In this short article I would like to present some details of the composition of this interesting arm of the Later Roman Army.
First it is necessary to identify the units and their location:
Title Command Equites Catafractarii Biturigenses (2,3) Magister Militum Preasentalis I Equites Catafractarii (2) Magister Militum Equites Catafractarii Ambianenses (2,3) Preasentalis II Comites Catafractarii Bucellarii Iuniores Magister Militum per Orientem Equites Catafractarii Albigenses (2,3) Magister Militum per Thracias Ala Prima Iovia Catafractarium (1) Dux Thebaid Cuneus Equitum Catafractarium (1) Dux Scythia Equites Catafractarii Iuniores (2) Comes per Brittanias Equites Catafractarium (1) Dux Brittanniarum
1 Associated with garrison town
2 Field Army unit
3 Originally based in Gaul
Also of interest are the locations of the state arms factories (Fabricae).
I have listed only the factories associated with the production of horse armour (ie, Clibanaria).
In the east:
Antioch, Caesarea (in Cappadocia), Nicomedia.
In the west:
Augustodunenis (in Gaul).
The list that the above units are taken from is generally accepted as being dated from the end of the fourth century. It would seem from the list that the Eastern Empire had an overwhelming majority of these units. However, if we analyse the titles of the units, we see that many of these have been stationed in the west (and in Gaul in particular) for many years and may have only recently transferred to the eastern area after Julian’s campaign in Persia (approximately 363AD).
With regard to the strength of these units it is now generally accepted that most units listed as Equites (ie, vexillations) would have a paper establishment of 600 men until the reign of Theodosius, when they were split to provide many other units (usually identified by Seniores or Iuniores attached to the title).
The units listed as Ala would presumably follow the earlier organisation of Roman Auxilary cavalry and have a paper strength of approximately 480 men. The Cuneus is a much smaller unit and may only represent 200 men. As in many other periods of warfare it is very likely that these strengths were never reached in practice and, in fact, when the units were halved in strength this may reflect official recognition that 300 men units were about the average strength possible to maintain for a cavalry detachment of the period. (In the reign of Diocletian pay records from Egypt show an Ala with a paper strength of 480-500 men was recorded with approximately 370 men available for service. Admittedly these records date from nearly a century before the list; however, the manpower shortage of the Late Roman Army will only have grown worse as the fourth century progressed).
As to the description of these units:
Firstly the rider. From descriptions in Ammianus, they seem to be extremely well protected. A long coat of mail or scale armour protected the body. Forearrns and legs were protected with laminated armour or vambraces. Gloves and shoes of mail protected the hands and feet. The helmets normally depicted do not seem to me to represent accurately the form used by these cavalry. Most Late Roman helmets found have a strong neck guard and extremely deep protection for the cheeks and face. The helmets also have a strong reinforcement for the crown. The descriptions in both Julian and Ammianus imply a facemask of iron or bronze, however, this may be simply artistic licence or possibly may represent armour similar to that used in the sports display helmets.
The horses are generally protected by a textile housing upon which are stitched bronze or iron scales. The neck of the horse was protected in a similar fashion (surprisingly, in the find of the horse armour at Dura only the body housings were found). The face and head of the horses were protected by leather chamfrons (as found in excavations in Scotland). These chamfrons included pierced bronze domes to protect the horse’s eyes.
Offensive weapons consisted of a large headed spear or contos approximately 3 to 3.5m in length and, probably, a spatha-type sword. Officers, by analogy with later Byzantine practice, may have carried a heavy mace.
The tactics adopted by these units were fairly simple. To achieve the maximum effect from their charge it was necessary to ‘pin’ the target with other units to prevent the target evading which would render the attack useless. In virtually all actions when faced by attack from Catafracts the standard Roman defence was to evade and then counter-attack as the Catafracts rallied back on their blown horses. The above tactic was used to great effect by the Roman Army of Aurelian in the campaign against the Palmyran Army in 272AD and by the Army of Constantine in the battles at Verona and at the Milvian Bridge in 312AD.
The tactic of ‘pinning’ the enemy was used to great effect by the Army of Constantius in the fight against the troops of Magnentius at the battle in the plains of Mursa in 351AD. In the final stages of this battle the horse archers in the Army of Constantius were used to ‘pin? the heavy legionaries preparatory to the charge of the Catafracts. At the battle of Strasburg in 357AD the Catafracts forming a flank of the Roman Army under Julian fled from the German infantry who had infiltrated their lines and started to hamstring their horses. This perhaps illustrates the weakness of Catafracts. Due to their complete armour their vision is badly impaired and this can become a crippling disadvantage in a close melée.
Certainly, the Emperor Julian used this technique to combat the Persian cavalry in the campaign in Mesopotamia in 363AD.
It should not be thought that the use of Catafract cavalry declined too quickly. In the reign of Valentinian I a unit of Catafracts was used in a battle against Saxon pirates in Armorica. In conclusion therefore, we can see that these horsemen, when used correctly, could provide a commander with a virtually unstoppable force to break the enemy line of battle. It is a pity that the generally used rules for Ancient Wargaming (WRG) do not adequately reflect the power of these units. I contend that under all but the most exceptional circumstances a Catafract unit could not fail to break through a line of enemy infantry. As we have seen from the above the best defence against a Catafract charge is to evade (or run!) If this is not possible then Generals should defend the front of their line with ‘wolf-pits’ or caltrops.
Graffito, Clibanarius. 2nd-3rd centuries AD, Dura-Europos, Yale University Art Gallery
Reconstruction of a fourth century Roman Catafract using third century drawing as a basis. Helmet from a fourth century example from Holland. Chamfron and leg defences from examples found in Scotland. Horse armour from Dura in Mesopotamia. Note socket on reverse end of lance. From the third century drawing it seems likely that the body is protected by lamellar or scale armour with mail to protect joints. If body protection is lamellar the scales would overlap in an upward direction to prevent penetration by thrusting weapons from below waist level. The sleeves shown could be replaced by leather Pteruges and the mail skirts could conceivably by replaced by leather lamellar thigh defences (an example of which was found at Dura). The sword is shown on the left side. I strongly suggest the inclusion of a sword amongst the Catafract’s weapons since the loss of the lance would presumably render him useless in a melée.
Armour of the Roman Legions.
What Soldiers Wore on Adrian’s Wall by H. Russel Robinson.
Notitia Galliarum by Otto Seeck.
Ammianus Marcellinus Books 16-18.