we are, happily, not left to conjecture, for many moulds, principally of stone, in which they were cast have from time to time been found.

Other implements of bronze consist of a socket, or rather ferrule, the upper part of which is armed with a number of deadly spikes (Fig. 99), usually in three rows, alternately. This, firmly fixed at the head of a wooden shaft, would, when wielded by a powerful arm, be one of the most terrible of weapons. An example in my own collection has been attached to the shaft by a rivet, and three of its spikes (of which there had originally been twelve) have, at the time when it was in use, been broken off on one side, with great force.

Bronze daggers vary in length from two and a half or three, to five and a half or six, inches on the average; the larger ones being an inch and a half to three inches in breadth at their broadest part,

Fig. 100.

where the handle has been attached, from whence they taper gradually down to the point. They are sometimes ribbed or fluted. In most instances the handle has been attached by three rivets; in some cases, however, as in Fig. 100, only two have been used, and occasionally there is evidence of the

attachment being effected by thong or other ligature. The handles were of horn or wood, and were usually semi-lunar where attached to the blade; in one instance, however, the blade has a "tang," or "shank," which has fitted into the square-ended handle, to which it has been fastened by a single peg. The blades occasionally present incontestable evidence of long use, having been worn down by repeated sharpenings. In the instance of the dagger found at Stanshope, which had been fastened to the handle by a couple of rivets as well as by ligatures, evidence existed of its having been enclosed in a sheath of leather; and this example also presented the somewhat curious feature of impressions of maggots, which had probably made their way from the decaying body into the inside of the sheath, between it and the blade, and had there remained, and thus gradually become marked upon the corrugated surface of the bronze.

Bronze swords, which have frequently been ascribed to the Ancient British period, are now generally admitted to belong to Roman times, as are also spear and lance heads, both socketed and otherFig. 101. wise.

Examples of bronze spear and arrow heads are shown in the group, Figs. 82 to 94, and on Figs. 102 and 103; a sword on Fig. 101: these will be sufficient to indicate their characteristics.

CHAPTER V.

AMONG ROMAN ROADS, TOWNS, AND VILLAS; TESSELLATED PAVEMENTS, TEMPLES, ALTARS, SEPULCHRAL INSCRIPTIONS, MINING OPERATIONS, ETC.

TRACES of Roman roads, and of British roads adopted by the Romans, remain to the present day in many parts of the kingdom, and are still in several instances used as the high-roads on which we yet travel. The roads were usually made in a straight line, seldom turning out of the way even for a high hill, and were sometimes raised considerably above the common level of the ground. They are classed as of five kinds :-viæ militares, or military roads; viæ vicinales, branch roads; viæ privatæ, private roads; via agrariæ, country roads; and via deviæ, by-roads. According to Vitruvius the forming of a road was conducted as follows:First, two parallel furrows were cut so as to mark out the intended width; the surface-earth and loose stones between these were then removed until a solid foundation was reached, and this being levelled was

covered with earth and beaten down quite hard; this was called the pavimentum. On this was laid a stratum of small stones, generally squared, over which was usually poured a quantity of mortar or concrete; this was called the statumen. On this a layer of small broken stones, mixed with lime, in the proportion of one part of stones to two of lime, and forming a stony concrete, was laid; this was called the rudus, or ruderatio. The next layer consisted of chalk, lime, and broken tiles, or earth, mixed together, or of gravel, or sand, and lime, mixed with clay; this was called the nucleus. Upon this was laid the surface layer, the finishing stroke of the work, which consisted sometimes of a regular pavement of squared flagstones, and in others of a firm bed of gravel and lime; this was termed the summum dorsum, or summa crusta-the whole structure, thus formed, being called agger. Of course it must not be understood that all Roman roads were formed in this manner, but still in most instances that have been examined there is an approximate adherence to these general rules. Such roads as those described were the grand military roads of the kingdom, along which the lines of the Itineraries were traced. At the end of each Roman mile (1000 paces, or 1611 yards, being 149 yards less than our English mile) of these main roads a milestone (milliarium) was placed. These were usually plain cylinders or short pillars of stone, bearing an inscription denoting the distance