rious pafsion glistened in his eyes, and was visible in every word. Have we not seen that he was the tutelar genius of all those with whom he lived. When he perceived himself dying, all his regret was to leave so many happy mortals behind: "Oh !" exclaimed he, “if it was but given to a mortal to make happiness a legacy." The Abbé Elanchet died at St Germains en laye, the 29th January 1784. Good men regret him, his friends bewail his lofs, and the poor bless his memory.

FARTHER REMARKS ON THE

*

CIRCULAR BUILDINGS

CALLED DHUNES IN SCOTLAND, WITH CONJECTURES

CONCERNING THE USES FOR WHICH THESE HAVE

BEEN ORIGINALLY INTENDED.

Continued from vol. vii. p. 289. BUILDINGS of this sort, have never yet been discovered in England; nor have I heard of any on the east coast of Scotland farther south than Rofshire. Many vestiges of these are to be found among the Western Isles, and along the western coasts of Scotland far south. Dun M'Swene in Kilmartine parish, and Dun Aula in Craignifh, both in Argyleshire, are the southermost I have yet heard of. Mr Pennant has described, with great accuracy, the remains of two structures of that sort in Glen-elg; and Dhuncdornadilla, in the parish of Rea in Sutherland, has been accurately described by the reverend Mr Pope. They are also frequently to be met with in the Shetland isles, as appears from the following extract of 2 letter from Arthur Nicholson, younger, of Lochend So I find this word thould be spelt.

After mentioning them in gene

esq; near Lerwick.
ral, he thus proceeds:

"I fhall give you," says he, "an account of the principal one, which I conjectured to have been the residence or strong hold of their prince for the time. It is situated on a small island, about half a mile long, lying off the S. E. end of is country. This castle stands about twenty yards from the sea, and seems to be of a different kind from any others in the country, there being no ditches round it; but I can trace a slight stone dike that has encompassed it. This castle is by far the most entire of any iu this country, it being still forty-five feet high. It is built round a circular court, twenty feet diameter. You enter through the wall, from the side next the sea, by a low door into the court; on the opposite side of which, is a door raised three feet from the ground, which leads to the stair. The stair is placed in the heart of the wall, and leads up by high narrow steps to the top of the building. The thicknefs of the wall is sixteen feet. The whole height is divided into stories about five feet high. Each of these stories or galleries go round the building, in the heart of the wall, except where the stair interrupts. In the inside of the building there are three ranges of square holes, each range going from bottom to top, which divide the whole structure into three unequal segments. These holes are separated from each other by one or two stones' thickness; and are from eight inches to a foot square. I imagine they have been intended for throwing mifsile weapons from, in case of the enemy getting pofsefsion of the court.

The galleries are divided from one another by broad thin stones, which form so many floors and roofs. The building on the outside appears to taper considerably, till within one-third of the whole. height, when it goes up perpendicularly; and rather seems to fall without the plumb *. The inside wall is plumb. What makes me imagine that this has been the seat of the prince is, that the place next adjacent. to it is called Conigsburg, which is the Norwegian term for king's seat." All these buildings, he remarks, are of dry stone.

The structure here described, resembles those of Dornadilla, and of Glen-elg, (which last I have seen since the former part of this paper was written, and examined with great care,) much more than it does those of Dunagglesag and Dunrobin, though they all agree in their leading features. The stairs, and the galleries in the heart of the wall, being the most striking peculiarity of these buildings, and what is most difficult to comprehend, I fhall take some pains to render these intelligible.

SECTION,

Representing the stairs and galleries in the heart of the wali of a Dhune laid open.

*This is a peculiarity not observable in any of those I have seen er Acard described in Scotland.

t

1

The above sketch may be supposed to represent a section of one of these circular walls, in which the inner surface is taken down, so as to lay the stairs, and galleries open to the view; you may be supposed to enter the lowermost stair at A, ascend the flight of steps, and you land on the horizontal gallery E. Before you can reach the other flight of steps you must move forward to F; and so on round the whole building, till you reach the foot of the next flight of steps at H. This you ascend till you reach the second gallery at I. Here you again proceed forward round the whole area till you arrive at the third flight of steps K; and by a similar process, repeated at every gallery, you at last gain the top of the whole. If the paper on which the engraving is made be bent in a circular form, you will thus have a most perfect idea of these stairs and galleries, which you will easily see are altogether unlike to those in any other building on the globe, and clearly indicate that they must have been appropriated to some particular purpose, very different from any of the ordinary uses of life. I fhall next endeavour to discover what these purposes

Conjectures concerning the uses to which the buildings above described have been appropriated.

WHEN the manners of a people, and the customs to which these gave rise have changed, in a country where the art of writing was unknown, it must happen, that if any works of art have been so strong as to resist the ravages of time for a long period, every memorial of the uses for which they were originally intended, may be totally lost; and it may become a

matter of very great difficulty, to form even a probable conjecture on that head. Such is actually the case at present with regard to those buildings which form the object of our present inquiry; no record, no tradition even is preserved of their origin and uses. Formerly, it was customary to ascribe every stupenduous undertaking to the ancient race of giants, with which traditionary history had peopled every country of the globe, in remote times; but we do not find that this universal bias takes place here. The only circumstance that bears the appearance of tradition in this case is the vulgar name usually appropriated to these structures, viz. the druid's house, as the druid's house of Dun-agglesag the druid's house of Glenelg, &c. But it is well known, that most things that are extraordinary in Scotland, havé been attributed blindly to the druids, so that little reliance can be had upon this circumstance.

Modern antiquarians, as might be expected, are divided in opinion concerning the uses of these structures; but these conjectures may be all reduced to the following, viz. first, that they have been intended to serve the purpose of watch towers; or second, places of defence; third, habitations for the princes or grandees of the land; fourth, places of religious worship; each of which fhall be considered in order.

1. Watch towers.

MANY persons believe that these structures were erected, like the watch towers on the southern coasts. of Spain, to serve as beacons, on which fires might be lighted, to alarm the country in case of an invasion VOL. viii.

H

#