height above the surrounding country. Hence | conquests of Agricola, he took no pains to the Watling-street and other Roman roads, maintain them, and contented himself with which strike northward into Scotland, are fixing the limits of the province along the found to connect a series of Roman encamp- line of forts which guarded the lower isthmus. ments, many of which in process of time ger- But this line of forts he strengthened by minated into cities. But if they were uncer- a continuous rampart; and the question tain about the permanence of their advanced arises whether the whole, or what part, of conquests, they were in the habit of drawing the existing remains of fortification between a transverse line of communication, to which the Tyne and Solway are due to his energy the same name of limes was also given, to con- and prudence? nect their outposts laterally; and this line they sometimes protected by a rampart of earth, a ditch, and a palisade. This fortification seems also to have borne sometimes the name of limes; but this was not a correct use of the term; the proper word was vallum. The first Roman general who penetrated into the region of the isthmus was the renowned Agricola, in the year 79: after advancing, in his second campaign, to the limits of the Brigantes, he planted a series of camps or castella from east to west, and opened, we may be assured, a communication between them. This was called a prætentura; there was no connecting line of rampart, but the castella were near enough to afford support to each other. We may believe that many of the existing camps in this district were constructed originally by this commander. But Agricola carried his victorious arms further north. He drew a second prætentura between the Clyde and Forth to secure the province within the upper isthmus; and from this vantage-ground he again issued forth, and we may still perhaps trace the line of his further advance by the remains of his encampments in Fife and Forfarshire. Still Agricola erected no continuous rampart. His time was limited, and so were his means.

The system of vallation was carried out more effectively in the next generation. Trajan executed great works of this kind in Hungary and Bulgaria, and commenced at least the extraordinary bulwark of mound and ditch, which ran for nearly four hundred miles between the Rhine and the Danube. Trajan's successor, Hadrian, was celebrated for the development he gave to this system of fortification. The barbarians hitherto, says a Roman writer, were kept aloof by rivers or by limites; but Hadrian fortified these balks of turf with palisades, after the manner of a wall-fence. The German vallum, which he continued and perhaps completed, was crowned with such a palisade, but modern explorers have found no traces throughout its length of the employment of masonry. Hadrian came in person into Britain; but he was not anxious for military distinction. He had already relinquished some of Trajan's acquisitions in the East; without perhaps formally abandoning the

millia

first wall,' or

6

On this subject the direct testimony of ancient authors is explicit up to a certain point, but is provokingly imperfect beyond it. Dion Cassius, speaking of the time of Commodus-that is, after Hadrian and before Severus-affirms that there was then a raixos dividing the Romans in the island from the barbarians; but reixos, though we commonly construe it a wall,' is used by the best Greek writers indifferently for a wall of masonry or a mound of earth; nor is it certain whether this author alludes to the works on the upper or the lower isthmus. Herodian speaks of xwuara in this place; but this word is properly applied to earthen rather than to stone ramparts. Spartian, the next on our list, says plainly enough of Hadrian; Murum per 80 primus duxit' He built the was the first to build a wall;' as if there had been another wall built at a subsequent period. Now, as certainly as there was no other stone wall either here or at the upper isthmus, we are led at first sight to question whether by murus a stone wall is meant at all. The suspicion that murus here, as often elsewhere, is used for a rampart generally, with no limitation to a rampart of masonry, is confirmed by a passage of the same writer, in which he attributes a murus to Severus also: Muro per transversam insulam ducto utrimque ad finem oceani munivit.' If Hadrian built a wall here, then Severus did not, and vice versa. There is no other such wall in the island. A third passage of Spartian, speaking also of Hadrian, Post murum apud vallum (or aut vallum) missum in Britannia,' is too corrupt to make any use of. Nor can we draw any inference from the words of Capitolinus, who says that Antoninus drew 'alium murum cespititium,'' another wall of turf,' that the previous rampart of Hadrian was of stone. The phrase is ambiguous in Latin, and may mean either a second wall of turf like the first, or a second wall, not of stone like the first, but of turf only. All we can say positively so far is, that Hadrian erected a rampart of one kind or the other.

The next builder claimed for the Wall is the Emperor Severus. Spartian's testimony, we have already seen, is nugatory. Eusebius

verus

(in Jerome) says that this Emperor erected a Vallum, 132 miles in length, the error in the numbers being perhaps that of the transcriber. Eighty-two miles (LXXXII. for CXXXII.) would answer very well to the width of the lower isthmus from Tynemouth to some point on the Solway, and thirty-two (xxxII. for cxxxII.) equally well to the upper: but neither this nor any other writer tells us which of the two was the isthmus Severus fortified. Aurelius Victor, indeed, says, 'Severus muro munivit; but again, vallum per 32 millia passuum; which looks like a correction of Eusebius, and would refer to the rampart of the Forth and Clyde. Eutropius follows Victor: "Severus vallum per 32 mil. pass. . . deduxit,' while Cassiodorus adopts the vulgar reading of Eusebius: "Sevallum per 132 m. p. &c.' Still there is nothing definite about the Wall. Supposing Severus to have raised a rampart in the same district as Hadrian, we still have no direct evidence that either one or the other erected a wall of stone. But besides the doubt arising from the number of thirtytwo miles, an expression in Eutropius may incline us to fix the rampart of Severus at the Forth, rather than at the Tyne. The barbarians, we are told, had penetrated into the Roman province, but the Emperor not only drove them out, but advanced far into the heart of Caledonia beyond it. He eventually made peace with the barbarians, and withdrew in sickness to York; but he built, it is said, a rampart of thirty-two miles in length, 'ut receptas provincias omni securitate muniret i. e. to give double security to the territory he had recovered. There is no reason to suppose that the barbarians had wrested from Rome the populous and powerful tracts south of the Tyne: the recovered provinces, it would seem, were the lands imperfectly subjugated, between the Tyne and Forth. On the other hand, it must be added, though there are two, or in fact three lines of rampart in the vallum of the lower isthmus, of which Hadrian may have constructed one, and Severus added another, we have no distinct trace of a double fortification in the remains, very imperfect, it must be confessed, of the upper, while we know from indubitable testimony that one such rampart was there raised by Antoninus. We can only say, that, as far as direct evidence goes, Severus's claim to the wall is more doubtful than even that of Hadrian.

In default, however, of testimony of the first class, we may find perhaps indirect evidence, from inscriptions, from the style of construction, or from the general character of the fortifications before us.

The advocates of the Elian hypothesis rely much on inscriptions. Many inscribed stones have been discovered, since the time of Horsley, in the camps contiguous or near to the Wall, and some even in the mile-castles, bearing the name of Hadrian, and dated in his reign, while comparatively few have been found bearing reference to Severus near the Wall, and none immediately upon it. There is no doubt indeed that the camps were generally of the age of Hadrian, or earlier, and the discovery of his name thus recorded in them can have no weight in the question before us. The stone walls of the camps may be of much later construction than the camps themselves. Two inscriptions of Hadrian, however, have undoubtedly been discovered in mile-castles; unfortunately the position in which they stood cannot be ascertained. They were excavated among the débris near the gates; but it is not necessary to suppose that they were erected in a conspicuous place over them.* Such large slabs as those which present these particular inscriptions correspond with the masonry of the gates, which was much more solid than that of the walls; and if the wall was erected at a date posterior to Hadrian, it might be found convenient to seize upon such massive blocks wherever they were to be found. There seems reason to

Dr. Bruce lays great stress on the discovery of such slabs, inscribed with the name of Hadrian, in holm (Vindolana), and also in the mile-castles the camps at Great Chesters (Esica) and Chesterat Milking Gap and Cawfields. It is the last only that can have any importance for the question before us, and the account he himself gives of their port the notion of their having been originally discovery would tend rather to discredit than superected on the spot where they were found. Of the slab at Milking Gap he says, it was found in taking up the foundations of the mile-castle' (p. 234), and again repeats the statement at p. 383. It would seem from these words that the slab in question was taken from some other place, and built into the foundations of the wall or gateway, where the largest stones were ordinarily used. We have had an opportunity, however, of making inquiries on the spot, and of the identical labourer who excavated the inscribed stones both at Cawfields and Milking Gap; and he asserts that they were found lying among the débris with which the interior of the castles was filled, one close beside, the others not far from, the gateway. The actual foundations were not meddled with. The stones must have occupied some elevated place in the masonry of We give this as the most natural interpretation the walls; but still it is not necessary to suppose that of the words of Eutropius. It may, no doubt, be they were placed there by Hadrian's officers. We urged on the other hand that Eutropius speaks admit, however, that Dr. Bruce has a reasonable rhetorically, and supposes all the four provinces of presumption in his favour, and we hope that in a Britain to have been momentarily lost, when Seve-future edition he will corroborate or correct the

rus came to the rescue.

statement we have here made of the circumstances.

believe that these large stones were often | The construction is also very rude; the conveyed to considerable distances. Inscribed courses indeed are tolerably even, but there slabs have been found at Hexham, four miles is no regularity in the setting of the stones of from the nearest known station, at Corbridge one course above those of another. There is (Corstopitum), and others at Lanercost, none of the squared and chequered masonry nearly as far from Walton or Petriana, we observed in the remains at Rome itself, or There seems therefore no difficulty in sup- even in some parts of Colchester, Richboposing that the inscription found in the mile- rough, and Lymne. Except at the gateways castles at Milking Gap was transported from the materials used both in the Wall and in the camp at Chesterholm (Vindolana). the camps required no machinery to transWhile, however, we give this as a possible port them. They seem to belong to an age explanation of the circumstance, we acknow- when there was abundance of rude manual ledge the evidence for Hadrian from these labour, but great lack of science and machinery. inscriptions to be of the gravest importance. Nor can we find any traces of the wellOn the other hand, the partisans of Seve known durable mortar of the best Roman rus point, not to inscribed slabs along the builders; here again the northern wall seems Wall, or in the camps, but to numerous re- very inferior to the great castles of our southferences to the date of his reign in names eastern counties. These are presumptions graven on the rocks, which seem to have only, but as far as they go they seem to tell served for quarries to the Roman builders. against Hadrian, and in a less degree against Indirect evidence may also be drawn, we Severus also. think, against the claims of Hadrian, from an But if we look to the general design of the inscription found at Kirkandrews, near Car- engineer, the Elians bid us observe the palisle, south of the barrier, 'legati legionis vi. rallelism between the wall and vallum, the ob res trans vallum prospere gestas.' Ob- one always keping a little to the north of the serve here the phrase vallum, not murus. It other, and never cutting it, as a proof of a would seem that the mound was in existence common designer. We think that Mr. before the Wall, and therefore that the Wall Maclaughlan's accurate survey has shown was built after Hadrian. But the sixth legion that this parallelism is not so complete as it appears to have been removed to York before has been represented, and that in some places A.D. 190; this inscription therefore does not the wall and the vallum form their angles interfere with the claims of Severus, A.D. 210. respectively (the vallum never admits a curve, If, however, there be any significance in the wall only in following the ridge of the the use of the word vallum, the language of cliffs) without any reference to each other. the Notitia would be fatal to both the rival As for the grandeur of the design of the Emperors; for in that document, dated circ. Wall, we fully admit that nothing can be A.D. 410, the barrier is still called vallum, more imposing than the view of this mighty and not murus. So also the Antonine Itine- bulwark in some of its strides across the cliffs, rary speaks always of the vallum; and it but the impression this circumstance makes may be taken as one of the many suspicious upon us is rather unfavourable to the Elians circumstances attending the work which goes than otherwise. We cannot conceive the by the name of Richard of Cirencester, that most powerful of all the Emperors crouching in this Itinerary we find the barrier desig- behind such a barrier, when his legions, withnated as the murus. out mound or wall of any kind, must have been amply capable of defending the province against any barbarian aggressors. We very much doubt whether the first constructors of the great limitary ramparts resorted to such expedients as important measures of defence. We believe they were more concerned to erect permanent and visible boundary lines to the empire, and that it was no unimportant consideration with them to find means of employing their soldiers without exercising them in perpetual warfare. would be well for mankind if certain modern

As regards the masonry of the Wall, the presumption seems to be against Hadrian. None of the existing limitary ramparts of the Antonine period present any masonry at all; they were merely mounds and stockades.*

Two or three places on the line of the Antonine barrier have derived their name from camps, as Castle Cary; and we presume that Falkirk is so called from the palisade: but there is no place that retains in its name the tradition of a wall. So also along the line of the Limes Transrhenanus, there are numerous places denominated Pfahlsburg, Pfahls-heim, &c.; but wall never enters into the composition of their names. On the other hand, the names on the line of the lower Isthmus refer to a wall, not to a palisade; as Walls-end, Walton, Thirlwall, and a great many others. The barrier in Germany has been lately examined with

It

care by our experienced antiquary Mr. Yates, whose memoir, the only one we possess on the subject in England, has been produced in the Transactions of the Archæological Institute, at their Newcastle Meeting in 1852.

emperors could invent such innocent employ- | The Caledonians had no engines of war, and ment for their own restless armaments.*

Unfortunately we have from henceforth very meagre notices of the affairs of Britain. Between Rome and her subjects the harmony seems to have been complete, but the province was harassed by marauders in the north, and by pirates in the south and east. Carausius, when he raised the standard of an Undoubtedly regular soldiers were at a disad vantage as opposed to savages before the invention independent sovereignty in Britain, took of fire-arms. The heavy-armed legionary could vigorous measures for the repression of the neither overtake nor intercept the Caledonian gal-Saxon corsairs, but we do not hear of his atlowglasses. Accordingly, a mound and palisade might be a very useful adjunct to the defence of a frontier, but it would seem preposterous to oppose a massive stone wall to the raids of half-naked

predatory hordes were encountered by Theo- | have occupied, from first to last, fifty years dosius, the general of the Emperor Valens, in building. in the neighbourhood of London, in the year 368. The invaders were routed and driven back beyond both the limitary ramparts, and Theodosius restored, as we are expressly informed by a reputable historian, the camps, castles, and prætenturæ, or chains of forts in the north, and reconstituted the province beyond the Solway under the designation of Valentia. As, however, no prudent general could hope to retain the permanent occupation of this exposed district, it might be judged expedient to take this opportunity of securing the lower and more important line of defences by the strongest fortifications. If, hitherto, the bulwarks of the lower Isthmus had been confined to the camps and mounds of Hadrian and Severus, it was now, we may suppose, that the stations were fenced with masonry, and the Wall designed and at least partly executed, with broad openings at every mile for the temporary shelter of the exposed provincials beyond it. After the retirement of Theodosius the frontiers were again assailed by the restless savages. Stilicho, about 400, issued orders from Gaul for putting the island in a state of defence against the Saxons, the Picts, and the Scots, and if we may rely on the evidence of the poet Claudian, his designs were carried fully into execution.* We may at least admit that his engineers continued and extended the plan of Theodosius. Finally, after the withdrawal of the Roman garrison by Maximus, the Picts and Scots repeated their attacks, and the single legion which was sent from Rome in 414, and again a few years later, may have assisted or at least advised the natives in putting the finishing stroke to their defensive works, and thus the Wall, the remains of which we now see, may

Such is the latitude which we would recommend in interpreting the well-known passages of Gildas and Bede, which supply us with the only direct assurance that the Wall was not built by Severus, to whom, says Bede, it is erroneously attributed, and still less by Hadrian, while they expressly ascribe it to the Romanized Britons of the fifth century under the direction of a Roman legion. In this assertion, as it stands, we do not place much faith. The authority of Gildas, though, perhaps, nearly contemporary, is extremely slender, and Bede, two centuries later, seems to have readily acquiesced in his predecessor's assertions. But writers of this class are prone to fixing to a single occasion, or on a single person, the authorship of works which really belong to a period, and we may fairly interpret the legend referred to as the mythical way of expressing the fact that the Wall was a barrier executed during a series of years, not against the Caledonians, but their successors the Picts and Scots. It was the work, we can easily believe, of the natives themselves, under Roman superintendence; for the district seems to have been at the height of its population and productiveness, and the execution of the work is just of the rude character we should expect from the unskilled labour of the children of the soil directed by scientific engineers. For the tradition of the Wall having been the work of Severus, we may account on a similar principle. Severus, the most resolute enemy of the Caledonians, was the eponymus of Roman invasion, the Hercules of the later empire; it is said that both the upper and lower ramparts have been known to the Gael within times quite recent as the Gual Sever, or wall of Severus.*

We can hardly imagine that such formiWe can hardly consent to regard Claudian's vi- dable ramparts, if defended by the disciplined gorous lines as mere rhetoric:

The existing remains of Roman fortifications at Lymne, Richborough Colchester, Burgh Castle, and Caistor, on the south-eastern coast, seem all to belong to one period, and that not earlier than the reign of Carausius, if they are not, indeed, to be ascribed to an age as late as Stilicho's. Roman pavements of an earlier date have been discovered under the foundations of the walls of Colchester. The 'Saxon shore' was furnished with many other fortifications besides those enumerated, which have now disappeared, as at Brancaster and Reculver. Other Roman stations, such as Ituna on the Essex coast, have been swept away by the sea.

bands of Hadrian or Severus, could have been so repeatedly broken through by half-armed barbarians. But when the Roman arms were finally withdrawn, no strength of natural or artificial defences could avail for the protection of the timid and helpless natives. The Wall was speedily penetrated, and from the middle of the fifth century it ceased to afford

* Pinkerton asserts, we know not on what authority, that the Antonine barrier bore the name of Gualsever. This is the traditional title given to the lower wall by Spenser :

'Next there came Tyne, along whose stony bank
That Roman monarch built a brazen wall,
Which mote the feebled Britons strongly flank
Against the Picts, that swarmed over all,
Which yet thereof Gualsever they do call.'
See Bruce's Roman Wall,
p. 378.