Re: Jastorf - Przeworsk, tidied

'Chapter III

Some Problems of La TÃ¨ne Chronology The Late Pre-Roman Time or NELT

The summer of 1854 happened to be a hot one and the Lake of NeuchÃ¢tel in Switzerland had become very shallow. At the small place of La TÃ¨ne a whole pile of ancient weapons had collected on the shore - long iron swords, wide spear tips and also bronze and silver fibulae.

Whether this was treasure which had been concealed in the water or a cargo of wares which had sunk while being transported across the lake, or votive gifts thrown into a deep place to placate the god of the waters, is still unclear, although recent research lends support to the idea that all these objects belonged to Celtic warriors who had fallen into the lake from a collapsing bridge.44

In 1874 the Swedish archaeologist Hans Hildebrandt, who was researching European antiquities from the Iron Age, proposed that the term Hallstatt be given to the earlier period and La TÃ¨ne to the later period. From then on the concept of La TÃ¨ne as a chronological concept for designating the latter centuries BC came into being.

Nine years later, in 1885, Otto Tischler, basing himself on the typology of the development of fibulae and swords, subdivided the La TÃ¨ne into three stages - early, middle and late, or La TÃ¨ne I, La TÃ¨ne II and La TÃ¨ne III. The fibulae of the Early-La TÃ¨ne were bent from the blank in the following order: the needle, then the spring with the bowstring was wound round, the arc (back) of the fibula was bent, and the remaining end again bent upwards, the lower part of the fold was used to make the catchplate of the needle, and the upper part (the leg) remained free. The leg which did not carry a structural load often became a decorative element. When the craftsmen began to decorate it with large bronze balls in the plastic style, the leg often started to break so they began attaching it to the arc. Sometimes the end of the foot partially or fully covered the back and only thus was attached. And so the Mid-La TÃ¨ne fibula was born. In the Late-La TÃ¨ne period, the design changed completely, the catchplate, foot and back were made as one piece. These changes in fibula design which correlate to changes in the shape of swords and scabbards, formed the basis of Tischler's system.

In 1902 Paul Reinecke, continuing this system of European chronology from the Bronze Age to the Roman times and basing himself not so much on the typology of objects as on the combinations of their combinability in closed complexes, primarily in graves, proposed subdividing the La TÃ¨ne into four stages.46

His La TÃ¨ne A did not find correspondence in Tischler's material, but subsequent stages corresponded to Tischler's. He reckoned definite dates as follows: La TÃ¨ne A - 500-400 BC; La TÃ¨ne B - 400-300; La TÃ¨ne C - 300-100; D - 100-0.

The Tischler-Reinecke system - and many of their conclusions, despite differences in methods, coincided - served as a basis for all further research and their terminology was widely accepted. The basis of their system still remains unshakeable today, although many amplifications have been made to it.

Other proposals for subdividing the La TÃ¨ne culture have been put forward, of which the best known are Josephy Dechelette's work devoted to Gallo-Roman antiquities and the work of the Swiss archaeologist David Viollier who constructed his system on the horizontal stratigraphy of the burial ground at MÃ¼nsingen.48 Much later, in 1968 Frank Hodson addressed himself to the materials of this burial ground,49 introducing a number of amplifications, but the terminology of the Swiss and English researchers never became widespread.

The art critic Paul Jacobsthal made a significant contribution to the study of Celtic art by subdividing material according to the stylistic features of the articles. Archaeologists had to take his conclusions into account.50 Vincent Megaw, who has compiled an extensive catalogue of Celtic works of art, has continued in the same vein.51

1956 saw the publication of an important work by the Czech researcher Jan Filip,52 dedicated to the Celts of Central Europe. Having pulled together a huge amount of material, he attempted to subdivide a number of politico-socio-economic stages in the development of Celtic society - the same stages that are already known to the reader from the preceding chapter. Observation of archaeological material led him to distinguish horizons in which objects had come together in complexes which have been named according to leading types of objects - "the Duchcov fibulae horizon", "the horizon of belts with stamped designs", etc. The value of the observations on these horizons and their mutual placement in time is not transient, but as further researchers have shown, Jan Filip had a tendency to raise the absolute datings of objects somewhat too high, and therefore one can only make use of this work at the present time with certain corrections added.53 There were two reasons why Jan Filip dated too high. Firstly, because he somewhat overestimated the length of time that the Late-Hallstatt features survived; secondly because of a mistaken, or at any rate, doubtful dating for the destruction of the oppidum of Manching.

In 15-14 BC, at the time of the Alpine campaign, the Romans seized the lands on the right bank of the upper Danube. The important Celtic oppidum of Manching has now been excavated here with its painted pottery and Mid-and Late- La TÃ¨ne fibulae - among the latter were some characteristic long fibulae with narrow triangular backs - Nauheim type. The Roman camp of Oberhausen was situated close by with objects that were typical of Augustus's time. Comparison of these data suggested itself, seemed irreproachable, obvious and was widely accepted. The whole collection of objects from Manching was given an ultimate date of 15 BC and served as a bench-mark in determining their absolute date. Only latterly has it become clearer that the situation was more complex, that the date given for the downfall of Manching may be wrong. When Filip was writing his work it had been an axiom, causing no particular doubts and remained so for a long time. Practically all the authors who were writing about the Late-La TÃ¨ne period either based themselves directly on this date, or silently approved it, or even if they had doubts about it, did not put forward any specific suggestions for changing it. At the same time they made innumerable valuable observations in their works. Nearly all the stages of Reinecke's system were subdivided into sub-phases and the uneven development of the La TÃ¨ne culture in different territories became manifest.54

The lack of allowance for the uneven nature and specific character of the development of culture in the Celtic world in Northern and Eastern Europe led to yet another massive delusion. All these cultures - stages of the Seedorf culture of the Jastorfs in Germany, the Przeworsk and Oksywie cultures, the PoieneÅti-Lukashevka and Zarubintsy cultures and the memorials in Scandinavia were synchronous with La TÃ¨ne Stage D and were regarded as part of this stage, although Rolf Hachmann in 1961 introduced the term "Late Pre-Roman period" to designate them.55

Hachmann's work on the chronology of this Late Pre-Roman period had great significance and was widely used. A vast amount of material was generalised in it and submitted to scrupulous analysis. A large number of important burial grounds with rich collections of fairly standard inventories in the graves had by then been excavated in northern Europe. These were RzÄdz (Rondsen) with 828 burials, WymysÅowo with 336. The WilanÃ³w with 103, 670 in Grossromstedt, 820 in Hornbek, 266 in Tostedt-WÃ¼stenhofen 306 in Schmiedeberg-Aalgast, etc.56 All these burial grounds, and others not enumerated here, gave Hachmann rich material to work with. Using the method of horizontal stratigraphy, that is observing differences in collections of types of objects in the graves and their distribution within the area of the burial ground, he isolated stages in the development of each necropolis. Each stage, each group of objects probably corresponded to one or two generations of the buried, inasmuch as each generation had its own subculture, its own collection of adornments, fibulae, buckles, etc. Whether this mechanism worked in ancient times like our contemporary changes in fashion, or whether there were other psychological and social reasons for this phenomenon is hard to say, but the fact that such diversity took place is revealing. The existence of a Celtic veil, when many objects originated from Celtic prototypes, as well as the existence of quite vigorous links between various groups of European peoples, led to work on international and interregional objects which made it possible to synchronise the temporal groups revealed through horizontal stratigraphy. As a result, Hachmann managed to isolate three stages within the framework of his Late Pre-Roman period: the early period, or a, the middle which in a number of burial grounds falls into b and c, and the late, which in certain cases has been subdivided into d and e. Phase "e" was isolated later by R. WoÅÄgiewicz for a series of burial grounds belonging to the Oksywie culture.57 evolution of combinations of objects in complexes coincided here with evolving formes of fibulae, weaponry and other objects, which had been traced earlier by J. Kostrzewski58 and D. Bohnsack.59

Taking the "Late Pre-Roman period" of Northern Europe as synchronous with the Late-La TÃ¨ne period of more southerly regions, through the absolute dates of the latter, Hachmann did a time count. His conclusions in this respect were not very consoling: "An unbiased survey of all the possibilities of establishing absolute dates for the beginning of the Late-La TÃ¨ne period and separate earlier stages of it have finally led to the conclusion that it is still not possible to give any definite date. This does not mean, in principle, that such a date is impossible to find but that a whole series of memorials which, probably, could have yielded good dates, have not up to now been excavated as thoroughly and carefully as might be desired or considered possible.60

Hachmann was thinking of those memorials which although they did not yield absolute date-marks for the very beginning of the Late Tene, still served as guide-posts.

These are the northern Italian burial grounds of San Bernardo and Persona near Ornavasso. Here coins of the Roman Republic were found in 140 of the 331 burials. But only 152 coins from 90 burials were described and defined and there was no certainty about the accuracy of E. Bianchetti's definitions. The source was incomplete and Hachmann slated Moberg for excessive faith in it.

Then there is the treasure from Lauterach, where two silver fibulae of Late-La TÃ¨ne construction were found, close in design to the widespread Nauheim-type fibulae and, possibly, representing a prototype of the latter, together with 23 dinarii of the Republic, the most recent of which was minted by the master craftsman T. Cloulis around 100 BC. There were also two Celtic coins of the Tektosagi tribe which, however, did not typologically date back to the earlier minted versions. And so some researchers have put forward the proposition that the treasure was buried in connection with the events that served to destroy the oppidum at Manching, that is to say in connection with the Roman occupation of the Prealpine zone in 15 BC.63

Of more than 300 fibulae found in Manching, around 70 were Late-La
TÃ¨ne, of which at least 30 were Nauheim-type with a narrow, triangular back, but, on the other hand, the later, so-called bent fibulae (geschweiften) with the strongly curved arc64 were completely missing.

Although we have no direct evidence that it was the Romans who destroyed Manching and this in 15 BC, and although this destruction could have happened earlier during events not authenticated by written sources,65 Hachmann still feels that this date can be used.

Two other places, whose dates are supported by historical data are Alesia and Bibracte in France. Alesia was the capital of the Mandubi tribe, the last stronghold of Vercingetorix, the leader of the Gauls, who had risen against Caesar and perished in 52 BC. Excavations started there in the last century, revealed a ditch which, at one time, had been filled with bodies of fallen Gallic warriors, and which yielded a very rich collection of Celtic weaponry in use in 52 BC.

Bibracte - the capital of the Aedui was an oppidum whose inhabitants in 5 BC moved 27km to Augustodunum. Finds of Nauheim fibulae are rare in Alesia but are known in Bibracte. This induced K. Schumacher to date such fibulae to the second half of the 1st century BC.67

Another important place which might also help in establishing an absolute chronology for the Late-La TÃ¨ne could well be Entremont in the south of France - the capital of the Saluvi, destroyed by the Romans in 121 BC, if there were more finds and a better defined stratigraphy.68

All these not very reliable pegs of absolute chronology which were holding together a network of relative chronology for the Late-La TÃ¨ne fell, as it were, within a period. Generally they told us when this period "was", but they did not tell us when it began. The lower portion of the chronological network was left dangling.

Hachmann understood this and we do not find any attempts in his work to fix absolute dates for his stages in the Pre-Roman period. But at the end of his work there is a conclusion which has been accepted as an axiom by many researchers:

"If we examine once again all the reasonably possible points of view and then attempt to evaluate which dates may be acceptable for the beginning and which for the end of the Late Pre-Roman period in northern Central Europe and Scandinavia, we would come to the following conclusion:

I. The beginning of the Late Pre-Roman period is between 120 and 100 BC.

II. The end of the Late Pre-Roman period is the latest in the post-Augustinian time..."69

Thus, in the final analysis, the intuitively determined date for the beginning of the Late-La TÃ¨ne and synchronous northeast European pre-Roman period looked highly attractive, because in this case historical justifications for cultural changes in Europe had been found i.e. - the transition of the Jastorf culture to the Seedorf stage and the formation of the Przeworsk and Oksywie cultures. All these events could be linked with the campaigns of the Cimbri and Teutons from Denmark along the Oder to the Danube where changes can be observed. The Cimbri advanced first of all to the middle Danube where they clashed in 113 BC with the Boii and the Scordisci, then into Northern Italy and France, until they were beaten in 102 by Marius.

By the beginning of the 1960's, the edifice of European chronology although, as we have seen, not resting on a sure foundation, still seemed an unstormable fortress - the divergences and defects in its construction being almost unnoticeable. Of all the Late-La TÃ¨ne memorials which found support in absolute chronology, only Manching was systematically researched by specialists and yielded a large collection of materials.

It was not La TÃ¨ne specialists who began excavating under this stronghold, but researchers on the Early Roman period, whose freshly accumulated material ceased to correlate with that of Manching.

Certain doubts on the dating of Manching although not always openly expressed, never left the researchers and these must have been strengthened by O. Uenze's work, in which he established that the amphorae of settlements, analogous to Manching insofar as the composition of their finds was concerned, dated back to the pre-Caesarean period. Rainer Christlein's brief article on this subject must have been revolutionary.71

He observed that if Manching had been destroyed in 15 BC and then the Roman camp at Oberhausen had been built, there would not have been enough time for an entire horizon of finds to develop usually accompanying bent Late-La TÃ¨ne fibulae, unknown in both Manching or Oberhausen. This horizon is well represented in almost all of Europe and the Grossromstedt burial ground is one of its most representative and striking sites.72

After re-examining the horizontal stratigraphy of this burial ground, R. Christlein came to the conclusion that R. Hachmann had not correctly determined the successive stages of development in the burial ground. The southern group of graves turned out to be more recent than the northern ones as was thought by Hachmann. Several burials already from the Roman period, Stage B1 according to Eggers, were found among graves of the southern group. Christlein also considered it possible to move the beginning date of the stage B1 back, inasmuch as the Almgren 236 type fibulae which were typical then are known at Oberhausen, ie. around 14 BC. Thus Christlein arrived at the following deduction: the southern group of graves at Grossromstedt and those analogous to them are earlier than 14 BC, the northern group is still earlier, and the most northerly extremity of the burial ground which has generally not been excavated, could contain another 300 graves or so with representatives of an entire generation. These are even earlier. And just in front of them were objects of the type found in the upper layers of Manching, whose destruction, in Christlein's opinion, occurred at some time around the middle of the 1st century BC.

Christlein's reasoning was not irreproachable, but judging from further work on the chronology of Late-La TÃ¨ne, the essence of the problem was clearly perceived - the date for the destruction of Manching can and must be lowered.

Peter GlÃ¼sing worked in parallel with Christlein and came to similar conclusions. It seems that during the Alpine campaign of 15-14 BC, the land immediately surrounding Manching was generally not occupied by the Romans. Their line of posts went further south, standing on the upper reaches of the Iller, the Lech and Isar rivers, controlling approaches along their valleys to the Danube. And only Oberhausen was made an outpost in the Lech valley, like the summer camps at Oberaden or Haltern on the Rhine. But this land did not belong to the local population of Raetia either. The disposition of their burial grounds with inhumations coincides with the zone of occupation. The banks of the Danube between the Lech and Inn, where Manching was situated, were occupied in the second half of the 1st century BC by the Germans by Suebia with a culture close to that of Grossromstedt. These are the burial grounds of Kronwinkl and Uttenhoffen. By the time they appeared here, Manching no longer existed or had ceased to exist. And the time of their appearance has been determined by the find of a belt hook and a mid-La TÃ¨ne fibula with a strongly curved back, variant G/H according to Kostrzewski, ie. objects which are typical of the middle phase of the pre-Roman period according to Hachmann and are often still found together with early Late-La TÃ¨ne type K or J fibulae. The transition of the Manching culture to the Kronwinkl-Grossromstedt culture occurred around the middle of the 1st century BC.

The burial grounds of the Middle Rhine and Luxembourg also yielded materials for the period between the horizon of the Nauheim fibulae of Manching and the Roman camps.75

Christlein drew the attention of researchers to the very simple, hinged fibulae, the study of which would reveal great possibilities and be useful in determining the first appearance of the Romans. This wish was fulfilled in 1975 by Sabina Rieckhoff during work on fibulae from the castella at HÃ¼fingen and the La TÃ¨ne settlement which existed there. She examined the materials from HÃ¼fingen against a very extensive background of cultural changes in the entire Rhenish-Danubian region. The following facts were established: between the horizon of the Nauheim fibulae, which were widely represented at Manching, and the horizon containing the fibulae of the Augustan camps on the Rhine and Danube there lies a whole stratum of fibula finds, some of which represent prototypes of the forms which would develop in the Augustan period and later. Among these one finds the simple, iron fibulae of Late-La TÃ¨ne design made from wire of rectangular cut which closely resemble the Kostrzewski 15 type. (These are sometimes called "soldier's" or "legionary" fibulae77); prototypes of the Aucissa type fibulae used later by legionaries in the Augustan-Tiberian period; the so-called Alesia type fibulae with the wide, triangular back; versions of bent fibulae with projections covering the spring; early versions of the so-called Diestelfibel); pincer type fibulae and some others. They were mostly dispersed along the inner side of the limes have typological predecessors in northern Italian, Dalmatian or Gallic forms and appear to be the beginning stage of the next provincial Roman culture.

The principal time covering the combination of these forms would have to be 50 - 20 BC. The preceding Manching horizon would thus be moved again to the middle of the 1st century BC. J. Graue was conducting research at the same time as Sabina Rieckhoff and in 1974 published newly reworked material on two burial grounds at Ornavasso. Of the six stages of development he isolated in the San Bernardo and Persona burial grounds, the second stage corresponded to the Manching horizon; the Nauheim fibulae, the spoon-shaped J variants, (per Kostrzewski) the Almgren 65 fibulae, the fibulae similar in design to the G/H variant but with a long spring (the Ornavasso type) and swords with bell-shaped, curved quillon etc. Twenty-eight coins of the Roman Republic were found in graves from this stage, the latest of which was 76 - 71 BC.

Fibulae from the collection distinguished by Sabina Rieckhoff, including the simple Late-La TÃ¨ne Almgren 15 variants, found their way to Ornavasso for the third stage with the most recent coin dating to 37 BC, and the early Roman Diestelfibel and Aucissa fibulae for the fourth stage with coins dating from 18-2 BC.78

Thus, the horizontal stratigraphy of these two northern Italian burial grounds not only confirms the earlier date for Manching and the rhythm of development of the early provincial Roman culture, but gives reliable, absolute dates. The somewhat earlier dating as compared with the Gallic, southern Germanic and Czech dates is either random or may be explained by the disposition of the burial grounds in Italy, where the provincial Roman culture may have begun forming earlier.

And so data have accumulated forcing us to disregard the date of 15 BC set for the destruction of Manching, and to move this date back to the middle of the 1st century BC. This dating can also be justified from a historical point of view, because it was in 60 BC that the Celts suffered defeat at the hands of the Dacians under King Burebista and the tribes of Boii and Helveti moved westwards. Then, in 58 BC, Germans of the Suebian alliance of tribes under Ariovistus advanced towards the Rhine. Manching was in the area of operations of these forces. E. Major linked the destruction of a La TÃ¨ne settlement in Basel with painted pottery, analogous to that of Manching, with the movement of the Helveti in 58 BC79. A layer of destruction dating to the same time has been noted in the large settlement of Magdalensberg in Noricum.80

It is still hard to evaluate fully the consequences of shifting the Manching date and the debate continues. But we are dealing with a system and changes to part of it must of necessity lead to changes to the entire system. A changed date for the destruction of Manching alters the date of the Nauheim fibulae which were considered one of the basic chronological indicators for the second half of the 1st century BC. Now it appears that all, or at least the majority of memorials, which have been dated from these fibulae, should be put back 50 years to the first half of the 1st century BC. Confirmation for such dating has been provided by the find of a Nauheim fibula in Talamon in northern Italy, where a sanctuary had been destroyed in 82 BC81. Correspondingly, preceding chronological groups will also have to be either moved back, or consolidated in time.

The absence of Nauheim fibulae in Alesia can be explained by the specific nature of this site. Excavations here revealed a field of battle and finds represent a masculine culture whereas the Nauheim fibulae most probably formed part of the female attire.

Thus, on the Rhenish-Danubian borderlands and in Europe generally a distinctive cultural situation was developing in the second half of the 1st century BC. One could say, with some over-simplification, that three distinct cultural circles were in existence, although without clear-cut borders (illus. 4).

To the west of the Rhine and south of the Danube the circle of the future provincial Roman soldiers' culture began to develop based on northern Italian, Dalmatian and Gallic forms. This was represented up to then by a simple, iron "soldier's" and different hinged fibulae. S. Rieckhoff was engaged on these and dated them to 50 - 20 BC. This is the very beginning of the Roman period, Stage A1.

After the Alpine campaign of 16 - 14 BC and Drusus' activities on the Rhine and in Germany in 12 - 9 BC, when centres of Roman culture were built in the form of camps such as Oberaden, Oberhausen, Xanten, Neuss A, Haltern and Dangstetten, the second stage began - A2.

Parallel with this, on the lands adjoining the Rhine from the east and the Danube from the north, in southern Germany and, to a lesser extent, in the Czech lands the culture of the Grossromstedt horizon was developing with its bent Late-La TÃ¨ne fibulae, abundant weaponry and attractive black-polished situlae vessels.

In the heart of the Germania Libera and right up to Scandinavia and the Dnieper a circle of La-TÃ¨neicised cultures was situated - the Jastorf culture at the Seedorf stage, the Przeworsk and Oksywie, and the Zarubintsy cultures, all preserving their former La TÃ¨ne aspect.

R. Hachmann described the corresponding stage in the development of these cultures, which was synchronous to the Late-La TÃ¨ne period in southern regions, as "the Late Pre-Roman period", emphasising in this way its difference from the La TÃ¨ne culture of the Celts.

But these cultures were so steeped in Celtic influence, which had become an organic and integral part of them, that the removal of the word "La TÃ¨ne" from their designation emasculates the very essence of this phenomenon. It is not accidental that the majority of researchers could also not desist from using the adjective "Late-La TÃ¨ne" when referring to memorials and finds of the Przeworsk, Oksywie and Zarubintsy cultures. Perhaps it would be expedient to use the term "North European La TÃ¨ne" (NELT) in this instance, although it is not very exact from the geographical point of view - the Zarubintsy and PoieneÅti-Lukashevka cultures lay to the north-east and east of the Celtic world.

For the time being the debate around Manching has rocked the upper stages of the Late-La TÃ¨ne and NELT chronological system, a breach in its till unshakeable (as the majority of researchers had believed) construction has begun to form, as in the lower stages of the eastern part - in the chronology of the La TÃ¨nicised cultures of the Ukraine and Moldavia. It all began with the amphorae. In the Zarubintsy settlement of Pilipenkova Gora the handle of a Cossian amphora stamped 230 - 220 BC was found, in some settlements of Rumanian Moldova - fragments of Greek amphorae stamped 220 -180 BC and in Lukashevka II - fragments stamped 220 -146 BC.82 And this
is 100 years earlier than the earliest datings for the burial grounds which are being established from the fibulae using R. Hachmann's and Jan Filip's chronological systems. In the old controversy between the proponents of the short chronology of the Zarubintsy culture and PoieneÅti-Lukashevka, on the one hand, these cultures were thought to date to the turn of the 2nd -1st centuries BC and to have come from Central Europe83 and, the proponents of the long chronology, on the other hand, they were thought to go back to Scythian times and were dated to the 3rd century BC.84 The amphoral finds have come down heavily in favour of the latter. The opponents can now only doubt the authenticity of these finds.

The contradiction between the chronology of the fibulae and amphorae and the chronology of the burial grounds and settlements nevertheless remains. Attempts have been made to resolve matters by introducing a special "fibulae-less stage" in the development of these cultures, but they were not successful - the "fibulae-less stage"* turned out to be fictional.86

In the meantime, the course of the debate around Manching and the doubts about R. Hachmann's and Jan Filip's absolute datings have compelled archaeologists to consider the possibility of a shift downwards in the entire chronological system. Then Kasimierz GodÅowski proposed a clever solution to the problem.87

The earliest forms of fibulae in the NELT are variant A according to Kostrzewski, long, of mid-La TÃ¨ne design with little balls - thickenings on the foot. (illus 4; 37, 38, 76). These are quite well known in the Celtic world. Jan Filip considered them one of the earliest mid-La TÃ¨ne variants ("united" in his terminology) and not very successfully called "disjointed" or "segmented". The little balls or other decorative elements divided up, as it were, the foot. The "disjointed" fibulae are typical for Stage C of European La TÃ¨ne, and more so for C1 than C2.

Variant A fibulae were regarded earlier as the subsequent typological link of the Celtic disjointed types. GodÅowski saw them as one of the variants which was developing in parallel. Thus the beginning of NELT would synchronise with La TÃ¨ne C, rather than La TÃ¨ne D.

Just as the Celtic disjointed fibulae served as primordial prototypes of local fibulae in the eastern part of NELT, the Celtic fibulae with balls played the same role in the western regions. Their local variants survived for quite a while longer in the Jastorf culture.89 We can synchronise NELT with Stage C thank to the swords with the bell-shaped, curved guillon, which were so widely represented in the Przeworsk and Oksywie cultures.

GodÅowski cites yet another argument. All the NELT cultures appear to have quite a large amount of varied objects placed in the graves which were not at all characteristic of their predecessors and forebears, bearers of earlier phases in the Jastorf cultures and the Pomeranian culture in Poland. It is possible that the new model of burial ceremony was also borrowed from the Celts. But in that case it could have happened no later than Stage C1, because later the Celts themselves dropped this custom.90

A large part of the international datable objects of NELT have something in common with finds in Celtic oppida, but this particular stage, when the Celts changed to a semi-urban way of life in the oppida and adopted the burial ceremony that is elusive to archaeologists - the oppida horizon -usually correlates with La TÃ¨ne D. Just lately, a considerable amount of material has been accumulating which shows that this horizon was begun earlier. Such centres as Stare Gradisko in Moravia, Stradonice and GraÄane in the Czech lands and Manching in Bavaria already existed in La
TÃ¨ne C. Caesar also informs us that the Cimbri and Teutons had already clashed in Gaul with the oppida-citadels.91

Thus, the beginning of NELT is pushed backwards, to La TÃ¨ne stage C, and even the last of Stage B2 (illus 4). Sometime within the boundaries of this stage the peoples of Northern and Eastern Europe received a powerful dose of Celtic influence. Henceforth, although contacts with the Celtic world of La TÃ¨ne culture were maintained to some degree or other, the NELT cycle cultures evolved in their own way.

How can we determine in absolute figures the period of most intensive contact between the Northern European peoples and the Celts? This is not a simple task and still today almost insoluble, because, unfortunately, period C itself has altogether no reliable, absolute leads. It is merely "squeezed" between the beginning of period La TÃ¨ne D, the dating of which is uncertain as has been discussed above, and the even shakier date for the end of period La TÃ¨ne B. Only one thing is clear, that La TÃ¨ne C can not be much later than 100 BC, if we accept the dating of the treasure at Lauterach from the Roman coins to be later than 82 BC, if we proceed from the dating of the Nauheim fibula from Talamon as also later than 76-71 BC, and if we draw on the dating of the second period of burial grounds at Ornavasso. It is possible that the campaigns of the Cimbri and Teutons could have played a certain role in the reshaping of La TÃ¨ne C into La TÃ¨ne D, and of the roughly synchronous phase NELT-b into phase NELT-c.

The boundary between La TÃ¨ne C and La TÃ¨ne B is still less clear. On the famous bas-reliefs of the Athena Nicephora temple in Pergamum, dated around 225 BC and no later than 180 BC, Galatian weaponry typical of La TÃ¨ne C1 stage is shown. This gives according to K. GodÅowski the terminus ante quem of stage B.** In Central Europe the transition from La TÃ¨ne B to La TÃ¨ne C could have occurred, in GodÅowski's opinion, much earlier than the end of the 3rd century BC, perhaps even around the middle of that century. But he can only draw here on his intuition***. The dating of the division of Stage C into C1 and C2 are no less intuitive**** -sometime during the first half of the 2nd century BC

On the plate we have made up, the beginning of Stage C1 has been carried over approximately to the second half of the 3rd century, to 225 BC, based on the consideration that the re-forming of the Celtic culture could have begun in connection with their reverses in 225 and 195 BC in Italy and in 212 BC in Thrace, and in connection with the subsequent concentration of the Celtic population in Central Europe.*****(illus 4).

If we look for historical events that might have led to the formation of NELT cultures, it could be supposed that an impetus might have been given by the hypothetical movement of a group of peoples of eastern variants of the Jastorf culture and part of the bearers of the Pomeranian culture of Poland to the south-east, along the eastern slopes of the Carpathians, a population which had appeared on the historical arena around 220 - 216 BC under the name of the Bastarnae. The events of 179 - 168 BC in the Balkans might similarly have served as a definite impetus to the forming of the PoieneÅti-Lukashevka and Zarubintsy cultures. More about that later.

The last quarter of the 3rd century BC and the first quarter of the 2nd century BC thus represent a fairly likely period, when a suitable situation might have developed for the forming of Stage C1 of the La TÃ¨ne culture and the arising of NELT.

But historical events and a historical situation cannot of themselves, however convincing this may seem, serve as guides for dating archaeological phenomena. An archaeological lead to absolute dates, absent until now, is needed. Only the amphorae of PoieneÅti-Lukashevka and of the Zarubintsy peoples indicating the same period of time can serve this purpose. But we do not have enough to go on to be completely sure of these finds from the settlements and our constructions remain purely hypothetical.

The chronological table we have composed draws on GodÅowski's data and illustrates his basic theses with a small exception. In the absence of the usual such tables, the lines separating one stage in the development of the material culture from another, and one layer of co-finds from another, are drawn on a time-scale obliquely rather than horzontally. This reflects the fairly obvious fact of unevenness in the changes of stages in different territories.

Besides, in such a graphic expression the stages partially overlap each other all the time, which obviously reflects the real situation more accurately - that the beginning of each new stage is to be found in the preceding one. But, additionally, it should be remembered that any delimiting line in the flow of time, in the fluid process of modification of objects and the way they combine in complexes can only be conditional. Naturally, the table does not by a long chalk include the whole collection of objects typical of each stage, but only the most significant and interregional. It is also natural that their place on the time scale only really corresponds in an appropriate way.

In GodÅowski's system a number of objects have been shifted backwards in time. Apart from those objects already mentioned earlier, others have been affected such as the Duchcov fibulae******, the fibulae with balls, and items of the Waldalgesheim and Plastic styles. Further research will show to what extent this is justified.

In determining absolute dates, GodÅowski based himself on the following unfortunately very meagre data. Only one burial in Stage A led to an absolute date. This is Klein Aspergl in WÃ¼rttemberg with its classical black-figured vessels, dating to around 450 - 425 BC. A rich burial in Vix, Burgundy, with black-figured pottery dating to 530 - 520 BC and Etruscan and Greek bronzes from the end of the 6th century BC limits the date of the beginning of Stage A from below. Vix still dates back to the Halstatt D3 period. The upper boundary of Stage A is not clear. But a rich burial in Waldalgesheim, where a collection of objects relating to both Stages B1 and B2 was found, ie. which lay on the time boundary of these stages, contained a bronze vessel from the workshops of Tarent made around 380 - 370 BC.

True, there is another opinion regarding this vessel: it appears to be only an Italian copy of a Tarent design, made later, in 340 - 320 BC and surely not earlier than 332 BC.94 Since a contradiction has arisen with the early dating of the Tarent situla in that the boundary between B1 and B2 has been shifted too far back, so that Stage B1 has almost been reduced to nil, B. Schiering's dating seems more real and logical than the date accepted by K. GodÅowski according to Gisella Zahlhaas.95

The fact that objects of the plastic style, typifying Stage B2, particularly the bracelets of hollow hemispheres, were made around 279 BC, was confirmed by a find of such items in Isthmus, Greece.96 They could have found their way there during the Celtic invasion of 279. But the date is not very reliable, since it is only linked to historical events. These bracelets might have arrived at the sacred source on the Isthmus as votive gifts or through some other circumstances. If the Isthmian bracelet can determine an earlier dating for the appearance of the plastic style, then the existence of objects made in the Waldalgesheim and Disneyesque styles during those seven decades (280 - 212 BC), when the Celts held sway in Thrace, can be confirmed by finds of a necklace in Tsybor Varosh97 and a Celtic chariot with Disneyesque bronze decorations in Mezek.98

The sculptures from Pergamum, as has already been said, give a terminus ante quern for Stage B2 - around 180 BC.99 or around 240-220 B.C.** Stage C1 began earlier than this date, but when - is still unknown.*** The end of Stage C2 and the beginning of D1 are also indeterminate. D1 has a whole string of connections to absolute dates: Manching, Alesia, Bibracte, Lauterach, Talamon, and lets add the treasures containing the coins of the Republic and the Late-La TÃ¨ne fibulae in Dacia100 - but none of these help define a boundary with the preceding Stage C2. After the publication of E. Graue's book one might be able to turn again to Ornavasso with its massive finds of Roman Republican coins. The authenticity of the complexes has now been verified. But another danger lurks in the use of these materials. The burial grounds in San-Bernardo and Persona have been left by the northern Italian tribe of the Leponti with a Celticised culture, but one which developed apart from the main evolutionary line of Celtic culture. Their rhythms may have been different.

And so there are still many unclear elements in the European La TÃ¨ne chronological system. The need for a detailed monographical review of this at a new level of knowledge is becoming imminent. Neither K. GodÅowski's book nor the present work is dedicated especially to its chronology and therefore cannot fulfil this task.

Now the author together with V.E. Yereomenko is preparing an article for "Acta Archaeologica Carpathica" entitled "The Celtic Antiquities of the Transcarpathian Ukraine and Some Problems of the Chronology of the La TÃ¨ne Period" in which the current state of the chronological system will be reviewed. Plate from this article is also published here (Ill. 5a).

...

* In 1986 in the Faculty of Archaeology, at Leningrad University, V. Ye. Yeremenko, a former student of this author, defended his diploma work in which, among other things, he analysed the chronology of the Zarubintsy culture. Having used the correlation method in his research on the cemetery at Korchevatoye near Kiev , V. Ye. Yeremenko established three phases in that cemetery. The first phase proved to be a non-fibula one. However, that does not confirm the opinion of the adherents of the wide dating that there should be a sequential connection between the preceding Scythian pottery, but is analogous to the vessels of the Bodenbach-Podmokly group in the Czech lands and to the pottery of the second phase of the cemetery at Luboszyci in Poland . These analogies allow us to compare the early phase of the Korchevatoye cemetery with the LT-C1b and Sate "b" of NELT and therefore date it to the early 2nd century BÐ¡. Thus, the datings of the earliest Zarubintsy burials and settlements coincide and the existing antinomy is removed.

V.Ye. Yeremenko also notes that the amphorae from Rhodes could not have reached the bearers of the PoieneÅti-Lukashevka and Zarubintsy cultures later than in 183 BÐ¡. It was in that year that the king of Pontus Pharnaces captured Sinopa and started a war with Bithynia and Cappadocia. Rhodes, Pergamon and Rome opposed Pharnaces, while the Greek towns - from which the amphorae reached the Bastarnae - supported him. The trade of the Bastarnae with the opposing coalition would have been practically impossible; besides that, navigation of the Aegean and the Black Sea during the war was dangerous. It is also highly unlikely that the Rhodes amphorae could have reached the Bastarnae after 179 BÐ¡, since it was in 179 BÐ¡ that Philip V formed an anti-Roman, hence - an anti-Rhodes union with the Bastarnae. Moreover, the latter at that time were at war with the Greek towns situated on the western coast of the Black Sea. In the years 202-197 Ð¡ Philip V had also been at war with Rhodes and Rome, thus supplies of wine from Rhodes would most probably have reached the barbarians of the northwestern Black Sea coast during the years 197-183 BÐ¡

** V. Ye. Yeremenko noted another interesting thing. During the construction of the temple of Athena Nicephora (between 225 and 180 BÐ¡) Pergamon did not achieve any significant victories over the Galatians. On the contrary, in 183 BÐ¡, it was defeated by the king of Pontus Pharnaces I and his allies. True, in 190 BÐ¡ the legions of Manlius marched victoriously through Galatia. However, that victory could not be attributed to Pergamon, but to its ally, Rome. The victorious wars of the king of Pergamon Attalus I against the Galatians had taken place earlier, in 241 BÐ¡ and then in 229-228 BÐ¡, and most probably the Celtic weapons depicted on the triumphant reliefs of the temple date back to just that time. The construction of the temple itself was most likely a propaganda act intended to raise the morale of the population of Pergamon by reminding them of their former victories. But in this case the terminus ante quem LT-B2 also related to the years 40-20 of the 3rd century BÐ¡. This agrees better with the other data available.

*** This idea of K. GodÅowski is confirmed by the materials from the Celtic graves near Bologna in Italy. Here the fibulae that closely resemble Variant A come from the complexes relating to the mid-third century BÐ¡, while the burial grounds themselves can not be dated to a later time than 192 BÐ¡, since they belong to the tribe of Boii, which migrated that year to Central Europe. A couple of fasteners which can be regarded as prototypes of Variant A fibulae were found in the rich burial at Ceretolo (dated to the end of the first quarter or to the second quarter of the 3rd century BÐ¡). These Celtic burial grounds contain some other objects (chain-belts, swords, shield bosses8), which in Central Europe are typical of stage C1, while in Italy they belong to the first half or to the middle of the 3rd century BÐ¡.

**** The latest amendments to the relative chronology of the La TÃ¨ne epoch were introduced by J. Bujna9. Analysis of the combinations of objects in the Celtic complexes of the Carpathian region allowed him to establish the following phases: B2a, B2b, B2/C, C1a, C1b. During Phase B2/Ð¡1 the plastic style was still in existence, for instance, it may be seen on the small balls decorating the feet of the fibulae and on bracelets. Bracelets of sapropelite were still in existence, as well as bronze anklets made of semi-spheres, though the number of joints in these anklets decreased and the semi-spheres became larger. During Phase C1a there appeared the first fibulae decorated with figure-of-eight whorls. During C1b the fibulae began to be decorated with a small ball or a disc not only at the foot, but also at the point where the foot of a fibula is joined to the arch. Bracelets of sapropelite were replaced by glass ones, enamelled chain-belts were made, and the hook-shaped fasteners were often styled as animal heads.

During Stage C2, besides pottery of grey clay, graphite pottery appeared. The middle La TÃ¨ne fibulae lost their "jointed" look. Besides Variant A, Variants B and Ð¡ appeared. The characteristic type of fibula for the beginning of Stage C2 is the MÃ¶tschwil type fibulae, which is well represented in the latest stage of the MÃ¼nsingen burial ground.10 No finds of such brooches are known in the Carpathian basin, which, as J. Bujna believes, is proof of the absence of remains belonging to Stage C2 in that region. It remains unclear, whether the Celtic burials in the Carpathian region disappeared at the beginning of Stage C2, or whether the Celts continued to live there, retaining forms of C1 up to the beginning of Stage D.

No new data have yet appeared that would help to establish the absolute dating of Stages C1 and C2 and phases within them.

***** H. Polenz assigned the transition from Stage B2 to C1 to that same time, having noted that the development of the La TÃ¨ne culture proceeded with an average periodicity of about 75 years. One can assume that in Central Europe and the Carpathian basin Phase B2/C1, on the whole relates to the span of time between 225 and 192 BÐ¡. The objects made in the style of Stage Ð¡ were probably introduced into those areas by the Celts after their expatriation from Italy.

****** The chronology of the Duchcov fibulae was studied by V. Kruta.13 He established a specific phase "Duchcov-MÃ¼nsingen" covering the period from the second quarter of the 4th century to the beginning of the 3rd century BÐ¡. The lower date was established in the following way. The prototype of the Duchcov fibulae, which were very close to them in construction but with different proportions, are the so-called Marzabotto fibulae. One example was found in the Etruscan cemetery Marzabotto near Bologna, which was in use - judging by the finds of Attic red-figure vases - in the late 5th- very early 4th centuries BÐ¡. Its abandonment is related to the time (387 or 390 BÐ¡) when Rome suffered from the Celtic invasion. It was then that the Celtic cemetery was established on the ruins of the Etruscan town of Marzabotto. Its burials have yielded examples of Duchcov and MÃ¼nsingen fibulae. A similar picture is revealed by the neighbouring cemeteries Arnoaldi and Benacci, the burials of which yielded also the forms of fibulae transitional from the Marzabotto to the Duchcov and MÃ¼nsingen types.

To establish the upper date of the "Duchcov-MÃ¼nsingen" phase is a more difficult task. As J. Bujna has demonstrated, in the Carpathian basin the Duchcov fibulae are present in the complexes of Phase B2a while in Phase B2b they are replaced by the so-called Paukenfibeln . It cannot be ruled out that the cessation of the Duchcov-MÃ¼nsingen style with its characteristic use of corals is connected with the following events. In 283 BÐ¡ the Romans conquered the areas inhabited by the Celtic Semnones situated to the south-east of Bologna (the modern San Marino and the north of the Marche province). They defeated the Celts and Etruscans near the Vadimonis lacus, and in 268 BÐ¡, founded their colony Ariminum (modern Rimini) on the coast of the Adriatic Sea. Since the Po river valley east of Bologna is very swampy, the Celts lost a convenient outlet to the sea and thus could no longer obtain their corals. A general change of style also occurred.

It may just be the years 80-60 of the 3rd century BÐ¡ which should be regarded as the borderline between phases B2a and B2b. During the Celtic expansion to the east, into Greece, Thracia and Asia Minor (an expansion which may have been caused by their defeats in Italy) the objects of Phases B2a and B2b coexisted. This explains the strange combination of finds on the Isthmus, in Mezek, and in Tibir Varosa. The question of how long the Duchcov fibulae continued to exist in the territories of Poland, Rumania and the Ukraine remains an open one.

Significant material for the more precise setting of absolute dates for the La TÃ¨ne chronological system is provided by dendrochronology Analysis of the remains of an oaken shield discovered at the site of La TÃ¨ne has allowed us to date that shield to 229 BÐ¡. The shield boss is of the type which, according to the correlation tables of J. Bujna, relates to the border between B2b and B2/Ð¡1.

From burial 96 of the cemetery at Waderath on the Rhine come the fibulae with small balls on the arch relating to Stage C1b and the charred wood from that same grave is dated to 208 BÐ¡.

Another example: analysis of the logs of the bridge in Tille (Switzerland), where all the accompanying finds belong to Stage D1, has shown that the trees were cut between the years 120 and 116 BÐ¡. (See Illus. 5a).'

Perhaps you should look up the article he announces?

Torsten