^{page 270 note *} The people engaged in this work have their houses in the moss, These are at first sometimes built of sod, supported by a frame of wood, laid on the surface of the moss; but as soon as any progress is made in clearing the ground, they are cut out of the moss itself. For that purpose, a drain is cut through the moss, and at least a foot deep into the clay, as far as the intended house is to reach ; a space from two to six yards wide is then cleared all round it ; and lastly the area of the house is also cleared, leaving a wall of moss on every side, about four feet and a half thick, at bottom, and three feet thick at top. The feet of the cupples which are to support the roof are inserted into this wall, but do not rest upon it, as they reach as low as the clay, from which they rise up, nearly perpendicularly, as far as the top of the wall. The gables are completed with sod or mud. As the moss-walls dry, and are consolidated, what was originally ten or twelve feet high, sinks down to the height of five or six feet.

^{page 275 note *} Tacitus in Vit. Agric. cap. 31.

^{page 275 note †} Σεβῆξος δϊων την Καληδονίαν, ἀμύθητα πξὰγματα εχε, τας τε ὕλας τεμνων, τα τε ἕλη χωννυων, χαι τους ποταμους ζ υγνὑων. Dio. Cass. Lib. lxxvi. cap. 13. The works he enumerated were attended with such difficulty, that though, according to the same historian, Severus was never met by the British army in the field, he lost fifty thousand men in the course of this expedition. Ibid.

^{page 276 note *} That the river Carron was the boundary of the Roman province is rendered probable by the situation of Arthur's Oven, as it was called, which is supposed to have been a temple dedicated to Terminus, and erected near the Roman frontier. It stood on the west side of the river Carron, or between that river and Kinnaird, There is also a passage in Herodian that favours the same opinion. That historian mentions the army of Severus passing τα πξοβεβλημενα ξἑύματιἰ τε χαι χώματα της Ρωμαιων αξχης. He adds, that on this frontier the Barbarians easily made their escape, and concealed themselves in the thickets and marshes. Herod. Lib. iii. cap. 48.

^{page 277 note *} The chain of posts between the Forth and Clyde is mentioned by Tacitus, Vit. Agric. cap. 23. as the work of Agricola's fourth campaign, which coincides with the year 81 of our æra. See Horseley's Britan. Book i. chap. 3. It was about fifty years afterwards that the wall of Antoninus was built, nearly in the same line. The age of the moss cannot therefore be estimated at much less than 1700 years.

^{page 278 note *} See Lord Cromarty's paper on Peat-moss, Phil. Trans. vol. xxvii. p. 296.

^{page 278 note †} See an Account of Hatfield Chace near Doncaster, Phil. Trans. vol. xxii. p. 980. It may be proper to observe, that the mosses of Kincardine, &c. being placed above the level of the adjacent plain, are of the kind that might be expected to break out and overspread the lower grounds, which however they are not known to have done, while they remained in their natural state. They do not indeed abound very much in water, insomuch that the floating off of the peat, when it is carried to such an extent as it is now, requires an artificial supply of water. This supply is accordingly procured at present by an engine which Mr Drummond has caused to be erected for raising water from the Teith, and which is one of the most material improvements that has been made in the husbandry of the moss.

But though there is no memory of the moss having flowed while it remained in its natural state, on the 21st March 1792, it burst out on the west side, near the southermost cottage, to the height of its side-wall, covering fifty-six yards in breadth, and about the extent of an acre of ground that had been cleared, and early in the morning of the same day of 1793, (since the first communication of this paper), it was discovered to have flowed again, and to have reached the northermost cottage of the same line of houses. The inhabitants escaped by a window on the opposite side of the house. The moss afterwards bore down the side walls of the house that were built of stone, and continued to flow slowly forward eight feet in depth at the middle, and 1200 feet in breadth, until nine o'clock in the morning of the 23d, when it had advanced 600 feet, and covered twelve acres of ground that had been cleared. It would undoubtedly have flowed much farther, had not a great number of men been employed night and day, in giving vent to the water mixed with the moss that had flowed, and in intercepting that which continued to discharge itself from the main body of moss.