of view, that not one of them was ever taken or surrendered. They were placed so as to command the principal passes into the Highlands. They form a ring-fence round the territory hastily overrun by Agricola in this third campaign. Beginning in the west with Bochastle, at the Pass of Leny, near Callander, we come successively to Dalginross, at Comrie; Fendoch, at the mouth of the Sma' Glen; the camp at the junction of the Almond and the Tay; and, Ardargie, in the parish of Forgandenny, on the River May, commanding an extensive prospect of the Ochils, and along the course of the road from the Tay to the great camp at Ardoch. Here was evidently the base of operations, with accommodation, if need arose, for the entire Roman army in Scotland.
Having thus viewed the land and pegged out his claim by means of forts, Agricola returned to winter quarters. In the following summer—the summer of 81 A.D., he made no forward movement. But he was meditating a great enterprise—no less an enterprise than to penetrate beyond the Tay and break the power of the Caledonians in their remote fastnesses. It behoved him to be cautious, so he constructed the chain of forts which afterwards became the Wall of Antonine—from Borrowstonness, on the Forth, to Old Kilpatrick, on the Clyde. Meantime he was laying his plans with admirable foresight. He entrusted the forts in Strathearn to the courage of their slender garrisons, and the issue proved that he could do so safely. But there was unexplored territory westward and eastward. Nobody knew what dangers might be lurking there, ready to assail him in rear the moment he left the security of his fortified place. So we find him in the summer of 82 A.D., in Argyll and Kintyre, with a small force, not fighting so much, as simply exploring, at one point feasting his Roman eyes, greedy for conquest, upon the coast of Ireland, seen dimly in the distance, and perhaps scheming in his heart and head to add it also at a fitting time to the Roman domains.
Returning from the west country, Agricola entered upon the campaign of three years' duration, which issued in the Battle of Mons Grampus, the crowning glory of his arms in Scotland, and the immediate occasion of his recall. He chose the alternative route this time. From his chain of forts he could see the broad expanse of the Firth of Forth, the coast of Fife, the central Lomonds, and the distant hill-country, whose acquaintance he had already made. That distant point was the goal of his endeavour, and the shortest way to it was across the Firth of Forth, through the western division of Fife, and on by one of the Ochil passes—Glendevon, or some other further east into Strathearn. There was no time to lose. For, while he was engaged in the subjugation of Fife, the fleet, after exploring the harbours, had doubled the East Neuk, passed safely through St. Andrews Bay, and entered the Firth of Tay. Its unexpected appearance caused the greatest consternation among the Caledonians. The immediate result was to greatly increase the peril in which the devoted garrisons in Strathearn stood. So great was their danger, and so well was it known, that there were those with Agricola who advised a retreat to the chain of forts between the Firths. But Agricola was not to be shaken in his resolve, which was to finally break the power of the tribes who dwelt to the north and east of the Grampians, and who, so long as they remained free and unchastised, were a standing menace to the Strath of the Earn and to the garrisons who held it at the hazard of their lives. He formed a camp to winter in at a place called Grassy Walls, on the east side of the Tay, near to Perth. But there was still time, before the winter set in, for a little exploration and a brush with the enemy to revive the courage of his soldiers, which had begun to droop a little. Advancing northward on the left bank of the river, Agricola reached the Isla, and not caring to cross it so late in the year, in the face of the enemy who were massed upon the Hill and Muir of Blair beyond, he diverged to the right, following the course of the Isla until he came to the place where Coupar Angus stands now. Here he paused. He had marched from Perth in three divisions to prevent surprise, and in this neighbourhood there are three positions marked by Roman remains that correspond with these divisions. The main force was stationed at Coupar Angus; the Ninth Legion at Lintrose, two miles south-east; a third small body at a place two miles south-west, overlooking the Tay, and guarding the passage. These details are important, as helping to determine the true site of the Battle of Mons Grampus. It may be taken for granted that the Roman General made good use of his opportunity to survey the ground upon which the decisive battle was fought. Before retiring to winter quarters at Grassy Walls, the Roman soldiers had a chance given them of testing the strength and valour of the Caledonians. The Ninth Legion was stationed at Lintrose, and here the enemy delivered their attack under cover of night. They had penetrated into the camp ere they were discovered, and it might have gone hard with the Legion if help had not been at hand. But the alarm quickly spread to where Agricola was stationed with the main body. On his arrival the Caledonians took to flight. With the first touch of winter the march southward was begun, and when the summer came the legionaries and the auxiliaries clamoured impatiently to be led northward to the final encounter.
The theory maintained in this paper regarding the last campaign of Agricola, and the site of the Battle of Mons Grampus was first broached in the Statistical Account of the parish of Bendochy, published in 1797. It has been since adopted by Skene in his classical work on Celtic Scotland, to which I desire in this place to acknowledge my great indebtedness. Other sites have been fixed upon, but there are none that can fairly be put in comparison with the neck of land at the junction of the Isla with the Tay. What may be called the traditional view of the site of the battle locates it at Ardoch, in the vicinity of the great camp. No doubt, good authorities can be quoted in favour of the correctness of the traditional view. But there are several reasons which render it highly improbable that the great battle was fought at Ardoch. The very name Mons Grampus implies the existence of some conspicuous eminence in the near neighbourhood. There is no such eminence near Ardoch. Further, we know that Agricola's scheme of operation embraced joint action on the part of the fleet and land forces. There could have been no such co-operation if the movement of the legions had been west of the Tay. And it is a fatal objection to the Ardoch site that there are not three stations corresponding to those which we have seen the three divisions of Agricola's force occupied on the night of the surprise. General Roy, indeed, has tried to turn the edge of this objection by placing the Ninth Legion at Dalginross, the main body under Agricola at Ardoch, and the other division at Strageath, overlooking and guarding the Earn. But it has been retorted upon him that Agricola could have made no worse disposition of his forces, from a strategic point of view, than to have stationed his weakest division at Comrie, nine miles distant from the main body, in the very heart of the enemy's country, close to the hills, from which they could rush down upon any favourable opportunity, and to which they could retreat in the event of a repulse.
Besides, help came from the main body in the course of a few hours—between night and morning. It would be a difficult task even now for a body of men to cover the ground between Ardoch and Comrie in the dead of night; and we must remember that in the time of Agricola the country was a pathless wild, rough with woods in the higher parts, and covered with treacherous morasses in the valleys. The Damnonii—within whose territories Ardoch, Comrie, and Strageath lay—were more highly civilised than the Caledonians beyond the Tay and the Grampian range. They had towns, as we have seen; they probably engaged to some extent in agriculture; their food did not altogether consist of fish, milk, and the produce of the chase. But their towns were few and far between, and the means of communication very imperfect. The native tribes were not road-makers, and the Romans had not been long enough in possession, nor had leisure been granted them to form the solid and straight lines of communication upon which, everywhere, their power was based. We have Roman roads in Strathearn, and I daresay a careful student of the district could walk every foot of the way from Ardoch to Perth along these roads—street-roads, they are called locally, as I discovered one night to my surprise on making inquiries as to the shortest route between the manses of Gask and Trinity-Gask. But these roads were not in existence when the Battle of Mons Grampus was fought. Much rough pioneering work had to be gone through ere it was possible to lay them down. Meantime, the respective positions of the Romans and Caledonians had changed somewhat. The tide of conquest did not remain at the high-water mark of Agricola's advance. The Roman garrisons were withdrawn from Strathearn and from Ardoch. The Wall of Hadrian, between the Solway and the Tyne, was substituted for the frontier chain of forts between the Forth and Clyde, and, in consequence, the native tribes kept pressing ever southward. Hadrian's Wall checked their progress, but their presence in ever-increasing numbers was a danger to the province. They had now come to be known
