to pass through, which was afterwards brought down to the shaft and securely fastened (Fig. 26). This variety shows no rivet-hole. Ornamentation is by no means rare upon these spear-heads; it generally takes the form of open work, such as circles and ovals perforating the blade, and of filed or cast patterns upon the sockets, some even showing traces of gold inlaying. Barbed spear-heads are extremely rare, and were probably only used in the chase. [Illustration: PLATE I* Shield of Italian Workmanship, Sixteenth Century _A. F. Calvert_] _Arrow-heads._--Arrow-heads in bronze practically do not exist in this country, although they occur on the Continent and in Egypt, where they are generally of the types shown in Figs. 27 and 28. It is highly probable that the flint arrow-head was in use through the whole, or nearly the whole, of the Bronze Age, being retained because of its efficiency and cheapness. Bronze must have been a comparatively rare and dear alloy, and the weapons exhibit as a rule the minimum of metal in their construction compatible with efficiency; arrows from their very nature are continually being lost, and this fact alone would render their use expensive. [Illustration: FIG. 27.--Bronze arrow-head.] [Illustration: FIG. 28.--Bronze arrow-head.] _Shields._--Among primitive races the shield was invariably of wicker-work or of wood, and as the examples in bronze which have been unearthed are of a high order of skill in workmanship and design we may naturally infer that they were of comparatively late introduction, and only appeared when the expert artizan of the age was capable of producing plates of considerable area and of uniform thickness. In the British Museum are several very fine examples of shields, one of which we illustrate to show the general form and shape (Fig. 29). It was dug up not far from the river Isis, in the vicinity of the Dyke Hills, near Dorchester in Oxfordshire. It is circular in form, about 13 inches in diameter, and ornamented with two concentric rings of bosses which encircle an umbo. All these bosses have been repousséd in the metal except four, which are used in two instances as rivet heads to fix the handle in position, and in two others to fasten buttons to the interior of the outer rim. It is probable that a guige was fastened to these buttons. So thin is the metal that it can hardly have served as a shield without some auxiliary strengthening, and this was conjecturally afforded by a lining of leather moulded into the depressions of the shield when wet. There is no reason for supposing that the metal now seen was the size of the original shield; in fact there is a probability that it was larger, and that the metal merely formed the centre. A bronze buckler found near Aberystwith was formerly in the Meyrick Collection and preserved at Goodrich Court, whence it was transferred to the British Museum. It is about 26 inches in diameter,