Overlooking the point near the castle of Crawford, where these three roads may be supposed to unite, we find another camp (No. 3), situated on a small spur of the hill, of a low elevation, and separated from the mass of the mountain by a very gentle hollow. It is only fortified by a single rampart which, on the south side looking towards the Clyde, is now very indistinct. It is an irregular oval of 349 feet by 120; and, like the last, is destitute of

water.

From the point of junction the united road bends a little to the right till it reaches the mouth of the Raggen, or Ragged Gill, a pass of considerable elevation, which cuts off the projecting portion of the hill. Traversing this, it again descends to the banks of the Clyde, near the boundary of the parishes of Crawford and Lamington. Before tracing its progress through the latter parish, I may, however, complete the survey of the former by describing two camps which are situated on the side of the hill overlooking the Clyde. The first of these (No. 4, and plate 2, fig. 3) is within sight of the one last mentioned, and occupies one of a number of small knolls at the foot of the hill. It is remarkable from the small detached fortification on the west side, and also from the fact that on the north the ground in the interior rises abruptly much higher than the rampart. It does not possess the advantage of any supply of water within its precincts.

The second camp (No. 5, and plate 2, fig. 4) lies further round the hill, and is situated in a small cleft or corry about half way up the ascent. There is a faint trace of a second rampart at the upper part and on the south side, but the whole camp has been tampered with in search of stones for a march fence. There is a small and feeble spring at the apex of the second rampart, but as a military post this is certainly the least defensible I have met with, being completely commanded within easy arrow flight by the high ground which encircles it on the south and east. During the quarrying operations referred to, a small cairn was taken down, and beneath it two rude stone cists were found. One of these contained an urn in capital preservation, about half full of ashes and calcined bones. (Plate 3, fig. 5.) It is of a yellow earthenware, very slightly if at all glazed, and measures six

inches in height, five in diameter at the top, four and six-eighths at the bulge, and three at the bottom. The same pattern as at the bulge is repeated inside the top to the depth of half an inch. In the same cist, if not within the urn, were also found two spear heads and a bracelet of bronze; the latter especially being a splendid specimen of verd antique. The exterior surface is round, the interior flattened. Its internal diameter is two and three-eighths inches; the external, two and seven-eighths. These curious reliques are preserved in the interesting and extensive collection of the eminent local antiquary, Mr. Sim, of Cultermains, who has assisted me in the kindest and most obliging manner during the progress of my investigations.

In Lamington parish, opposite the mouth of the Raggen Gill, Arbory hill rises isolated in a conical form to the. height of about 500 feet above the level of the Clyde. On its top is a circular fortification, formed by three concentric ramparts, with an interval of ten yards between them, the diameter of the interior being 264 feet. There is no well or other supply of water.

After entering the parish of Lamington, the Roman road continues to descend the right bank of the Clyde between that river and the hill. A short distance from the boundary of the parish, Forrest, in his map of Lanarkshire, places "the site of a Roman chapel." Of this saccellum I have been unable to find any trace; and although this map was published as late as 1813, I have not met with a single person who can recollect having observed anything of the kind.

Near the farmhouse of Coldchapel there are two camps, (Nos. 7 and 8) on the opposite sides of the Hawkwood Burn, from which they must have been supplied with water. The first is an irregular square, measuring diagonally 165 by 150 feet, situated on a small piece of level ground, and its rampart is in excellent preservation. The other is very irregular in form, measures 254 by 220 feet in its longest dimensions, is less distinctly marked, and lies on the first gentle rise of the hill. These camps appear to have been connected with another on the opposite bank of the Clyde, forming with it an equilateral triangle. This camp will be afterwards described. During the formation

of the Caledonian railway, a number of small urns were dug up near the farmhouse of Coldchapel. One of these (plate 3, fig. 6) is now in the possession of Mr. Sim. It is rudely formed of coarse yellow and unglazed earthenware. Its dimensions are: height, two inches; diameter at top, two and a half; at middle, two and five-eighths; and at bottom, one and a half inches. In an old map I find that this farm was sometimes called Cat chapel, which appears to denote that it had been the locality where some battle has been fought.

A little further on, the name Causeway house gives us, if we required it, another proof that we are on the line of a Roman road. In fact, this word in Scotland may be always considered as an indication of there having been an iter in the vicinity.

At Harthope hill, on the north side of Wandel Burn, Forrest lays down a camp in his map; and it is mentioned in the statistical account as "a ring of earth and stones, about twenty yards in diameter, supposed from its form to be the remains of a British camp. There is certainly an old green turf dyke at this place, without any regular shape, but from its whole appearance I am inclined to doubt its ever having been a camp at all, and should rather consider it the remains of an old sheep fold. I am also of opinion that several of the other camps mentioned in the statistical account of this parish may be considered apocryphal, ex grege those at Hartside and Braehead."

On the shoulder of Devonshaw hill, about half way from the top, there is a camp of an irregular form, measuring 330 by 251 feet (No. 9). There is no water in its immediate vicinity, but it commands a most extensive view both up and down the river.

A bronze jug or goblet (plate 3, fig. 7) was found on the farm of Loanhead in this parish, near the line of the Roman road which is here again known as Watling street, and also indicated by the use of the word Loan.

Almost at the northern extremity of the parish the Whitehill rises out of the alluvial land on the side of the river, and commands an extensive view. The remains of the camp (No. 10, and plate 3, fig. 8) are situated on the southern face of this hill. There is an appearance of an old well having been sunk near the north-east angle of

VOL. X.

2

this camp. The hill continues to rise at this side above the camp, and there are some traces of either one or two more fortifications, but not so distinctly marked as to enable me to insert them in the plan.

From this camp the Roman road appears to have left the bank of the Clyde, and to have continued through the parish of Culter in a valley which is separated from that river by a range of hills of no great elevation. Shortly after its entry into this parish, it passes a place on the farm of Low Hangingshaw, called the Cat Craig or the Rock of the Battle. Several curious reliques have been found at this place, and now form a part of Mr. Sim's valuable collection. 1. A bronze dagger, similar to that engraved in Wilson's Archeology of Scotland, p. 264, but of larger dimensions, being eight inches long by two and a quarter broad. 2. The fragments of a number of rude urns. 3. Several granite querns or hand mills. These most probably were brought from a distance, granite not being one of the rocks of the district, which consist of greywache, green stone, amygdaloid, and porphyry. At Culter Park, a short distance further on, a bronze celt, similar to that in Wilson, p. 257, class VII, and also to that, fig. 1, plate 37, in the Journal, vol. viii, was found, and is now in Mr. Sim's possession.

At this point it is necessary to diverge from the line of the Roman road in order to embrace a series of camps in the valley of the Culter Water. The first of these (No. 11) is situated on the top of the hill which overlooks Culter Park on the south, and commands an extensive range of country. It has been but slightly fortified with a single rampart, which is now very much destroyed in places, and generally very faintly marked. It is an oval of 248 by 102 feet. A road communicating with this camp is quite perceptible on the south side of the hill leading in the direction of Culter Water.

The second (No. 12, and plate 2, fig. 9), occupies the top of the adjoining hill, on the opposite side of a small valley, but of greater elevation. It has been more strongly fortified, with two, and in some parts three, ramparts. Both these camps must have drawn their supply of water from a small stream in the valley between them. Descending from this camp to Culter Water, we find a third camp, (No. 13), on the slope of the hill, defended by a single en