iron found in this Broch was a dirk-like blade greatly corroded. The pottery was abundant, but coarse and fragmentary, and destitute of ornamentation. ----- Footnote 84: It belongs to the class of fibulæ which are often described as bow shaped and cruciform, and is represented in _Archæologia Scotica_, vol. v. plate 16. ----- [Illustration: Fig. 194.—Ground plan of the Broch of Yarhouse, Caithness, with its Secondary Constructions, on a peninsula in the Loch of Yarhouse, cut off from the land by a ditch.] [Illustration: Fig. 195.—Circular Brooch of Brass, found with a burial in the mound covering the ruins of the Broch (2½ inches diameter).] In 1866 and 1867 I excavated the Broch of Yarhouse, situated in the south end of the loch of the same name, about six miles south of Wick, in Caithness. The ground plan of the structure is shown in Fig. 194. Its appearance before excavation was that of a conical grass-covered mound, 200 paces in circumference, and 18 to 20 feet high. It stood on a low flat triangular projection of the shore of the loch, and was cut off from the land by a ditch now silted up, and varying from 25 to 30 feet wide. In the upper part of the mound we found portions of two human skeletons, at a depth of from 2½ to 3 feet under the turf; and at different places on the sides of the mound, lower down, the remains of three other skeletons were met with. Near one of those first found was a flat circular brooch of brass (Fig. 195), of about 2½ inches diameter. It was rudely inscribed with letters which appear to be a blundering attempt at the formula ISVS NAZAR [ENVS], a common and popular talismanic inscription on the brooches of the thirteenth and fourteenth centuries. These skeletons were not enclosed in cists, but simply embedded in the earth and stones of the mound. They were not deep enough to have any determinable relation with the structure of the Broch below. They were all incomplete and the bones in disorder, though this might perhaps be accounted for by the movement of the loose material of the slope of the mound in the course of ages. The inference appeared to be that they were casual interments made in the mound long after it had become a grassy knoll. This was also the conclusion to which Dr. Joass came with respect to the burials in the mound at Kintradwell. It is easy to see how such a practice might have arisen in remoter districts, where burial-grounds connected with ecclesiastical sites were distant and roads were few. In point of fact, there is evidence which seems to connect the custom with the later Paganism of these northern parts. Mr. Petrie found a small cemetery of stone cists, containing interments after cremation, overlying the ruined Broch of Okstrow, in Orkney. In this case, the mound which covered the ruins must have been chosen as a place of heathen sepulture because it was a mound. A grave containing two oval bowl-shaped brooches, and therefore belonging to the heathen