Image Source : Istock
The loss of the Ninth has now been pigeon-holed into the archaic and ambiguous ‘Legend’ category, where it probably really shouldn’t be. Those armored eagles that have become entangled in posterity’s hunger for mystery are now the most famous advocate for historical fancy.
There are now two primary theories as to what ultimately happened to the Ninth Hispana. The first is the somewhat anachronistic theory of annihilation in the mist laden landscapes of Scotland, and the re-location theory of the legion being removed from Britain in the second century AD, and being lost somewhere in the wilds of the Empire near after. The latter re-location theory is now widely accepted, and on the surface justifiably so, as it is lucid and appears to be laden with evidence. However, in the search for black and white answers, there are always shades of grey determined to put the brakes on. What follows is a re-address of the balance of the theories about the fate of the Ninth legion.
The archaeology and written records of Ancient Rome are paradoxical. The nature of the evidence that survives can be misleading, because although the Roman world left behind a staggering wealth of written records and colourful archaeology, this combined narrative is far from comprehensive.
The theories that are made and re-made are subject to this information, and therefore it is always open to more than one conclusion. We are missing vast treasure troves of information from this idolised part of history. It is all tangled within historical theoretical debates of what are facts, and the ideas of objectivity. The mystery of the Ninth legion would undoubtedly be lost upon the Romans themselves; it is our contemporary world that does not have the relevant text or material evidence that can attest to its exact destiny.
The last surviving remnant of evidence that truly attests to the Ninth Legion still being part of the military garrison of Britannia is an inscription dated c.108 AD, found in Eburacum (modern York). The inscription, found in the mid nineteenth century, is a fragment of a tablet that commemorated the building of a gateway into the legionary fortress. It states that the gateway was built under orders from the Emperor Trajan (53-117 AD). It was created in the twelfth year of Trajan’s reign, which would put it in the bracket of 107 AD and 108 AD.
‘The Emperor Caesar Nerva Trajan Augustus, son of the deified Nerva, Conqueror of Germany, Conqueror of Dacia, pontifex maximus, in his twelfth year of tribunician power, six times acclaimed emperor, five times consul, father of his country, built this gate by the agency of the Ninth Legion Hispana.‘
After this date, the Ninth legion went on to live in mysterious infamy. Later that century in around 165 AD, the Emperor Marcus Aurelius commissioned a pair of columns listing every Roman legion and its location throughout the empire. The Ninth ‘Hispana’ and Twenty-Second ‘Deiotariana’ were nowhere to be seen. This evidence simply provides a window of time. Between the years 108 AD and 165 AD the Ninth appears to have marched off the records. Under the normal circumstances of the Roman treatment of disgraced and destroyed units, the legion would be subject to the ‘damnatio memoriae’. This is an act of removing the legions name from all the monuments it had been inscribed on, however, to date there is no evidence of this happening to any monuments relating to the Ninth(1). Where did the Ninth Legion Hispana go?
The story of the Ninth can be reconciled with the archaeology and this is at the very heart of the re-location theory. Excavations at the Njimberg (‘Noviamagus’) Roman fortress at Nijmegen in the Netherlands in 1959 yielded a find that was to change the course of the Ninth’s mystery, and propel it into the turmoil of academic debate. The archaeologists were astonished to discover a roofing tile (tegula) with a stamp that looked like none other than the Ninth Hispana, it carried the weighty suggestion that the Ninth legion was there.
Another vital piece of evidence is several roof tiles with ‘VEX BRIT’ stamped upon them. The ‘Vex’ is an abbreviation of ‘vexillatio’ and ‘Brit’ is interpreted as ‘Britannica’. A ‘vexallatio’ is a common statement from the army; it means a detachment of a specific unit or legion, usually commanded by a centurion (2). Essentially the stamp proclaimed ‘a detachment from Britain’ made this.
In 1938 at the pottery-works site of De Holdeurn, less than three miles from the fortress at Nijmegen yielded a fragment of a ‘mortarium’, a very heavy and thick dish used by the army, principally for grinding food. This one was stamped ‘LGVIIIIHISP’, a supposedly clear abbreviation of the Ninth Hispana. However this discovery was not recorded in the initial report (3).
The roof tile found at Nijmegen is delicate and fragmentary, and it is possible to raise lucid conjecture as to the nature of this evidence, just by the visual quality alone. For example, the tile is broken at the end, leaving only the partial fragment of the last numeral. It reads: ‘LegVIII….’ And the last numeral is broken, not only this, the remaining broken numeral is slightly diagonal. This has led some to believe that it in actuality, it is the tile stamp of the Eighth Legion ‘Augusta’. Evidence from the Eighth has been found at Nijmegen, in the form of an inscribed silver medal found at the site, from a centurion of the Eighth, which could suggest at the very least detachments of the Eight were there at some point. Thus the diagonal and broken numeral at the end could actually be the ‘A’ from the cognomina ‘Augusta’. This can simply be seen as illustrative of the ease of which interpretations can be made.
A tile with the Ninth stamp that is virtually identical to those found at Nijmegen and De Holdeurn, and the inscription in York, with the stamp ‘LEGVIIIIIH’ was found in the tile works at Scalesceugh, south of Carlisle in Northern Britain, in 2010. They suggest a related manufacturing process, due to the use of ‘LEGVIIIIH’ instead of the ‘LEG IX HISP’ that is usually found on material from York.
However, ultimately all of these pieces have not been conclusively dated, being placed in the date range of the early second century (4). Because they are archaeologically durable building tiles; they are difficult to date without affirmative contexts.
Archaeological data recovery is a complex and precarious business; it relies on methodological extraction and recording from verified strata.
The evidence that is supplied is however, open to interpretation and this appears to be the case here. Jules Bogaer, the principle archaeologist in the discovery of the tile at Nijmegen, writes:
‘“This piece comes from the Period III level or from the destruction layer lying above it, and in fact from the third officer’s house north of the large stone gate-building on the east side of the fortress”. (Die Besatzungstruppen des legionslagers von Nijmegen IM 2. Jahrhundert Nach Christus. Bogaers, J.E, 1967, p.64)
It is suggested in the report that it is unclear which context the tile belongs to. Immediately this evidence can be essentially rendered defunct. It appears the tile was found out of context, as its exact position is unknown. The report goes on to suggest that the destruction layer is largely compiled of topsoil. This means for archaeologists, it cannot be taken as stratified evidence. Topsoil can contain finds and debris from all periods of a sites inhabitation; it’s like trying piece together a jigsaw with no edges. The crux of this is simple, no tiles or debris pertaining to the Ninth have been found in conclusive contexts, like as part of a building for example, or found underneath a piece of veritable dating evidence, a coin for example.
It is unlikely that a seasoned archaeologist such as Bogaer would base the theory of the Ninth in Nijmegen in the second century on this seemingly flimsy material record, yet understandable. The ‘Phase III’ level he describes is almost certainly second century in date due to his interpretation of the evidence. Bogaer then states that the two tiles found at Nijmegen and De Holdeurn were the only two pieces found relating to the Ninth at the sites and their surroundings (5). This partial and very slight material record of the Ninth is a shaky foundation to build a theory around.
He also supports the theory of the Ninth being at Nijmegen in the early second century, with other detachments of other legions. It would not explain the detachments manufacturing the tiles with the Ninth’s stamp if it was a mere detachment. If this was the case, it is far more likely that the Ninth was in Nijmegen in its entirety, and was thus the sole manufacturer. The principle question of when this was however remains, it could have been the early second century as the theory suggests, but the vital archaeological evidence is inconclusive.
The evidence for the Ninth could be explained via the historical context and known re-location behavioural patterns of the Roman army. Principally, the revolt of the Batavi tribe in in 69 AD (6) prompted severe and swift action from the Empire, and detachments of the Ninth from Britain could certainly have been involved in the suppression of the revolt. The victorious Roman army then planted a base in the heart of the Batavi territory, at Nijmegen. Could the tiles be from this period?
Detachments or ‘Vexillations’ were common and a feature of what made the Roman military system so extraordinarily efficient. Contrary to popular belief, legions were not static entities, and they were certainly not at full strength at any one time, this could be the result of a number of things, such as illness, leave for example. It is assumed they were at full strength at the beginnings of campaigns of conquest, yet even this is still in doubt. They were elastic in their usage, and in this flexibility resides the main ingredient for the Roman army’s success. They could be broken up, and these detachments could then be placed in theatres of conflict, and industrial usages, even in a civic policing role. A prime example is an inscription found on the Tyne, from c.158 AD. It records detachments from Britain of the VI, II, and the XXth legions being used in the two Germanic provinces as reinforcements (7). Why does the tile evidence have to be from the second century? They could equally have been a product of diffusion, or as a result of earlier events.
The evidence for the specific archaeological contexts the tiles were found in is elusive and should be treated with trepidation and abject caution. The absence of correct chronology and other evidence at this time means it cannot bear verifiable witness to the fate of the Ninth. The re-location theory is a strong one, but on this occasion it would appear too much faith has been placed in an unreliable material record. What can be said is that the Ninth legion in its entirety or detachments of which were in the area, critically however, it is unknown when.
The nature of its presentation has become dangerously close from ‘theory’ to ‘fact’. As an archaeologist, as there were only two partial and broken fragments of material, it would suggest they were merely archaeological detritus, possibly as a product of production elsewhere, and then re-used. Until there is new evidence relating to the Ninth, in the form of material remains or documentary evidence, its fate will remain in perpetual shadow. Yet therein lays the magic, because all that is left is the doorway for discussion about the possible ideas and scenarios. As long as that keeps happening, that means the Ninth Legion Hispana has never truly disappeared at all…..
In Part Two, the theory of the romantic Scottish annihilation at the hands of the tribes of Caledonia will be investigated, and an attempt to suggest the most logical solution as to what happened to the lost legion of Rome.
A special thank you to Meir Edrey of Mainz University for his translations of the texts, and to James Spry for his valued and unique insights.
Due to spatial constraints, it has been difficult to include the whole plethora of loosely associated evidence and interpretations relating to Nijmegen and its surroundings; I have attempted to stick to the central archaeological evidence rendered by Bogaers and his excavations, as this is the crux of the theories and approaches.
However, I omitted an inscribed horse harness pendant found at a villa site nearby, however this was also an artefact found out of context, and thus wasn’t included. Please see the excellent article by Duncan B. Campbell entitled ‘The Fate of the Ninth: The Curious Disappearance of Legio VIIII Hispana’ for a comprehensive evaluation of evidence, presented in a balanced and thoughtful way. It can be found here: (http://www.academia.edu/856165/The_fate_of_the_Ninth_The_curious_disappearance_of_Legio_VIIII_Hispana)
1) Cowan, R, 2007, p.227, ‘For The Glory of Rome: A History of Warriors and Warfare’, Greenhill Books, London
2) De La Bedoyere, G, 2006, p.109: ‘Roman Britain: A New History’, Thames and Hudson, London.
3) ‘Bogaers, J.E , 1967, p.64. ‘Die Besatzungstruppen des legionslagers von Nijmegen IM 2. Jahrhundert Nach Christus.’, (‘The occupying forces of the legion camp of Nijmegen IM 2nd Century after Christ’) Translated kindly by Meir Edrey, Mainz University.
4) Cowan, R, 2007, p.226, ‘For The Glory of Rome: A History of Warriors and Warfare’, Greenhill Books, London.
5) ‘Bogaers, J.E , 1967, p.64. ‘Die Besatzungstruppen des legionslagers von Nijmegen IM 2. Jahrhundert Nach Christus.’, (‘The occupying forces of the legion camp of Nijmegen IM 2nd Century after Christ’) Translated kindly by Meir Edrey, Mainz University.
6) Goldsworthy, G, 2003, ‘In the Name of Rome’, Orion Books, London.
7) De La Bedoyere, G, 2006, p.109: ‘Roman Britain: A New History’, Thames and Hudson, London.
8) ‘Bogaers, J.E , 1967, p.64. ‘Die Besatzungstruppen des legionslagers von Nijmegen IM 2. Jahrhundert Nach Christus.’, (‘The occupying forces of the legion camp of Nijmegen IM 2nd Century after Christ’) Translated kindly by Meir Edrey, Mainz University.