Interpreting the evidence

The Republican period: gladius Hispaniensis

Evidence for the use of the gladius in the Republican period comes in two principal forms. First, we have literary accounts describing the way in which the Romans employed it. Second, we have graphic archaeological evidence of what happened to the victims who came up against the weapon. An element of caution is necessary in both cases, as we shall see, but the overall picture is consistent.

Valencia, 75 bc (overleaf)

The legionaries of Pompey break into the town of Valencia in Spain in 75 BC, held by men loyal to Sertorius. Bitter street fighting ensues before the Sertorian legionaries are defeated and the survivors tortured and massacred. Here, legionaries and cavalry in the two armies are using the long gladius Hispaniensis both to thrust and to chop. The equipment of the opposing forces was virtually identical, and the widespread use of body armour and helmets meant that limbs and faces were most vulnerable to the sword.

The Greek historian Polybius, who had fought against the Romans, been captured by them, then later became the friend and confidant of the general Scipio Aemilianus, was an eyewitness to the efficacy of the gladius: ‘Besides the shield they also carry a sword, hanging on the right thigh and called a Spanish sword. This is excellent for thrusting, and both of its edges cut effectually, as the blade is very strong and firm’ (Polybius 6.23.6-7). In one passage, he highlights the importance of the sword as a thrusting weapon: ‘The Romans, on the contrary, instead of slashing continued to thrust with their swords which did not bend, the points being

9-S-

very effective. Thus, striking one blow after another on the breast or face, they slew the greater part of their adversaries’ (Polybius 2.33.6). The passage refers to combat against the Insubres in 223 BC - a time when the xiphos was supposedly still in use by the Romans. It is possible that Polybius was projecting back contemporary use into the past, however, this being something ancient historians were prone to doing. A dual usage of the sword can be seen in a later passage, where Polybius is describing the spacing of a legionary battle line and the style of fighting employed:

Now in the case of the Romans also each soldier with his arms occupies a space of three feet [89cm] in breadth, but as in their mode of fighting each man must move separately, as he has to cover his person with his long shield, turning to meet each expected blow, and as he uses his sword both for cutting and thrusting it is obvious that a looser order is required, and each man must be at a distance of at least three feet from the man next him in the same rank and those in front of and behind him, if they are to be of proper use. (Polybius 18.30.6-8)

This, from Polybius’ own time, in the middle of the 2nd century BC, was well after the adoption of the gladius Hispaniensis. Livy was writing much later, under the Emperor Augustus, but the problem of backward projection with him is even greater. His history of Rome was compiled entirely from earlier writers, when dealing with the Republican period. His graphic description of the effect of the gladius (see above, p. 4) evokes its psychological, as well as actual, effect. Excavations at the Spanish sites of Cerro de la Cruz (150-125 BC) and Valencia (75 BC) have borne out Livy’s account, with skeletons at both sites showing clear evidence of the brutal attacks to which the victims had been subjected, including limbs hacked off. They are all-too-obviously evidence of the sword having been used for cutting blows.

At Cerro de la Cruz, two articulated skeletons (UE 1401 and 1402) were found face down in the street with clear signs of trauma. Skeleton 1401, a young male in his 20s, had had his right shoulder blade sliced through cleanly by a blade, possibly the result of a failed attempt at decapitation. There was another injury to his spine, inflicted from behind and probably with a downward motion (Quesada Sanz et al. 2014).

At Valencia, the excavated skeletal evidence was even more graphic and almost certainly the result of post-combat atrocities, rather than indicative of combat itself. One individual had been spitted lengthwise on a pilum and then had all four of his limbs hacked off. Another had had his severed head placed between his legs. There seems little room for doubt that these incidents were the handiwork of the gladius Hispaniensis (James 2011: 100).

Writing under Augustus, but describing events in the 4th century BC, Dionysius of Halicarnassus describes the use of the Roman sword:

For while their foes were still raising their swords aloft, they would duck under their arms, holding up their shields, and then, stooping and crouching low, they would render vain and useless the blows of the 36 others, which were aimed too high, while for their own part, holding

their swords straight out, they would strike their opponents in the groins, pierce their sides, and drive their blows through their breasts into their vitals. And if they saw any of them keeping these parts of their bodies protected, they would cut the tendons of their knees or ankles and topple them to the ground roaring and biting their shields and uttering cries resembling the howling of wild beasts. (Dionysius Rom. Antiq. 14.10.2)

A Mainz-type gladius from one of the Bay of Naples sites, possibly Pompeii (Italy), probably the latest datable example (AD 79). (Miks A588) (Photo: © Salvatore Ortisi)

It is likely that Dionysius too is anachronistically describing contemporary use of the gladius and simply projecting that back into the past. The long blade of the gladius Hispaniensis, together with the flat neckguards of contemporary helmets of the Montefortino and Coolus types, have even been suggested as indicative of a crouched style of fighting among the infantry, something the passage from Dionysius could be taken as supporting (Connolly 1991b). The use of a longer blade would certainly have suited cavalry as well as infantry.

We are reminded again of Plutarch relating how Cassius Scaeva, one of Caesar’s most experienced centurions, lopped off the shoulder of an opponent with his sword, despite being grievously wounded.

The Early Imperial period: the Mainz type

One of the most important sources for the way in which the gladius was used within the Imperial Army is the Tate Roman writer Vegetius. However, it is necessary to understand that, as an epitomator, he primarily gathered and reproduced earlier sources to form his work. This helps explain why his text is sometimes contradictory, particularly when describing how the sword should be used. In his first book of the De Re Militari, originally a stand-alone document intended to persuade a Tate Roman emperor to re-adopt the old Roman ways of waging war, we may recall the injunction that ‘they learned to strike by stabbing, not by cutting’ (Vegetius, DRM 1.12 trans. author). It has been argued that this passage derives from a lost portion of the work of the Early Imperial encyclopaedist Celsus, who would presumably have been using other writers in turn. That the comment was nevertheless applicable to his time is a not unreasonable deduction to make, given his likely audience - Romans of the senatorial and equestrian classes, many of whom would have seen service as officers in the Army, so unlikely to have been impressed by inaccuracies.

The Mainz-type blade retained the long point of the gladius Hispaniensis but was much smaller in overall size. It seems likely that this reflected a desire to change some of the less desirable characteristics of the bigger, earlier swords, most notably the weight. There was a price to pay for this, however, as the shorter weapon was of less use to a cavalryman, who needed a long reach, hence the adoption of the spatha. This was simply a Romanized version of the long Iron Age blades used by the Gallic and Iberian cavalrymen opposing and in the service of the Roman Army. This new, shorter gladius could still be used for chopping where necessary, but was now becoming much better suited to an economical and less open

\ i,J I5f fjjjgjfx		fry/ /f £ • J . ^FZJtL «*«, mj
	: ' >v.	\EM^WhW JEsotJT*''

Dacia, AD 101-02 (previous pages)

An encounter during the first Dacian War of AD 101-02 sees legionaries in hand-to-hand combat with Dacian warriors, who have been driven back to their baggage train. The Romans are using their gladii (mostly of the newer Pompeii type but with a few Mainz types visible) in a variety of ways - thrusting, chopping, and stabbing like a dagger. All of these poses are depicted on the Tropaeum Traiani at Adamclisi, a monument probably produced by the Army, unlike the more formal (and arguably more fanciful) helical frieze on Trajan's Column in Rome. The Adamclisi monument shows the reality behind the cut/thrust debate evident in the Roman literary sources: in the field, the Army employed whatever technique suited their purpose.

style of hand-to-hand combat centred around stabbing. Shorter blades and deeper neckguards on helmets may be indicative of a more upright fighting stance for infantrymen.

A fine series of tombstone reliefs from the Rhineland make it plain that both auxiliary infantrymen and legionaries used the Mainz type of sword.

The High Imperial period: the Pompeii type

The classic fighting stance of the gladius-wielding legionary is depicted on a sculpted column base from the headquarters building of the legionary fortress at Mainz-Mogontiacum and dating to the second half of the 1st century AD. One man stands with his shield raised in front of him and his left foot advanced, while his Pompeii-type gladius is held horizontally at the ready by his side, poised to strike. In another passage in Vegetius - almost certainly derived from the writings of Iulius Frontinus, who commanded armies in both Britannia and Germania Inferior (Pliny the Younger said he was one of the two most respected men of his time) - we find troops being drilled in swordplay for ‘stabbing and cutting’ (DRM 3.4). Frontinus was writing in the same period as the Mainz sculpted relief, but appears not to have favoured a change in the style of Roman fighting. He clearly felt that the chop was as good as the stab, but in this he may have been old-fashioned (and possibly influenced by writers describing the Republican period).

The Pompeii-type sword was the result of this evolution in fighting styles. Parallel-edged and with a short point, it was clearly influenced by the form of the spatha and hence the last vestiges of the old Celtiberian sword encountered in Spain were evolved out.

OPPOSITE Adamclisi Metope 9 depicting a short sword ready to stab. (Photo: © J.C.N. Coulston)

Arguably one of the most important conflicts for the Romans of this period culminated in Trajan’s two wars in Dacia (AD 101/02 and 105/06), where the interface between Celtic, Germanic and steppe peoples made for some interesting opponents. Ironically, the monument that so aptly epitomizes the Dacian Wars - Trajan’s Column, in his forum in Rome - is of little value to the study of the short sword and its use. Most such weapons depicted on the Column were evidently supplied as metal attachments, now long since vanished, and what remains was largely generically depicted, and of dubious evidential value anyway. Fortunately, the other monument related to this conflict, the Tropaeum Traiani at Adamclisi in the Dobruja

	- - i’ - . V- - ' W '

' - *•

■. ~-;Sv	S^vip!
	3J*S?
	Sp-S* c'j

Adamclisi Metope 32 depicting a short sword stabbing. (Photo: © J.C.N. Coulston)

(Romania), is far more useful. Probably produced by military craftsmen, rather than the metropolitan sculptors responsible for Trajan’s Column, it is the nearest thing we have to eyewitness depictions of the frontier troops involved in these wars. Here the short sword is depicted stabbing, chopping and even employed for an inverted (thumb upward) dagger-like thrust from above (similar to that used by gladiators to despatch defeated opponents in the arena who had unsuccessfully asked for mercy). This last action might seem somewhat implausible, were it not for the fact that a cherub is shown using exactly the same grip on a relief from the Temple of Venus Genetrix in the Forum Iulium in Rome (now in Naples Museum, Italy).

The men depicted on the Adamclisi metopes all fight in a truly upright stance, but one incidental detail reflects upon the style of fighting of Rome’s opponents: the legionaries all wear the articulated armguard (manica). Armguards had been used in gladiatorial combat since at least

The swords of Pompeii

It is worth asking why so many swords have been found in Pompeii and other Bay of Naples sites when many Roman military bases produce very few by comparison. Much depends on understanding what these finds represent: are they a snapshot of everyday life, forever frozen in time by the eruption of Vesuvius, or do they reflect the unusual circumstances surrounding that natural disaster? One possibility is that the weapons may have belonged to inhabitants within the towns and villas, while another refines this to focus on veterans who may have retired to the area and kept their weapons.

One of the more intriguing aspects of the official Roman response to the disaster, however, was the deployment of the nearby fleet at Misenum under its commander, Pliny the Elder. Every ship in the fleet would have had its own contingent of marines, equipped in much the same way as land-based infantry and commanded by a centurion. Indeed, on at least two occasions, legions (Iand IIAdiutrix) were raised from marines, since they provided a ready-trained, ready-equipped fighting force.

The Bay of Naples swords may therefore have come from marines active in the area during the eruption (Ortisi 2015). One at least - the sword on the beach at Herculaneum - was found with its owner, who was also carrying a bag of tools and whose physique suggested he rode horses and had well-developed musculature in his right arm. If just a craftsman, then why was he carrying the sword and other military equipment? We will never know for sure, but it is at least possible that he was a marine from the Misenum fleet.

The sword (and other military equipment, including helmets) found at the gladiator barracks in Pompeii suggests that the site may have been used as a focus for the marines' activities while assisting the town in its time of need.

the first half of the 1st century AD and are depicted in graffiti from Pompeii. Clearly the fear of injury to or even loss of that sword arm had led to the adoption of this type of armour by the Army around the middle of the century, as evidenced by tombstones, but armguards were certainly known before and used after the Dacian Wars, so it is not true that (as the archaeologist Sir Ian Richmond once suggested) they were a particular countermeasure adopted against the Dacian falx. Indeed, the largest single find of armguards to date comes from a Hadrianic deposit in the fort at Carlisle (Cumbria, England), on the northern British frontier.

Both ferrous and brass plates were used for armguards (although they were not mixed), riveted to an underlying base of (usually three or four) longitudinal leather straps to provide an extremely flexible defence. Just as the hand guard plate caught stray weapons blows to the sword blade, so the armguard deflected blows off the forearm. Because the armour sat on top of the arm, where it offered the most protection when the sword was employed for the thrust, its plates overlapped upwards from the wrist, and - due to natural bunching - were at their most concentrated over the inside of the elbow joint, a blow to the wearer’s sword arm would skip up to the elbow where it was stopped dead. Roman soldiers clearly felt that their sword arms were vulnerable and the articulated armguard was equally obviously a neat solution to their concerns. Thus, while the Adamclisi reliefs show that legionaries were still quite happy to raise their sword arms to deliver a cut, the manica subtly encouraged them to keep the arm down and thrust instead.

The Carlisle armguards had all been repaired several times and at least one was a ‘cut’n’shut’ formed from two donor defences. Such repairs suggest that Roman soldiers in Britain during the early 2nd century AD were only too happy to have their sword arms protected in this way.

THE GLADIUS IN SERVICE Ownership

Roman soldiers were expected to provide their own equipment for service. This could be bought from the Army (some new; most probably bought back from retiring soldiers) or from private suppliers (most of whom were presumably retired military craftsmen). This meant that soldiers were keen to indicate their ownership on equipment, and swords were no exception to this. While manufacturers tended to stamp blades, owners would mark their possessions in a different manner. The classic means of indicating a soldier’s place within the Roman military system was by giving his centurion’s name and his own name. On some equipment this might be done by scratching the letters into the metal, but the most common method was by forming the letters out of a series of dots using a punch, a technique nowadays known as punctim. The first technique is found on part of a Mainz-type scabbard found at Kalkriese and associated with the defeat of Varus in AD 9 (or, some argue, its aftermath, when Germanicus ‘cleaned up’ the site), naming a Doxitus or Roxitus as centurion. The second technique was used on the Pompeii-type sword from the Rhine at Mainz now in the Royal Armouries in Leeds. The name of C. Valerius Primus occurs four times on the blade (at least two of them with the centurial symbol >) and C. Ranius once (Schoppa 1974). This suggests that Primus was not only the centurion but also the owner of the sword at one stage, perhaps selling it on to Ranius, a member of his century. The ownership of weapons is reflected in Roman military law, too, for the Digest records (49.16.3.14) that theft of another’s weapons resulted in reduction in rank. Discarding body armour, shield, helmet or sword were treated as desertion (49.16.14.1). On this point, Frontinus (Stratagems 4.5.17) relates how the loss of his sword drove Marcus Cato, son of Cato the Censor, to wade back into the midst of the enemy to retrieve it in order to avoid disgrace. Despite being wounded, he succeeded in doing so.

Since swords were the personal property of soldiers and not owned by the Army, it is worth pondering why so many intact swords and scabbards come from rivers, while excavations on sites tend only to produce components. Such riverine deposits, like those in the Rhine at Mainz, occur at locations that have long traditions of deposition from prehistory right up until the early medieval period. Many archaeologists see these deposits as having ritual origins, and there is certainly indisputable evidence for the dedication of weaponry and equipment in a religious context during the Roman period (as happened in Germany and Gaul with supposed examples of Julius Caesar’s swords, or at the temple at Klein-Winternheim already mentioned). These deposits would

also seem to imply that a soldier offering a sword had already purchased a replacement, if he were not to fall foul of Roman military law. Indeed, model swords were sometimes offered in place of the real thing. In any contract religion, it is of course cherished items that tend to be offered, so the finds from rivers tell us how much these weapons mattered to their owners. In the same way, the detailed representations of swords and scabbards on tombstones of the 1st century AD reflect the soldiers’ pride in their appearance and their military equipment.

Information on the cost of swords is sparse, to say the least. A papyrus letter (P. Giss. 47) dating to the second decade of the 2nd century AD from an unknown individual to Apollonios, a strategos (local governor) in Egypt, records a price of 80 drachmae (equivalent to 20 denarii) for an ‘Italian sword’, this being just under a quarter of the price of a brass cuirass he mentions in the same letter. Our anonymous correspondent, possibly an officer, clearly viewed these as bargains, presumably offered by a private craftsman in Koptos. At that time, a Roman legionary could expect to earn 300 denarii per annum. While this Egyptian price may have reflected the market value for a sword, we might anticipate that (as with other equipment) a soldier buying his sword directly from the Army would have paid less, since the price would have been nominal (and, effectively, subsidized).

The quality of the scabbards of

Front and rear views of a Mainz- type sword and scabbard from the Rhine at Mainz (Germany) (Miks A481). (© GDKE-Landesmuseum Mainz; photo: Ursula Rudischer)

lst-century AD swords has often led modern commentators, swayed by art-historical considerations, to suggest them to be officers’ weapons, the Sword of Tiberius being the classic example of this. However, analysis of the materials used (tinned brass plate) and the methods of production (stamping) suggest that even ordinary soldiers boasted fine weapons; and this is supported by both the tombstone evidence and comments from various writers, including Pliny the Elder.

Carriage

For most of its period of use, the short sword was carried in a scabbard on the wearer’s right-hand side. Scabbards were generally positioned so that the mouth was above the level of the waist, and reliefs such as the Adamclisi metopes frequently show the pommel at near-armpit level. This led at least one modern commentator to complain that it would not have been possible to draw a sword worn thus - an assertion bemused re­enactors have always been only too happy to disprove. In fact, the gladius can be easily drawn by inverting the right hand, thumb downwards, then grasping the handgrip and pulling straight upwards. It is a quite natural progression to continue this movement forwards to bring the sword down to the side, point forwards, in the characteristic ‘at the ready’ position depicted on relief sculpture. It should not be overlooked, however, that the actual process of drawing the sword presents different aspects of the weapon to its user: first, as the sword emerges, there is the pommel, with its ‘skull-crusher’ top nut; then comes the raised blade for a quick chop; finally, as the blade comes down to the side and is drawn back, it is ready for a more leisurely and considered thrust.

One of the ways in which centurions distinguished themselves from

lesser infantrymen was by wearing their swords on the left. They merely

had to draw the weapon across themselves to achieve the same end. Neither

technique presented any problems. The sword and its belt could, however,

prove to be a liability at times. The Herculaneum ‘soldier’ was carrying his

decorated sword belt wrapped around his sword, for some reason, while Scabbard fittings belonging to _# . ....

Mainz-type _sword_s (D_rawing_s- Trajan’s Column depicts a legionary fording a river with his body armour,

M.C. Bishop) tunic and sword belt held aloft on his shield to keep them dry.

In the Republican and Early Imperial periods, the gladius scabbard always seems to have been suspended from a waist belt. Although Roman

belt plates are common in the Imperial period, from the Republican era there is a marked dearth of them. This may be partly explained by the tombstone of the centurion Minucius, which shows a sword belt with no plates; and in this it is matched by the (arguably more stylized and perhaps less reliable) so-called Altar of Domitius Ahenobarbus relief, with its depiction of a military levy of the Late Republican period under way. Minucius’ tombstone provides further detail of the manner of sword suspension, with three straps wrapped once around the scabbard below the upper suspension band. It is unclear how this was then attached to the belt, but the dagger worn next to it is suspended by means of surplus material from the belt terminal wrapped around it and tucked back over the belt itself. Since the suspension loops are not prominent on the sword or the dagger, they

were presumably held flat against the scabbard by the wrapping process, while the straps must also have threaded through them in order to secure the weapons.

The more familiar use of the four-ring system of suspension is found in depictions of the Early Imperial period. Here, as shown by tombstones and finds from Windisch and Herculaneum, large decorated buttons attached to the belt evolved into frogs that formed an integral part of belt sets. Initially two belts were needed, one for the sword and another for the dagger, frequently worn crossed (‘cowboy-style’) or one above the other. By the end of the 1st century AD, it seems to have become common to see only one belt being worn, with the sword now suspended from a baldric. The baldric, sometimes studded if tombstone depictions are to be believed, was looped over the left shoulder (right, if a centurion). Re-enactors have discovered that, if it is left unfettered, the baldric has an alarming tendency to move and runs the risk of the scabbard becoming entangled in the users’ legs. Their solution is often to secure the baldric under the waist belt, but whether the Romans used a similar approach is unknown (and probably unknowable). No fasteners were used with the baldric (it just slipped over the head) and objects sometimes identified as ‘baldric fasteners’ in fact come from cavalry harness.

A variety of Pompeii-type scabbard fittings. (Drawings: M.C. Bishop)

Possible die for stamping decorated scabbard plates from Nijmegen. (Drawing: M.C. Bishop)

Since both the Delos and Herculaneum swords were found with small buckles, this has led to attempts to work out precisely how the suspension rings were used to attach the scabbard to the belt (Connolly 1991a). It has long been known that sets of belt fittings included frogs, but precisely how the sword was attached to these has remained uncertain. To the rear of some surviving Pompeii-type scabbards, there is often an arched fitting between the upper and lower suspension bands, ostensibly just holding together the side guttering, but it is in fact raised in such a way that suggests it had another function and that straps could be looped under it. Since frogs were fixed to the belt, in some cases riveted as belt-plates were, any adjustment had to be provided for in the way the scabbard was

Mainz-type scabbard from the Rhine at Mainz (Germany) (Miks A479). (© GDKE-Landesmuseum Mainz; photo: Ursula Rudischer)

attached to the frogs, and that could allow for a simple loop to pass through the rear scabbard plate, through the rings, and around the frogs. Some of the Mainz-type scabbards have small loops in this position, and these may have been used in combination with the suspension rings to secure the scabbard to the belt or baldric. Tombstones clearly show that the top pair of suspension rings were the ones attached to the belt during the early 1st century AD; but the four-ring system potentially allowed a lot of flexibility for sword suspension, and no one system need have been in use all the time or even everywhere at the same time.

Pliny the Elder described how soldiers liked to personalize both their swords and their scabbards: ‘But of what use is it to collect all these instances, when our very soldiers, holding ivory even in contempt, have the hilts of their swords made of chased silver? When, too, their scabbards are heard to jingle with their silver chains, and their belts with the plates of silver with which they are inlaid?’ (Pliny the Elder, Natural History 33.152). The silvered hilt on the Rheingonheim sword springs instantly to mind, as do the elaborate chains on the Herculaneum sword scabbard (also depicted on Rhineland tombstones).

The fact that Roman historians liked to highlight the unusual can serve to underline normal usage. A splendid example of this can be found in the account of some Roman legionaries creating a diversion during the siege of a fortress by Marius during the Jugurthine War (112-106 BC), doing so by strapping their shields and swords to their backs in order to make it easier for them to climb (Frontinus, Strat. 3.9.3). These were, the subtext runs, not normal things for Romans to do, and thus worthy of note.

Once the gladius was replaced by the spatha during the 2nd century AD, sword carriage was universally switched to the left hip, because the longer blade could not so easily be drawn from the right (although cavalrymen, who had been using the weapon for longer and wore it lower on the right hip, had coped). The four-ring system of suspension was at the same time replaced by a scabbard slide on the front of the sheath, through which the baldric passed before looping around the scabbard.

Care and maintenance

Even the finest sword needs regular attention to keep it in peak form. As Fronto, tutor to and friend of the joint emperors Marcus Aurelius and Lucius Verus, observed: ‘If one must fight with a gladius, it makes a difference whether it be rusty or bright’ (Fronto, On Eloquence 1.16).

Decoration and display

The advent of the principate, founded by Augustus, saw new forms of display introduced into state art and into the decoration of Roman military equipment; and nowhere was this more evident than in the adornment of sword scabbards. The Sword of Tiberius is perhaps the most obvious example of this, with an embossed scene depicting Tiberius receiving a triumphant Germanicus on its locket plate, a head of Tiberius on a medallion lower down, near the centre of the sheath, and at the bottom a chape with two more embossed scenes. One of the latter scenes represents the shrine of the standards in a legionary headquarters building containing an eagle standard (Germanicus had recovered two of the three eagle standards lost in the Varian disaster of AD 9); the other shows an axe-wielding female figure thought to be a personification of Vindelicia (the area around Augsburg, now in southern Germany), recalling Tiberius' and his brother Drusus' suppression of the region in 15 BC. Germanicus, Tiberius' nephew, was revered by the Army - so much so that his birthday was still celebrated two centuries after his death - while Tiberius himself was a successful general under Augustus. Standards were viewed as sacred within the Army and the place where they were kept and displayed was the focus of every military base. All of the images on this one sword scabbard reflected, and perhaps were designed to engage, the loyalty of the soldier. Such decorative schemes have often been viewed as crude state propaganda, but this perhaps underestimates the role of the tastes of the soldiers themselves.

The Imperial family features on other Mainz-type scabbard fittings, with locket plates depicting Augustus' daughter Julia and her two sons, Gaius and Lucius (Augustus' chosen but doomed heirs), as well as Tiberius addressing assembled troops. There are also scenes of kneeling captives on locket plates, as well as trophies of captured arms and horsemen trampling barbarians on chapes, to inspire the soldier idly inspecting his scabbard. The she- wolf and the twins Romulus and Remus from the foundation myth of Rome helped bolster the general feeling of military success on the Fulham Sword (along with further more fragmentary examples), while unit identity and perhaps even esprit de corps were evoked by depictions of the thunderbolt-and-lightning rods on scabbards from Windisch and Strasbourg; devices used to adorn the faces of legionary shields. Many of these designs are echoed on contemporary belt plates, reminding us how important the sword and belt were to the identity of the soldier.

The focus of Imperial imagery had shifted subtly by the time Pompeii-type scabbards were being produced. The Imperial family were no longer central to the decorative themes, but rather Roman military might and mythological scenes. Mars and Victoria are

ABOVE Detail of the tombstone of Annaius Daverzus from Bingen (Germany) showing his Mainz-type sword and scabbard with chain (catena) ornament. (Photo: M.C. Bishop)

frequently depicted, with captives and trophies also being among the familiar motifs.

It is unlikely that these state-derived images were imposed upon the troops from on high - there was no discernible centralized control of Roman weaponry in the period when the gladius dominated - and they almost certainly reflect the aesthetic tastes of the troops themselves. That they were receptive to both the subtle and not-so-subtle messages is possibly a measure of the success of the Imperial 'project'.

Hone or whetstone from Carlisle (Cumbria, England). (Drawing: M.C. Bishop)

OPPOSITE Column base from the legionary headquarters building at Mainz (Germany) depicting two legionaries, one with a drawn Pompeii-type gladius in the classic stance. (Photo: © M.C. Bishop)

We do not know for sure how blades were maintained, although some combination of sand and olive oil may have been used as an abrasive to keep them clean and polished. Mineral oil was certainly known of, Pliny the Elder mentioning that smiths applied it to exposed nail heads and that it had ‘many other applications’ (Pliny the Elder, Nat. Hist. 35.182). We are better informed about how their edges were kept sharp, however, because hones or whetstones survive from many Roman military sites. Normally rectangular in shape and of a size to fit comfortably in the hand, they were also rectangular in section and made of fine-grained stones. Excavated examples frequently have concave faces caused by long use. Repeated sharpening of a blade with a hone would inevitably lead to some erosion of the steel edges. It is possible that this to some extent enhanced the waisted shape of Mainz-type blades, but there seems little doubt that the basic shape was a product of the forge and not a result of over­enthusiastic sharpening, not least because the broadest part of the blade above the point might be the bit to do the cutting.

Blades could also be blunted or nicked and might require repair at the forge, while those that are found without any apparent defect may well have been set aside to have their organic handle components renewed or repaired. Those organic components would themselves have required light care, in the form of cleaning and drying after use in order to prevent them drying out or rotting, and occasional oiling. Encased hilts, like that of the silvered Rheingonheim example, may have been more problematic in this respect.

Training

Training was key to the success of the Roman Army and practice with the sword was a fundamental part of the process. In 209 BC, having captured Carthago Nova, Scipio Africanus introduced new training methods for his army on campaign in Spain that included characteristic gladiatorial components such as sword drill with wooden and blunted weapons (Polybius 10.20.1-7; Livy 26.51). The use of buttons to blunt the wooden swords is mentioned by Polybius and implies first that they were shaped to include a pointed tip and second that stabbing was practised even then. It is important to note that these were not raw recruits and no mention is made of sparring against the stake.

A similar ploy was used by Publius Rutilius Rufus in 105 BC, using instructors from a gladiatorial school to teach his legionaries to be more flexible in parrying and delivering blows (Valerius Maximus 2.3.2). Soon after, Marius employed soldiers trained by those same instructors to suppress the Cimbri, so it seems there is a pattern to this (Coulston 1998: 2-3).

Moreover, the implication to be drawn from this close relationship with gladiatorial sword-handling skills is that, in battle, legionaries actually fought hand-to-hand in a similar manner to gladiators. It was not identical, however, since gladiators generally only fought in matched pairs and were intended to provide an entertaining display and not a quick kill. Nevertheless, they needed to know which parts of an opponent’s body

were the most vulnerable in order to avoid them until a kill was required. The impression of similarity is further reinforced when it is noted that the pose depicted in the reliefs on both the Mainz column base and the metopes of the Tropaeum Traiani is in fact echoed on numerous representations of gladiatorial combat, both in sculpture and in mosaics: left foot advanced, shield held close to the body, sword held horizontally at the ready by the side. Roman military specialist Adrian Goldsworthy (himself a fencer) noted that this pose placed the weight of the soldier behind his shield and not behind his sword arm, which may seem counter­intuitive to a modern swordsman.

Vegetius preserves more than one account of the type of sword training that was undertaken by recruits and known as armatura (Vegetius, DRM 1.11-12; 2.23; 3.4). The common theme is that new recruits must practise at the stake (palus) using a double-weight wooden sword or singlestick (rudis) and carrying a double-weight wicker shield. Why double-weight? The logic was simple: ‘Moreover, they are given that double-weight shield frame and singlestick, so that when a recruit took up real, lighter, weapons - as if freed from the heavier weight - he therefore fought more safely and more quickly’ (Vegetius, DRM 1.12 trans. author).

Even in the 20th century, the advantage of a weight differential in training weapons was not lost on singlestick swordsmen (Allanson-Winn & Phillipps-Wolley 1905: 60). This methodology is borrowed directly from gladiatorial training, so by the Imperial period, the Republican experiments had clearly become embedded in the military system. This sword drill taught specific aspects of combat:

The ancients, as is recorded in the books, trained recruits in this way. They wove rounded shields of wicker like basketry, in such a way that the frame should be double the weight of a battle shield. And likewise they gave the recruits wooden foils, also double weight, in place of swords. And next they were trained at the stakes, not only in the morning, but also in the afternoon. For the use of stakes is particularly advantageous not only for soldiers but also for gladiators. And neither arena nor field ever proved a man invincible in arms, unless he was carefully taught training at the stake. However, single stakes were fastened in the ground by each recruit, in such a way that they did not wobble and protruded for six feet. The recruit practised against this stake with the wicker shield and singlestick as though with a sword and shield against an enemy; so, he might aim for the head or face, then he is threatened from the sides, then he strained to cut down at the hams and shins; he retreated, attacked, leaped in, as if the enemy were present; he assaulted the stake forcefully, fighting skilfully. In doing this, care was taken that the recruit rose up in this way in order to wound, but did not lay himself open to a blow anywhere. (Vegetius, DRM 1.11 trans. author)

OPPOSITE Relief from Pula (Croatia) depicting a Pompeii-type sword in its scabbard. (Photo: © 52 J.C.N. Coulston)

Three principal target areas are defined in this passage: the head and face, the sides (or upper body) and the lower limbs. There is more, however, on how this must be done:

Wmm.

The gladius in the arena

The curious two-way relationship between the Roman Army and the world of the gladiator is more than just the stuff of fiction. For just as gladiatorial training made a huge contribution to the operations of the Roman Army, so studying their equipment can help fill in missing details for the military use of the gladius.

No certain example of a gladiatorial short sword has ever been recovered. One of the Pompeii swords certainly came from the gladiatorial barracks in that city, but since (like the others) it was found with its sheath, it is unlikely to have belonged to a gladiator; they had no need of a scabbard, after all. There is, however, a rich variety of depictions of the short sword in use in the arena on stone, metalwork, ceramics, in graffiti and mosaic, as fresco, and even painted onto glass.

One of the most impressive 1st-century BC sculpted reliefs comes from Lucus Feroniae in Italy. Part of this shows one man administering the coup de grace with a dagger in his left hand, while resting a gladius Hispaniensis on his right shoulder. A lanyard, which is tied around his arm, is clearly visible attached to the top nut on the pommel of the sword. This was presumably a measure to prevent him accidentally losing the weapon during combat.

It is interesting to note that one type of top nut from military Mainz-type gladii features loops on either side for small rings, possibly for the attachment of a thong or strap, while the Rheingonheim sword is said to have had a loop attached to the pommel when it was first found in the 19th century. This was subsequently removed during rather heavy-handed conservation of the weapon.

Another set of reliefs from Rome, now in the Munich Glyptothek, shows an armoured gladiator with his sword (with a fluted pommel and hand guard) raised above his head to deliver a downwards chop. This clearly implies that gladiators, just like the Roman Army, were as happy to cut as they were to stab.

Finally, a few depictions of gladiators (including funerary reliefs and a mosaic) provide us with a detail that is lacking from military iconography: depictions of the palus, the man­sized stake against which sword drill was carried out.

OPPOSITE Gladiators from Lucus Feroniae (Italy). While the victor executes his victim with a dagger in his left hand, he rests a gladius Hispaniensis (which is attached to his wrist by a thong) on his right shoulder. (Photo: © Sophie Hay)

Further, they learned to strike by stabbing, not by cutting. For the Romans not only easily conquered those who fought by cutting but mocked them too. For a cut, should it be delivered with any force whatsoever, frequently does not kill, when the vital parts are defended by equipment and bone. On the contrary, a point brought to bear is fatal at two inches [49mm]; for it inevitably penetrates whatever vital parts it is stuck into. Next, when cutting, the right arm and flank are exposed, but the point is delivered with the protection of the body and wounds the enemy before he sees it. (Vegetius, DRM 1.12 trans. author)

Clearly, the emphasis here is on stabbing, not cutting, but elsewhere (probably using another source) Vegetius is less fussy: ‘It is also appropriate to be trained at the stake or with singlesticks, when they should learn thoroughly to attack with a stab and a cut to the sides, the feet, or the head’ (Vegetius, DRM 2.23 trans. author).

The issue of continual training is even raised as one way of keeping troops gainfully employed prior to battle: ‘They are very frequently occupied in striking with wooden singlesticks in place of swords, stabbing

and cutting for many days until they sweat’ (Vegetius, DRM 3.4 trans. author). Here Vegetius is almost certainly using Julius Frontinus’ lost work on military science (to which Frontinus’ surviving Strategemata was intended as an appendix).

The attention to striking the face of the victim is found elsewhere in Roman literature. It occurs in Polybius (2.33.6) and in the tales of Cassius Scaeva in Caesar (and several other writers), who struck an opponent full in the face with his sword (Plutarch, Caesar 16.4) while incidentally hacking the arm off another! Another of Caesar’s centurions, Crastinus, was found dead on the field at Pharsalus having similarly been stabbed in the face (Caesar, BC 3.99). We also find it mentioned in Tacitus’ accounts of battles on the Weser (Tacitus, Annals 2.14) and at Mons Graupius (Tacitus, Agricola 36.2), and it inevitably caused the victim to flinch and could force them off-balance even if the weapon did not strike home.

Attacking the torso was of course harder, since even unarmoured enemies normally tended to fight with shields. Nevertheless, the Roman soldier would have looked for opportunities, particularly when a foe raised an arm to strike overarm with a sword, as this area provided the greatest chance of a fatal wound. The feet offered an unexpected and vulnerable target which, when struck, could distract sufficiently to allow a mortal blow elsewhere.

The second form of training was for the more experienced soldiers and this was the mock battle, still using wooden staves instead of swords (Onasander, Strategikos 10.4-6). Once the basic moves had been learnt, this second stage obviously introduced a more dynamic element, with moving opponents who could and would strike back.

No wooden training equipment has survived from the Roman period. A gladius-shaped wooden sword was excavated from the Roman fort at Carlisle, but it was certainly not double the weight of a battle sword and is often explained as a child’s toy (Caruana 1991), although a votive substitute weapon is also possible. The rudis presented to a gladiator upon retirement seems to have been just such a double-weight training stave.