The Romans on the Riviera and the Rhone; a sketch of the conquest of Liguria and the Roman Province

Front matter

5*

THE ROMANS

ON THE

Preface

Having enjoyed for the last twelve years the privilege of wintering on the north-western shores of the Mediterranean, I have turned them to account by following up in every direction the tracks left by the Romans on the French and Italian Rivieras. Nor have my explorations been confined to the coast, but have been extended north of the Apennines as far as the Po in Italy, and in France have included the seldom-visited valleys of the Durance and the Drome. It is in the hope of throwing some light of local colouring on to an obscure field of Roman history, that I am tempted to offer this sketch to the public. For mainly owing to the irreparable loss of the books of Livy's history treating of the conquest of Provence (the Roman Provincia), we are left very much in the dark on that important episode in Roman history. Nothing but the intimate acquaintance with the locality which the writer possesses has rendered it possible to eke out the slender materials available, for the construction of any kind of bridge over the gap in Roman history between the narratives of Livy and Caesar ; and some bridge is indispensable, to connect the broken ends of the two narratives. For Caesar's Gallic War, with which the student usually begins his acquaintance with Gaul, is really only the second act of the drama of the Roman conquest of that country. It is indispensable, for instance, to be forewarned that (i) the ' Omnis Gallia' of

Vlll PREFACE.

Caesar means Gallia with the Rhone Valley left out, as already conquered and reduced to a Province ; (2) that the inroad of the Helvetii into Gaul, so ruthlessly arrested and repulsed by Caesar, was looked upon as threatening a repetition of the Cimbro-Teuton invasion, which had so nearly overwhelmed Italy less than half a century earlier ; (3) that the driving of Ariovistus back across the Rhine showed that Caesar's eagle-eye saw at a glance where the real danger lay, and that it was better policy to encounter Germany in Alsace than on the Rhone or in Provence.

Passing my winters as I do within sight of the ruins of Forum Julii, which is really a Rome in miniature, I have been imbibing an atmosphere as completely Roman, as if I had been living on the outskirts of the Eternal City.

In fact I have often had to ask myself, can it be that the long line of mellow arches, which I see there striding across the plain, are other than those of the Claudian Aqueduct, and that those brown-hooded shepherds, sheltering with their flocks from the fierce rays of the midday sun under spreading stone-pines, are anything else than the familiar figures and features of the Roman Campagna?

When, escorted by the arches of the aqueduct — those avant-coiireurs of the ruined city, — you approach the entrance, where the Porta Romana once stood, and gaze on the picturesque outline of the walls still standing, though in ruin, and the Roman towers, cutting clear against the western sky, you have before you a scene of concentrated Roman ingredients, unsurpassable even at Rome.

One has often to ask oneself, why is it, that when once the Romans have set their seal on a place, it is redeemed for ever from dullness, and that its desolation becomes more eloquent than words can express ?

Somewhat to the author's own surprise, his work, which originated in musings over the ruins of Forum

PREFACE. IX

Julii, has taken the form of a sketch of the expansion of Italy into Gaul through the Roman conquest of Liguria, which prepared the way for Caesar's Gallic Wars.

It was only by degrees that the author became aware how much wider was the ancient Liguria than the strip of Italian coast to which it is limited by modern geographers. He soon discovered that at Frejus he was in a centre, which was Ligurian before it became Roman, and that he was located at the most convenient point on the coast for exploring the whole of ancient Liguria, which extended from the Rhone on the west across the Alps as far east as the line of the Trebia, prolonged across the Apennines by the valley of the Magra to the Mediterranean.

In Gaul, it reached from the coast inland as far as the level of the Isere, and in Italy up to that of the upper half of the course of the Po, and even beyond it1.

The references in the text and notes will make it clear that the author has had occasion to draw more on foreign than on British sources for assistance in the composition of this sketch. While French, Germans and Italians have dealt with the different branches of his subject, the author is unacquainted with any English work dealing specially with it.

Although reluctantly compelled to point out the serious errors into which M. Lentheric has fallen on the subject of the Via Aurelia, the author is glad to acknowledge the pleasure he has derived from the rest of his work which is known to the French as La Provence Maritime, and to the British reader as The Riviera Ancient arid Modern.

For a detailed account of Forum Julii the author refers his readers to the " Histoire de Frejus" by J. A. Aubenas, to which the author is himself much indebted.

1 In the days of Justinian, Milan was considered the capital of Liguria. See Gibbon, Milman's Edition, Vol. iv. p. 57.

X PREFACE.

the student who wishes to go deeper in the subject will find an admirable and scholarly work.

The author offers his best thanks to Mr Jenkinson, Cambridge University Librarian, for unfailing courtesy in supplying him with works of reference, as also to Mr A. S. Murray and to Mr W. R. Wilson of the British Museum. He has also to express his obligation to Mr E. S. Shuckburgh for the use of his translation of Polybius and other assistance, and to Professor Rhys of Oxford for valuable hints. He is indebted to Mr A. Rogers of the Cambridge University Library for the compilation of the Index at the end of the volume.

The author has finally to tender his best thanks to Mr H. J. Edwards of Trinity College, for reading over the proofs while the work was passing through the Press.

LIST OF MAPS AND PLANS.

The Riviera, Provence (Roman Province) and regions adjacent.

The mouths of the Rhone.

Plan of Marseilles. Ancient and Modern.

Plan of Frejus (Forum Julii). Ancient and Modern.

Section of the Table of Peutinger.

Fi-ontispiece {after Preface)

THE ROMANS ON THE RIVIERA AND THE RHONE.

CHAPTER I. Introductory

Before bringing the Roman actors on to the stage, it will be well to take a preliminary glance at the early distribution of races in Liguria and Gaul. For, as far as I know, there are very few persons on this side the Channel, who are at all aware of the extremely subversive views on the subject, which are rapidly gaining ground in France. Had I been writing only thirty years ago, it would hardly have been necessary even to refer to the presence of traces of pre-historic man in the regions under review. But the recent discoveries in the field of pre-historic archaeology have been too important to be left entirely unmentioned, even in a work professing to deal exclusively with history. The future indeed seems to belong to the physical investigators, who brush aside in a somewhat high-handed manner the old fashioned history, which is based mainly on texts.

Largely in consequence of physical evidence the ideas on the subject of the earliest inhabitants of Gaul, which were embodied in the works of the late Amedee Thierry, and which prevailed in France so recently as the date of the publication of the Life of Julius Caesar by Napoleon III., are undergoing a radical change. H. I

For there is no trace in those works of any perception of the theory that the Gauls never really formed the groundwork of the population in the country we call France1. It is now beginning to dawn upon the French mind, that as an invading minority of German Franks have given their name to modern France, so its ancient name Gallia was derived from invading Galli — a race resembling Germans in type — who subdued, without displacing, the indigenous population.

It seems evident that, if the majority of the inhabitants of France were of Gallic origin, they ought to be tall and fair. Whereas — and that more especially in the generally believed-to-be purely Celtic regions of the west of France, they are as a matter of fact found to be small and dark. The skulls too and skeletons derived from the tombs in western France are mostly round and short, instead of being long and tall.

So whatever race-feud exists between French and Germans is not to be attributed to the Celtic element in the French people. For, in spite of their difference in language Celts and Germans are of kindred blood. Strabo, p. 1962.

The first modern writer to draw public attention to the above anomaly was the Baron Roger de Belloguet, whose remarkable work, Ethnogenie Gauloise, was published in the year 1869. In Vol. II., p. 309, the Baron sums up in the following terms his conclusions, arrived at, be it observed, prior to the production of the convincing physical evidence, since discovered in caves, lakedwellings, and tombs, "nous concluons, que deux types d'une constitution physique aussi differente, que les grands blonds et les petits bruns, ne peuvent etre sortis d'une meme souche. II est faux que les hommes a

1 We find for instance the statement that "nineteen-twentieths of the French are descended from Gauls" in Duruy, History of Rome, Chap. LIII. p. 83 (English translation). Edited by Prof. Mahaffy.

2 In references to Strabo, the page of Casaubon's edition will always be given.

Chap. I. 3

famille celtique II en resulte, que les Celtes ne

Startling enough to the ordinary British student, nourished in the traditions of Caesar and Livy, as these conclusions will appear, they seem moderate in comparison with the latest views advanced by the eminent Celtic scholar, M. d'Arbois de Jubainville. For this distinguished member of the Institut goes the length of asserting that there were never in Gallia at any one time more than 60,000 Gauls or Celts (for the terms are synonymous1) including non-combatants. In his Premiers habitants de V Europe, Vol. II. p. 7, M. d'Arbois de Jubainville, referring to the observation of Caesar, " nam plebes paene servorum habetur loco," De Bello Gallico, VI. 13, expresses himself as follows :

"Les membres de l'aristocratie gauloise ne pouvaient pas mettre sous les armes plus de quinze mille cavaliers. Vercingetorix fit cette evaluation d'effectif lors de la grande insurrection dont il fut le chef. Or, une regie de la statistique de ce temps etait que le nombre des hommes capables de porter les armes, formait le quart de la population totale.

"Les quinze mille cavaliers, avec leurs femmes, leurs

1 See p. 5.

Had such a paradoxical conclusion as the above been enunciated by a less eminent author than M. d'Arbois de Jubainville, it would have been at once dismissed as undeserving even of consideration. But, whatever adventitious assistance the opinion may derive from the name of the author, it becomes at once evident that, in its exaggerated form, it necessarily falls to the ground when confronted with the plain text of Caesar's Gallic war, and the general consensus of the classical authorities. (From what I gathered during a personal interview with M. d'Arbois de Jubainville,, I have reason to hope that he is prepared to modify his extreme views.)

For in Lib. II. c. 4, we find the total of the Belgic contingents alone amounting to 266,000 combatants. In Lib. VII. c. 76, the outside Gallic host brought up for the relief of Alesia, inclusive of Belgae, numbered, according to Caesar, 240,000 infantry and 8000 cavalry1, while 80,000 more were beleaguered inside with Vercingetorix.

Although M. d'Arbois de Jubainville considers the infantry unworthy of enumeration, as not of Gallic blood, he still does not actually deny that they were present, and draws attention to the fact that the ' plebes,' from which the infantry was presumably recruited, spoke Gallic. It does not appear to strike M. d'Arbois de Jubainville as at all improbable that a minority of 15,000 families should have succeeded in imposing its language on a people numbering three millions.

M. d'Arbois de Jubainville is perhaps not necessarily in conflict with Caesar as to the numbers, only as to the

1 Of the 15,000 cavalry, the number given by M. D'Arbois as estimated by Vercingetorix, only 8000 appear to have obeyed the summons.

CHAP. I. 5

nationality of the combatants before Alesia. Whatever its exact composition may have been it is hardly surprising that Caesar described the Gallic-speaking host gathered around Alesia under the general name of ' Galli,' and that he failed to enter into any ethnological subtleties on such an occasion.

That Gallic blood in the Gaul of Caesar's time was confined to the cavalry or fighting aristocracy is a paradox, which—pace M. d'Arbois de Jubainville — will not bear serious examination. If we admit that recent discovery proves that the groundwork of the population in western Gaul has always remained Iberian, there must still have been a large admixture of Gauls in the inferior population or "tiers etat," which was apparently of no more account in Caesar's day than in France before the Revolution. For we are expressly told by Caesar, Lib. VI. c. 13, that it frequently occurred amongst the Gauls that independent individuals broke down, overburdened with debt or fiscal charges, and losing caste, descended in the social scale to swell the ranks of the poor clients of the nobles.

In the opening chapter of the first book of his Gallic war, is contained the only general reference to distinctions of languages existing in Caesar's Gallia, from which the Rhone valley, formed 60 years earlier into the " Provincia Romana," was excluded. The passage is known to every schoolboy. "All Gaul is divided into three parts, of which the Belgae inhabit one, the Aquitani another, while the people whom we call Galli, but who call themselves Celtae, inhabit the remaining part. All three differ from each other in language, institutions and laws1." Although unfortunately Caesar tells us nothing as to the nature of this difference, whether it was of kind, or only of degree,

1 If Caesar had visited Gallia a century or two earlier, he would have found Galli occupying two out of his three divisions. For he expressly states (De Bell. Gall. 11. 3) that the Belgae, who of old came from beyond the Rhine, expelled Galli then in possession of Belgica.

much light is thrown upon this important point by Strabo.

For in Lib. IV. p. 176, which deals with Gallia, Strabo writes, "Some make a threefold division of it into Aquitani, Belgae and Celtae, of whom the Aquitani differ totally from the other two, both in language and physique, resembling Iberians much more than Gauls. The Belgae and Celtae are both Gauls, to outward appearance, but there is a slight difference of language." We must therefore infer that both Celtae and Belgae spoke Celtic. But as Caesar tells us (L. II. c. 3) that most of the Belgae were sprung from Germans we are led to infer that the Celtic spoken by the Belgae was not as widely different from the German as the two languages became later. When (Lib. I. c. 47) Caesar required to confer with Ariovistus, the negotiations were carried on in Celtic, which the German chief had acquired during his occupation of the country of the Sequani. The difference in language spoken by Celtae and Belgae, referred to both by Caesar and Strabo, is perhaps represented by that between Goidelic and Brythonic later.

With regard to the distribution of the population in ancient Gallia, it seems most probable that what is true of our day was true of Caesar's, viz. that the further you advanced southwards and westwards, the darker and smaller the population became, being further removed from the fair, tall, German type of the north-eastern region, the Belgica of Caesar. That there ever existed anywhere a type of small dark Gauls or Celts is now considered in France a complete delusion.

For the type of Gaul, made known to us by the consensus of classical authorities, was tall, fair and blue-eyed — only differing as Strabo informs us (p. 290) from the German or Teuton in being less fierce, gigantic and ruddy. We read in his life by Suetonius that when Caligula ran short of German prisoners for

CHAP. I. 7

his sham triumph over that nation, he was obliged to put up with the biggest Gauls he could find. He compelled them to let their hair grow long, dye it red, assume barbarous names and learn German. Again on p. 196, Strabo, comparing Gauls and Germans, writes " they resemble each other in their nature and constitution, and are akin (crvyyevels) in blood." Such was the type of the Gauls, who in the year B.C. 390 sacked Rome and are portrayed in Virgil's well-known lines :

Of this type too were the Gauls, who a century later invaded Greece, and pillaged the treasure of Delphi, passing on thence through Macedonia and Thrace into Asia Minor, where they carried fire and sword into its most fertile provinces1.

These were the centuries when the Gauls were at their prime. But the history of their exploits has been unaccountably neglected by British historians. It is given in full detail and in a most readable form by Amedee Thierry in his Histoire des Gaitlois.

When Caesar invaded the country, to which they had given their name ' Gallia,' the Gauls were already a degenerate people, as Caesar himself points out. From living remote from the enervating luxuries of the 1 Provincia,' and nearer to the Germans, the Belgae alone escaped degeneracy. In De Bell. Gall. VI. c. 24, we read, " But there was formerly a time, when the Gauls surpassed the Germans in valour and carried war into their country without provocation."

As a matter of fact it was from Belgica that the Romans mainly derived their knowledge of the Gauls. It was there that Caesar's toughest fighting took place.

1 The most famous of these Gauls were the Tectosages, who finally settled down in Galatia, and were the ancestors of St Paul's " foolish Galatians."

The adjoining Rhine Provinces, Germania Superior and Inferior, where the tall, fair type of Gallicised Teuton has'always predominated, were the only parts of Gallia ever permanently garrisoned by the Romans. The Romans, in the opposite sense, anticipated the Germans in keeping the "Wacht am Rhein." So it is not surprising that, writing about A.D. 350, in the reign of Julian, Ammianus Marcellinus should observe, L. XV. 12, "The Gauls are almost all tall, fair and ruddy, with a terrible gleam in their eyes."

It was not till comparatively late that the Romans had much intercourse with western Gallia, where, in the reign of Gallienus, Bordeaux became an imperial residence.

To Amedee Thierry the anomaly of attributing the small and dark Bretons and Welsh to a Celtic or Cymric stock seems to have presented no difficulty. But the opinion is rapidly gaining ground, both in France and Britain, that in spite of their Cymric name and speech, the population of Brittany and Wales is to be regarded as mainly Iberian. Professor Boyd Dawkins in his Early Man in Britain is a strong exponent of this view as regards the Silures, quoting Tacitus1 to support it.

In those regions of the far west the strain of invading Celts, who imposed their name and language on the conquered indigenous population, has gradually become absorbed and disappeared. As far as language is concerned, the same process has been steadily going on in Ireland, where the descendants of English settlers have become more Irish than the Irish, and yet the English language has prevailed.

While the language has undergone at least two great changes in Ireland, i.e. from Iberian to Celtic and from Celtic to English, the blood in the south and west of the island is probably still mainly Iberian. If the theories of Roger de Belloguet, D'Arbois de Jubainville in France, and Professor Boyd Dawkins in England,

1 Agricola, c. 11.

CHAP. I. 9

hold good, it will soon be hard to find a purely Celtic race left. In the same chapter where Tacitus refers to the tradition of the Iberian descent of the Welsh Silures, he suggests the likelihood of a German origin of the Caledonians. There is a natural tendency on the part of readers, and especially in the case of countries which ^ they know only by name, to assume that the bulk of the inhabitants belongs to the race which has last given its name to the country. But such assumption frequently turns out to be utterly unfounded. The conquest of a country often involved little more than s a change of landlords.

There have from time immemorial been two main classes of invaders, namely that of organized bands of armed adventurers, bringing no women ; and that of tribes in search of fresh settlements and consisting of whole families, with all their belongings. If the armed adventurers settled in the conquered country they would intermarry with the native women, and their descendants gradually become merged in the native population which, notwithstanding, would probably have adopted the name and language of their conquerors.

As far as we can gather from the classical authorities the Galli, with the notable exception of the Helvetic emigrants arrested by Caesar, belonged to the first class, and the Germans to the second class of invaders. In the case of the Cimbri and Teutones we know that they moved in huge caravans of covered waggons, as the reader will find further on when I deal with their invasion of Provence.

Of the three races, Iberian, Ligurian and Celtic, found in Gallia Transalpina it appears that Iberians and Ligurians divided between them the possession of the North-Western Mediterranean coast line prior to the appearance of the Celts. For in the Ova Maritima of Festus Avienus, which though itself of the 4th century X A.D. is based on the Periplus of Himilco, the Carthaginian navigator of the 4th century B.C., we find the Rhone assigned as the earliest authentic dividing line between yc- the Iberians and the Ligurians —

and Strabo, p. 166, mentions the former extension of Iberia up to the Rhone.

By the light of recent investigation it indeed appears probable that not only did the mouths of the Rhone separate these two primitive nationalities on the Mediterranean coast, but that in prehistoric times a dividing line between them was carried up northwards into Gallia1. Into the midst of this Ibero-Ligurian population, formidable bodies of Gauls or Celts, armed with their then irresistible long swords, began to pour, at dates of which we have no positive record. The contests between Celts and Ligurians are referred to as being of frequent occurrence by Festus Avienus, v. 132 of the Ora Maritima :

While our attention is drawn to encounters between Celts and Ligurians, it is to be observed that nothing is said about any resistance being offered by the Iberians within the limits of Gaul.

Driving a Celtic wedge between Iberians, forced

westward, and Ligurians, south-eastward back into the

x/ Rhone Valley, the main body of Gauls occupied in

force the north-eastern and central plateaus of France,

the whole of which they ultimately dominated.

When history opens with Herodotus, about the middle of the fifth century before Christ, we find that bands of Celts had already forced their way right through western Gallia across the Pyrenees, and formed

1 In Chap. v. this question is more fully treated.

CHAP. I. II

a settlement on the Atlantic sea-board of the Iberian peninsula. But they had not yet reached the Mediterranean. For Herodotus informs us in Lib. II. c. 33, "that the Celts are beyond the Pillars of Hercules," thus locating them somewhere on the coast of the Atlantic Ocean to the extreme west of Europe. While in Lib. VII. c. 165, Herodotus enumerates both Iberians and Ligurians amongst the mercenaries collected from all parts of the Mediterranean by Terillus, the expelled tyrant of Himera, in his final effort to recover his throne, B.C. 480, Celts are conspicuously absent from the list. Inasmuch as in the succeeding centuries — and notably in the first and second Punic Wars in the service of Carthage — the Celts became famous as professional mercenaries, we may consider it proved that in the 5th century B.C. they were as yet unknown to the powers bordering on the Mediterranean.

We must therefore certainly reject Livy's date, B.C. 600, the reign of Tarquinius Priscus, as being at least a century and a half too early for the appearance of Celts in Italy, and it seems doubtful whether his account of the emigration of the two main swarms of Celts under Bellovesus and Sigovesus from a parent hive in France into Italy and Germany respectively will bear serious examination.

In spite of Mommsen's qualified acceptance of the main facts without the date1, Messrs Alexandre Bertrand and Salomon Reinach, in their recently published joint work Les Celtes dans la Vallee du Po et du Danube, go very near to rejecting Livy's narrative altogether. For these exponents of the latest theories on the subject, the Roman Province of Gallia Transalpina was never the parent hive of Celts, of whom there were never enough in the west to have swarmed back over the Western Alps into Italy or Germany.

They argue, too, that if the Cisalpine province had been overrun by Gauls from the Transalpine, we ought

1 History of Rome. English translation. Book II. c. iv.

i to find the names of the Bituriges and Arverni figuring there in the place of Insubres and Boii, the latter of whom are scarcely heard of in the Transalpine Province, till we read of them in Caesar's Gallic war, as invading it in conjunction with the Helvetii.

There certainly seems a great deal to be said in favour of the contention that Gallia Cisalpina derived her main contingents of Gauls direct from the valley of the Danube, by which route all authorities are agreed in bringing them originally into Europe. For the Gauls seem to have dropped contingents all along their road through Austria, Germany, and Switzerland to the Rhine, some of whom probably poured over the Rhaetian and Julian Alps into Italy. But the theory of a re-flux of Gauls from Gallia Transalpina is not to be rejected altogether.

The first and only Celtic confederation to establish itself permanently on the Mediterranean sea-board was that of the Volcae. This branch of the Celtic family, from whose name Volcae our word Welsh is derived (as I am reminded by my friend Professor Rhys of Oxford), was the latest to separate itself from the Germans, amongst whom it had found a temporary domicile before crossing the Rhine.

The establishment of the Volcae in the plains between the Rhone and the Pyrenees, occupied previously mainly by an Iberian population, may be fixed approximately at 400 B.C., but nothing is known of the circumstances under which it was effectuated. The Volcae were divided into two branches — Arecomici and Tectosages. The Arecomici, who appear to have been early brought under the civilizing influence of their neighbours the Greeks of Marseilles, were the founders of the great Celtic emporium at Narbonne, which, long before Roman intervention in Gaul, became the principal mart of exchange for native products brought down to the coast from the interior, including tin from Britain.

We learn from Strabo (p. 190) that before the

CHAP. I. 13

date of the Roman conquest of Provence (B.C. 122), the over-lordship of the Arverni extended to Narbonne and the Mediterranean sea-board. Thus the Volcae were clients of the Arverni. While, on the right bank of the Rhone, the Celtic confederation of Volcae displaced or enslaved the former Iberian population about B.C. 400, the Ligurian Salyes, or Salluvii, succeeded to the last in barring Celtic access to the Mediterranean on the left bank.

Thus, the country — now known as Provence — lying between the Rhone, the Durance, the Alps and the Mediterranean, although officially regarded by the Romans as part of Gallia Transalpina, was never occupied nor conquered by Gauls. The bulk of the population of this little-known south-eastern corner of France has always remained Ligurian, just as in its south-western corner the Iberians have never been dislodged from Aquitania. To this day, the south-western quarter of France is largely Iberian in its affinities, the Pyrenees having never served as an effectual barrier between races.

This preliminary sketch of the distribution of races in Gallia Transalpina would be incomplete without some mention of the introduction of a Greek element into the population via Marseilles. If the reader will refer to the section of the Carte de Peutinger reproduced at the end of the volume, he will observe that the name 'Gretia' is given to the region between Marseilles X and the Durance.

And first, as regards the foundation of Marseilles, there has been much discussion as to whether the Phocaeans took possession of an existing Phoenician settlement, or whether they were the original founders. The discovery at Marseilles in 1845 of a stone tablet, now preserved in the Museo Borely, engraved with a long Phoenician inscription, embodying minute regulations as to the functions and remuneration of the priests of the god ' Belen ' (the sun), was for some time thought to strengthen the argument of those who contended for a pre-existing Phoenician settlement. But it having been ascertained that the stone on which the inscription is engraved is of African origin, it cannot be held to prove anything decisive on the point in dispute.