The Roman spatha

The spatha was a type of straight and long sword, with a length between 0.5 and 1 meter, and a grip of 18 to 20 centimeters. It was used in the territory of the Roman Empire from the 1st to the 6th century AD. Later swords from the 7th to the 10th century, such as the Viking swords, are recognizable derivatives and are sometimes also referred to with the term spatha.

The spatha was widely used by Germanic warriors. It is unclear whether the sword originated from the gladius or from the longer Celtic swords. The spatha remained popular during the Migration Period. By the 12th century, it had evolved into the arming sword of the High Middle Ages.

The word comes from the Latin spatha, which is derived from the Greek word σπάθη (spáthē), meaning "any broad blade, of wood or metal", but also "broad sword blade". The Greek word σπάθη was used in the early archaic period for various types of Iron Age swords. The word 'spatha' does not appear in the Ancient Greek of Homer, but is mentioned in the works of Alcaeus of Mytilene (6th century BC) and Theophrastus (4th century BC). It is likely that spatha is the Latinized form of the Doric Greek word σπάθα (spáthā). The word survives in Modern Greek as σπάθη and σπαθί. The Latin word became épée in French, espasa in Catalan and Occitan, espada in Portuguese and Spanish, spada in Italian, spadă in Romanian, and shpata in Albanian — all meaning "sword". The English word spatula comes from the Latin spat(h)ula, the diminutive of spatha. The English spade, derived from Old English spadu or spædu, is a Germanic cognate word, derived from the Common Germanic spadō, which ultimately traces back to a Proto-Indo-European root word sph₂-dh-.

For cavalrymen, the spear was always the primary weapon. Thanks to the greater reach and the power generated by the speed of the horse, the spear could be used as a massive thrusting weapon. As a secondary weapon, riders typically carried a sword.

In the early phases of Roman history — around the 8th century BC — we already find indications of the use of swords by cavalrymen. For example, a bronze antenna sword  was found on the Esquiline in Rome, indicating northern influences. However, swords in Central Italy seem to have become rare in the following century. The emphasis at that time was clearly on spear and javelin, which were used for fighting from a distance. For hand-to-hand combat, some warriors carried daggers, possibly also the chariot fighters.

From the 6th century BC, Greek influence on armament gained ground. However, following Greek military relations, there seem to have been no specific sidearms for riders for a long time. Instead, cavalrymen used the same swords as the infantry. This included the Greek xiphos (ξίφος), a double-edged sword of about 60 cm, suitable for both cutting and thrusting movements. Less common, but recommended by Xenophon for cavalry, was the machaira (μάχαιρα), also known as kopis or falx: a single-edged chopping sword with a wide blade curved towards the edge. This type was also widely used by the Iberians.

After the Gallic sack of Rome around 390 BC, a significant weapon development seems to have occurred. Celtic swords were introduced into the Roman army, both in the infantry and cavalry. The Romans adopted these as a direct response to their military defeats against the Gauls. From these interactions later developed, from the 3rd century BC, the famous gladius — a straight sword with a wide blade of 50 to 60 cm, designed for both cutting and thrusting. Whether the gladius was ever used on a larger scale by the cavalry remains uncertain.

In the late republic, the originally Roman riders were largely replaced by Celtic and Germanic auxiliary troops. These cavalrymen used La Tène swords , a type of Celtic sword that was widespread in Gaul. It is often claimed that this type served as a model for the late Roman spatha (σπάθη), the first specific cavalry sword of the imperial army. However, this is incorrect—the spatha developed independently of the La Tène-sword.

It is unclear whether the spatha was originally a Celtic invention or a Roman development as a cavalry variant of the gladius. In any case, the spatha was widely used from the 1st century AD and gradually replaced the older Celtic La Tène swords within the Roman auxiliary cavalry.

Contrary to a widespread misconception, the early imperial spatha, like the gladius, was primarily a thrusting weapon. Although the spatha with its approximately 15 cm long blade was significantly longer than the infantry sword of the Pompeii type, it was also noticeably narrower and therefore particularly handy—almost rapier-like in nature. Both the gladius and the early spatha were ideally suited for thrusting, but if necessary—such as during pursuits—powerful and effective slashing blows could also be delivered.

The spatha from the 1st and early 2nd century AD typically had a straight blade of about 60 to 70 cm length, with a diameter of approximately 63 cm (presumably meant to be grams, but I leave it as in the source), and a width of about 3.5 cm. The blade had a diamond-shaped cross-section, with parallel cutting edges that only tapered to a point in the last 5 cm. The hilt, made of bone or wood, closely resembled that of the gladius: a large, usually spherical or cross-oval pommel, a grip with a hexagonal cross-section and four horizontal grooves for the fingers. A simple hand guard with a thrust plate offered some protection, although a true crossguard was absent.

From the later 2nd century AD, the design of the spatha began to change. The blades became wider, the characteristic central rib often disappeared, and blood grooves appeared. These adjustments indicate an increasing use of the spatha as a slashing weapon. During this period, the spatha also began to definitively replace the gladius in the infantry.

To date, no completely certain components of an early imperial spatha-scabbard have been found. However, some archaeological finds possibly point in that direction: a scabbard end piece with a knob closure from Rottweil (Germany) and a mouthpiece almost certainly belonging to a spatha from Ptuj (Slovenia).

What we see in detailed images on equestrian tombstones suggests that the scabbard of the spatha in that period in terms of design strongly resembled that of the gladius of the Pompeii type. This consisted of two wooden boards, covered with leather, equipped with a long mouthpiece of bronze or brass. This mouthpiece could be decorated with repoussé or openwork decor, and was equipped with two carrying rings on each side. At the bottom, there was also a metal band, often richly decorated, that ended in a decorative knob.

The spatha was originally carried on a belt on the right hip – just like in the infantry. The sword had to be drawn with an upward-facing palm and the thumb down. However, on a galloping horse, it proved difficult to insert the sword into the scabbard in the same manner.

At the end of the 1st century AD, this carrying system changed significantly: the traditional belt was replaced by a balteus – a shoulder strap from which the spatha was hung. 

In the late 2nd century, this balteus became wider and more richly executed, often with perforated and worked bronze ornaments. At the same time, the shape of the scabbard end changed: instead of knobs, they acquired a rounded shape from the Antonine period. 

Until the early 3rd century, basin- and heart-shaped scabbard ends were common; thereafter, disc-shaped Dosenortbänder came into fashion. These were usually made of iron, often with inlay work, but specimens of bronze or bone also occur.

The attachment system to the balteus also underwent changes. Instead of the earlier carrying rings, a bracket-shaped sword belt holder was now used, which was attached vertically to the outside of the scabbard. The crossing straps of the balteus were thereby tied to the scabbard. At the same time, the position of the sword shifted to the left side of the wearer – a carrying method that would remain in use until the end of the Roman period.

In the 3rd and 4th centuries, the spatha underwent a clear development. The blade length increased medium to about 70 cm, with some specimens even being 80 cm or longer. Especially in the 4th century, this length standard was exceeded: many blades from that period are longer than 75 cm and surpass those from the 3rd century. The blade width showed great variation at this time, from 4 to as much as 7.5 cm.

The differences in size indicate functional specialization. Shorter, wider blades (such as the Lauriacum/Hromówka type) were likely intended for infantry, while the longer, narrower variants (such as the Straubing/Nydam type) were typically cavalry swords. Many of these swords feature multiple fullers, indicating refined forging techniques.

Although some grips still show strong resemblance to those from the early imperial period in shape, short crossguards also appear in this period. However, the overall design of the weapons from this time indicates a clear shift in use: where earlier models were suitable for both thrusting and cutting, the later spatha was primarily intended as a hacking weapon. This emphasis on cutting technique would remain characteristic of swords well into the Middle Ages.

A remarkable specimen of the spatha of the Straubing/Nydam type was found in Canterbury Castle (England), along with a second similar sword and two human skeletons, both dating from the second half of the 2nd century AD. It is a pattern-welded blade measuring 69 cm long and 5.6 cm wide. Notable is the relatively long tang of 22.5 cm, which suggests that the weapon could occasionally be wielded with two hands.

The sword was in a scabbard, presumably made of apricot wood, with a characteristically pelt-shaped bronze fitting. The find was made in direct association with the skeletons of two young men, measuring 173 and 181 cm tall, respectively. Both likely fell victim to violence, possibly in a military or ritual context.