occupied these western stations ever heard the Caledonian harp, or listened to a Celtic bard, or received an embassy, as we are expressly told they did, from men like Ossian as ambassadors—the difficulty requires no farther explanation. The Romans were neither blind nor senseless, and knew well enough how to represent the poetical genius of the country which they were attempting in vain to conquer, as well as the wild boars of its woods, and the sea-dogs in its estuaries; and have thus left behind them, in rude but significant sculpture, as true a picture as could be imagined of the men on the soil, and the beasts in the field, and the fish so-called in the sea, and the bird in the air—between Simmerton and Duntocher, in absolute conformity with the text of Ossian. Nor is there any possible reply to this by our antiquarian friends. The Roman Wall itself, to which they constantly appeal, supplies the evidence, and they are bound, without a murmur, to accept it.

III. But the levels of the Wall, it may be said, as now ascertainable by actual survey—what other sort of evidence do they afford? This question implies—(1) A range of observation from the Kelvin at Simmerton westward to Duntocher in the first place, and then to Chapel Hill between Old Kilpatrick and Dunglass. The intermediate forts on that line are separated by equal distances, nearly as follows:—From Simmerton to New Kilpatrick, 1¾ miles; from New Kilpatrick to Castlehill, 1¾ miles; from Castlehill to Duntocher, 1¾ miles; the lowest point in which range at Duntocher is from 155 to 200 feet above the level of the Clyde, leaving sufficient room, therefore, for the Wall above the highest level assumed in the text of Ossian. From Duntocher to Chapel Hill there is a distance of 2½ miles, with no trace whatever of the Wall between. Chapel Hill is considerably lower than Duntocher, undoubtedly; but why is there so great a gap there, and no trace of a wall in the interval? Either, because there never was a wall so close to the tide; or because the tide itself washed the wall away, having been built too close to its confines; or for some other more probable reason yet to be assigned. The fort at Chapel Hill itself, indeed, is the most indistinct of them all; and if a regular fort of any importance ever existed there, it must have suffered either partial inundation, or some other serious shock, unquestionably.

(2) It implies also a corresponding survey of the ground intermediate between the Wall and the river. Now the intervening ground along the banks of the Clyde, from Chapel Hill to the Pointhouse at Glasgow, is a low-lying flat with a gradual rise inland, at the present moment, of not more than 25 or 30 feet. But according to Professor Geikie's latest survey, the Clyde must have been about 25 feet higher in the time of the Romans than it now is—and Professor Geikie, we presume, is an authority on such subjects, who may be quoted along with Hugh Miller and Smith of Jordanhill:—therefore the whole of that strath, and the strath on the opposite side, from Renfrew to Paisley, on this assumption, must have been submerged at the same time; and there could be no dwelling-place for human beings—neither local habitation nor a name—within the entire compass of that now fertile and populous region. But two or three Gaelic names survive on the northern verge of it, which not only indicate the presence of the sea there, but fix the very limits of its tide. Dalmuir, for example, which means the Valley of the Sea; and Garscadden, which means the Bay of Pilchards or of foul herring, must, in fact, have carried the waters up their respective streams to within less than a mile of the Roman Wall at Duntocher and Castlehill. It was in such retreats, then, that both salmon and herring (as the name of one of them imports) would take refuge in the spawning season; it was into such retreats also, they would be pursued by the seals; it was on the shore of such inlets the seals themselves would bask, when the Romans saw them; and it is at the two forts respectively at the head of these inlets—Duntocher and Castlehill—that they have been actually represented in Sculpture. Could anything be more conclusive as to the proximity of the tide, and very character of the shore, within a bowshot or two of the Wall in that neighbourhood, where there is now a distance of more than two miles between it and the river? and yet even more conclusive, in connection with this, is the fact that on the southern verge of the strath, right opposite to these, are other Gaelic names equally significant—such as Kennis, the Head of the island; Ferinis, the Hero's island; and Fingal-ton, which speaks for itself—at the same or a similar level with Dalmuir and Garscadden, that is from 100 to 200 feet above the present level of the Clyde, which seems to demonstrate beyond doubt that the whole intervening space of seven miles in breadth, with the exception of such small islands as those named above, was then an arm of the sea to the depth of 50 feet at least, if not more.

(3) Our survey is thus narrowed to a single point—the existence and alleged position of the fort at Chapel Hill, between Old Kilpatrick and Dunglass, on the banks of the river; and here it should be observed as between the two extremities of the Wall, east and west, that where it touches the Frith of Forth at Carriden the height of its foundation ranges from about 150 to 200 feet above the level of the sea, and where it approaches the Clyde at Duntocher it is nearly the same—which was probably its terminus. There is scarcely a vestige of it now traceable beyond that, and that it was ever carried farther in reality is a matter of acknowledged uncertainty. But scattered fragments of masonry, as we have seen, and the dimmest indications of a fort deep down in the earth have been discovered or imagined at Chapel Hill to the westward, which seems to be about 50 feet above the level of the Clyde—leaving still a very large margin beyond Professor Geikie's estimate; and a great deal of conjecture about what might, or might not have been there, has been indulged in by antiquarians. For the present, however, until proof to the contrary has been shown, let us accept as a fact that some military station had really been established there in connection with the Wall—then, how have its fragments been so widely scattered? how has it been so completely entombed that it can only be guessed at under the soil? and how has the connection between it and the Wall, more than two miles distant, been obliterated? No other fort on the line, that we know of, is now in the same condition; and therefore, we repeat, either the Romans were foolishly contending with the tide, by building too close to its confines, and the tide drove them back and overthrew their works; or the fort itself was originally on a higher level, and the shock of an earthquake, or a landslip from the mountains, or both together, carried the whole mass of masonry and earthwork at this particular point down to their present level, where they would be washed by the tide and silted up in their own ruins. This is a view of the matter, indeed, which no antiquarian, so far as we are aware, has hitherto adopted; but any one who chooses to look with an unprejudiced eye, for a moment, at the enormous gap in the hills immediately behind, reaching down to the shore and including this very region, must be satisfied that the case was so; and recent discoveries—one of a quay-wall or foundation of a bridge at Old Kilpatrick, about 4 feet deep in a field; and another of a causeway, more than 20 feet submerged and silted up under sea-sand, on the same side of the river, near Glasgow, will most probably confirm it.

One other question, however, yet remains, touching this mysterious fort, which we may be allowed to say only "Ossian and the Clyde" can enable us to answer—Why was such a fort ever thought of there at all? It was either to receive