During the Great Ice Age, when the physical features of Scotland were moulded into almost their present form, the extensive plain of the River Forth was filled by a giant glacier, which swept down from the Highland hills to the lower land on the south and east, clearing the softer rocks from its path and exposing the hard basalt of igneous sheets and old volcanic necks. These great eruptive obstructions withstood the pressure of the eastward-moving mass of ice, and so prevented the ground on their lee sides from being subjected to the scouring action that hollowed out the land on the north and west and south. Numerous examples of this “crag and tail” formation are to be found in the track of the ancient glacier, but two of the rocks stand out with striking prominence; on one is built the Castle of Edinburgh, on the other that of Stirling. It is strange that of such natural strongholds early history has so little to say, for these fortresses were afterwards to have their names writ large on almost every page of Scotland’s romantic story. The third sister castle, Dumbarton, came earlier to the front. It was a stronghold of renown in the days of the Strathclyde Britons; but as time wore on its importance diminished, and the place which it had held in the principality of Strathclyde was taken by Stirling and Edinburgh in the consolidated Kingdom of Scotland. On the Gowan Hills, to the north of Stirling Castle, traces of an ancient fort show that the Britons considered it more important to defend the rising ground overlooking the River Forth than to occupy the crag, with its precipitous south-west face. When the Romans under Agricola attempted the conquest of northern Britain they constructed a chain of forts across the country, between the Firths of Forth and Clyde. The untrustworthy Boece asserts that Stirling was fortified at the time of those campaigns, but no real traces of their work have been discovered to prove that the Romans occupied the castle rock under Agricola in A.D. 81-82, or when Lollius Urbicus, Governor of Britain for Antoninus Pius, erected the wall on the line of the earlier forts. Near the Pass of Ballengeich is the so-called Roman Stone, with its indistinct, almost unintelligible letters. Antiquaries of a former day--Camden, Sibbald and Horsley--considered the inscription genuine, but recent scholars are of opinion that the letters were carved many centuries after the departure of the legions from Britain. Again, the existence of a Roman causeway has not yet been proved. The natural supposition that a military road, connecting the camp at Ardoch with the south, passed near Stirling led to the belief that the highway crossed the Forth at Kildean, or higher up at the Ferry of Drip. No vestiges of a causeway of undoubted Roman origin have, however, been discovered either at the river or on the Field of Bannockburn, through which it was thought to have passed on its way to the station of Camelon.