and dividing England from Wales. Ancient camps and intrenchments, Sir John Lubbock tells us, crown the greater number of the hills of England. General Pitt-Rivers explored several of these camps in the county of Sussex. Many extend over considerable areas, and all contain numerous worked flints and other relics of prehistoric industry. These relics are met with in great numbers at the base of the intrenchments, so that we may justly conclude that they date from the same epoch.
The most celebrated of these camps is that of Cissbury, three miles north of Worthing. We may also mention that of Hod-Hill in Dorsetshire, which greatly resembles the one at Cissbury, but we will describe the latter in some detail.17 It is situated on a somewhat lofty plateau of irregular form, its site having been chosen with great skill as one offering great facilities for defence. The earthen ramparts and the fosses protecting them cover an area of sixty acres, and their importance varies according to the relief of the ground; thus the thickness of the walls is very much greater on the eastern side where an attack would have been most fraught with danger; four doors give access to the interior, and on each side of these doors are ruins of rectangular structures strengthening their defence. Archæologists, however, are of opinion that these redoubts, though their construction is exactly similar to the rest of the fortifications, are of more recent date. In fact Roman tiles have been found amongst the ruins, but these really prove nothing, as every one is agreed that Cissbury was occupied by the Romans after the subjugation of England by them; and the only point at issue is really whether the walls of which the ruins still remain date from the Roman period, or from times prior to their arrival. We ourselves lean to the latter opinion, as drinking-water is absolutely wanting; a very important point, as the Roman generals always made it their first care to pitch their camps near a good water-supply. On the western slope at Cissbury on each side of the ramparts are fifty funnel-shaped depressions, some of which are as much as seventy feet in diameter and twelve feet deep. These holes may have served as refuges, and the larger ones were certainly lived in, as is proved by the charred stones of the hearths and the pieces of charcoal found near them; moreover, Tacitus18 tells us that the Germans lived in similar habitations. Whatever, however, may have been their ultimate use, these hollows were in the first place dug out with a view to obtaining flints in the marly chalk forming the bill; and recent excavations have revealed the existence of galleries connecting the depressions. When they became later human habitations some of the inside openings were blocked up with lumps of chalk, carefully piled up so as to make entrance extremely difficult, greatly adding to the security of the inmates.
Thirty of these shafts were excavated in succession; and amongst the rubbish of all kinds with which they were filled were found some well cut celts, showing no trace of polish, and some weapons or tools of the Moustérien type. The number of half-finished implements, and the even greater quantity of chips, points to these shafts having formed a centre of manufacture. Many of the implements were made of stag-horn, and amongst them we must mention some picks which, curiously enough, exactly resemble those of Belgium and the south of France.19 Similar wooden picks are found in the copper mines of the Asturias, in the salt mines of Salzburg, and in a petroleum well recently opened on the frontier between the United States and Canada. In all these localities traces can be made out of ancient mining operations. But to return to Cissbury: from amongst the prehistoric ruins there were also taken, numerous fragments of pottery, not at all like Roman ware, with the bones of the horse, goat, boar, and ox, all still represented in the fauna of England; with oyster-shells, and the shells of both land and sea mollusca, of species still to be found in Great Britain. But no trace has so far been discovered of metals, and neither the flint implements nor the bones of animals have any of the marks of rust so characteristic of the Bronze and Iron ages. Must we not then conclude that these shafts were sunk at a time long prior to the earliest historic period?
The walls of the subterranean galleries of Cissbury bore not only cup-shaped ornaments, strive, and curved or broken lines, recalling those on the megalithic monuments of Scotland and Ireland; but Park Harrison has made out some regular runes, or written characters, of which a reproduction was shown at the Paris Exhibition in 1878. This last fact is the more curious, as Sayce discovered in a passage giving access to a cave near Syracuse some characters somewhat similar in form, to which he assigns a proto-Phœnician origin. We may add that certain characters made out at Cissbury, differing but little from the modern letter b or the figure 6, are also found in the most ancient Palmyrian, Copt, and Syrian alphabets. Were this fact completely established, still more, if it were corroborated by other analogous facts, we should in it have a very valuable indication of the relations of England with the most ancient known navigators.
Germany also contains some ancient fortifications, of which the most remarkable are the Heidenmauer of Saint Odila, near Hermeskiel, between the Moselle and the Rhine. Huge stones, piled up without cement, form a triple enceinte, but there is nothing to connect these remains with prehistoric times. It is the same with the intrenchments in the Grand Duchy of Posen, the existence of which was announced at a meeting of the Anthropological Society of Berlin.20 Many of these defensive works, notably those of Potzrow and of Zabnow, bad been erected on piles. In the district between Thorn and the Baltic are numerous mounds of the shape of a truncated cone, the platform of which is surrounded by an embankment some 590 feet in diameter.21 Near many of these were picked up many broken human bones, mixed together in the greatest confusion with weapon, hatchets, and hammers, resembling Neolithic types. Everything bears witness to the struggles of which these mounds were the scene.
Similar relies of a past still obscure are met with in the south of Europe. Cartailhac has brought into notice the citanias, which are strange fortified towns in Portugal. On the plateau of Mouinho-da-Moura, southwest of Lisbon, were found numerous polished hatchets, associated with shells of marine mollusca and the bones of mammals belonging to species still extant.22 This station was protected by intrenchments of so great an extent that it has been impossible to examine the whole of them. There are also near the same place several caves, now nearly choked up. One of them was originally a regular tunnel; the cutting leading to the entrance was made of earth and small stones; it contained the bones of animals, some cinders, and four large vases of coarse workmanship. It is difficult to make out what this cave was used for, the great confusion in which the bones lay excluding all idea of its having been a tomb. Ribeiro had already made out at Lycea an intrenched camp protected by clumsily constructed walls. Inside the enceinte he picked up numerous fragments of ornamented pottery, with polished hatchets, shells, and a good many bones of animals. He also made out several sepulchres.23
General view of the station of Fuente-Alamo.
The prehistoric station of La Muela de Chert in Maeztrago reminds us of those of Portugal. It is situated on a little eminence, protected on the north and east by the natural escarpment of the plateau, and on other sides by a wall of some height made of stones without mortar. Some foundations of an oval shape, on which doubtless were built the homes of the inhabitants, can be made out in the middle of the enceinte. We can, however, but repeat here what we have said so often elsewhere, that it is impossible to fix the exact date at which these intrenchments were made. The discovery, however, of polished flint hatchets, diorite lance-heads, and a few bones of ruminants and cerviæ unknown in Spain in prehistoric times, would appear to point to a very considerable antiquity. Lastly, two young Belgian engineers24 have lately made out between Almeria and Carthagena a considerable number of prehistoric stations in which can be traced successively the different Stone ages and those of Copper and of Bronze. Several of these stations (Fig. 85) are regular fortified camps, protected by thick stone walls cemented with a thin layer of clay. The fire which destroyed the habitations has left behind, beneath the ashes and cinders, numerous objects, with the aid of which we are able to form a picture of the life led by the men who built the fortifications, and we know that they were agriculturists, for the very stores of grain have been found charred and agglutinated by fire. In the more recent stations flint, which was in the earliest time the one material used, has disappeared and is replaced by the copper, of which a plentiful supply was found in the rich mines riddling the mountains. Excavations have even brought to light the workshop of the metallurgist, with its moulds and vases converted into crucibles, its essays at new forms, its scoriae, and lastly its finished weapons, showing real skill in their production.
Although it is impossible to assign to them a definite date, we must, to make this part of our work complete, say a few words on the earthworks met with in Roumania. A former minister of that principality, M. Odobesco,25 classes them as valla, tumuli, and cetati de pamentu or citadels.
The valla include important works. One of them cuts across Valachie parallel with the Danube and loses itself in Southern Russia. Another crosses the north of Moldavia and Bessarabia, following a direction convergent with the former. These valla, although they are known in the country in which they occur as Fossés de Trajan, are certainly of earlier date than the Roman occupation, and in fact Roman roads cut across the intrenchments or fosses which have been levelled or covered over to make way for them. Excavations of the large tumuli are not yet sufficiently advanced for us to hazard an opinion about them. The smaller ones, however, are seldom of Roman origin. The funeral vases of calcareous stone which they contain bear witness clearly enough to their destination, and also to the rite with which they were connected.
The cetati de pamentu are regular earthen fortifications set up within short distances of each other on all the heights overlooking the torrential rivers of Roumania. These intrenchments, generally of round or oval form, are protected by deep fosses, parapets, and palisades. Masses of cinders and burnt earth bear unmistakable evidence to the cause of their destruction. All about, excavations have brought to light coarse pottery, grindstones for crushing grain, stores of millet which had been damaged by the flames, and a few primitively constructed bronze idols. When the vanquished Roumanians were driven from their intrenchments, they had evidently learned to use bronze, but were still, as we have already remarked, unacquainted with iron, as no object in that material has been found, nor does anything bear any trace of rust.
Thus, throughout Europe, man, in the presence of the many dangers surrounding him, endeavored in the very earliest times to protect by similar means his family, his flocks, and his wealth. In America we are able to quote facts of even more importance. The vast territory comprised between the Alleghanies and the Rocky Mountains, between the great lakes of Canada and the Gulf of Mexico, is intersected with truly colossal fortifications, almost all of them made entirely of earth. The ancient Americans knew how to protect every height and every delta formed by the junction of two rivers with redoubts, walls, parapets, fosses, and circumvallations. Not without astonishment we make out a regular system of fortresses connected with each other by deep trenches and secret passages, some of them hewn out beneath the beds of rivers, observatories on the heights, and concentric walls, some actually strengthened with casemates protecting the entrances. All these works were constructed by the so-called Mound-Builders, of whose ancestors or of whose descendants absolutely nothing is known.
All the strongholds of the Mound-Builders rise near abundant watercourses, and the best proof that can be given of the intelligence which guided their constructors in their choice of sites, is the fact of the number of flourishing cities such as Newark, Portsmouth, Cincinnati, Saint Louis, Frankfort, and New-Madrid, etc., which were built upon the ruins of various earthworks.
It would take us too long merely to enumerate all the ancient fortifications still existing in North America. Moreover they all resemble each other so much that the description of a few of them is really all that is needed to prove their importance.
Fort Hill (Fig. 5, p. 39) rises from an eminence overlooking a little river called Paint Creek; the walls vary in height from eight to fifteen feet, and exceed thirty feet in thickness.26 Several doors facilitate entrance, and one of them leads to a square enciente, the walls of which have been almost entirely destroyed. This enclosure probably contained the homes of the people, which may have been mere cabins of adobes or sun-burnt bricks, or buts covered with rushes, interlaced branches, or the skins of animals; on this point we are reduced to guesswork. In the centre of the principal enclosure can be made out, in almost every case, several much smaller enclosures, each containing in their turn one or more mounds. Some think these were consecrated to religious rites, but this is a mere conjecture, for nothing is really known of the form of government or of the religion of the Mound-Builders.
Forest trees have grown up on these abandoned ruins, succeeding other vegetable growths; the huge girth of the decaying trunks proving their longevity. Man, impelled by motives we cannot fathom, had abandoned the districts where everything bears witness to his power and intelligence, and the vigorous vegetation of nature once more has it all its own way.
The most remarkable group of prehistoric fortifications in North America is perhaps that near Newark, in the valley of the Scioto. It includes an octagonal enceinte eighty acres in area, a square enceinte of twenty acres, with two others, one twenty the other thirty acres in extent. The walls of the great circle are still twelve feet high by fifty feet wide at the base. They are protected by an interior fosse seven feet deep by thirty-five feet wide. According to measurements carefully made by Colonel Whittlesey,27 the total area covered by these intrenchments is no less than twelve square miles, and the length of the mounds exceeds two miles. The large entrances protected by mounds thirty-five feet high, the avenues leading to them which are regular labyrinths, the quaintly shaped mounds—one, for instance, represents the foot of a gigantic bird—all combine to strike the visitor with astonishment. We give a representation (Fig. 86) of a group, not unlike that we have just described, which is situated at Liberty (Ohio), and includes two circles and one square. The diameter of the great circle is 1,700 feet, and it encloses an area of forty acres, whilst that of the smaller enceinte is 500 feet; the area of the square, each side of which measures 1,080 feet, is twenty-seven acres. The walls are not strengthened by any ditch, and, contrary to general usage, the earth of which they are made was dug out from the inside of the enciente itself. We may also mention Old Fort (Greenup County, Kentucky, successively described by Caleb Atwater, Squier, and J. H. Lewis. It is situated forty feet above the river, and the total length of the walls exceeds 3,175 feet. Six entrances give access to it, and in the centre rises a mound representing some animal, a bear probably, measuring more than 105 feet. Several small mounds, beneath which were found human bones, cluster about the larger one.
Group at Liberty (Ohio).
Trenches at Juigalpa (Nicaragua).
We must not omit to name an extraordinary system of intrenchments at Juigalpa, in Nicaragua, which so far as I know is quite unique. This is a series of trenches extending for several miles (Fig. 87), varying in width from nine and a half to thirteen feet; at equal distances are oval reservoirs, the longest axis of which measures as much as seventy-eight feet. In each reservoir are two or four mounds, probably serving as watch-towers. We know nothing either of the people who erected these singular structures or of the enemy from whom they formed a protection. Nor can anything be guessed as to the way in which the defence was conducted. All is involved in obscurity, and at every turn we are compelled to repeat that prehistoric studies are weighted with uncertainty, long and arduous study being necessary to bring ever so little order into the chaos in which everything connected with them is involved.
We must cursorily refer to some other fortifications which really scarcely belong to our subject, though certain archæologists claim for them a prehistoric origin. We refer to the vitrified forts, which are strange structures in which stones, such as granite and gneiss, quartzite and basalt, have been subjected to a heat so intense as to produce vitrification.
These vitrified forts are enceintes, generally of round or elliptical form, carefully erected where they were most needed for defence, and protected by one or more ramparts.28 The ramparts all bear traces of vitrification, more or less complete, which has, so to speak, cemented them together. The vitrification is very unequal, being complete in some parts and scarcely noticeable in others. It is evident that the builders did not know how to direct their fire uniformly.
Ever since 1777 vitrified forts have been known in Scotland, and until 1837 they were supposed to exist nowhere else. About that time, however, Professor Zippe called attention to similar ruins in Bohemia, and later it was announced that discoveries of the same kind had been made in various parts of France, Denmark, and Norway. Virchow speaks of the Schlaken Wälle, or ramparts of vitrified scoria, near Kern29 and Schaafhausen, and gave an account of them at a meeting of German naturalists at Ratisbon. It would be easy to multiply instances. Vitrified walls are known in the Puy-de-Dôme, in which the facing is of clay, and draught flues, for regulating and fanning the flames, have been made out. At Castel-Sarrazin is a camp refuge with similar dispositions,30 and recently Daubree presented to the Académie des Sciences a piece of porphyry artificially vitrified from the prehistoric enceinte of Hartmannswiller Kopf in Upper Alsace.31
It is in Scotland, however, that are situated the most remarkable vitrified forts. A few years ago no less than forty-four were counted. The most celebrated are those of Barry Hill and Castle Spynie in Invernesshire, Top-O-Noth in Aberdeen, and a small fort which rises from a lofty rock in the midst of the Strait of Bute. Vitrified cairns also occur in the Orkney Islands, notably on the little isle of Sanday, but the most interesting structures of the kind are Craig Phœdrick and Ord Hill of Kissock, which rise up like huge pillars on the hills at the entrance of Moray Firth, at a distance of three miles from each other.32
Craig Phœdrick is now covered with a luxuriant vegetation of broom, furze, and fern, with groves of firs and larches, amongst which the explorer makes his way with difficulty to the fortifications, or rather to the piles of massive blocks to which that name has been given. These blocks form an acropolis of oval form, the upper part of which is a flat terrace encircling a central basin some six and a half to nine and a half feet deep, which may be compared to the craters of the extinct volcanoes of Auvergne. The sides of the mound are strewn with cyclopean blocks of vitrified granite, which evidently originally formed part of the fortifications. It is on the eastern side, overlooking the valley of the Ness, that the buildings are of the greatest importance; two terraces can be made out, the lower projecting beyond the upper, forming a double series of almost perpendicular fortifications, constructed of vitrified blocks cemented together with thin layers of mortar, spread without any attempt at regularity. The blocks form, with the mortar, a conglomerate so compact that when struck with a hammer they break without separating. Examination of fragments under the microscope prove that they have gone through important mineralogical transformations, under the influence of what must have been an extremely high temperature. The heat must have been indeed intense which could cause mica to disappear entirely, and feldspar to melt almost completely.
The hill known as Ord Hill of Kissock is crowned, as is Craig Phœdrick, with ruins still standing, but the vegetation about them is so dense and thorny that it is difficult to make out the condition of the remains. The ruins, which can only be seen from one side, appear however to have formed part of fortifications, dating from the same time and serving the same purpose as those of Craig Phœdrick. Were they forts? There is certainly no sign of their having been used as habitations. Or were they, as some archæologists are disposed to think, beacon houses used for warning the people of the approach of the Norman pirates or Scandinavian Vikings, whose depredations were not discontinued until the eighth century of the Christian era? Hypotheses are always easy, but proofs of these hypotheses are difficult to find, and we confess we have none to bring forward.33
Passing to France, we find the greater number of vitrified forts in the Département de la Creuse. At Châteauvieux is an enceinte of oval form, 416 feet wide at its broadest part.34 An earthwork, 22 feet wide at the base, serves as foundation to a wall, the outer and inner portions of which consist of small granite stones, arranged in regular layers. The space between the two series of small stones is filled in with a sheet of melted granite, some twenty-four inches wide, resting on calcareous tufa. The whole mass is completely vitrified, and regular geodes or nodules lined with crystals and draped with pendent drops of melted rock have been produced.
The ancient fortress of Ribandelle, of circular form, rises above the Creuse, opposite Châteauvieux. It was successively occupied by the Celts, the Romans, and the Visigoths, but we are unable to fix the date of its erection or the name of the people who built it. There remain but a few ruins at the present day, but we can make out in them the same mode of construction as that followed at Châteauvieux. The walls are faced with unhewn stones, the outer side of which still retains a natural appearance, while the inner is corroded and disintegrated. In the wall itself, separated from the facings by beds of peat mould, are great blocks of vitrified granite. The traces of the action of fire are specially noticeable in the upper part of the walls, so that they were evidently finished when the fusion took place.
The site of the furnace in these forts is difficult to determine. It was evidently not situated under any of the blocks, for the earthworks on which they rest retain no traces of the action of fire. Nor was it situated at the side, for the outer facings have retained alike their original form and consistency. Nor can the furnace have been lit on the blocks, as heat exercises its action by radiating in every direction. We are therefore forced to the conclusion that the fire was spread with the aid of spaces left in the inside of the construction at various points, for the vitrified mass is divided into blocks, about nine and three fourths feet long, at very short distances from each other.
These few examples will be enough to give some idea of the strange vitrified forts. Many of them retain traces of Roman. occupation. The Guéret Museum possesses a fragment from the Ribandelle walls in which a Roman tile is completely imbedded; and M. Thuot picked up other tiles in a similar condition amongst the ruins. This is a very decided proof that the vitrification took place after the arrival of the conquerors of Gaul. The weapons and tools discovered would appear to confirm this idea, and to suggest similar explanations of vitrification elsewhere. If so, we shall have to admit that vitrified forts date from the earliest centuries of the Christian era, and are not prehistoric at all. We have, however, noticed them here on account of the grave doubts in the matter, and because they furnish a striking and valuable illustration of the relations existing from the most remote tunes between widely separated races, and maintained until the present time. In no other way can we account for the practice of the extremely difficult and complicated operation of the vitrification of bard rocks in districts so far apart as Norway and Scotland, Germany and the midlands of France.
The more we think of the difficulties vitrification presents, the greater is our astonishment. How was the fusion achieved of elements so refractory alike in their structure and in the resistance offered by accumulated masses of material? By what processes was heat brought up to the 1300 degrees necessary for the fusion of granite? The incineration and fusion of the materials of which the vitrified forts are made, especially the granite ones of La Creuse and the Côtes du Nord, bear witness, says Daubrée, to a surprising skill and knowledge of the management of fire in those who burned them, but these qualities were manifested also in extremely ancient metallurgical operations. It is quite impossible to suppose the vitrification to have been the result of a conflagration. No fire, whether accidental or the work of an incendiary, could be powerful enough to produce such results. The use of petroleum in the most terrible conflagrations of our own time—those of the Commune in 1871, for instance—did calcine and disintegrate stone, but I know of no case of vitrification.
The Keramic Museum of Sèvres contains several specimens which present very notable differences to each other. Those from Château-Gontier are formed of very close-grained quartzite granite of a greenish color streaked with black. The conglomerate welding there together is a vitrified scoria full of very small bubbles made by the escape of gas which had not had sufficient strength to get out. The block from Sainte-Suzanne (Mayenne) consists of quartz mixed with half calcined grains of feldspar, bleached by the action of fused glass, which once introduced filled up as it congealed all the vacant spaces with a vitreous substance of light greenish-white color. The fractures are green and bright, and are dotted with white points, which are all that is left of the stones after their disintegration in the grip of a heat that was alike intense and rapid in its action. The fragments brought from Scotland differ from those just described. They consist of small pieces of granite completely merged in a thick paste with which they form the mass, the whole breaking together when it does break; and the melted matter seldom has any bubbles in it.35
The process employed in cementing the materials of the vitrified forts was then perfectly unique. The processes employed to obtain the necessary heat varied according to circumstances and according to the nature of the materials used. At Sainte-Suzanne and at La Courbe marine salt was used as a flux. Captain Prévot36 thinks that the walls were smeared with a coating of clay, and that as in the baking of bricks spaces were left between so as to produce more intense heat. M. de Montaiglon is of opinion that the buildings were in the first instance erected without the use of any calcareous or argillaceous material, and that glass in a state of fusion was poured over them afterwards, this glass consolidating them and forming with them one indestructible mass. M. Thuot seems much disposed to share this last opinion, but he thinks that some chemical materials such as soda or potash were also used. Yet one other possible solution may be mentioned, a solution which is becoming more and more generally accepted, namely that the granite was not after all really melted, but that the vitrification should either be attributed to the fusion of the argillaceous mass, which has been subjected to an igneous transformation, such as that which often takes place in furnaces for baking bricks and in lime-kilns.37
Whatever explanation we may accept, however, the processes employed certainly bear witness to a much more advanced state of civilization than was acquired in the earliest ages of humanity. We have been led by the great interest and mystery of the subject to dwell longer on it than we intended, and we must hasten to return to prehistoric times with a determination not to transgress again.
Fortifications are a proof of combined action leading to a common end; they imply social organization, chiefs to command, workmen to obey. A recent discovery enables us to form a very accurate picture of prehistoric men gathered together not only for purposes of defence, but in a society already rich, industrious, and, if we may so speak, learning to cultivate the arts of peace.
The Ægean Sea has ever been the theatre of igneous phenomena, and the three little islands of Thera, Therasia, and Aspronisi, which shut in the Bay of Santorin, are built up chiefly of volcanic materials.38 In 1573 an eruptive cone suddenly appeared; in 1707 the inhabitants of Santorin saw rise up a short distance from their shores a rock that increased in size for several days and then suddenly split up. This splitting up was succeeded by a great eruption of incandescent materials; an eruption which lasted for no less than five years, forming at the end of that time an island some 400 feet high by 3,279 feet in circumference. In 1866, after many violent shocks of earthquake, the ground was rent asunder on this island and masses of volcanic matter were belched forth, whilst on the other side of the island the soil sank to such a degree that canoes were used to get to houses which but the day before were nine feet above the sea-level. This eruption went on until 1870, and the quantity of scoriæ vomited forth during its continuance welded three islets, which had hitherto been separate, to the principal island, of which they now form part. On entering the Bay of Santorin we see on every side banks of lava, beds of scoriæ, and piles of cinders of a purplish-gray color rising in cliffs to a height of more than 1,312 feet. All these materials are the result of innumerable eruptions, and the central crater of the volcano is probably situated about the middle of the bay. It is supposed that at one time a conical mountain, from 1,958 to 2,600 feet high, rose where soundings now give a depth of water of over 1,300 feet. A sudden break up of the mountain probably produced this abyss, and formidable eruptions have led to the pouring forth of immense quantities of pumice-stone. The three islets mentioned above would be the remains of the old central cone, and a bed of pumice-stone from 98 to 131 feet thick is spread over the whole of their surface, telling of a violent cataclysm of which neither history nor tradition has preserved the memory.
The letters of Pliny the Younger39 say that the eruption of Vesuvius which caused the destruction of Portici lasted five days, and we know that the houses are covered with a uniformly distributed bed of pumice-stone some thirteen feet thick, and of cinders about three feet thick. Everything points to the conclusion that a very similar catastrophe overtook Santorin; there too whole villages were buried beneath cinders, stones, and molten lava, belched forth by a volcano in action; there too men were the witnesses and the victims of the eruption, as is proved by an accidental circumstance which took place some twenty-three years after.40
The removal of the pouzzolana, so called after the volcanic ashes of Pozzuoli in Italy for the works on the Isthmus of Suez, necessitated important excavations, and the cuttings revealed the existence of dwellings which had been bidden away from the light of day for many centuries. The masses of rubbish hiding these prehistoric ruins were some sixty-five feet high, and consisted chiefly of volcanic ashes piled up, for some accidental reason, in comparatively modern times. Beneath the pouzzolana a thin layer of humus contains fragments of pottery of Hellenic origin; which marks the close of the historic period, and covers over the mass of pumiceous tufa vomited out by the volcano. It was in this tufa, which is eight feet thick, that the first signs of buildings were discovered. Further excavation brought to light two houses with doors, windows, and bearing walls. In one of these houses there were five different rooms. Other discoveries rapidly succeeded each other, alike in the island of Therasia and at Acrotiri, the principal island, which has given its name to the group. The plan of these houses is an irregular parallelogram, the angles of which are rounded and the sides more or less curved. This arrangement differs greatly from that adopted in Greece as well as from that in use at Therasia after the time of the volcanic eruptions. The houses too are quite different in their mode of construction. The walls consist of great blocks of lava placed one above the other, without any trace of cement or of lime, and are merely kept in place by a reddish earth mixed with chopped straw or marine algae. Large branches of olive or cypress trees, still with the bark on, are imbedded in the masonry. These pieces of wood, the size of which varies considerably, were probably added to give the necessary solidity to the walls in the numerous earthquakes, the disastrous effects of which were only too well known to the ancient inhabitants of Santorin. It is curious and interesting to note the use of the same expedient among the inhabitants of the islands of the Archipelago who are still exposed to the same danger. The doors and windows are clumsily arched, and the roof seems to have been a low vault. It was made of stones and coated with clay and supported by the trunks of olive trees, the charred remains of which lay upon the floors of the crushed homes. These trunks show no sign of having been touched with metal tools; not a metal nail or clamp has been found, and we cannot but conclude that the remains belong to the age when stone alone was employed.
The inside walls were not glazed or decorated in any way, except in one instance, that of a house at Acrotiri, from which the rubbish has been cleared away, revealing on the walls a layer of lime on which was some colored ornamentation which still retained an extraordinary brilliancy when it was discovered.
In all the houses and in every room of each were found beneath the tufa burying them masses of lava and volcanic scoriae, forming a most eloquent witness of the cause of their destruction. Near one of the houses of Therasia is a little cylindrical structure, about three feet high; which cannot have been a well, as it rests directly on impermeable lava, and was certainly not a cistern, as it is too small for that. May it, as some think, have been an altar? We cannot tell, and though the religious sentiment was probably no more absent among these primitive races than it is among the barbarous peoples of our own day, it does not do to express an opinion in the absence of positive proof.
Successive excavations have yielded a number of objects which throw a new light upon the manners and customs of the inhabitants. Terra-cotta vases are more numerous than anything else (Fig. 88), and among them preponderate large yellow vessels capable of holding about one hundred quarts. Most of them have a clumsy brim, and a rough attempt has been made at ornamentation by the potter with his fingers on the damp clay. Other vases of finer clay, colored red or yellow, are covered with ornaments and graceful arabesques; the garlands of fruit and flowers are often of remarkably beautiful workmanship. Cups with well-shaped rounded handles, made of some kind of red ferruginous earth, others of gray material, were picked up in all the houses. These various vessels were used for many different purposes; some to cook food, the marks of the hearth being still on them, whilst others retained some of the chopped straw with which the domestic animals had evidently been fed. The most curious of all are those which are supposed to represent a woman; the front part projecting and surmounted by a narrow neck bent backwards, with two brown prominences supposed to stand for breasts, and dots round the upper part representing a necklace, while ear-rings are indicated by elliptical bands of different colors. We shall have to refer again to these curious vases when we speak of the discoveries made at Troy; we need only add now that the pottery found at Santorin differs completely, alike in form and ornamentation, from the Greek, Phœnician, and Etruscan specimens, of which the museums of Europe contain so many. They are evidently therefore not of foreign origin, but of native manufacture. The absence of clay in the island of Santorin has thrown some doubt on this, however, but the researches of M. Fouqué have revealed the former existence of a large valley, at the base of the principal cone, which valley ran down to the sea-shore near the island of Aspronisi; and in which probably was found the clay which the potters of the district soon learned to turn to account.