which, from its position on the top of a hill of considerable elevation, is evidence of no small engineering skill. There are also two mounds, which will be found in the plan, but I am not prepared to say that they are artificial. As a military post, this is one of great strength and importance. It commands a most extensive view, not only of the course of the Clyde, but of the road into Annandale, and from the nature of the ground must have been almost impregnable before the invention of gunpowder. The camp is certainly not a rectangle, but of an irregular form, and from this it may be urged that it is not a Roman fortification. But I think this is clearly rebutted by the fact that an undoubted Roman road leads directly into it, and we must not forget that it occupies the whole of the plateau, and that the attempt to inscribe a rectangle within the latter would have destroyed the security of the camp; because had this been done, the extreme suddenness and steepness of the descent would have enabled a light-armed enemy to have established himself in a perfect and secure cover within a few feet of the base of the rampart. I have been told by one of my servants that, about twenty years ago, some stone dykers, employed by my father, found either on this hill or in its vicinity, a grave in which, to use the words of my informant, were contained "an auld swurd and the banes of a Christian". If so, they kept their discovery concealed; and there is now no possibility of recovering any of the arms or antiquities they may have found.

In connection with this camp, I may mention a most amusing instance of stupidity and ignorance, by which this fortification has been metamorphosed into a worthy and respected country gentleman. In Paterson's Roads, London, 1826, page 230, I find in the list of seats on the road from Carlisle to Glasgow the following entry: "Elvanfoot Inn before-Bodsberry, - Camp, Esquire".

Descending from this hill the Roman road, clearly and distinctly marked, continues in a north-west direction along the right bank of the Clyde, at a little distance from that river. In the bed of the Shilling Cleuch Burn, one of the small tributaries which it crosses, a Roman camp kettle was found about forty years ago, but having been appropriated by the finder to domestic purposes, it has

long ago been lost sight of. The road continues in the same direction for about three miles, till after crossing the Midlock and Camp waters, it reaches the base of a hill which projects from the main range, and causes a considerable deflection in the course of the river. At this point, which is in the immediate vicinity of the castle of Crawford, it is rejoined by the branch which separated from it at Dryffe church in Dumfrieshire.

This branch, after crossing the country from the valley of the Annan to that of the Nith, ascends the latter river and its tributary the Carron, with its feeder the Durisdeer Burn, and enters Lanarkshire by the Wellpath. On the Dumfrieshire side of the pass, and not far from the top, there is a camp, the situation of which is marked on Plate 1. On entering the upper ward, it descends the left bank of the Potrail, the Daer, and the Clyde, which, however, may be considered as a continuation of the same river, till it arrives at the village of Crawford, where a portion of it is still known as Watling street. Here it crosses the river by the castle ford, and joins the other branch, as mentioned above.

It is highly probable that the road was here also joined by another branch from the upper part of Peebleshire. Although traces of this are found in several places, I have not been able accurately to trace its course; its general direction is, however, indicated by a double-dotted line. It follows the valley of the Camp till that stream separates into two branches, when it ascends to the high ground between them, and inclines to the left along the heights above that branch which is known as the Grains Burn. On the side of the hill immediately above the junction of the two streams, and on a sort of knob or prominence, a very strong and elaborately fortified camp is met with. (No. 2, and plate 2, fig. 2.) There is no water within the precincts of it, and the road passes to its right, but bends, as above stated, after passing it. About two miles further up the Grains Burn, and at the bottom of the glen which passes the shepherd's house on the farm of Crimp Cramp, a coin of Vespasian was found some years ago. This place is certainly not on the line of road, but it is by no means improbable that it may have been washed down from the vicinity of it.