VITRIFIED FORTS solidated by the fusion of the rocks of which they are built. This fusion, which has been caused by the application of intense heat, is not equally complete and regular in the various forts, or even in the walls of the same fort. In some cases the stones are only partially melted and calcined ; in others their adjoining edges are fused so that they are firmly cemented together ; in many instances pieces of rock are enveloped in a glassy enamel-like coating which binds them into a uniform whole ; and at times, though rarely, the entire length of the wall presents one solid mass of vitreous substance. Since John Williams one of the earliest of British geologists, and author of The Mineral Kingdom first de scribed, in 1777, these singular ruins, about fifty examples have been discovered in different parts of Scotland. The most remarkable are Dun Mac Uisneachain, the ancient Beregonium, north of Oban ; Tap o I^oth, in Aberdeen- shire ; Craig Phadraic and Dun Dhardhail, in Inver ness ; Knockfarrail, near Strathpeffer ; Dun Creich, in Sutherland ; Findhaven, near Aberlemno ; Barryhill, in Perthshire ; Laws, near Dundee ; Dun Gall and Burnt Island, in Buteshire; Anwoth, in Kirkcudbright; and Cow- denknowes, in Berwickshire. Dun Mac Uisneachain is the largest in area, being 250 yards long by 50 yards broad. The strongest and most cyclopean is the Tap o Noth : here the walls are about 8 feet high, and between 20 and 30 feet thick. In Dun Mac Uisneachain, Barryhill, and Laws the remains of small rectangular huts or dwellings have been found. For a long time it was supposed that these forts were limited in their range to Scotland ; but they are now known to exist in Londonderry and Cavan, in Ireland ; in Upper Lusatia, Bohemia, Silesia, Saxony, and Thuringia ; in the provinces on the upper banks of the Rhine, especi ally in the neighbourhood of the Nahe ; in the Ucker Lake, in Brandenburg, where the walls are formed of burnt and smelted bricks ; and in several places in France, such as Chateauvieux, Peran, La Courbe, Saint Suzanne, Puy de Gaudy, and Thauron. They have not been found in England or Wales ; and Worsaae, Herbst, Rygh, Hilde- brand, and Stephens assert that they do not exist in Denmark or in Scandinavia. All the examples yet described present a general simi larity in form and structure. In some of the Continental forts the vitrified walls are supported by masses of unverified stone built up on each side. This, in all probability, constituted an essential feature in the Scottish forts. Except on the hypothesis of buttresses of a similar kind, it is impossible to explain the vast quantities of loose stones which are found both inside and outside many of the vitrified walls. The method by which the fusion of such extensive fortifications was produced has always excited much interest and conjecture. Williams, when he first directed attention to the subject, maintained that the builders of the forts, whoever they were, found out, either during the process of smelting bog-ore, or whilst offering sacrifices, the power of fire in vitrifying stone, and that they im proved upon this discovery by using it for the purpose of cementing and strengthening their strongholds. This view has been keenly controverted, and other theories have been suggested. It has been held that the vitrified sum mits were not forts at all, but the craters of extinct vol canoes (West, Pennant, and Cordiner), an hypothesis long since abandoned as unscientific ; that the vitrified summits are not so much vitrified forts as vitrified sites, and that the vitrescence was produced by beacon fires lighted during times of invasion, or by bonfires kindled on hill tops in religious celebrations (Sir George Mackenzie, Dr S. Hibbert, and Principal Daniel Wilson); and, lastly, that if they were forts they must have originally been built of wood and stone, and that their present vitrified appearance is not due to design, but to their being set on fire by a besieging enemy (A. Fraser Tytler, Forbes Leslie, Von Cohausen, and Dr Joseph Anderson). The theory of Williams which has, with modifications, been accepted by all the principal British and Continental authorities, such as MacCulloch, Hugh Miller, Virchow, Schaafihausen, Thuot, and Montaiglon is likely to hold the field. It is supported by the following facts : (1) The idea of strengthening walls by means of fire is not singular, or confined to a distinct race or area, as is proved by the burnt-earth enclosure of Aztalan, in Wisconsin, and the vitrified stone monuments of the Mississippi valley. (2) Many of the Primary rocks, particularly the schists, gneisses, and traps, which contain large quantities of potash and soda, can be readily fused in the open air by means of wood fires, the alkali of the wood serving in some measure as a flux. (3) The walls are chiefly vitrified at the weakest points, the naturally inaccessible parts being un- vitrified. (4) When the forts have been placed on materials prac tically infusible, as on the quartzose conglomerates of the Old Red Sandstone, as at Craig Phadraic, and on the limestones of Dun Mac Uisneachain, pieces of fusible rocks have been selected and carried from a considerable distance to the top. (5) The vitrified walls of the Scottish forts are invariably formed of small stones which could be easily acted upon by fire, whereas the outer ram parts, which are not vitrified, are built of large blocks. (6) Many of the Continental forts are so constructed that the fire must have been applied internally, and at the time when the structure was being erected. (7) Daubree, in an analysis which he made on vitrified materials taken from four French forts, and which he sub mitted to the Academy of Paris in February 1881, found the presence of natron in such great abundance that lie inferred that sea-salt was used to facilitate the original fusion. (8) In Scandi navia, where there are hundreds of ordinary forts, and where for centuries a system of signal fires was enforced by law, no trace of vitrifaction has yet been detected. A great antiquity has been assigned to vitrified forts, but without sufficient proof. Articles of bronze and iron have been found in the Scottish forts, while in Puy de Gaudy a Roman tile has been discovered soldered to a piece of vitrified rock. In a few of the German forts Prof. Virchow found some of the short logs of oak used as fuel in vitrifying the walls, and he concluded from the evenness of their cut surfaces that iron and not stone implements must have been used. These results indicate that these structures are the products of a high civilization, and were possibly in use as late as the early centuries of our era. It has been suggested that they were built as temporary refuges against the invasions of the Norsemen in Europe. There is much in the situation and character of the forts which favours this supposition. This is especially the case with reference to the Scottish forts. Here the vitrified summits are invariably so selected that they not only command what were, as we learn from the sagas, the favourite landing places of the vikings, but are the best natural defences against attacks made from the direction of the sea-coast. In Saxony and Lusatia the forts are known as Schwcdenburgen, and in the Highlands of Scotland as the fortresses of the Feinne designations which also seem to point to an origin dating back to the times of the. vikings. It may be interesting, as throwing light on the influence of high temperature on rocks, to notice some of the forms exhibited by the fused substances. Schistose gneiss is resolved into a kind of green glass ; mica is converted into a grey glassy stone of great hardness ; quartz is reduced to the condition of pumice; trap rock becomes glazed like bottle glass ; sandstone is transformed into a compact quartzite ; and the felspar of granite is changed into coarse porcelain. Some of the materials have assumed, probably during the process of cooling, the prismatic structure, with four or six sides ; while others, owing to the iron peroxide being reduced, have become distinctly
magnetic.