IV

FLAVIAN WARS AND FRONTIERS

X.

THE FLAVIAN ACHIEVEMENT

The inglorious issue and, as it turned out, the delayed decision, of the Danubian crisis cannot obscure the achievement of the Flavian emperors in other lands. The reorganization of the eastern frontier and the pacification of northern Africa were matched by a great advance in Germany: solid results continued to be attained with an economy of men and of money that would have gratified Vespasian himself.

Nor was there less apparent cause for satisfaction in the state of Britain. Wales and northern England had been subjugated, and the Roman arms had been triumphantly carried far into Scotland. With the departure of Agricola, silence envelopes the island for nearly forty years. When the veil lifts again it reveals the presence of a Roman emperor in Britain and the construction of Roman frontier-works between the Solway and the Tyne. What had happened in the interval? Only archaeology can provide an answer; and that answer is still faint and faltering.

Perdomita Britannia et statim missa. On the strength of this observation, it was long believed that all the conquests made by Agricola were immediately abandoned. There is exaggeration not merely in one member but in both members of the Tacitean phrase. The supersession of Agricola did not represent any change of policy with regard to Britain, for even had he remained a lull would presumably have followed the great advance initiated by his third campaign. Agricola's successor, whoever he was, was the author neither of an advance nor yet of a retreat.

The greater part of Agricola's gains in Scotland were held for some years, for that is the conclusion that appears to be indicated by the study both of the Roman coins found in Scotland and of the remains, structural and other, of certain military sites in Scotland, not merely Newstead and Camelon, but Ardoch and Inchtuthill, beyond the Forth. A definite frontier has not been found and may never have existed—for it is often misleading to apply that term to chains or groups of forts built for the purpose of penetrating and subjugating a refractory region. How long the Romans kept their hold on southern Scotland is not known—a withdrawal in the early years of Trajan or even towards the end of Domician's reign is not impossible. In default of more evidence from Scotland, knowledge may be augmented by the result of excavation on or near the frontier-works of Hadrian, where there are forts or posts that are proved, either by their situation, their structure or their remains, to be earlier than the Wall of Hadrian and therefore perhaps contemporary with the Stanegate.

The occupation of the line between the Solway and the Tyne was a necessary part both of the conquest of the Brigantes and of their subsequent repression, for these recalcitrant tribes were to be heard of again. As in Mauretania, the Romans in Britain were unable to find a single and satisfactory line behind which peace could reign undisturbed. It is no accident that the only wars that troubled Pius, the most Antonine of emperors, were waged in Britain and in Mauretania. Seen in its true light, the retention or the abandonment of Scotland has a local rather than an imperial significance—it concerns merely the depth of the military zone of control in northern Britain,

An advance in the north did not obviate the need for forts south of Hadrian's Wall and in Wales: the difficulties that confronted the Romans are indicated by the fact that it was necessary in the Antonine period to keep garrisons not merely along the more important of the roads across the Pennines at points like Bowes and Ilkley, but even on the southernmost fringe of the Brigantian territory, for example, at Brough in Derbyshire and at Templeborough. The civilizing methods of Agricola which extorted the grudging admiration of Tacitus were pursued with success in the south. There was no place for them in the north. Beyond Aldborough there were no cities.

In peace or war the army of Britain was imposing in size, especially in its contingent of auxiliary regiments, indispensable for open warfare, of which there was abundance, and for garrisoning forts. While Agricola was still in Britain, the legions had sent detachments for service in Germany: and several years later, apparently in 86 or 88, II Adiutrix departed for ever to the Danube. Three legions remained, II Augusta at Caerleon, IX Hispana at York, XX Valeria victrix at Chester. For their tasks they were none too many, and they acquitted themselves nobly. Britain continued to be a fighting province, and her legions were thought worthy of comparison with the best in the Empire, the Danubian troops.

Germany, however, suffers a change and a degradation. For more than a century, from Domitian to Antoninus Caracalla, the peace of this frontier does not appear to have been seriously disturbed. The German armies decline in numbers and in prestige. They had been the arbiters of empire: they now no longer play a decisive or even an independent role.

It had long been desirable, and it now became possible, to reduce the formidable total of the legions stationed on the Rhine. To this frontier Domitian had brought his new legion I Minervia: but he had withdrawn in succession three legions, I Adiutrix, XXI Rapax and XIV Gemina. None of these ever returned to Germany, and after the departure of XIV Gemina to Pannonia in AD 92, the two German armies number three legions apiece: in the course of the following generation each army surrenders yet another legion, and the camps of Vindonissa and Noviomagus are abandoned.

For Lower Germany the Rhine provided a secure frontier. The tribes beyond it had been persuaded or intimidated into submission. But that river no longer marked the eastern limit of Upper Germany. Shortly after 89 the frontier that had been won as a result of Domitian's war against the Chatti was prolonged northwards beyond the Lahn and touched the Rhine at Rheinbrohl, opposite the boundary of the two Germanies. From this point a new frontier by land ran in an irregular line south-eastward to reach the Danube a little above Regensburg. At the death of Domitian it had probably not been clearly delimited along its whole course. But the process may be described as completed. Indeed, such modifications as were made later, even the advance east of the Odenwald and the valley of the Neckar in the time of Antoninus Pius, bear a local rather than a general significance. There was a change, however, in the form of the defence—in the Antonine system all the forts were strung out on the line of the frontier itself. Hadrian erected a wooden palisade. Later (perhaps in the time of Caracalla) the Raetian frontier received a stone wall about eight feet high, Upper Germany, however, the mound and ditch known as the Pfahlgraben.

The Flavian advance had secured a route between the Rhine and the Danube, from Mainz by Stuttgart to Ulm. It might indeed have been expected that the Roman advance would cut yet deeper into southern Germany, to incorporate the land of the Hermunduri in the valley of the Main and win not only a shorter frontier but a shorter line of communications from Mainz to Regensburg by way of Nuremberg. But this was not to be, and a salient of free Germany still faced the Roman frontier on the west and on the south. In this region southwest from the territories of the Hermunduri extended a broad belt of virgin forest: it presented no threat to the Romans and promised no advantage from annexation.

To an advance of the Roman frontier in Germany and an annexation of territory beyond the Rhine there is a solitary reference in the imperfect records of history. In his account of the nations of Germany, Tacitus inserts, while refusing them the right to appear there, the inhabitants of a district which he designates as decumates agri. They were Gauls and immigrants, subsequently annexed to the Empire—“mox limite acto promotisque praesidiis sinus imperii et pars provinciae habentur”. The meaning of this term is quite un­certain: it never recurs—and was perhaps obsolescent when Tacitus wrote, for the regions beyond the Rhine had become part of a province. The military territories of Upper and Lower Germany had not hitherto in official language been dignified with the name of provinces. That title appears for the first time in the reign of Domitian, an emperor enamored of precision and uniformity. The change of title first attested in AD 90 may have followed close upon the suppression of the revolt of Saturninus: an earlier date is not excluded. However that may be, the new lands beyond the Rhine soon received an organization based upon tribal communities, such as already existed among the Mattiaci and the Suebi Nicretes: for the rest, however, the civitates appear to be new creations, with names derived from their locality, a fact which justifies Tacitus' refusal to number them among the nations of Germany. In the upper valley of the Neckar, indeed, around Sumelocenna (Rottenburg) a large region became Imperial domainland: but the natives subsequently developed into a self-governing community.

From the brief and confused remarks of two epitomators of a later age, this process of organization is sometimes assigned to Trajan, an emperor who is never allowed less than his due: but it may well go back to the institution of a province of Upper Germany in the time of Domitian. For Trajan there remained little or nothing to be done along or within the frontiers of Germany and Raetia; and no military activity is recorded. Even a remorseless panegyrist like Pliny the Younger must confess himself defeated.

The advance of the frontier in Germany and Raetia had been completed in its essentials by AD 96. The date at which certain forts were first established cannot always, it is true, be closely determined: whether some forts belong to the last years of Domitian, to the brief reign of Nerva, or to the beginning of that of his successor is a problem that belongs to topography rather than to history. The process of annexation was completed, so Tacitus records, by the drawing of a limes. This was the term which soon came to be applied to each and all of the frontiers of the Empire, at first perhaps only when they were lines of demarcation or defence on land, but later to rivers as well. The original meaning of the word, a straight path, hence a boundary, might be taken to suggest that it developed by an easy and natural transition to signify the limit first of a province, then of an empire. But this is not so: the word had also a narrower sense—it was a technical military term, designating the straight clear path along which a column of troops moved forward to attack in a battle or in a campaign. Limites were constructed to penetrate hostile territory and were subsequently maintained to control it. These military lines of penetration had been employed by the Romans in their invasion of Germany in Augustan days: and Domitian operated in this way against the Chatti over a wide front. A military road, accompanied in its course by fortified posts or watch-towers, might thus be used to isolate difficult territory and might sometimes correspond more or less to the limit of effective control. In northern England the Stanegate may have fulfilled this function for a time: and at an earlier date the Fosse Way was perhaps the earliest frontier of the Roman province of Britain.

The essential of a limes, then, is a road with watchtowers or forts along it. It is not necessary that it should be provided with any other defenses. As has very properly been observed, the essential feature of Hadrian's Wall is not the stone wall itself, which is best regarded not as a barrier but as an elevated sentry-walk.

In this sense, all the limites of the frontier provinces of the Empire embody the same principle. But here the resemblance ends. Just as army varied from army in composition and functions, so did limes from limes. Great differences may indeed be observed along the same frontier at the same time. The system of defense designed by Domitian after the annexation of the Wetterau has already been described—a chain of watchtowers, with here and there a small post, running along the rim of the Taunus and sweeping around the northeast of the Wetterau, to join the river Main near Hanau. Here there was an enemy to be feared, the Chatti: and so the forts that housed the auxiliary regiments were situated in the rear. South of Hanau, however, things were different, and the regiments could be placed on the line of the frontier, for the patrolling of which they supplied the troops. This frontier followed the bank of the Main for some twenty miles, as far as Worth, where it struck southwards, keeping to the line of a ridge, and descended to the Neckar at Wimpfen. Thence the Neckar provided the frontier as far as Cannstatt, with a chain of forts, constructed, like those on the rest of the limes south of Hanau, of earth. East of Cannstatt the point of junction with the limes of Raetia at this date is uncertain, for the limit of the Roman advance had not yet been clearly marked by any natural or artificial line.

In the form which it was ultimately to receive, the Roman frontier in Germany was a visible and an imposing barrier. But none of the features which gave it this character, in Upper Germany the palisade of Hadrian and the later mound and ditch (the Pfahlgraben), in Raetia the palisade and then the stone wall, were present in the original scheme. This scheme, indeed, could have served its purpose adequately enough without them, for the new frontier was designed to be, not a line of defense, but a line of patrols, to watch the natives and prevent their crossing without leave the limit that Rome had set, whether that limit was a river or a line drawn across the dry land. This function is illustrated by the inscriptions which Commodus set up to commemorate his repairs along the Pannonian limes of the Danube—“ripam omnem burgis a solo extructis item praesidis per loca opportuna ad clandestinos latrunculorum transitus oppositis munivit”. Even before the Flavian reorganization of the frontiers of Germany and Raetia, it was the rule that natives should not cross the boundary rivers how and when they pleased. This interdiction was now reinforced by a stricter control. The tranquility of the frontier is illustrated by the fact that Hadrian transferred the garrisons from the forts in the Wetterau to the line of the limes itself: of a German invasion there appears to be no danger, and Roman soldiers usurp the duties of gendarmes and customs officials.

The full significance of the measures adopted by the Flavian emperors in Germany and of the changes thereby effected or portended was not at once apparent to contemporaries. In the year 98 Tacitus published his Germania. Of its character and purpose there has been much debate. Though it may very properly be denied that the tract was written to serve a moral or a political end, Tacitus would not be its author if it did not betray some indication of his personality and his opinions. At the time when the Germania was made public, Trajan was on the Rhine. A policy different from that of the Flavians might be deduced from his character and his career. From the earliest encounter of the Romans with the nations of Germany down to the second consulate of Trajan, more than two centuries of history had been filled with the record of their wars: and the latest triumph celebrated over them had been false and futile.

It may be inferred that Tacitus hoped in secret for that conquest of Germany which he did not dare openly to advocate. He recounts how an offending people, the Bructeri, were pitilessly massacred by a confederacy of their neighbors for the advantage of Rome. In the comments which this edifying spectacle has moved him to record, it is perhaps permissible to read, not so much solicitude for the future destiny of the Empire and hope that the enemies of Rome may ever be divided thus, as irony and indignation that so ignoble a policy should in the present be recommended. But Tacitus was deceived and disappointed: Trajan sought his laurels in other lands. Tacitus should have assigned more space and more significance to the Danubian Germans, the Marcomanni and the Quadi: he should have compiled for his contemporaries and for posterity some account of the peoples of Dacia and Sarmatia.

To the changed and calamitous situation on the Danube Tacitus had already in his biography of Agricola borne emphatic testimony—disaster upon disaster, continuous and unmitigated, four Roman defeats in Moesia, Dacia, Germany and Pannonia. As later in the days of Marcus Aurelius, from Bohemia to the Pontus all was confusion. The needs of the Empire summoned Domitian three times to a frontier which no emperor before him had ever troubled to visit. He adopted the methods which tradition and common-sense recommended: he had not designed the conquest and annexation of Dacia, but had sent Fuscus and then Julianus across the river to restore the prestige of Rome and secure peace for the future by humbling and weakening Decebalus. It was now advisable to come to an understanding with Dacia: a strong Dacia, with a monarch who could keep his own subjects in order, and check the nomad tribes on either side, might become an integral part of the system of frontier defense. For this reason Domitian paid Dece­balus a subsidy and lent him engineers to build forts. In a later age, when Rome could no longer hold and defend Dacia as a province, Aurelian yielded this territory to the Goths and acquired for Rome a century of peace along the Lower Danube. To choose, delimit and garrison a frontier is only a small part of frontier defense: more important are the relations with the tribes beyond it: instead of continuous unrest, of repeated punitive expeditions against an elusive or inaccessible enemy, empires before or since have not disdained to enlist by the payment of money the cooperation of the more civilized or the neutrality of the more turbulent tribes along their borders. Trajan subsidized the Sarmatae Roxolani beyond the Lower Danube: in the first year of Hadrian they complained of a reduction of the money paid to them.

Rome had hitherto paid subsidies to the Marcomanni and Quadi; now Dacia occupied that privileged status. That a Roman victory in Dacia should have been followed by peace without conquest was distasteful or inexplicable to contemporaries who were more familiar with history as it appeared transfigured in literature than with the stern requirements and the sober methods of imperial frontier policy: and even if Domitian had not been detested and his memory condemned, the choice which he took must have appeared ignoble when confronted with the glorious achievements of an emperor who revived the wars and triumphs of ancient days. Yet it might be urged that a policy adopted in the face of a sudden emergency should only be judged with reference to that emergency. Time might have refuted Domitian's policy by its results: that it was folly and a failure is not at once proved by the fact that it was reversed by Trajan. Moreover, should the power of Decebalus appear to have been unduly augmented, the ephemeral empires and rapid ends of Burebista and of Maroboduus gave some grounds for confidence in an issue other than that of war and conquest.

The change in Roman foreign policy was accompanied by a re-organization of the defense of the long and imperiled frontier of the Danube. Though evidence is scanty, it is clear that ample compensation must have been made for the neglect that had prevailed hitherto. Additional auxiliary regiments and new forts would be required. The Column of Trajan depicts wooden watchtowers along the Danube, like those built by Domitian in the Taunus; and repairs were made on the road hewn in the rock on the southern bank of the Danube in the narrow gorge, called the Pass of Kazan, a road begun by Tiberius, but for which Trajan was to have the ultimate and enduring credit.

What can be inferred of the movements and distribution of the legions provides an indication of value. In the time of Vespasian six (or perhaps seven) legions comprised the garrison of the provinces of Dalmatia, Pannonia and Moesia: and two of these were still stationed in the interior. By the end of Domitian's reign there were probably nine. Three legions had arrived in succession from the Rhine, I Adiutrix, XIV Gemina and XXI Rapax, one from Britain, II Adiutrix: none went back, and one had perished.

Moesia had been divided in 85—63. Uncertainty about the boundary between the two provinces contributes to the difficulty of determining which of the Moesian provinces had three legions, which two, for the total appears to have comprised five. The camps of I Italica, V Macedonica and VII Claudia were probably, as before, Novae, Oescus and Viminacium: about the camps of the new arrivals, probably II Adiutrix and IV Flavia felix, there is no certain evidence.

As for Pannonia, the line of the Danube from Carnuntum (east of Vienna) to its confluence with the Save at Belgrade had long been neglected. The defection of the German and Sarmatian allies of Rome laid bare what was perhaps the most vulnerable section of the whole frontier, between Vienna and Budapest—in the time of Hadrian it was held by four legions, stationed at Vindobona, jCarnuntum, Brigetio and Aquincum. By the end of Domitian's reign XIII Gemina had probably moved from Poetovio to Vindobona; and XV Apollinaris garrisoned Carnuntum as before. There how appear, however, to have been two more legions in Pannonia, I Adiutrix and XIV Gemina, possibly at Brigetio and at Aquincum.

The garrisons of Britain and the German provinces have now fallen to three legions apiece, and Pannonia, with four legions, holds pride of place among the military provinces of the Empire. Long neglected, the Danube comes into its own, with nine legions as against six in the Rhine. By the time of Hadrian the Rhine armies have shrunk to four, and ten legions in the Danubian provinces attest and guarantee the importance of that frontier in peace and in war.