ANN HYLAND
HISTORICAL INTRODUCTION: GREECE1
In classical Greece warfare was dominated by the clash of infantry phalanxes and cavalry tended to play a peripheral role. For example, cavalry could be deployed successfully when the phalanx formation had been broken. During the Peloponnesian War Athenian and Spartan cavalry acting alone had little impact, though Athens shipped a small number of horses (two to three hundred) in special transports when raiding the Peloponnese and elsewhere. Some Spartan allies were strong in cavalry: Boeotia could field a thousand, a force matched by the Macedonians and Chalcidians. At Delium (424 B.C.) Boeotian horse contributed to the Athenian rout, and in Upper Macedonia Brasidas’s Macedonian and Chalcidian cavalry overwhelmed the Lyncestians (Thuc. 4.89, 102, with Spence 1993: 9–30).
Philip and Alexander of Macedon were the first Europeans to make consistent, extensive, and successful use of hard-hitting cavalry, which on Philip’s assassination Alexander took deep into Asia Minor, Iran, and ultimately India. Alexander inherited an army of some 32,000 infantry and 5,100 cavalry; the horses came principally from Macedon and Thessaly, both renowned for their cavalry, to the tune of 1,800 each. Greece supplied another six hundred, Thrace and Paeonia nine hundred. Another one thousand already in Asia Minor raised the total to about 6,100. Alexander owned an exceptional, but vicious, large, black Thessalian stallion called Bucephalus; the famous story has it that the twelve-year-old Alexander saw that it shied at its own shadow—a common occurrence with young, green horses. Skillfully positioning it Alexander successfully rode it. It stayed with him as his favorite charger until dying aged thirty ostensibly from old age or battle wounds (Arr. Anab. 5.19).2
Under Alexander Greek cavalry reached its pinnacle and chroniclers have left more information on this phase of cavalry growth than any other. From 336 to 326 he fought four major battles, and made frequent assaults on tribal territories, in all of which cavalry played a prominent part. The huge enemy numbers and their losses, quoted in multiples of thousands, were often plainly fictitious, set against Alexander’s moderate numbers. For example, at the Granicus (334 B.C.) Alexander’s Companion Cavalry lost twenty-five, other horsed units over sixty, infantry thirty, set against one thousand Persian cavalry and infantry and two thousand Persian prisoners. Sources offer comparable figures for the battle of Issus (cf. Arr. Anab. 2.8, with Diod. 31.2 and Just. 11.9, Curt. 3.2.4–9). Among the spoils at both was a huge haul in Persian horses, largely Nisaeans from the battlefield and the baggage park, as no high-ranking Persian went to war with only one charger. Lower-grade remounts and sumpters swelled the spoils.
By 331 Alexander’s cavalry reached seven thousand, plus units left in strategic garrisons; his infantry about forty thousand. Opposing him at Gaugamela Darius probably had about 200,000 infantry and 34,000 cavalry. Darius’s horses came largely from Turan in which lay Bactria, Sogdia, and Arachosia. Median, Hyrcanian, and Parthian cavalry used horses from south of the Caspian Sea. Other horses were of Cappadocian, Armenian, and Syrian stock, while the Scythian tribes rode fully armored horses.3 Although some Persians were drafted into Macedonian ranks after Issus, it was after Gaugamela that the equid composition changed even more radically.
As Persia’s master Alexander had the best of the ancient world’s horses. Babylon welcomed him with herds of horses and cattle and spoil from the baggage park included camels and elephants. Some idea of Babylonia’s equid wealth is given in Herodotus who noted the satrap Tritantaechmes’s stud of 16,000 mares and 800 stallions (Hdt. 1.192; cf. Arr. Anab. 3.15). The Nisaean herds of the Median plains were reckoned at fifty thousand (Arr. Anab. 7.13.1), Bactrian remounts also fell to Alexander (Arr. Anab. 3.30) as well as those of other mounted units. In addition, five hundred cavalry arrived from Macedonia and Thessaly (Arr. Anab. 1.24; 29; 3.17). Five years later in India, welcomed by the people of Taxila, Alexander faced Porus at the Hydaspes River. Alexander fielded about five thousand cavalry among whom were his elite agema and Companions, the superior cavalry of Bactria, Sogdia, and Scythia, plus mounted archers of the Daae; the cavalry played a crucial part in Alexander’s last decisive set-piece battle. At the Hyphasis (Beas) Alexander’s army mutinied (Arr. Anab. 5.8–27) and the long exodus from India through the hellish Gedrosian desert began. Alexander’s army was ravaged by nature’s most hostile elements. The cavalry, built so meticulously over many years was wasted, animals dying from thirst, hunger, and exhaustion, butchered for meat, or drowned in a flash flood (Arr. Anab. 6.17; 5.21–8).
HISTORICAL INTRODUCTION: ROME
Rome’s cavalry went through the same process of small beginnings, massive expansion, and eventual decline, but with a legacy that outlived its decline as it reemerged as a strong element in the Byzantine army. In Rome’s early years horse ownership was mostly confined to the equestrian class, men wealthy enough to support such a luxury. As her conquests expanded Rome adopted many of the practices of her opponents, especially those of military equestrianism. The rise from city-state to master of Italy increased scope for cavalry recruitment and access to noted horse-breeding areas. Etruria had a traditional equestrian society, an efficient balanced-seat riding style, and maintained cavalry. Campania’s famous cavalry predated Rome’s expansion. In Magna Graecia and Sicily there was Greek influence in horsemanship and breeding (Frederiksen 1968: 3–31). Under Roman rule Sicily became a location for large stock-raising latifundia. Greek expertise in cavalry and in raising and acquiring suitable animals must have given Rome’s fledgling cavalry a boost.
After the Roman victory at Pydna (168 B.C.), the Achaean-born historian Polybius was taken to Rome as hostage along with other leading figures. As an active cavalryman he was well suited to report on the current state of Rome’s cavalry structure, and that of her enemies. In the conflict between Rome and Gaul in 225/4 B.C., levies on Italian allies provided most cavalry mounts. Rome allegedly fielded conveniently rounded figures of 700,000 foot and 70,000 horse (Polyb. 2.23–4). Nevertheless Italy offered numerous remounts. Certain areas were rich in horses, as illustrated by six thousand requisitioned by Hannibal’s men in Apulian and Sallentine lands for cavalry use after being broken in (Livy 24.20). In fact after crossing the Alps, Hannibal arrived in Italy with at least six thousand cavalry, including Spanish and Numidian forces. In general, Carthaginian horsemanship was superior to that of the Romans. At the battle of the Trebia (218 B.C.) Hannibal’s Numidian horsemen, riding small, agile horses, terrified the Romans with their hit, run, re-form, hit and harry tactics. At Trasimene the next year Hannibal’s cavalry pursued and cornered the Romans, driving them into the lake where many were killed or drowned. At Cannae (216 B.C.) ten thousand Carthaginian cavalry, commanded by Hasdrubal, Hannibal’s brother, played a notable part in the victory as they defeated the Roman cavalry and turned to attack the infantry in the rear. Only 370 Roman horsemen survived out of six thousand (Polyb. 3.113–17). Scipio Africanus’s eventual defeat of Hannibal at Zama (202 B.C.) showed that lessons had been learned. Africanus used the cavalry of an ally, King Massinissa of Numidia. Henceforward Numidian cavalry frequently served in Roman armies. They are mentioned as fighting under Fulvius Flaccus against the Saluvii (Ridgeway 1905: 261), and in Caesar’s Gallic War (B Gall. 2.10). Caesar’s cavalry was very cosmopolitan as he levied units from several pro-Roman Gallic tribes, and purchased numerous remounts from Italy and Spain (B Gall. 8.55). Furthermore, better weapons were supplied: a robust lance pointed at both ends, a larger and better shield, both on the Greek pattern.
By Augustus’s time auxiliary cavalry were formed into alae (wings), either of 512 men (quingenaria), or 768 men (milliaria), including officers. In a mixed unit (cohors equitata quingenaria) 128 were cavalry, and in a milliaria unit 256. A cavalry turma (troop) was thirty men, plus two officers. Auxiliary cavalry was drawn from allies, client kings, mercenaries, and peoples conquered then attached to Rome. As cavalry increased in number so did the ethnic content of auxiliary alae, fighting after their own methods, using native weapons and indigenous horses, until the hazards of war or old age dictated fresh remounts, culled from an ala’s current posting.
Ethnic units were posted throughout the Empire. For instance, Thracian cavalry saw service in Britain, Germany, Raetia, both Pannonias, Syria, Egypt, and Mauretania Caesariensis (Hyland 1990b: 191–2.). In the east, Syria, a Roman province since 64 B.C., supplied a steady stream of remounts for Vespasian and Titus’s Jewish War, and was to be an important supplier of horses through the Imperial period. Antiochus, Soeamus of Emesa, Agrippa, and the Arab king Malchus sent cavalry contingents. Many Syrian allied cavalry were cataphracts, protected by armor from head to foot and using couched lances. It had long been eastern practice to have heavy cavalry horses at least partially or fully armored, and in some cases the armor consisted of leather or padded cloth—common down the centuries for eastern and Indian horse. Cataphracts started appearing in Roman cavalry from at least Hadrian’s day, as separate from the eastern allied horse noted above. From A.D. 166 to 175 there were repeated conflicts with the Iazyges, Sarmatians, and the Marcomanni (Dio 72.7.1–5); peace was concluded in 175, the price to the Iazyges being eight thousand cavalrymen, making Sarmatian mounts the second largest influx of foreign horses. Fifty-five hundred were sent to Britain (Dio 72.16.2), where Ribchester (Bremetennacum) had a full Sarmatian unit. Their presence in the north of Britain was long-lived, shown from an inscription from the time of the Gordians (Richmond 1945: 15–29).
With the Emperor Gallienus the horse’s role changed. Vast distances and the need to deal with many separate incursions called for speedier movements that could only be accomplished by cavalry. To cope with this Gallienus created a fast, mobile force (fifty-sixty alae) composed largely of Dalmatian, Moorish, and Osrhoenian horse, armed with their own missile weapons (Williams 1985). From their base at Milan, these forces were able to launch rapid attacks sometimes covering fifty miles per day). It is significant that Moorish and Osrhoenian horsemen were used, as their indigenous horses (as are many eastern horses) were light framed and noted for endurance, better suited to sustained speed and able to recover rapidly from excessive exertion under heat conditions.
In the late Empire under Diocletian, the army possibly expanded by 30 to 50 percent, cavalry being about a third of the projected 600,000 total. Under Diocletian vexillations became permanently detached and eventually were termed legions at one thousand, while alae became known as vexillations (Jones 1964: 679ff.). The Illyrian emperors were beset by revolts and between A.D. 287 and 298 moved massive numbers of troops, especially cavalry. In the vulnerable territory five hundred strong vexillations were placed at twenty-mile intervals to render rapid aid to beleaguered forts. Fit, well-fed horses can easily sustain sixteen miles per hour, and at speed can travel in excess of twenty (Hyland 1990b: 193).
Constantine created the posts of Magister Peditum and Magister Equitum in overall command of the two military spheres. The number of mobile cavalry vexillations was increased. His crack cavalry were five hundred strong units (Scholae): seven in the east, five in the west. From these his forty mounted bodyguards were chosen, initially from elite cavalry with Roman citizenship, but later mostly from Germanic peoples (Jones 1964: 613). In respect of providing remounts, the Theodosian Code specifies army horse acquisition from Constantine to Honorius. Several methods were used: exaction of cash from taxpayers liable to contribute to the “horse fund” (11.17.1, A.D. 367); direct levy of a certain number of horses per district (11.1.29, A.D. 401); levy on persons up for government preferment (7.23.1, A.D. 369). Some horses came with cavalry recruits (7.22.2.1, A.D. 326). Later the law changed to cash only, but some, defying the law, still brought horses, as shown by Flavius Abbinnaeus, a thirty-year veteran cavalry officer stationed in Egypt in A.D. 342. One of his duties was tax collection. When he sent a batch of horses to the exactor Plutammon, he received an irate reply complaining of receiving horses instead of cash (Bell 1962: Papyrus 13).
GREEK AND ROMAN CAVALRY TRAINING
Xenophon advises on training, and much of his advice is present in modern horsemanship techniques. His treatise The Cavalry Commander (Eq. mag.) gives commonsense advice any good trooper could follow. Athenian cavalry archives show troopers owned their mounts. Appraised values ranged from 100 to 1,200 drachmas. The fourth-century B.C. average was about 700 drachmas, the third century about 400. Values fell 50 to 100 per annum, troopers regularly changing horses. However, Cleochares of Cephisia’s unbranded chestnut remained valued at 600 drachmas for four years. It has been suggested it might have been several chestnut horses, as they were common. I see no reason for this as a superior animal bonding with his rider and the avoidance of mental and physical trauma would be invaluable. Many 1,200 drachmas horses were not devalued as their true value was considerably higher than the maximum allowed (Kroll 1977: 83–140; Spence 1993: 191–211).
Xenophon advocated trooper recruitment by persuasion, or if necessary by court order. It was to be ongoing to replace those vacating the ranks (Eq. mag. 1.2: 8–12). Horse recruitment would have been similar. Exemptions from service were: infirmity, age, lack of means; unfit recruits were to be dismissed. If chargers lost condition, the owner/cavalryman was fined the cost of its feed. Those that could not keep up with the squadron, or jibbed at keeping in line, were to be dismissed, branded on the jaw with the sign of the wheel (Arist. [Ath. Pol.] 49.1, 2). Brilliant horsemanship was encouraged to promote enthusiasm. Mounting was “from the spring,” many owing their lives to this ability as they could remount after a fall, or after being unseated in battle (Xen. Eq. mag. 1.11, 1.5). A spear aided mounting, akin to pole vaulting without the run-up; grasping the mane, pushing off the withers with the left hand, and leaping was another means, and was possible wearing full panoply (for documentary proof in full medieval panoply, see Hyland 1999: 108).
Xenophon implies Athenian cavalry and care of horses had deteriorated, reminding cavalrymen of even the basics: horses were to be well fed to enable them to work at speed both harrying and escaping from the enemy; hoof care was essential to maintain soundness; chargers were to be docile (i.e., obedient) as fractious horses and kickers were a nuisance and helped the enemy (Eq. mag. 1. 3–4). Dominant horses, especially stallions, could have been aggressive against neighbors, disrupting cavalry cohesion: “the men too were to be made obedient, or their good horses, horsemanship and fine armor would be useless.” This is slightly comical, perhaps indicating privileged Athenian youth resented orders. By Xenophon’s day cavalry numbers had dropped from about 1,000 to 650. They fell again to below two hundred in the early third century, rising to three hundred by mid-century. Eventually by raising the number of tribes from ten to thirteen cavalry numbers rose, fixed by law at one thousand (Xen. Eq. mag. 2.2; Arist. [Ath. Pol.] 1.11.5; Kroll 1977: 83–140, at 95ff).
Rather neglected in battle accounts are one-to-one combats which happened once a charge’s cohesion was lost. The Arthashastra ascribed to Kautilya, a work from Mauryan India of the late fourth and early third century B.C., sheds some light on such combat. After the agreement (ca. 306) reached by Seleucus I and Sandracottus (Chandragupta to the Greeks), founder of the Mauryan Empire, some Indian cavalry tactics related by Kautilya were absorbed into Seleucus’s own military (Shamastry 1951). Significantly many of these continued to be practiced into Mughal India (Hyland 1998: ch. 9). Indeed many are mirrored in modern dressage, showing how highly trained a well-proportioned, balanced, and intelligent horse can be—three requisites for a good warhorse. Equestrian historians link haute école to the Italian and French masters of the sixteenth century and later, but they should be credited to ancient India. Understanding equitation fundamentals illustrates what it took to produce Greek and Roman cavalrymen.
Arrian’s Ars Tactica
Arrian describes complicated cavalry maneuvers, the details of which military horsemasters knew how to impart to their equid and human recruits. The maneuvers have certain affinities with some modern equestrian competitions, notably the American Quarter Horse Association’s reining patterns and India’s hog-sticking attacks, so it is possible to trace essential elements in Roman cavalry training. Crack Roman alae rode in the spectacular Hippika Gymnasia. All maneuvers had their practical combat aspect. Horses were highly trained and troopers used weapons, especially missiles, with deadly accuracy. At first reading, the maneuvers seem simple. In our mechanized age it is difficult to appreciate, without extensive equestrian understanding, how complex they were. Arrian describes each move stride by stride, each weapon use angle by angle. The full implication did not become really clear until each move was worked slowly, section by section, on horseback—reins in one hand, text in the other, and later at a more recognizable cavalry speed of fast canter. Incorporated in the moves were rapid missile launches by troopers moving at blistering speed; contact aggression of the Cantabrian Gallop as troopers circled, one by one meeting their individual adversary head on; the Celtic Toloutegon, the rapid rollback over the hocks from the gallop. These “Roman” maneuvers linked with the specifics of Xenophon, and indeed Arrian referred to himself as Xenophon.
There is a strong thread of Roman horsemanship in modern riding, particularly in American Western riding which initially reached the Americas from Spain whose own early equestrianism owed much to the Arabs. It has to be remembered that much of Roman cavalry came from Spain. The round pen breaking circle evolved from the Roman gyrus. A British gyrus has been reconstructed at The Lunt, near Coventry in Warwickshire. The high sides concentrated a horse’s mind on his work, giving a sense of containment and control, and the curved track kept the movement flowing. In a rectangular arena some horses have a favorite place to balk—repeatedly. At some permanent camps there was a Basilica Exercitatoria, or indoor riding school. Inchtuthil had a timber structure 140 by 90 feet, very close in size to a modern school, and Chester a stone built one 250 by 80 feet. At Newstead at first troopers made do with the parade ground but in the second century A.D. a 160 by 50-foot indoor school was constructed (Davies 1968: 73–100). In theGyrus and Basilica Exercitatoria basic horsemanship would have been drilled into horses and riders, and selection made for different grades according to obedience, willingness, temperament, and ability of both men and horses. We know from Hadrian’s adlocutioat Lambaesis in North Africa that alae and cohortes equitatae had different grades of horses.
It is difficult for the cavalry of the cohorts to make a good impression … after the exercise of the auxiliary cavalry. They have a larger ground coverage, a larger number of men throwing javelins; their right wheeling is in close array, their Cantabrian maneuver closely knit, the beauty of their horses and the elegance of their equipment is in keeping with their level of pay.…
The performance of the Ala I Pannoniorum warranted an unusual extra cash reward (ILS 2487 and 9134; see Speidel 2006).
Common sense dictates that hopeless cases were rejected, the horses sold or demoted to the baggage train, appalling riders sent to the infantry. Neither would have been any use in an ala, except to the enemy. Superior combinations would have been chosen for team spectacles, brilliant ones for solo performances.
Ideally the usual Roman parade arena was square. The Tribunal (including an area for spectators) was sited centrally for a clear view of the whole proceedings. However, some of Arrian’s moves needed the whole space clear; the Tribunal was best sited on one side, in front of which the men performed. The best surface was a fine tilth (Tact. 34.1), giving better traction, especially on the fast turns. It prevented concussion which could produce a variety of lamenesses. Of course parade space differed and compromises were made. On smaller grounds I suspect a full complement of riders was not always possible. Two turmae, sixty-four men, working on a three-acre plot, such as that at Hardknott Castle, meant a very crowded arena with no room for maneuvering well at speed. Individual missile launches at extended gallop would not have been possible. A 15-hand horse’s stride at only extended canter is 11 feet 6 inches to 12 feet.4 I would hesitate to put any horse into a full gallop in such a confined space.5 Lambaesis had a 220 square yard area, one mile west of the legionary fortress. In its center was the monument on which Hadrian’s adlocutio was inscribed, hinting that this ten-acre site was a parade ground. Legionary forts had much larger grounds than those for cohortes equitatae(Davies 1989: 97–100).
After basic grounding troopers tested their horsemanship on route marches. In his Epitome Rei Militaris, Vegetius notes cavalry conducted twenty-mile marches three times a month. En route riders practiced their evolutions, pursuing, retreating, charging, and countercharging. Terrain used was to be a mix of flat and hill work to fit horses for all eventualities. The main Ars Tactica moves listed were:
the charge from concealment the testudo and petrinos
the Cantabrian Gallop javelin and lance handling and delivery
Missile launch and the Touloutegon hand-to-hand combat
The Charge from Concealment
For parade purposes concealment was hypothetical. In reality it would have been an ambush. The en masse entry was explosive, men shouting war cries (Tact. 37.1), hooves drumming, standards whipping in the wind created by the horses’ speed. Standards were riders’ reference points. Arrian says they helped control maneuvering and prevented teams becoming mixed. Signifers (standard bearers) dictated speed, turns, changing patterns, the charge and its velocity, this also being signaled by trumpet (Tact. 36.3). A signifer needed exceptional horsemanship, his instructions to his horse second nature, concentrating on giving orders via the signum. The horse would have recognized trumpet calls and acted accordingly. Horses do recognize sounds that mean orders, and even if the rider is slow the horse reacts correctly.
This charge was aggressive. A vital requirement was sustained control, horses well balanced, working off their hocks, forehands lightened, enabling rapid turns in all evolutions. The epistrophai (Tact. 35.6) defined by Professor Frank Walbank as “a movement by which a whole body of troops swings round, pivoting on the left or right file leader, so as to finish at a 90° angle to its former position,” clarifies one of the preparatory moves (Walbank 2: 226). Opposing teams rode in single file toward the Tribunal, side by side but a considerable distance apart. Keeping to a controlled canter the team to the Tribunal’s right kept horses on the right lead, while that on the Tribunal’s left kept horses on the left lead. As the signifers reached just in front of the Tribunal and signaled, each team wheeled 90 degrees to face their opponents, launching a head-on charge. It required split-second timing and placement and echoes the Athenian cavalry’s antihippasia sham fights when troops divided into opposing teams, three times interleaving at increasing speed, until at full gallop the final charge (Xen. Eq. mag. 3.10–14). These widely time-separated moves prepared for battlefield action, interleaving at impact point.
The Testudo and Petrinos
After a halt from the gallop two turmae, for convenience A and B with A the superior, went smoothly into the complicated Testudo and Petrinos movement (Tact. 36–9). It was a severe test of horsemanship, split-second timing, adroit and fast handling and delivery of javelins from any angle. In battle it translated into relentless missile attacks coming from all directions. A trooper with the ability to hurl successive javelins at any angle, maintain control of his mount, defend himself, change course, and come in repeatedly was a valuable addition to any unit.
A and B lined up opposite the Tribunal, horses side by side but with a space between teams. In the interspace two target horses and riders from B lined up head to tail but with several strides between them, allowing 15 to 16 feet per stride. The teams each formed a testudo with, as Arrian said, “the horses’ heads turned away.” This meant the horses stood obliquely, the 4-foot cavalry shield covered the rider’s side and the side and back of his horse. When tried out on horseback using two horses placed in close echelon it worked out that the second (and subsequent) horse’s head was also tucked in behind the shield of the first rider. The shield I held on the lead horse had its center covering my side and back, the bottom half covered the second horse’s head, part of his neck, his chest, and my horse’s back behind the saddle. Horses chosen would have been of equal height and similar stride extension or the line would have been ragged with movements lacking synchronicity.
Each team had its turn at missile delivery, the numerous javelins held in the concavity of the shield and transferred from left to right hands at delivery, the troopers managing reins, shield, and javelins in the left hand, the edge of the shield pressing on the left of the neck functioning as a rein for a turn to the right. The horse quickly equated shield edge with rein pressure. The petrinos’s first stage was for A riders to come out in quick succession bearing left on a curve as aggressors and hurl as many javelins as possible at the B targets. Meanwhile B riders moved singly to the right trying to hit A riders’ shields before each A member completed his curve and launched into his missile run. To avoid collision this section had to be perfectly spaced. Before the crossover a B rider had clear sight of an A’s shield to aim at, but exposed his unprotected right side to oncoming A riders who had not yet crossed over into their dead run at the targets. The rider aimed to protect his right side while loosing his javelin(s) to the right. As he then swung left toward the arena’s edge a B rider had to twist in his saddle to hurl a javelin over his horse’s rump at an oncoming A. As each B loosed his last javelin at an A he swung his shield over his back to protect it from A’s javelins. These moves needed exceptional dexterity so the poor horse did not get clouted in the process. With two turmae in action this rapid fire exercise exhibited superb horsemanship. The target horses were sure to have been carefully selected. A horse would stand for one or two blows on a shield, especially when he realized the headless javelins were doing him no harm. However, with the constant barrage of thirty-two yelling troopers hurling missiles they had to have been very trusting and phlegmatic.
The petrinos’s finale allowed the best marksmen from A to shine. After their turn at the targets each galloped as far as the Tribunal, his horse on its right lead, and at the Tribunal circled right, at the same time loosing a javelin under the eyes of the dignitaries, aiming across the horse’s neck at the oblique angle that took it to the farthest part of the arena. The best two continued galloping on the circle, each aiming his next javelin behind him at a tangent to his arc of travel. The speed gave a safety margin for the next trooper coming up behind for the final launch (such maneuvers have been analyzed stride by stride and illustrated in Hyland 1993).
The Cantabrian Gallop
The teams lined up as in the testudo: team A from the right, wheeling right into a circle; B team charging from the Tribunal’s left, also wheeling into a circle (Tact. 40.1–7). The description of the opening of the Cantabrian Gallop is a rather confused set of orders, no doubt understood by those used to such maneuvers. It might be expected that opposing teams would take opposite directions, but this would negate the purpose of the exercise which was to have an A and B rider meet at precisely the same spot on the arcs of both circles, putting them in position to deliver a thrust at their opposite number’s shield, with the more dangerous “whittled spear” replacing the headless javelin. This spot was directly in front of the Tribunal, the dignitaries receiving a perfect view of the front of the attacking rider, his weapon handling, and the rear view of B rider, defending himself by parrying A’s spear. Both teams had to be on the right lead and on the right rein. If B riders were to provide frontal views in front of the Tribunal at the point of impact the lead would have been the left, and the rein the left, that is, a reversal of roles with B attacking and A defending (see further Campbell 2004: 44–5).
This maneuver was also worked out on horseback at canter. It became clear that speed and size of circles were mirror images. It may sound easy to ride a circle but it is not. To have thirty-two perfectly spaced riders in each circle, without any horse falling into the arc, a tendency with most horses, needed superb riding, timing, and accurate leg aids. The spear’s impact was to be delivered with enough poundage to drive right through the opponent’s shield.
Javelin Throwing and Lance Exercises
Next came an exhibition of long-range javelin throwing by the most proficient horsemen (Tact. 40.8–10). With the Tribunal high up on their right each rider held his horse ready at one edge of the ground, then exploded into a dead run, throwing as many javelins as possible before reaching the opposite arena edge. Arrian says “a really good man will succeed in releasing javelins in the correct fashion before his horse leaves this area.” In speed terms and using the Lambaesis measurements of 220 yards/660 feet per side it took a horse with a 15-foot stride, the average for a galloping 15-hander, traveling at 20 miles per hour, 22 seconds to cover the distance and loose fifteen javelins. At 25 miles per hour it took approximately 17 seconds. It can be appreciated just how skilled these marksman were, and how honest their horses. At a gallop on a free rein each horse would have lowered its head, neck, and topline thus facilitating javelin launch. In hand-to-hand combat at vastly reduced speed, if any at all, a horse’s head, on a tight rein, was higher, so head and neck offered some protection to the rider. In skirmishing such expert javelin men were very valuable.
In lance exercises a target was set up to the left of the Tribunal. Now armored “with iron corselets, helmets and shields which are no longer light ones” (Tact. 41.1), riders had to contend with the unbalancing effect of a weighted torso and of the shield’s weight when wielding weapons, and had to counteract the added pull to the left when delivering a strike, especially in a turn at the gallop. Equilibrium was maintained by utilizing the saddle horns and thigh and lower leg muscles. There is a textual discrepancy in Ars Tactica 41.1 where lances are discharged, while in 42.1 they are referred to as javelins, before reverting to lances: but the operational section was 41.2. Individuals hurled lances at the target; the more skilled launched a second, even a third lance (Tact. 41.3). The second, and successive, charges were made with two lances, requiring some considerable dexterity in weapon handling. No mention of a quiver is made, but Josephus’ Bellum Judaicum mentions cavalrymen carrying javelins in a quiver, although he does not say where it was positioned (3.97). The best position was on the offside (right) rear saddle horn, attached so the javelin would have been ready to cast. With successive missiles the movements of withdraw, raise, draw back the arm, loose, and curve down to withdraw another weapon would have been perfected until fast and smooth. The only other position would have been on the offside front horn, but withdrawal would not have been so smooth. On the back is the worst position because of fouling with the shield unless withdrawal was super-accurate.
The horse had to be very honest and move straight or the lance(s) in the initial shots would have missed the target. The last exercise is with three lances, the first loosed from the arena’s edge, the second when level with the Tribunal, the third as the horse begins his turn to the right for a cast at an additional target. This final throw needed greater precision as it had to be made before the turn was completed. Arrian notes that this throw was the most difficult of all (Tact. 42.3). To add to the cast on the turn the horse had to decelerate, collect himself on the right lead, transfer his weight onto his quarters, and shorten his topline. Accompanied by a certain raising of head and neck it made the cast across the neck more difficult. In horsemanship terms it meant supreme coordination by both rider and horse. Exceptional marksmen discharged three, some four, lances before the turn to the right and the final shot. It is unclear if the second target is in line with the first, set in the corner of the arena, or at some distance on the side of the arena along which the horse travels after turning. On balance I suggest the second target was in line with the first, or in the corner, as Arrian clearly states the last throw was the most difficult. If it was the third option it would have been no more difficult than the earlier shots which did not need an across the neck aim. Arrian considered successful troopers were “truly well trained for warfare” (Tact. 42.5).
The finale involved a variety of team and individual mounted exercises that tested the proficiency of riders with all cavalry weapons, and their horses’ training to the full, including jumping. Such horses had to have had temperaments submissive to a rider’s dictat, but able to show aggression and power in executing maneuvers. This may sound contradictory but some horses appear docile but are also extremely courageous, and have the character to work in complete harmony with riders without in the least being cowed, at the lightest signal leaping into action, at the same time remaining under full control. Such would have been Hippika Gymnasia horses, especially those of the elite troopers. Brutal handling nearly always ends in a worthless horse that would in a crisis have left a trooper vulnerable to an enemy (figure 23.5).
Figure 23.5 Diagram detailing one of the Roman cavalry drills, the Cantabrian gallop maneuver (from Campbell 2004: 44–5). Examples of the indoor exercise halls discussed here may still be seen on the grounds of West Point, the United States Military Academy.
The Celtic Touloutegon
The majority of the exercises were with handheld missiles—javelin, spear, lance, pike, but in addition there were casts made with stones both from the hand and from slings. In the stave exercise Arrian is imprecise. He uses “they” for troopers and “horse” in the singular:
…they now charge carrying straight staves, first as if to provide a defense, and then as if reaching after an enemy in flight, next as if attacking another enemy. As the horse turns they raise their shields above their heads and move them over to cover their backs and they complete this movement, whirling their staves as if another enemy were charging. (Tact. 43.2)
This turn was the Touloutegon. Arrian does not say if this was a complete about-face, or only a very sharp turn to the right or left. As the shields were spun over the riders’ heads to protect their backs it would appear it was a 180-degree turn, and that the new enemy came in directly behind them. Done at speed this was difficult, since the rider had to check the horse onto its hocks and before he lost momentum spin him to right (or left) until he faced the direction from which he had come and was in a position to face the “new” enemy. Presumably the other unmentioned first attacker had been seen off. If the horse was well balanced he would have been able to jump off his hocks at speed, and give his rider the impetus for a renewed charge against the “other enemy.” This move is now called a rollback, done by superbly schooled horses and talented riders. It would have been invaluable for a single horse in chasing a particular enemy who was ducking and weaving, rather than for a concerted pursuit by a whole unit. For a Roman trooper the turn was exacerbated as it was combined with swinging his shield over his back with his left (rein) hand, while wielding his weapon with his right. The turn had to be communicated to the horse by leg and body weight signals. In pursuit it is an established fact that horses, especially stallions, can be really helpful using teeth, and on occasion a front hoof to strike at their quarry.
Sword Exercises
Another major weapon used by the Roman cavalry was the spatha (Tact. 43.3), longer than the gladius and intended for sweeping slash strokes to either side of the horse, or with its point into a fallen foe. There were a series of strokes incorporating attack, parry, riposte, and counterattack. When delivered from a static position while the horse maneuvered to keep head on to the opponent, thus offering a small target, turns on the haunches and turns on the forehand were useful movements, the extra spatha length keeping the assailant at bay. With a running side-by-side confrontation the trooper’s outside leg kept his mount close enough to deliver lateral strokes. In all sword strokes the front saddle horns aided the trooper’s seat retention, the rear horns prevented him being butted out of the plate by lance or spear. The front horns allowed the rider to put extra poundage behind his stroke. In pursuit with either lance or sword the fast horse with an ability to keep on the enemy’s tail was invaluable. Intelligent horses learn very quickly. Some are born with an innate sense of “boss horse.” It reflects wild behavior from when horses ran in herds and a stallion chivvied mares and youngstock along. Horses learn to follow and work cattle, to control a herd, block a breakaway, head it, return it to the herd, and generally push the bunch in the direction a rider wants. A cutting horse’s job is to bring a selected animal out of the herd and prevent its return. In pursuit the same scenario obtained, substituting man for cow. In harrying a good horse was a great help, often anticipating avoidance tactics before his rider recognized them.
Although Rome used units of mounted archers, particularly in the eastern provinces, Arrian specifically alludes to missiles shot from “a machine” (Tact. 43.1). This can only be a form of crossbow. China used crossbows as early as the fourth century B.C. It was one of many armaments the elite Hippika Gymnasia trooper was expected to be proficient with. A ballista hole in a piece of leather horse armor from Dura Europos attests to such a machine’s use, even if by an enemy (China: Loewe 1980: 249; Dura Europos: Rostovtzeff 1936: 430).
Lastly every trooper showed his upper-body strength, spring in his legs, and optimum timing as, fully accoutered, he sprang onto his galloping horse. In this the front and rear saddle horns formed an ideal hold to push against and swing high and clear over the horse’s back.
Arrian’s maneuvers show the depth of training given to the best Roman cavalry, but to reach these heights it is obvious the grounding in basic horsemanship was thorough. If not, the average trooper would have been a liability, as would his uncontrolled horse. We know from Aelian (NA 13.9) that lunging was practiced in the Roman sphere:
Indians compel them (the horses) to go round and round returning to the same point. Now if a man would do this he requires strength of hand and a very thorough understanding of horses.
This teaches obedience, smoothes out a horse’s paces, and gets it supple working on a circle. We know from sculptures that Roman saddle horses were long reined. This teaches obedience to the bit, gets the horse moving in a straight line, and the signals, especially on the turns, from the reins on its sides prepare a horse for the transference of these aids to the rider’s lower legs once the horse progresses to ridden work.
In conclusion neither Greece nor Rome was in the first league in terms of general horsemanship and horsemastership. The works of Xenophon and Arrian on cavalry training aimed at a pinnacle of expertise. But high-caliber horses that had received superb training and were matched with superb riders were not the norm in the ancient world. However, to offset this both the Greeks and the Romans had access to knowledge gained from significant equestrian nations and tribes. As is clear from ancient texts and treatises, they learned from their enemies. Greece frequently clashed with Persia, and the Persian army was renowned not only for its equestrianism but also its use of equestrian tribal allies. The Romans had the good sense to adopt the equestrian skills of peoples such as the Sarmatians, introducing armored cavalry and also employing Syrian mounted archers. Certainly from the third century onward cavalry was to play an increasingly important role in the imperial army as the Romans found themselves fighting skilled horsemen such as the Persians and the Huns.