[63]. Archæological Journal, vol. xxx. p. 268.

[64]. Proceedings of the Bristol and Gloucester Archæological Society, vol. v. p. 137, and Plate XIV., from which the figure here given is copied by permission.

[65]. These have been conjectured to be of Christian time, and to have been used in connection with the celebration of the Eucharist, but the evidence is insufficient to carry this conclusion. See the papers by Albert Way in Archæological Journal, vol. xxvi. p. 52; and by Rev. E. L. Barnwell in Archæologia Cambrensis, vol. viii. (Third Series) p. 208, and vol. x. p. 57.

[66]. A denarius of the Emperor Nerva was subsequently found close by the place where the armlets were discovered. The underground structure appears, like many of its class, to have been associated with an overground habitation, the site of which was marked by fire-burnt pavement, remains of querns, beads, etc., found near the present surface.

[67]. These armlets were analysed by Professor Church, and the composition of the metal determined as follows:—

Armlet No. 1.Armlet No. 2.
Copper 86·49 88·19
Tin 6·76 3·64
Zinc 1·44 9·13
Lead 4·41 —
Loss ·90 —
100·00 100·96

[68]. Another saucepan of this form found in the Loch of Dowalton, and bearing the maker’s name stamped on the handle, is described in connection with the relics from Crannogs in Lecture VI.

[69]. This arrangement of triple dots is a very characteristic feature of the illuminated Celtic manuscripts. It appears also on the monuments and metal work of the Christian time. This is the only instance of its occurrence on these balls, and though it may be held to suggest a possible connection, the suggestion is too feeble to imply distinct relationship.

[70]. This structure, which was explored by Mr. William Watt, consisted of several sub-rectangular chambers with rounded corners, having small cell-like constructions opening off them. The chambers were arranged on both sides of a long winding passage. Their door-ways had checks for the doors, and bar-holes behind them. The largest chamber was about 20 feet square. From 6 to 8 feet of the height of the walls remained. They were dry-built, and converged towards the upper part as if to form beehive roofs. Hearths of square form, surrounded by flagstones on edge, were found in the floors. Many implements of stone and bone were found in the chambers, and a large accumulation of bones and horns of animals, among which those of the red-deer and the Bos primigenius were abundant. Among the stone implements were several polished celts. The collection is preserved at Skaill House.

[71]. Dr. John Alexander Smith has discussed this point fully in his exhaustive notice of these Stone Balls in Proc. Soc. Antiq. Scot., pp. 56-62. Dr. John Evans remarks that “it seems probable that they were intended for use in the chase or in war when attached to a thong which the recesses between the projecting discs seem well adapted to receive.” He also states that "these Scottish Stone Balls seem to belong to a recent period, as compared with that to which many other stone antiquities may be assigned."—Ancient Stone Implements, etc., of Great Britain, pp. 377-379.