CHAPTER II

THE MATERIAL CIVILIZATION OF THE CELTS DURING THE LA TÈNE PERIOD

ABOUT the year 400 BC a flood of barbarian invaders crossed the Alps and advanced upon the cities of Upper Italy. These barbarians, whose invasions were to be such a menace to the Mediterranean world, were the Celts1, whose territories, now so shrunken, extended in the heyday of their glory from the British Isles to Asia Minor. The purpose of this chapter is to describe their early civilization and history and the causes which led to their migrations. The period which mainly concerns us stretches from the sixth to the end of the second century BC. The preceding phases have been dealt with in vol. II, chaps. 11 and 21; while the Celtic invasions of Spain, Italy, Greece and Asia Minor are described in chaps, 3, 16, and 24 of the present volume.

Prior to the second century BC, we are indebted to archaeological research for the greater part of our knowledge concerning the, Celts. To archaeology therefore we must turn, and see what light it sheds on their culture in this still somewhat obscure period.

The Early Iron Age in Central Europe is divided into two epochs: the Hallstatt and La Tene periods, which take their names from the sites where antiquities characteristic of these two cultures were first identified. Hallstatt is situated in Upper Austria, La Tene in Switzerland, at the eastern end of the Lake of Neuchatel. But the cradle of the La Tene civilization is not to be sought in Switzerland: the region in which that culture unfolded lay farther to the north. Broadly speaking, the La Tène culture was originally a Celtic culture, but this does not imply that it was shared by all the Celtic peoples in its earliest stages. Later, its influence spread into non-Celtic regions as well.

The credit for having first established the chronological subdivisions of this period rests with Tischler. His scheme has since been modified by others, notably by his compatriot, Paul Reinecke. Below is given Reinecke’s most recent chronology, together with those of Schumacher and Déchelette.

Reinecke1 Schumacher3 Dechelette3

La Tfene A

From the second half of the sixth century into the second half of the fifth century B.C.

La T£ne1

From end of the sixth century to c. 400 b.c.

La Tene I

c. 500-300

La T£neB

Beginning from the end of the fifth century B.C.

La T£ne2 c. 400-300

La Tfene C

Before and after 200 B.c.

La T£ne3 c. 300-100

La T£ne II

c. 300-100

La T&ne D

c. 121—15 B.C.

La Tine4

C. IOO B.C.-A.D. I

La Tfene III

C. IOO—A.D. I

The subdivision of La Tène I into two phases is of considerable importance, but, although accepted by leading German and Austrian archaeologists, it has not elsewhere met with the recognition which, in the opinion of the present writer, it undoubtedly deserves. This is due to a mistaken notion, either that La Tène is nothing more than the last phase of the Hallstatt period under another name, or that it is a mere transition from Hallstatt to La Tène. On stylistic and typological grounds the first assumption is unjustifiable, and the same applies to the latter, except possibly in eastern Bavaria and Bohemia. The art of La Tene A betrays features that are quite distinct from the art of the periods preceding and following it. Indeed, without recognizing this phase as a separate cultural and chronological entity, one cannot obtain a thoroughly accurate conception of the genesis and development of art in the second Iron Age.

A brief summary of some of the leading features of Reinecke’s phases A-C is necessary in order to give an idea of the material cultures of the Celts1. Phase D (Late La Tène) falls outside the scope of the present chapter.

Phase A, or the Earliest La Tène Period. The funeral rite is generally inhumation under a tumulus. The most typical weapons are: the short sword or dagger; long swords; heavy, one-edged, curved knives, often of considerable size; bronze helmets. Horse-trappings and parts of chariots are not infrequently found. The chief objects of adornment are: Certosa, mask- and bird-headed fibulae; girdle-clasps based on the palmette motif or on zoomorphic designs; armlets, torcs and finger-rings of bronze and gold; orange and blue-green stratified eye-beads. The most important series of objects are the imported Greek bronze vessels, which date, in Greece, from the sixth century: beaked-flagons, stamnoi, round flat dishes with and without (rigid) handles, large bronze cauldrons with iron ring-handles, and the like. Barbaric imitations of the flagons are not unknown. Attic black- and red-figured pottery occurs on a few sites in France and southern Germany. In the west native pottery is best represented in Champagne. It also occurs in the region of Trèves.

The zone covered by the La Tène A culture extends from France to Bohemia, and for the sake of convenience may be divided into a Central, a Western and an Eastern area. The Central area consists of the Middle Rhine region and the adjacent districts, in which the celebrated series of burials commonly known as the ‘Chieftains’ Graves’ are found. Inhumation is the dominant rite in this group of burials, the dead being mostly covered by tumuli. These, with their Greek imports and their gold ornaments, are the richest graves of the whole La Tene A group. It is probable that the La Tène culture first flowered in this region. South of a line through Hagenau, Rastatt, Stuttgart and Ulm, the Late Hallstatt (D) culture persisted down to c. 400 BC. La Tene A, therefore, is not found to the south of this line, nor is it represented in Switzerland. Next in richness to the Middle Rhine area is the Western area, eastern and north-east France. The rich chariot-burials (Somme Bionne, La Gorge Meillet, Berru, etc.) of the Marne and neighbouring districts belong to this phase, but as Reinecke’s fourfold division of the La Tène period has still to be recognized by French archaeologists, the limits of La Tène A in the Western area have not as yet been established. The Eastern group consists of east Bavaria, southern Thuringia and the western half of Bohemia (particularly the south-west districts). The more or less contemporary Greek influence, so strongly reflected in the imports and the native art of both the Western and Central areas, and in the shapes oi the native pottery of the Western area, had far less effect upon this Eastern group. The pottery, though distinctive, betrays little connection with Greece in form, though in decoration an older Greek influence, derived from Upper Italy, is occasionally to be discerned. The influence of Upper Italy is far more strongly felt than in the Central and Western areas. Farther to the east sporadic finds occur, but most of them, if they belong to phase A at all, date from the very end of that phase. In La Tene A the potter’s wheel was introduced into Central Europe. But hand-made pottery is very common in this and the succeeding phase: most of the Marne pottery (in Phases A—C) is hand-made.

Phase B is better known as the Early La Tène Period. It reveals a greater uniformity of civilization than the period immediately preceding it. The funeral rite is inhumation, but large cemeteries of flat-graves take the place of the older and more isolated tumuli, though in some districts the latter persist. The greater part of the Marne flat-graves, and the flat-grave cemeteries of the Boii in Bohemia and the Helvetii in northern Switzerland date from phases B and C. The most characteristic weapon is the (usually) shortish, thrusting sword, with sheath ending in an open-work or trefoil chape. The most typical objects of adornment are the Early La Tene fibula (with foot bent back till it touches or nearly touches the bow) and a great variety of armlets and torcs, among the most characteristic of which are those with ‘buffer’ and ‘seal-top’ terminals. Girdle-clasps also occur. In north-east France we find polychrome and unpainted pottery, while in Brittany, where painted pottery is not found, vessels occur with rich curvilinear designs (St Pol-de-Léon and Plouhinec). All three types seem first to appear in the preceding phase. But the last named may have survived for a long time, if Déchelette is right in connecting it with a ware, dating from the latter part of the La Tène period, found at Glastonbury and other sites in England. In the Rhine area the pottery is not painted, though vessels with ‘graceful’ designs are found at Braubach and elsewhere. East of the Rhine and Neckar pottery is rare. Compared to A, phase B is very poor in imported bronze vessels. A bucket, decorated with drooping palmettes and tendril motifs, was among the objects in the celebrated find at Waldalgesheim. Another specimen was unearthed in the Gaulish cemetery at Montefortino. These finds show phase B to have been in existence by c. 400 BC. The Italian cemeteries are important: they yield characteristic B objects in association with antiquities of the fourth century BC, but Hellenistic pottery, and metal work of the third century, are almost entirely lacking. This would show that, apart from certain isolated localities (where it may have lasted longer), La Tène B ended c. 300 BC. According to Schumacher, this stage witnessed a development in agriculture in the low-lying districts of south-west Germany.

The culture of phase B covers a wider area than that of A. It spreads as far east as Budapest, and beyond; we find it in Moravia and to the north of the Central German mountain-chain in Silesia (as far as the Oder), Saxony and Thuringia. These were frontier provinces between Celt and Teuton. It was probably in this period that the Helvetii occupied Switzerland and the Boii and Volcae Tectosages Bohemia and Moravia. But the most important event was the Celtic invasion of Italy c. 400 BC and their settlement in Cisalpine Gaul. Towards the close of this fourth-century period Pytheas of Marseilles made his famous voyage.

Phase C or the Middle La Tène Period. In this, as in the preceding period, the predominant rite was inhumation under flat graves. There is no evidence for a displacement of population in Switzerland and the greater part of northern Bohemia: the large flat-grave inhumation cemeteries persist undisturbed into this phase. The same is true for the Marne area. Cremation burials occur, however, both in south-west Germany and the southern half of the East Alpine region, but they are exceptional. The most typical weapons are: long swords, with slightly tapered point and bell-shaped guard (the scabbards often finely decorated, with strangulated and heart-shaped chapes); broad-bladed spear-heads; large oval or oblong shields with ‘triggerguard’ bosses. The most characteristic objects of adornment are: the Middle La Tène fibula (with foot clashing the bow); a variety of bracelets, among which may be mentioned the type with hollow semi-ovoid or semi-globular protuberances (Nussarmring'). The band-shaped glass bangle first appears. Torcs are rare. Perhaps the most typical object of adornment is the chain girdle composed  of bronze (sometimes of iron and bronze) members decorated with red enamel. It was in this phase that red enamel was first substituted for coral by the continental Celts; the latter was employed in Central Europe from Hallstatt D to La Tène B. Wheel-made pottery becomes more common; the shapes of the vessels sometimes reveal Hellenistic influences. Painted vessels still appear in burials of the Marne group (Montfercaut). But pottery of phase C is rare in this region. Graves of the Middle La Tène period occur in Upper Italy. Phase C was marked by a great expansion of Celtic power to the east and the south-east: the Middle Danubian and Dacian areas, together with Carniola, etc., began to be fundamentally affected by the La Tène culture, while the defeat of Ptolemy Keraunos, the onslaught on Delphi, and the first incursion of Celts into Asia Minor happened in 279-8 BC. The area to the immediate north of the Central German mountain-chain was abandoned. A topographical study of the different sites reveals that the propensity for agriculture, already manifest among the Celtic peoples in La Tène B, was further intensified in this phase. Classical authors, too, speak of the Cisalpine Gauls as an agricultural people.

Perhaps the chief advance in civilization of the Middle La Tène period was the adoption of currency. If the inferior early coins, forming part of the hoard from Auriol, are really barbaric copies, they were struck by the pre-Celtic inhabitants of southern France, for their area of distribution lies to the south of the Celtic zone. Although, at an outside estimate, the earliest gold staters of Philip of Macedon may have been copied by Celtic peoples prior to his death in 336 BC, the minting of coins by the continental Celts can hardly have become very widespread before the Middle La Tène period.

The earlier Celtic coins were exclusively imitated from Greek prototypes and a number of the latter have been found in the Celtic area. It is not possible here to allude to more than a few types. If one may generalize, silver predominated at first among the eastern Celts (the middle and lower Danubian regions); in the Western zone gold was common. The first Greek coin to be copied in the West was the gold stater of Philip II, while the silver coinage was based on the later (fourth-century) issues of the Greek colony of Massilia and on those of Emporiae and Rhode. The coins of these three cities were copied by the Celts from the third to the first century BC. The gold stater, apparently, did not reach the western Celts by way of the Danube; it came from Massilia: the Massiliotes, having no gold coinage of their own, used the stater of Philip. To the east, the earliest type copied was the silver stater of the same king. As a rule, the coins were first copied by the Celts nearest the Greek centres: Celtic coins of the third and second centuries do occur in south-west Germany, but the majority date from the first. In Britain minting began roughly about 100 BC. Celtic copies of Greek coins which had been issued one or more centuries earlier are quite common. The extent of barbarization is such that it is not always easy to recognize the classic prototype. The Roman denarius was first copied in southern France at the end of the third century BC, but farther north not until the second century BC. The concave coins known as ‘Regenbogenschitsselchen' date from phases C and D.

II.

THE ART OF THE LA TÈNE PERIOD

The two chief formative elements in the art of the La Tène period are the survivals of certain Hallstatt influences and the new influence of more or less contemporary Greek art. The former reveal themselves in technique and in ornamentation, the latter for the most part in the ornamentation only. The older barbaric survivals are geometric in feeling, the new influence of the Greek art naturalistic. The zoomorphic motifs are mainly derived from the Greek element, though in the more easterly districts of the phase A group they must be regarded as survivals of a style known as the Upper Italian situla art. The influence of plant motifs, for the most part Greek in origin, becomes more and more prominent: witness the derivatives, mainly geometric, of the palmette, the free tendril and other motifs. One should also mention the scroll patterns with thickened ends and the trisceles.

Viewed as a whole, the art of the continental Celts in the La Tène period may be regarded as a period of gradual decline. It attains its zenith almost at once—in phase A—but in each succeeding chronological subdivision it becomes more barbarized and conventionalistic. In the British Isles, this process was reversed, the zenith being reached in the later stages of the La Tène period. On the continent, the decline was perhaps due to a falling-off in the number of southern imports, perhaps also to a corresponding decline in the art of Greece itself. The cradle of the true La Tène style is probably to be found in the Middle Rhine region, for here lay the point of contact between the two above-mentioned formative elements. At all events, it is in this region that the Greek imports are best represented.

Reinecke was the first to understand the development of La Tène art. He observes that, although in Celtic art of the fifth century BC we occasionally encounter fairly faithful representations of Greek types, we find for the most part, both in plant and figural motifs, that a conscious stylizing of the Greek models has taken place. Certain unessential details are exaggerated, others suppressed, while disintegrations and combinations with other motifs are not infrequent. Finally, from these degenerate and barbaric reproductions, new forms arise. By phase C the classical tradition was scarcely recognizable: the art had now become purely Celtic.

Two divergent tendencies have been traced in La Tène art. The first, an inheritance from the Hallstatt period, is mostly in evidence during phase A. It consists in a ‘mechanical repetition of congruent motifs,’ little regard being paid to the design as an organic whole. The second, due to Greek influence, is found side by side with, and survives this repetitional tendency: the ornamental form—which is often highly complex—is treated as an indissoluble unity, a complex organism. As the La Tene period proceeds, the plant motifs become less representational and more geometric in feeling, though this does not imply that the arrangement of the design was less organic. One important characteristic of La Tene art remains to be mentioned: the practice of embellishing the whole ornamental field with primary and complementary curvilinear patterns (e.g. the helmet from Berru, the scabbard from Lisnacroghera). This entailed the abolishing of the differentiation between design and background, a tendency foreign to Greek classical art of the best period, and one which contributed also to the loss of the originally naturalistic character of plant motifs in Celtic art.

It remains to describe the channels through which Greek influence reached north-east and eastern France, and the middle Rhine area. The main route seems to have been Massilia, the Rhone and the Saône, the chief exports from the south being metal vessels, Attic pottery, and wine. The peculiarities of the Eastern La Tène A area have been discussed above. The most important route for this region passed over the Brenner. Scythian influence, though possibly of importance later, can hardly have played a part in the genesis of the La Tene style; even for later times, the possibility of a parallel development must be taken into account.

III.

EARLY TRADERS AND CONTACTS WITH THE GREEK WORLD

The first known classical author to mention Celts is Hecataeus of Miletus (r. 540—475 BC). His works are lost, but two fragments have survived that are of interest to us. In the first he mentions ‘Nyrax, a Celtic city,’ but of this Celtic city nothing more is known. In the second he speaks of ‘Massilia, a town in the land of the Ligurians, in the region of the land of the Celts, a Phocaean colony.’ The truth of his statement is borne out by the evidence of archaeology: the Ligurians, in his day (La Tene A), dwelt in the regions around Massilia, and the territory of the Celts lay to the north of the Ligurian zone. The redaction of the pseudo-Scylax periplus (c. 335 BC) mentions the presence in southern France of Ligurians, but not of the Celts. Yet when Hannibal marched from Spain to Italy in 218 BC, the only peoples mentioned as encountered in southern Gaul are the Celtic peoples Archaeological finds show that they can hardly have conquered that region prior to c. 300 BC.

After Hecataeus, the next author to mention them is Herodotus. He tells us that the Danube rose ‘in the country of the Celts’ near the town of Pyrene; he then adds that the Celts dwelt beyond the Pillars of Hercules, next to the Cynetes, who were the westernmost people of Europe. These remarks he reiterates in Book IV, 49. Attempts have been made to prove that his second statement is not essentially connected with his first, and that there may have been another town called Pyrene which was not situated in the Pyrenean area. But on the whole, the facts suggest that Herodotus blundered over the region in which the Danube rises, although he is correct in saying that the source of that river lay in Celtic territory, that there were Celts in the Pyrenean area and that Celtic peoples were already settled in Spain. During the fifth century BC, the Celts were in possession of all those regions. Timaeus (352—256 BC) alludes to the Celts as dwelling by the Ocean.

It is highly probable that the Greeks derived their first knowledge of the Celtic peoples from traders. It has been observed, in connection with the second of the two fragments of Hecataeus, that there was a lively trade in Greek bronze vessels and pottery between Massilia and the Western and Central areas of the La Tène A zone, a trade in which the Ligurians of south-eastern France appear to have acted as middlemen; even supposing the Greeks were not in direct touch with these Celtic peoples, they must have learnt of them through the Ligurians1.

The Greeks also seem to have derived their first knowledge of the more westerly and northern regions, including the British Isles, through commercial channels. It is probable that one of the chief materials in this sea-traffic was tin, though Irish gold may still have been of importance. Much has been written about the Cassiterides (‘the tin islands’), but owing to insufficiency of evidence, the definite knowledge which has accrued from these writings is slight. One scholar holds that kassiteros, the Greek for tin, is of eastern origin; but it seems more probable that in Arabic, for instance, the word for tin (kasdir) was borrowed from the Greek. Arguments for a Celtic origin of the word are also inconclusive. In spite of manifold attempts at identification, we are still unable to localize the Cassiterides. Whatever the origin and first significance of this word may have been, it is not improbable that it came to be used of any source (or sources) in western Europe from which tin was exported to the classic world. The most important of these lay in Cornwall, Brittany, Spain, and the adjacent islands. With the Bohemian deposits which figured largely in trade during the Early Bronze Age, we are not concerned. How great a part Cornwall played in the tin traffic, is hard to say, but there is reason to believe that the ores of this region were exploited in the Bronze Age. Cornish tin was apparently exported c. 300 BC, and it is possible that trading stations were at that time established in Cornwall by Celtic peoples from the western parts of the Spanish peninsula. If, at an earlier date, Cornish tin reached Tartessus, it is probable that the Oestrymnians acted as middlemen in the traffic, for there is no evidence for supposing Cornwall to have been in direct touch with that port. The deposits in Brittany and Spain were apparently worked in very early times, and the Breton tin probably reached Tartessus by sea. But, despite recent criticism, this metal may also have come to the south, by overland routes through Gaul, before as well as after 300 BC.

Greeks were already in touch with Tartessus during the seventh and sixth centuries BC. We know from Herodotus that Colaeus of Samos, about the year 620 BC, chanced by accident upon the port of Tartessus at the mouth of the Guadalquivir, in southern Spain, and brought back a rich cargo, and that the Phocaeans in their long penteconters established relations with that port.

Avienus’ poem Ora Maritima is of great interest in this connection: although written in the fourth century A.D., it is based upon earlier sources. It has been shown that chief among these is a lost Massiliote Periplus dating from the sixth century, but Ephorus and other early Greek authors were also employed; while, apart from additions of his own, Avienus seems to have used later Greek sources1. In view of the composite nature of the poem, it should be used with caution when chronology is involved, and checked with other evidence. The Periplus falls into three parts, one dealing with the navigation between Tartessus and Massilia, another with the Atlantic coasts between Tartessus and the Oestrymnides, while the third gives an account of Ierne, Albion and, apparently, the Frisian coast, an account probably derived by the Tartessians from the Oestrymnians, which the Massiliotes in their turn derived from the Tartessians.

We find a group of names connected with the latter people. Scholars are not in agreement in their identifications. But most authorities are of the opinion that the northernmost Oestrymnis (there are two in the poem, one being situated in the Spanish peninsula) is Brittany. There are two days and nights sail between the Oestrymnides and Ireland; so the former can hardly be—as some have supposed—the British Isles, which are mentioned under other names in the poem. The Oestrymnides are probably small islands off the coast of Brittany. They are mentioned as being rich in tin and lead3. The tin deposits of the Breton mainland we know to have been exploited from a yet remoter period. We are told that the Tartessians traded with the Oestrymnides, as later did the Carthaginians

The sacra insula mentioned is undoubtedly Ireland. The name, if one may say so without offence, is probably due to a mistake, the early form, Iwerio, being confused by the Greeks with ‘holy.’ The insula Albionum  refers to Great Britain, of which it is the earliest name. This is regarded by some as a pre-Celtic word. It occurs, however, in Celtic sources. In the earlier of these, it is applied to Britain generally; in the later, to the districts north of the Clyde. It survived in the expression ‘the Kingdom of Alba’ and some authorities hold it to be derived from a Celtic word meaning ‘white,’ ultimately connected with the Latin albus. At all events, its pre-Celtic origin is not proven.

There is an allusion to ‘a Northern Liguria’ whose inhabitants had been driven out by the Celts. This region Schulten places on the Frisian coasts, and he derives the passage from the Periplus. It is very surprising to find the Ligurians so far north and, until satisfactory evidence of a philological character can be produced in support of this passage, it appears to the present writer safer to regard it as due to some misunderstanding1. At all events, we should remember that even if the above passage was derived from the Massiliote Periplus, its author probably had the information at second hand from the Tartessians, who learned of it from the Oestrymnians.

Shortly before the end of the sixth century the Straits of Gibraltar fell under Carthaginian control, and the connection of the Greeks with Tartessus and the Atlantic sea route was severed . Some regard this severance as the chief cause of the opening of a land-trade across Gaul by the Greeks, and the establishment of such a route they date to the fifth century. Objections to this theory, based upon lack of numismatic evidence, are hardly valid. At the same time, while we have every reason to believe that trade relations at that time extended along the Rhone-Saone route between the Greek colonists and the Celts (with the Ligurians acting as middlemen), it is not so easy to say whether the more westerly route—from southern Gaul to Corbilo (at the mouth of the Loire) and thence to Cornwall—was in use at this particular time or not. In the present writer’s opinion, the route from Ictis (in Britain) to the mouth of the Rhône may have been opened before 300 BC. At all events some sort of trade connection between the Greek colonies in southern France and the Marne area was established by 500 BC.

So far as the Greeks were concerned, the gate to the Atlantic was barred by the Carthaginians from the last decade of the sixth century to the time of Pytheas, and even the latter’s voyage was followed by no immediate resumption of traffic along the old sea-route. Pytheas of Massilia was the first Greek known to have visited these islands. His celebrated voyage probably took place about the years 325-3 BC. We cannot say for certain whether its purpose was scientific. Unfortunately his account of it has perished, but fragments of it are quoted by Strabo, Pliny and other writers. Polybius and Strabo regard him as a liar and a charlatan, but time has shown their opinions of him to be ill-founded. He sailed to Brittany, for he notes that the Ostimii dwelt on the headland of Kabaion, perhaps Pointe du Raz or Pointe St Mathieu. Some hold that the Oestrymnians of Avienus are to be identified with the Ostimii of Pytheas, the former being the corrupt form. Pytheas speaks of an island, Uxisama, which is alleged to be the island of Ouessant. These identifications are somewhat hazardous, even though some of the places are mentioned by the geographer Ptolemy: for all we know, the latter may have attempted to localize names of places ultimately derived from Pytheas. Pytheas apparently visited the stannaries of Cornwall— he mentions Belerion (Land’s End)—and it is not improbable that Diodorus’ information concerning the route from Ictis to southern France was derived from him through Timaeus1. If Pytheas sailed eastwards up the Channel to Kent, his otherwise unaccountable statement that the latter was some days’ sail from Keltike is easily explained. He states that he travelled all over the accessible parts of Britain, and he would seem to have circumnavigated the island2. We probably owe to him the name Orcas, the northernmost cape in Great Britain (Dunnet Head ?).

Most spectacular of all was his report of the remote island of Thule, which he says was six days sail northward from Britain, an estimate which favours Thule being ‘the island of Scandza’ (i.e. Norway), and not one of the Shetlands which are too close. In view of the mention of the midnight sun, it is not likely that Pytheas’ Thule can be identified with Iceland, as that phenomenon is only visible from one or two points in the extreme north of that island.

From what has been said above, it is more than probable that the ancients derived the main part of their early knowledge of the Celts from traders and sea-farers, though it is open to doubt if, prior to the time of Pytheas, Greeks were in direct contact with the British Isles.

IV.

THE EARLY HOME OF THE CELTIC PEOPLES

The area of differentiation of the Celtic language probably lay in the Upper Danubian region. Both archaeological and linguistic evidence testifies to the presence of Celts in that area at an early date, while certain regions in eastern France belong culturally to the Upper Danubian province (Württemberg, etc.). But did the early Celtic home stretch farther afield ?

There is little reason for placing the cradle of the Celtic peoples in Gaul. The origin of the French river-names is against such an assumption. River-names (especially those of the larger streams) are apt to be older than names of towns. In France, many Celtic names of towns are found, very few of rivers; the majority of these last are of pre-Celtic origin. In south-west Germany, Celtic river-names are very common and they also occur, though somewhat less frequently, in north-west Germany. As the Celts seem to have abandoned this latter region at a comparatively early date, the occurrence of Celtic river-names there is not without significance. The evidence of place-names and river-names favours the view that the early home of the Celts lay, for the most part, to the east of the Rhine; and that the greater part of Gaul was not conquered by them until later. How far eastward their territories originally extended, it is not easy to say.

A recent hypothesis deserves special consideration, for it is more comprehensive and coherent than many of those hitherto advanced. The proto-Celts are identified with a people known by their pottery as the zone-beaker people, who inhumed their dead beneath tumuli, and whose territory, at the end of the Stone Age, extended from Central France to Bohemia. From ornamental features on the bell-beaker ware, the later chip-carving technique was evolved1. The development of civilization in the above zone-beaker area advanced practically without interruption until the end of the Bronze Age, when an Alpine people appeared. They brought with them the Urnfield Culture (Hallstatt A), Villanovan forms (destined to have a persistent influence in south-west Germany), the rite of cremation, and the disposal of their dead in large cemeteries of flat graves (or at all events under very low tumuli). For a time they were dominant over the proto-Celts, but the older population re-asserted itself and finally absorbed its conquerors. Inhumation again became prominent. From this time, once the absorption of the Urnfield Peoples was complete, it is permissible to speak of the population of southern Germany and a large part of France as being Celtic. North of this area, in the Marne district, Belgium, south Holland, the lower Rhine valley and the region to the east of it as far as the Weser, lived a different people, who were only externally affected by the Urnfield Culture and, unlike the majority of their more southerly neighbours, practised cremation. These are regarded as a more backward element which remained proto-Celtic, when the peoples farther to the south had become fully-fledged Celts. In both areas tumulus-burials prevailed. Such, in outline, is Rademacher’s theory concerning the early home and pre-history of the Celts.

Although there is much to be said for the theory as a whole, certain points are open to criticism. First, in the light of the distribution of river-names, the original home of the Celts seems to extend too far to the west. On the other hand, a few Celtic mountain-names and river-names do occur in south-west France, the most important of these being Cebennon, the old form of Cevennes (cf. Chevin, the name of a hill in Yorkshire). But these names are very rare, and how far back they take us is hard to say. When these are the names of rivers, they are not those of large streams, such as the Weser and the Rhine. In view of this, it is perhaps safer not to identify the proto-Celts with the late zone-beaker peoples, but to regard them as forming an element or elements in the late zone-beaker complex. Second,  some authorities hold the Urnfield peoples of south-west Germany to be of Illyrian stock. If it is true that isarno-, the primitive Celtic for ‘iron,’ was borrowed from the Illyrian, it would appear that the Celts and Illyrians were in close contact possibly early in the Iron Age. The second of these objections is not so important as the first, for the nationality of the Urnfield peoples is still an open question, and the Illyrian origin of isarno- is far from being definitely proved.

One fact seems to have been thoroughly established: the continuity of the Lower Rhenish Tumulus Culture. This culture, the result of a fusion between west European, Central-German and purely Nordic elements, developed more or less undisturbed from the end of the Stone Age until towards the close of the Hallstatt Period. Peoples from the Nordic zone began to appear in the Lower Rhenish area during the middle Hallstatt phases, and eventually, late in the sixth century, drove the greater part of the older inhabitants to the south and to the west. The period which immediately followed is an obscure one. But the remnants of the Celtic population to the east of the Rhine seem to have mingled with the Teutonic invaders, and from this fusion of Celt and Teuton the Belgic people came into being. There is little evidence for the hypothesis that the language of the Belgae was Germanic. But if the latter spoke some form of Celtic, by Caesar’s time at all events, it must at least have shown dialectical differences from that spoken by the Celts. During La Tène C, the Belgae began to push westward and, at the end of that phase, succeeded in bringing about the downfall of the four hundred year-old Marne culture. The period c. 100-60 BC witnessed the gradual penetration of south-eastern Britain by Belgic peoples.

The question of ‘Celts’ and ‘Galatae’ is a very disputed one. It has been suggested that the Red Celts conquered the Black Celts, who were the early population, and probably imposed their language upon them. If this is true, the Black Celts may not have been Celtic at all. Both from anthropological and linguistic standpoints, it is perhaps safer to dispense with the terms Black and Red Celt: they invite generalizations which, far from simplifying matters, mislead and obscure the true issues.

An examination of the skull-measurements from most of the Culture-groups within the Celtic area reveals the population to have been physically of a very mixed character, and shows that it is impossible to employ such an expression as ‘ the Celtic race.’

Among the ruling classes buried in the tumuli of Burgundy and Lorraine on the one hand, and the Alaise-Württemberg group on the other, long-headed elements have been detected. The former are orthognathic, the latter prognathic. These elements, which are regarded as intrusive, have been termed proto-Galatian and proto-Germanic respectively. Brachycephalic skulls, however, have been found in both the above groups. The Alaise-Württemberg tumuli often contained the remains of tall individuals. Farther to the east, in Bavaria, brachycephaly seems to have predominated.

The peoples of the Marne culture were a mixed population in which long-headed elements preponderated. Piroutet holds that among the peoples of the Marne culture, there was a Galatian element which came into that region from Normandy shortly before the close of the Hallstatt period, an element which soon grew to be the dominant one. Camille Jullian identifies the Galatae with the Belgae, but some of his arguments are based on a rather too arbitrary interpretation of classical authorities.

The majority of Greek and Roman authors did not clearly distinguish between Celt (in the narrow sense) and Galatian, although among the barbarians who invaded Italy and Greece in the fourth and third centuries BC, there was a tall, fair element which caught the eye of their southern adversaries and which is mentioned by more than one classical writer. This element many hold to have been Galatian, but in view of the complex physical anthropology of the Celtic-speaking peoples, one should be wary of laying too much weight upon generalizations of this nature.

There appears to be no evidence for supposing the Galatae and the Celts to have spoken fundamentally different languages, any more than the Irish and Welsh of later times; indeed, we have little reason to believe the original significance of the terms ‘Galatae’ and ‘Celt’ to have been more than northern and southern Celtic-speaking peoples respectively. Diodorus, whose account is largely drawn from an earlier source (possibly Posidonius) and Strabo support this view. Each, doubtless, had its peculiar characteristics, but a merging of the two groups was already taking place at the turn of the sixth and fifth centuries. In some regions it is probable that the Galatae may have asserted themselves at the others’ expense. By the fourth and third centuries, this process of fusion seems to have reached a sufficiently advanced state to prevent the classical adversaries of these barbaric invaders of Italy and Greece from clearly distinguishing between the two elements. If Chadwick is right, we find a somewhat analogous case in the Angles and Saxons at the time of their invasion of this country.

A passage from Ammianus Marcellinus remains to be discussed. It is apparently derived from Timagenes who lived in the reign of Augustus. ‘The Druids say that a part of the population (of Gaul) was really indigenous to the soil, but that other inhabitants streamed in from the islands on the coast, and from districts beyond the Rhine, having been driven from their old abodes by frequent wars, and occasionally by inroads of the raging sea.’ We have here an old native tradition, preserved by a learned priesthood, which as evidence far outweighs the late and largely fictitious Ambigatus story as given in Livy. As to the indigenous element, it is not at all unlikely that this was the pre-Celtic population, which had gradually become Celticized, though it is possible that the Druids may have included the Celtic peoples established in eastern France at a comparatively early date under this heading. The reference to the islands on the coast is not easy to explain. Greek writers of the fourth century hint or actually speak of the Celts’ proximity to the sea. The reference to part of the inhabitants coming from beyond the Rhine is of great interest, for it shows native tradition to be partly in agreement with conclusions deduced from philological and archaeological evidence: it was mainly from the east of the Rhine that the Celtic-speaking peoples expanded into Gaul.

It may be objected that this expansion took place at too remote a date for it to be remembered by the Druids. While not altogether agreeing with this view, one should remember that a large migration from the Lower Rhine area to the west and south-west occurred in the sixth century BC the Belgic invasion of north-east Gaul (also in part from beyond the Rhine) took place at a still later date; and similar movements occurred in regions farther to the south. The allusion to Celtic invaders being driven into Gaul by frequent wars may partly refer to the pressure upon the Celts caused by the expansion of peoples from the Nordic zone, whom many authorities regard as Teutonic.

V.

 CELTIC MIGRATIONS

South-western France and the Spanish Peninsula. With the possible exception of an early invasion of Britain, the first important movement of the Celtic Migration Period seems to have proceeded in a south-westerly direction and to have reached south-west France and the Spanish peninsula in the later part of the sixth century. They probably entered the latter region through the passes of the western Pyrenees. Our earliest source for the ethnography of Spain and Portugal is Avienus’ Ora Maritima. If the relevant passages are derived from the Massiliote Peripplus, Celtic peoples were already settled in the west and in the highlands of Castile, at all events within the last quarter of the sixth century.

An important contribution has recently been made on the Celtic invasion of south-west France and the Spanish peninsula1. In south-west France we find a group of tumulus-cemeteries extending from the Mediterranean to the Atlantic coasts, especially on the Ger plateau and the French Pyrenees region. They are not earlier than the last phase of the Hallstatt period (Reinecke D) and, in part, probably later. With their solliferrea, their peculiar antennae-hilted iron daggers and certain types of fibulae, they show marked affinities with a contemporary and somewhat later culture in Spain, although the burials of this Spanish culture are not under tumuli. On the other hand, the degenerate types of Hallstatt pottery yielded by the south-west French group, and, above all, the burial customs and the structure of the tumuli themselves, bear a most remarkable resemblance to the Hallstatt Tumulus Culture of the Lower Rhine valley and the regions adjacent to it. Further, Rademacher sees a connecting link between these two groups in the unique and isolated burial-ground at Haulzy (Marne). Déchelette is mistaken in describing the rite in the later graves of this site as inhumation; the burials are cremation burials throughout (cremations under tumuli). The earlier graves (1—69) date, typologically, from the Middle Hallstatt period. The structure of the graves and much of the pottery they contained testify to a migration from the Lower Rhine region. The later graves (70—78) mark the transition from the Late Hallstatt period (D) to La Tène I. They have yielded antennae-hilted daggers which closely resemble a ‘post-Hallstatt’ Spanish variant of that type of weapon.

These and other phenomena have led Rademacher to believe that, under pressure of peoples expanding from the Nordic area, the Celts in occupation of the north-west German zone moved westward and southward. This pressure began c. 900 BC, possibly earlier. In the sixth century a great Celtic migration took place from the Lower Rhine region, which, after various vicissitudes, reached south-west France. Here part of the new-comers settled in the districts above mentioned. But the more adventurous elements, crossing the Pyrenees, occupied those regions in the Spanish peninsula of which we have already spoken. Although adopting certain new customs, they remained in contact with their kindred in south-west France, influencing them and being influenced by them in turn. Cut off from the main body of the Celtic world, these ‘Spanish’ Celts during the fifth and fourth centuries developed what has been termed ‘a post-Hallstatt’ civilization of their own. This Celtic-Iberian culture is described below in chap. 24. It is to be noted that La Tène culture was not well represented in Spain until it had entered upon its later phases.

Italy. Livy repeats a fairly detailed account of an early Celtic invasion of Italy and the circumstances which led up to it: during the reign of Tarquinius Priscus at Rome, Ambigatus, supreme king of Gaul, found that country to be so over-populated that he summoned his sister’s sons Segovesus and Bellovesus; lots were drawn and the former led part of the superfluous population eastward into the Hercynian Forest, while the latter, at the head of a large host, invaded Italy. This is the story, stripped of its details. It has been proved to be a tissue of inaccuracies. Literary and archaeological evidence shows that Livy’s dating (c. 600 BC) is two centuries too early. With the exception of the Senones, none of the peoples mentioned as comprising Bellovesus’ host ever appear to have settled in Italy—the Senones are mentioned in the next chapter as being the last band of Celts to settle in that peninsula! Further, we have no evidence for a central authority being established among the Celts at so early a date; while the information given by Livy in this chapter on the political geography of the Celtic world is derived from Caesar’s Gallic War and in no way represents the ethnography of Gaul in the fourth—far less the sixth—century BC. Finally, some scholars regard the names of the principal characters in this chapter as fictitious. It is clear, therefore, that the story told by Livy in practically all its details is historically worthless. We shall see later whether there is a kernel of truth embedded in this mass of anachronisms.

It is difficult to decide by what pass or passes the earlier Celtic invaders of Italy crossed the Alps. Livy, whom as far as details are concerned we have found to be untrustworthy, makes the first invaders come by way of Mont Genèvre in the western Alps1. But archaeological finds and references from other classical authors show that the regions to the west of that pass—in fact most of south-eastern Gaul—were Ligurian territory until the third century, while the Celtic graves in the western districts of Upper Italy date from the later phases of the La Tène period. In view of this we are probably justified in ruling out all the passes of the western Alps from the Col di Tenda to the Little St Bernard, at least so far as the earliest Celtic invasion was concerned.

Before inquiring into the claims of other passes, mention must be made of climatic conditions. In the Early Iron Age, a deterioration of climate set in which considerably affected the higher Alpine passes (those over 6500 ft. high); the glaciers increased and the forest level subsided. Hallstatt and La Tfene pass-finds are much rarer than those of the Bronze Age. The two worst periods in this phase of cold wet weather seem to have been c. 850 and 500—350 or 300 BC. The Celtic invasion of Italy took place c. 400—390, and thus falls into the second of these two periods.

Livy speaks of some of the early invaders crossing the Great St Bernard. Although 8111 ft. high, this pass does not, like other passes of an altitude over 6500 ft., appear to have been affected by the ravages of glaciers. Finds of the La Tène period have come to light on the Great St Bernard, but they belong to the later phases. This tallies well with a passage in Polybius. He tells us that the Gaesati (a caste of Celtic warriors or mercenaries), lived ‘about the Alps and on the Rhone,’ and that they crossed those mountains in 225 BC and fought with the Cisalpine Celts against the Romans. The phrase ‘about the Alps and on the Rhone’ points to their being at that time settled in the Swiss canton of Wallis, a region in which finds of the La Tene period occur not infrequently. If they did come from Wallis, it is probable that they crossed the Great St Bernard. But evidence for this pass being traversed by the earlier Celtic invaders is at present lacking. There are no signs of any early case of crossing the St Gotthard (6936 ft.).

Whether the Bernardino (6769 ft.) was of importance in this respect cannot be determined. A number of cemeteries occur in the Ticino and Mesocco valleys, dating from the La Tène and ostensibly from the Hallstatt periods. These districts were extremely conservative, different types occurring in the same grave which in other districts would be four centuries apart. In view of this, it is impossible to say whether the graves which are typologically of an Early or Middle La Tene character belong chronologically to these periods. Reinecke is inclined to assign a late date to the whole group. Until the tangled chronology of these cemeteries has been unravelled, it is difficult to tell whether the earliest invaders crossed this pass or not.

We have little evidence for the routes from Como over the Splugen (6946 ft.) and Julier (7504 ft.) passes being used in the Early Iron Ages. The finds, in this area, are far less numerous than in the Ticino-Bernardino district, and the assumption is that the routes in question were less frequented, possibly on account of their being over 6500 ft. high and consequently more liable to be affected by the deterioration in the climate which is mentioned above.

The chief pass left for discussion is the Brenner, for it is hardly likely that the first Celtic hosts to invade Italy came by a more eastern route. One of the main trade-routes linking Italy with northern Europe went over this pass. It is of a sufficiently low altitude (4495 ft.) to have been unaffected by the above-mentioned change in climate. Some authorities hold that the close connection during La Tene A between Upper Italy and east Bavaria and Bohemia suggests that the Celtic invasion of Italy came from the Upper Danubian region rather than from France. If so, the normal route would have lain through the Tyrol, over the Brenner and down the Adige. Others contend that the dearth of pure Celtic La Tene finds in the Tyrol points to the first invaders not having come by that pass. But this may only mean that their transit through the Tyrol was a rapid one1. Reinecke, moreover, sees a hint of a catastrophe having taken place in the Tyrolean area c. 400 BC: the sequence of bronze hoards comes to an abrupt end, just at the time when the Celts first descended upon Italy. The fact that Venetia was not conquered by the Celts can hardly be used as evidence against their having come by this route: the point at which it debouched upon the Upper Italian plain barely touched the south-western corner of Venetia, the latter being protected from inroads from the north-west by the Alps. Nevertheless, strong Celtic influences made themselves felt at Este from c. 400 BC onwards, and this fact favours the assumption that the Celts and their Atestine neighbours were in close contact with each other from the time of the arrival of the former peoples in Italy. Finally—perhaps the most important testimony in favour of the Brenner—the earliest Celtic graves (phase B) of Upper Italy are confined to the eastern half of that region.

Of the four passes known to Polybius three lie in the west Alpine zone. Although these may have been important so far as local trade was concerned, we have seen that the evidence is against the earlier invaders having come from this direction. The fourth many identify with the Brenner.

Let us turn once more to Livy. It has been recently suggested that the Earliest La Tène population of Bohemia and east Bavaria, etc., was not of Celtic but of the old Illyrian stock, and that the Celts first appeared in those regions at the beginning of phase B. Others hold that the Celts were already in the Hausrück area by the fifth century. Whichever view is adopted, there is every reason to suppose that a change of population took place in east Bavaria, Bohemia and Moravia at the beginning of phase B (400 BC, or possibly a few years earlier). The new-comers, the people of the flat-grave inhumation cemeteries, were indubitably Celtic and seem to have come from the west. If the latter surmise is correct— seeing that Bohemia and the regions to the east and west thereof lay in the Hercynian area—there is a certain amount of truth in Livy’s account; two important Celtic migrations occurred about the same time: one moved eastward into the Hercynian region (Segovesus), the other southward into Italy (Bellovesus). We might argue from the fact that a Celtic people called the Boii dwelt both in Bohemia and the Bologna region that these movements to the east and to the south branched out from a common centre. Where that centre lay, it is hard to determine. Can it have been Gaul? We must remember that in Livy’s day the centre of gravity among the continental Celts lay in that province, but at an earlier time it was farther to the east. We know that a westward drift of Celtic peoples into certain regions in Gaul happened during the later Hallstatt and earliest La Tène phases. Is it likely, then, that they retraced their steps so soon ? Perhaps we have a hint of the truth in Appian, who tells us that a great part of the Celtic invaders of Italy came from the Rhine.

The south-eastward Expansion. Although later in time than the migrations of which we have just been speaking, the southeastward movement of the Celts was far greater in extension, for its ultimate waves spread beyond Europe into Asia Minor. The first recorded appearance of Celts in Greece was in 369—8 BC, when Dionysius I sent Celtic and Iberian mercenaries to aid his Lacedaemonian allies against the Boeotians in the Peloponnese. If we can trust Justin, Dionysius first came in contact with the Celts about twenty years earlier: when fighting against the Locrians and Crotonians in southern Italy, an embassy came to him from the Celts who had shortly before sacked Rome offering to fight for him or to harass his enemies in the rear In this connection it should be noted that Celtic graves of the fourth century have been found as far south as Apulia.

In 335 BC, when Alexander the Great made his expedition against the Triballi—then in Bulgaria—and their neighbours, the Illyrians, he received deputations from all the Danubian peoples. It was on this expedition that the embassy came to him from the Celts of the Adriatic. Unfortunately more precise indications of their whereabouts are lacking. Alexander’s expedition against the Triballi and the Illyrians was doubtless undertaken to prevent a repetition of the Triballian irruption of 376—5. From early in the fourth century, there had been an eastward drift of Illyrian and other tribes like the Triballi. This drift has been attributed to the upheaval caused by the arrival of the Celts in the region of the Save and the Middle Danube, and to their penetration into the lands between the Dinaric Alps and the sea-board from Istria to the river Narenta. It was possibly from the latter region that the Celts came who visited Alexander in 335. But the Celts did not take root in these Illyrian regions1: their penetration was purely of a warlike nature, and the native elements soon re-asserted themselves.

Archaeological evidence hardly supports the view that the Celtic occupation of the Middle Danube-Save region happened early in the fourth century BC. Objects from the counties of Turócz and Borsód, and the earliest finds from Munkacz, show that roughly about 400 BC the Celts had spread along the Upper Danube and beyond into the Upper Tisza (Theiss) area. The objects from Silivas are cited as evidence for the advance of Celtic bands up the Maros into Transylvania as early as that date2. Otherwise, very few objects of a La Tène B character have come to light in Transylvania. Indeed, apart from the finds mentioned above, there appear to be hardly any La Tène sites in this eastern area earlier than phase C, that is to say prior to c. 300 BC. In the present state of our knowledge, while not denying that Celtic raids happened in Illyria, Pannonia and Dacia during the fourth century or that isolated Celtic communities may have been dotted about here and there, it seems that the eastward drift of Illyrians and Thracians was mainly due to pressure from fugitive peoples, dispossessed of their lands by the Celts, rather than to immediate contact with the Celts themselves. Judging from the finds, it was not till the Middle La Tène period (c. 300-100 BC) that the area between the Danube and the Tisza together with the regions to the west and south of it became really Celticized. Dacia, too, began to be fundamentally affected by Celtic culture about this time; and, finally, it was during this phase, in 279 BC, that the Celtic hosts first appeared in Macedonia and then turned south and attacked Delphi. They were beaten off and soon after suffered defeat and dispersed. Some settled in Asia Minor, some in the Thraco-Illyrian area, while others are reputed to have returned to their old home.

VI.

CAUSES OF CELTIC MIGRATIONS

The Celtic migrations which have been described above were doubtless due to more than one cause. Justin ascribes the invasion of Italy to the quarrels of the Celts among themselves. This appears to be connected with the theory of over-population, referred to in the Ambigatus legend. But the overpopulation theory, though important if applied locally, is apt to be exaggerated. It is far more probable that there was a large warlike element that preferred living by plunder to a peaceful and industrious existence. We find this too among the Teutonic peoples of later times in an even more marked degree, for the latter had little inclination to till the soil, whereas the Celts even during La Tène I (Déchelette) practised agriculture extensively.

The attractions of the south proved a great magnet to these warriors bent on adventure and plunder. But the chief cause was apparently an external one: the pressure upon the Celts of Teutonic peoples. When Caesar arrived in Gaul in 58 BC, he found that a large body of Teutonic and other invaders had crossed the Rhine in the region of Mainz and Worms and were at war with certain sections of the Gaulish community; a few years later he drove back the Usipetes and Tencteri who had crossed the Lower Rhine and invaded Gaul. Caesar’s conquests put a stop to this encroachment upon Gaulish territory of peoples from beyond the Rhine. But there is reason to believe that a process of expansion of elements from this quarter at the expense of the Celtic peoples had a long history behind it.

In southern Sweden and Denmark, at all events from the start of the second phase of the Montelian Bronze Age (c. 1500 BC), we find a more or less ‘self-contained’ and indisputably ‘continuous’ culture, which, period by period, as the Bronze Age progressed, advanced into the north German area. For this people and their culture the neutral term ‘Nordic’ is here employed. There is, however, no archaeological evidence in favour of an invasion of the Nordic zone by a people who brought with them the Teutonic language subsequent to 1500 BC: indeed from that date, if not before, the evidence all points to the movements of peoples issuing from that zone and not entering into it.

Three stages have recently been distinguished in which the peoples of this Nordic culture encroached upon the Lower Rhenish area and north-west Germany1: c. 900—800 BC they advanced to the west of the Teutoburger Wald and occupied the valley of the Upper Lippe; c. 800-700 they made themselves masters of the right bank of the Rhine as far as the Duisburg region, while, not later than c. 500 BC, they had not only advanced south to the Siebengebirge but had crossed the Lower Rhine and the Maas and driven the Celts out of that area. A good deal depends upon the chronology of the Harpstedt culture which expanded from the Ems-Weser district to the Rhine and beyond. It is one of the leading features of these Nordic invaders. The influence of this culture upon the Lower Rhenish pottery dating from the transition between Hallstatt C and D (shortly before and after 600 BC) is very marked. But it probably originated as early as c. 900 BC.

There is little reason to doubt that this pressure from the north brought about the downfall of the Lower Rhenish Tumulus Culture. The second of these three stages seems to have caused some displacement in the population of the latter area. At all events, the earlier graves at Haulzy (Middle Hallstatt), both in their structure and the pottery which they contain, indicate a movement from that region. But Haulzy is an isolated phenomenon and these movements did not assume serious proportions until later. In the later graves at Haulzy (Late Hallstatt and La Tène I) and in the La Tène I burials of the Marne culture, we find pottery typical of the Late Hallstatt period in the Lower Rhine region. Just at the time when these vessels first appear in France (Haulzy), the Tumulus Culture in the regions from the Cologne district northward came to an abrupt end, i.e. shortly before the beginning of the La Tène period. The migration from the Lower Rhine into north-east France took place, therefore, a little before 500 BC. Part of these invaders settled in Champagne, where almost at once they were merged with other elements(from Normandy and, possibly, from the true Celtic area to the south-east); from this fusion, crossed with the Greek influence from southern France, the Marne culture arose. Other elements from the Lower Rhenish area passed on into south-west France, some of them advancing yet farther afield into the Spanish peninsula. If they abandoned their ancient homes shortly before 500 BC, their transit through north-eastern and central France must have been rapid indeed, for they seem to have reached Spain before the sixth century had come to a close. But on the analogy of the Vandal invasion of A.D. 406/9, this is by no means improbable. At all events, the Celtic migration into south-western France and the Spanish peninsula was ultimately due to Nordic pressure upon the Lower Rhine.

But part of the Lower Rhine tumulus-peoples were driven southward up the valley of that river. This movement took place at the same time as the above migration to the south-west. It has recently been shown that as the numbers of the population in the Lippe region declined (possibly c. 800 BC) there was a marked increase in the population of the Cologne area, while the sudden decrease in the population of the Cologne area which happened at the end of the Hallstatt period was accompanied by corresponding increase in the highlands of Eifel, Hunsrück and Taunus. This was maintained in the latter regions till the advent of the Teutonic peoples in the Middle La Tène period. Farther to the south, the Teutonic menace was mainly felt in La Tène D.

The Celts erected fortresses to stem the tide of invasion. The most important of these was the Odilienberg (near Strasbourg), its walls, some six miles long, composed of large blocks of stone joined by oak tenons, enclosing an area of over two hundred acres. As early as the second century BC we find Celtic houses in southwest Germany fortified with ramparts—another sign of the turbulence of the times.

But the regions further north seem to have lain deserted for some time. No Celtic finds of the La Tène period have been found north-east of a line from Andernach to Aachen. ‘Teutonic finds of an Early La Tène character’ occur to the east of the Rhine in the Cologne area, but to the west, they are mainly confined to the left bank of that river. The greater part of Belgium— apart from Eygenbilsen and a few other sites—would almost appear to have been a species of no-man’s-land during most of the La Tène period. But this may be due to the unobtrusive character of the graves, poor in tomb-furniture and without tumuli, only a few of which have as yet come to light. In the regions to the east of the Lower Rhine, and perhaps in Belgium too, the Nordic invaders mingled with the remnants of the older population, and from this intermingling the genesis of the Belgic peoples took place.

Such in brief was the march of events on the western boundary of the Celtic and Nordic zones. Let us now turn to the southern frontier of these two provinces which in the Iron Age lay in Central Germany.

It is held by Kossinna that the Harz district, and the region to the east thereof as far as the confluence of the Mulde with the Elbe, was occupied during the Middle and Late Bronze Age by various elements, two of which were Celtic and Teutonic (Nordic), the latter only appearing towards the end of period V1. In the Earlier Iron Age (Montelius’ Bronze Age VI) these Nordic elements become stronger, and the stone-cist, a form of grave so characteristic of the north, appears. The rite, of course, is cremation. But about 600 BC we find a mysterious group of inhumation graves, comparatively richly furnished with objects of adornment: tores, armlets, ear-rings, pins, girdle-mounts and amber beads. These some regard as evidence for a Celtic invasion; indeed, they seem to have come from districts where the Mehren culture prevailed, southern Eifel and Hunsrück, which were then presumably Celtic. This group of graves occurs both to the north and south of the Harz mountains and stretches as far east as the Mulde region, being best represented in the Halle district. Kossinna regards it as evidence for an eastward thrust of the Celts, at the expense of the peoples of the Nordic group (his Germanen). In this region, then, we have two cultures at this period, the Nordic and the Celtic, but they do not seem to have influenced each other.

In the fourth century, we find Celtic peoples established to the north of the Central German mountain-range in Thuringia, Saxony, Upper and Middle Silesia to the south of the Oder. This and the century preceding it were a troubled period for the latter province: not only was it invaded by Celtic peoples from the south and subjected to inroads of Scythians from the east, but about 500 BC (or possibly slightly earlier) an invasion from the north took place. The new-comers were the peoples of the Face-Urn and Stone-Cist culture which originated in the region to the west of Danzig about 800-700 BC, and spread south through Posen into Middle and Lower Silesia. Its peoples are generally regarded as the first Teutons to have invaded the latter province. Owing, doubtless, to the turbulence of the times, the old urnfield civilization of Silesia came to a sudden end. Whether the Face-Urn culture lasted in this region later than 300 BC is disputed. If it did not, we are forced to admit that Silesia, a remarkably fertile region, remained practically unpeopled for about two centuries: for the Early La Tene culture of Middle and Upper Silesia disappears and the succeeding phase (C) is only sparsely represented. Reinecke regards the end of the Early La Tène culture in Silesia as being due to Teutonic pressure and observes that the same process probably took place in Saxony also. Farther west, the Celtic population, settled in the Thuringian area during La Tène B, did not long survive the fourth century: during the third century, the Teutonic peoples conquered practically all of this region and beat the Celts back to their last great stronghold, the Steinsburg, on the Kleiner Gleichberg (near Romhild) which, in a sense, may be regarded as the key position to southern Germany. Farther east, about 200 BC, a fresh Teutonic invasion poured through the rocky mountain-gorge of the Elbe into northern Bohemia, the new-comers obtaining a firm footing in the land of the Boii. This is shown by certain grave-finds at Bodenbach on the Elbe, which, although revealing a civilization identical with a contemporary culture to the north of the Erzgebirge, differ from those of the Middle La Tène inhumation cemeteries of the Boii.

It seems probable, therefore, that the evacuation by the Celts of the territories to the north of the Central German mountains happened c. 300 BC and was due to pressure from the southward advancing Teutons. Indeed, it is not inconceivable that the Celtic invasion of Greece in 279 BC may have been ultimately connected with this pressure. If the latter conjecture prove correct, it would be an argument in favour of not regarding the Celtic expansion into Thrace, Macedonia and the Peloponnese as a later episode of the movement that resulted in the Celtic conquest of Cisalpine Gaul.

The gradual penetration of south-eastern Britain by the Belgae (Aylesford culture) which took place c. 100—60 BC may partly have been due to the expansion of the Teutonic peoples on the continent. But generally speaking, there were few Celtic migrations on a great scale in this later period. The way to the south was blocked by the Romans who established themselves in southern Gaul in 121 BC, and Teutonic aggression prevented movements to the east. A centripetal tendency asserted itself among the different Celtic peoples which culminated in the development of towns and town life. This is also true of the Celtic area east of the Rhine, where Teuton and Roman were soon to prove the upper and nether millstones. By the Late La Tène period (D) the Celts were no longer merely a country people but a town-folk as well. This development of town-life was the greatest contribution toward civilization in western and central Europe during the pre-Roman Iron Ages, and it also explains why the culture of the Late La Tène period exhibits a greater homogeneity than that of the phases which preceded it.

It remains to speak of the influence of climate. The dry warm climate of the Bronze Age was succeeded by a moist cool phase known as the sub-Atlantic Period. Correlations of the geological, botanic and other data with archaeological finds have shown that this phase began c. 850 BC. It was ushered in by a sudden increase in the rainfall; and c. 500 BC, a second rainfall maximum was reached. Generally speaking, the peoples of the Nordic culture-province, dwelling as they did in the more maritime regions, were to a greater extent affected than the Celts. It is tempting, therefore, to regard this climate-change as one of the factors which caused the peoples of the Nordic province to expand at the expense of their Celtic neighbours, and it was largely due to this expansion that the Celtic migrations began.

VII.

 THE LIFE AND CHARACTER OF THE CELTS

Our chief classical authorities for the mode of life and character of the Celts are for the most part late, but it is evident that they drew much of their information from earlier sources that have perished.

Judged by their art and their material culture, the Celts, though a barbaric people, attained a relatively high standard of civilization. The same may be said for certain less material aspects of their life. The first mention of Druids is not earlier than about 200 BC, although Druids were no doubt in existence before that time. By the middle of the first century BC, they had developed into a highly organized priesthood, and were men of considerable learning, with a knowledge of writing. Poetry was cultivated by the Celtic peoples, though the poems of the continental Celts have not survived. They were probably never committed to writing, but we know them to have been partly didactic and mythological, partly heroic and panegyric. Their religion was polytheistic. Strabo tells us that they held the soul and the world to be indestructible, but that ‘fire and water will one day prevail.’ There is a striking parallel to the last statement in the Norse doctrine of Ragnarok. It is hard to say whether their belief in metempsychosis was an old one. Some would trace it to Pythagorean influence derived from contact with the Greek colonists in southern France. Attempts have been made to connect the doctrine of metempsychosis among the Celts with the introduction of the rite of cremation—a rather difficult thesis to maintain. At all events, the rich tomb-furniture in such burials as the earliest La Tène graves of the Middle Rhine area would indicate a belief that the after-life was merely a prolongation of this life, that the chieftains who lie buried beneath those barrows thought that they would

               fight on, fare ever

There as here.

Our information concerning government and political organization among the Celts earlier than the first century BC is mainly confined to literary evidence and is very scanty. We know that a kingly government prevailed among certain peoples, but we cannot say for certain whether kingship was universal or not. From Polybius it would seem that a dual kingship existed among the Cisalpine Boii c. 236 BC. Whether this is also true for the Gaesati we cannot tell, for the latter seem to have been a conglomeration of mercenaries rather than a definite people. In an age of national migrations the existence of the comitatus— the band of retainers attached to the person of the chief or king— is not to be wondered at: the chief’s comitatus in a migrating people formed the nucleus of the new kingdom. The existence of this institution in this period is attested by Polybius. We have no evidence for the existence of a central political authority among the Celts; though circumstances or an ambitious king may have brought about alliances between various peoples, such unions were ephemeral. The sense of nationality was not highly developed among the Celts of our period, although Cisalpine Gauls recognized that there was some sort of kinship between them and their Transalpine brethren. Grave-finds in the Celtic area show that, as early as the Hallstatt period, there was a sharp division between the ruling classes and the lower orders, while the Marne chariot-burials and the ‘Chieftains’ Graves’ would indicate that, although these chieftains or kings ma not have ruled over very extensive territories, they were possessed of considerable wealth. The cessation of the migrations and the development of town-life no doubt led to considerable changes in the political organization.

In warfare, the Celts were noted for their cavalry, but they also had large numbers of foot soldiers. The latter are thought to have been drawn from the lower classes and from the subject population. Chariots were also used in warfare, but, as cavalry fighting developed among the continental Celts, they seem to have gone out of fashion. In the British Isles the use of the chariot survived later than on the continent. The chief weapons were the sword and spear; during the Middle La Tène period, the length of the swords rendered them cumbersome and they appear to have been so badly tempered that after a blow they had frequently to be straightened under the foot. Apart from the shield, defensive armour seems only to have been worn by the leaders, indeed the Celtic warrior frequently went naked into battle. Once the novelty of their appearance, their huge stature, strange cries and weapons had ceased to strike terror into their adversaries, the Celts were no match for the trained armies of the Greeks and Romans, though they retained their value as mercenaries.

Prior to the Late La Tène period, stone was not used in domestic architecture. The houses consisted of a single building or a group of buildings; their ground-plans were either round or rectangular. The art of fortification was highly developed, witness the murus gallicus, built of stone courses and wooden beams and the structure of the walls on the Odilienberg. Although the Celts were great agriculturalists, during the La Tene period they contributed greatly to the advance of industry, which developed further with the rise of towns and town-life in phase D.

All the evidence tends to show that the Celts reached a high level of culture in the La Tène period; and yet, in some respects, they were mere savages. They indulged in head-hunting, in the slaughter of wives, concubines and male dependants at funerals, and they practised human sacrifices. They had a passion for wine; to judge by what Diodorus tells us, their table-manners were disgusting. They had a marked liking for flashy clothes, and for bizarre fashions in hair-dressing. We know little concerning the position of women among the Celtic peoples. They appear to have been strong, handsome and ferocious; and were inclined to dominate their husbands. Strabo tells us that ‘the labours of the two sexes are distributed in a manner the reverse of what they are with us, but this is a common thing among very many other barbarians.’

The ancients have much to say concerning the character of the Celts. They were given to boasting, prone to flattery, quick to learn. They were capricious but devoid of malice, fond of money but extraordinarily hospitable, great fighters but easily defeated by strategy, truculent and fearless but unable to bear hunger or exposure. They were extraordinarily susceptible to the charms of poetry and from their works we know them to have been artists of a high order. Half civilized, half savage, they lived masterful, passionate lives in an atmosphere utterly remote from what literary men of today term the Celtic twilight.

CHAPTER III. THE NEW HELLENISTIC KINGDOMS By W. W. Tarn