The length of the saloon was generally double that of its breadth, and the height was equal to the latter, when the nature of the building would permit.

On the right hand of the saloon were the other apartments of the family, either for state or retirement, for amusement or repose; and on the left the servants apartments. As to wine it was buried in large vessels, some of them containing a quantity equal to our pipe, in subterraneous apartments, or simply in the earth, excavated and rammed full again with sand. In the country their accommodations were much more numerous and extensive. They had apartments for all the varying seasons of the year, some with stoves for the winter, and others for only admitting an extraordinary degree of solar heat and light, in spring and autumn; reserving the cool subterraneous vaulted apartments, with small windows, for the heats of summer.

Thus far with respect to the houses of the ancient Romans. In my next, I fhall endeavour to trace the luxury of the moderns, and subject it to the rules and criticism of clafsic elegance; without pretending to offer any opinions of my own, or venturing to excite the vengeance of the moderns against an admirer of the Greeks. I am, Sir,

Your humble servant,

VOL. X.

M M

B. A.

DESCRIPTION OF THE FORTIFICATIONS

AT DUN-O-DEER.

Continued from p. 212.

THE entry to both these castles has been from the west, near the stone and lime tower, where the hill is of more equal ascent than at any other place, although the descent towards the east, is, near the summit, much more gentle than towards the west, which. has occasioned some additional works there, the traces of which are still sufficiently distinct, and are delineated on the plan for your inspection, (see plate: fig. 1.) where A A A is the vitrified wall surrounding the whole hill, B B is the remains of another wall that has been drawn right across the hill, at that part where the descent to the eastward begins to be perceptible.. No marks of vitrification are discoverable in this wall. EE is the remains of a ditch, with a rampart,. stretching out beyond the vitrified wall still farther to the eastward. Beyond that, and considerably down the declivity of the hill, is the remains of another ditch of circumvallation, DDDD, seen in the perspective views of the hill at E & G fig. 2. and D.D. fig. 3. Below this in some parts of the hill, there are some indistinct marks of another ditch; but this is now in a great measure obliterated..

Besides these lines of circumvallation, which have been evidently intended for defence, there are several excavations in the hill, plainly artificial, the in

GROUND PLAN OF THE HILL OF DUN-0-DEER, WITH

THE FORTIFICATIONS ON IT.

f8.3.

BB

fig. 2.

EST VIEW OF THE HILL AND FORȚIFICATIONS, OF DUN-0-DEER, ABERDEENSHIRE

NORTH VIEW OF THE HILL AND FORTIFICATIONS,

tention of which is not quite so obvious; although

I think it probable that they also were meat for defence. Two of these are found between the cir cular rampart to the east, and the vitrifed wall at EE fig. 1. These hollows may be about five or fix feet deep, with an easy and smooth descent to the bottom from all sides. The hill is, at this place, only of a gentle declivity, which would render the works more difficult to be defended than where it was more steep. This gives room to conjecture that these cavities might either be intended to screen the defendants on ordinary occasions, from the mifsile weapons of the enemy, by way of a guard-house, or to conceal a body of men by way of ambuscade. Upon examining the face of the hill, along the east side, where the ditch of circumvallation runs along a more level surface than at any other part of the hill, and where, of course, it was much more liable to be forced, four more excavations of the same kind are discovered, which are marked beyond CC; nor could I perceive any more of the same kind, on any other part of the hill, save two on the west side, marked G Gfig. 1. & 2. and EE fig. 3. I leave others to conjecture what may have been the intention of these hollows. For although I do not forget that fame has placed gold in the bowels of this mountain, which might have induced ome perso s to dig in search of these imaginary treasures, yet there seems to be little reason to suppose that these cavities have been formed by that means, as they are greatly too wide for their depth, and as the rubbish that has been taken out of them, has been carried