Home   Genealogy   Forums   Search   Contact
MHO Home
MHO Home
 17th Century
 18th Century
 19th Century
 American Civil War
 World War I
 World War II
 20th - 21st Century

 Write for MHO
 Search MHO
 Civil War Genealogy Database
 Privacy Policy

Ancient Sections
MHO Home
 Ancient Home

Ancient Articles
Battle of Thatis River
The Battle of Megiddo
The Third Battle of Anchialus
Second Samnite War Phase 2
The Roman Disaster at Adrianople
Second Samnite War
War in So. Italy 342-327 BC
First Samnite War
Pompey and Ancient Piracy
Brasidas - Spartan Commander
Battle of Kadesh
Battle of Plataea
King Arthur
Roman Invasion of Anglesey
Agricola - The Final Invasion

Jeffrey R. Cox Articles
The Roman Disaster at Adrianople
End of the Battle of the Java Sea

Recommended Reading

The Fall of the Roman Empire: A New History of Rome and the Barbarians

Rome's Gothic Wars: From the Third Century to Alaric

Empires and Barbarians: The Fall of Rome and the Birth of Europe

Cascading Failure: The Roman Disaster at Adrianople AD 378 - Part 3 of 3
Cascading Failure: The Roman Disaster at Adrianople AD 378 - Part 3 of 3
by Jeffrey R. Cox

The Long Hot March

Before dawn on August 9, 378, as per usual Roman military procedures, the praesental army’s troops were awakened. They ate breakfast and got themselves ready for battle with their weapons, armor and whatever food and water they could carry. Camp was broken and the baggage was left just outside the walls of Adrianople, with a “suitable guard of legions” and under the cover of archers on the city walls. Inside the walls of Adrianople and thus protected were the civilian court of Valens, the almost all of the insignia of his imperial rank – the emperor would go into battle with a minimum of adornments – and that part of the imperial treasury normally carried by the praesental army in transit.[108]

The praesental army lined up to march in battle formation, meaning that once it arrived where it wanted the left wing of its battle line to be at the battlefield it could wheel to the right and march to where it wanted its right wing located, with, generally, the troops in the front of the column ending up on the right of the line and the troops in back on the left of the line. Then it could deploy for battle relatively easily. [109]

While on the march, cavalry would screen the front, rear and sides of the formation. At the head of the army would be the Emperor Valens and his bodyguard cavalry, the scholae. Since once the army was deployed for battle he would be at the traditional spot on the right of the battle line, the scholae was at the front of the column.

Thus, as Ammianus says, with extreme haste Valens and the Army in the Emperor’s Presence marched to their destiny.

With the battlefield as yet unknown, the route the army took is likewise unknown. Ammianus does give a few hints, however. At some point the army left the Roman road network and marched a very significant distance cross country, probably over the rough, hard, rock-strewn ground that is common in this part of the Balkans. Ammianus has the Romans eventually returning to a road. One rather obscure history has the total distance covered as 12 Roman miles. [110] Finally, “at the eighth hour” – between 2:00 and 3:00 in the afternoon – the Romans reached the Gothic laager, set up on a small hill.

The Army in the Emperor’s Presence began to deploy for battle, possibly on a small ridge south or southeast of the laager. [111] The Romans had deployment for battle down to a playbook, and they were running it page by page. The first page had the right wing cavalry at the head of the column, in this case the scholae, moved forward between the laager and the infantry to screen the army’s deployment, in this case presumably without their imperial charge. The second page had the left wing cavalry at the back of the column pass the infantry to join the right wing cavalry in screening the deployment.

Fritigern appears to have been surprised by the appearance of the praesental army. He immediately began deploying the Gothic infantry in front of the laager. With the wagon circle close at hand it would not have taken too long for the Goths, the vast majority of whom were wearing Roman armor and carrying Roman weapons, to line up and deploy for battle. While doing so they, in the words of Ammianus, “uttered savage and dismal howls,” which were some type of war chant. In the meantime, the van of the Roman infantry had wheeled to the right and was marching into position, banging their spears and swords against their shields and shouting taunts of their own. The lead of the column was far ahead of the rear, however, and it would take a few hours for them to line up and deploy.

The noise and the taunts were intended to intimidate. In this they succeeded. The screening activities of the scholae, which consisted of the shooting arrows, running around, kicking up dust, poking around the outside of the laager, launching fake charges to disrupt the Gothic deployment and generally making a nuisance of themselves added to the uneasiness for Fritigern and his troops. He could not counter the scholae because right now he had no cavalry; the Greuthungian-Alannic-Hunnic coalition of cavalry under Alatheus and Saphrax was away. They would be back, but that would take time.

Fritigern decided his best option was to re-open peace negotiations with Valens. He sent ambassadors to the eastern emperor. Ammianus describes Valens’ response: “The emperor was offended at the lowness of their rank, and replied, that if they wished to make a lasting treaty, they must send him nobles of sufficient dignity.” [112] In other words, Fritigern sent men who were not nobles and thus not men of “honor” with whom the emperor could conduct negotiations, so Valens responded by demanding that the Gothic chieftain send ambassadors of sufficient rank.

Ammianus and many, many other commentators call this offer to negotiate merely a ploy to buy time for the Gothic cavalry to return. And that was probably the case, but that does not mean it was only a ploy to buy time. While Fritigern certainly wanted his cavalry available, again, he knew that the best option for his people was the protection and stability offered by the Roman Empire, which was not served by continuing this war. This was also the first chance he had to discuss the situation with Valens personally. So this did represent an important opportunity.

But why would Valens, when he was supposedly on the verge of achieving his objective, agree to delay the start of the battle for negotiations? That has been another of the major questions surrounding the Battle of Adrianople. While Valens never had the chance to explain his actions, it is possible to reconstruct a number of considerations that could have driven him to delay battle or even seek to avoid it.

The strategic issues that existed when the Goths asked for entry into the Empire remained. Valens needed manpower, not just for the army but, if Frigeridus’ settlement of the Taifali is any indication, for agriculture as well. Once again, the Goths needed a home and protection from the Huns (or at least those Huns that had not signed on to Fritigern’s coalition). This war served neither of those interests. Killing the Goths in battle would cost Valens manpower, not just the potential manpower offered by the Goths but by Roman losses in the battle. If he could bring this clash to a peaceful resolution it might not bring much of a political benefit – anger at the Goths among the Roman citizenry in the Diocese of Thrace was strong – but it would bring practical benefits in relatively short order.

But Valens also had far more immediate concerns.

The praesental field army was having serious problems getting itself deployed for battle. The ecclesiastical historian Sozomen says that Valens attacked the Gothic battle line “before he had arranged his own legions in proper order.” [113] Zosimus indicates the army was in some disorder even as it left Adrianople that morning. [114] Ammianus only lists one specific problem with the deployment. [115] It has been argued that Ammianus’ relative lack of mention of disorder in the Roman ranks means that there was no serious problem with the line up and deployment of troops. [116] Except that argument ignores a basic fact of movement of armies, as the Greek playwright Euripides had noted and of which the veteran Ammianus was well aware: movements of large armies are always disorderly. [117]

Anyone who has watched or, better yet, marched in a parade is aware of the floats and marching bands stopping, starting, speeding up and slowing down, simply because people naturally move at different rates. The problem is exponentially worse with a 15,000-20,000-man army. On the move the army will tend to spread itself out, both in terms of width and especially length. Absent periodic stops for redressing the lines, of which there is no evidence here because of the perceived time pressures, by the end of the march the army column will be wider than when it first started, and, similarly, the back of the column will be farther away from the van than when it first started. As noted earlier, Gibbon says the troops had to “precipitate their pace.” [118] It need not be a major problem, and under normal circumstances it might just require a bit longer to get the army formed up in battle line and deployed for combat.

But here there were obvious problems. The first page of the Roman playbook was run properly with the right wing cavalry moving in front to screen the infantry deployment. The second page had the left wing cavalry at the back of the column passing the infantry to join the right wing cavalry in the screen. Except the left wing cavalry had, in the words of Ammianus, “greatest difficulty” in moving up, and a large number of them remained scattered on the road. [119] The scholae serving as the right wing cavalry had to perform the screening duties by themselves.

Ammianus does not give a reason for the left wing cavalry’s difficulties, but a likely reason is that the cavalry could not get around the infantry. The infantry spreading out on the march would have made getting around them more difficult, as the cavalry would have had a longer journey from the back of the column and the road would have been blocked. The infantry immediately in front of the left wing cavalry in the column would have been a unit of Batavian auxiliaries, the Batavi seniores, that was intended to constitute the second, or reserve, line of Roman infantry in the battle deployment. Some of the cavalry probably tried to push through, with the result that they became entangled with the Batavians. Others probably turned around and looked for another route to the front; this would explain the cavalry “scattered” or “straggling” along the road, and, indeed, one prominent work has advocated just this scenario. [120] Still others would have just waited impatiently for the traffic jam to clear.

The problems of the cavalry were not isolated and were not the cause, but rather the symptom, of the larger disorder among the Roman troops. Because the military tribunes, who were the Roman equivalent of today’s officers, had not been able to periodically stop the column and redress the ranks or, apparently, keep the men moving at a consistent speed, the army was more spread out. [121] It would thus take longer to get into position, longer still for the tribunes to get everyone lined up and deployed for battle. “The line of battle was formed with tedious confusion and irregular delay,” says Gibbon. [122]

The problems with disorder were complicated by a far worse issue: the condition of the praesental troops.

Valens’ moment in the sun had become some eight hours in the sun, in a region where the August temperatures regularly top 100 degrees. [123] The praesental troops had been forced to march cross country, over what Ammianus calls “broken” terrain, and, indeed, in this part of the Balkans the ground is often stony and hard, with little significant vegetation – and reflective of heat. [124] The Romans had fought in arid and semi-arid environments for centuries and so were experts at fighting in hot, dry conditions. But to speed along their march, the Romans had left their baggage behind, which included most of their food and water, and had marched since daybreak on dusty roads and rocky ground as the sun rose higher and the temperature soared. Soldiers wearing armor would have been broiling inside. The Eastern Roman Emperor Maurice would later suggest the troops take a pound or two of food and a flask of water. [125] The Roman troops here certainly took along some extra water because of the heat. But they had had no lunch, and Ammianus’ references to their extreme hunger suggest they took little if any food along.. And, sweating buckets in their armor, they would have quickly gone through whatever water they carried. Men and horses would have been fighting heat exhaustion and even heat stroke, and were probably dropping in the ranks.

While talking negotiation, Fritigern ordered the dry grassfields nearby set on fire. It may have been merely an innocent act to signal the Gothic cavalry, but the wind carried the smoke into the Roman lines, compounding the troops’ misery and possibly obscuring their vision.

To quote Gibbon, the Roman troops, “exposed without shelter to the burning rays of the sun, were exhausted by thirst, hunger and intolerable fatigue.” [126] And they still had a battle to fight. A big reason why Fritigern had the horde stay away from Adrianople was to make the Romans wear themselves out marching to battle in the event of a sortie. And Valens, apparently, did exactly what Fritigern had wanted.

It is here that the folly of the Roman plan at Adrianople becomes apparent. The praesental army had to march for eight hours in scorching temperatures with little water and less food, then deploy for battle and fight; then, if they won, march eight hours back to Adrianople, arriving after dark, and then set up camp again. Sixteen hours without food, at least eight without water.

During the Second Punic War, Hannibal had been a master at forcing the Romans into fighting him while fighting their own hunger, thirst, fatigue or even the weather. Rome had paid a heavy price for learning Hannibal’s lessons. But learn them it did – Publius Cornelius Scipio would use the same tactics against Carthaginian troops in Spain. Now the Romans were again about to fight while suffering from hunger, thirst and fatigue, but they had done this to themselves.

Ammianus himself has Julian on a campaign in Gaul, after leading his army on a long march, try – and fail – to convince his men to avoid battle in just such circumstances; the result was the Battle of Strasbourg, a Roman victory, to be sure, but a near-run thing. [127] Indeed, the practice had been for the army to march to the battle location, set up camp, rest and fight the next day. Vegetius, in his later military treatise to … someone (precisely whom is a matter of debate), discusses the importance of fighting a battle fresh. He cites an unidentified incident of what happens when you do not, an example that can only refer to Adrianople. [128]

It is hard to fathom what (or if) Valens and his staff were thinking when they developed this plan. While it is hardly unusual for soldiers of any nation in any war to deal with hunger, thirst and lack of sleep, a good commander will usually go out of his way to avoid it, and will willingly subject his troops to such circumstances only in the most dire of circumstances. Such circumstances were not present at Adrianople. Valens had other options besides rushing to battle. This plan, such as it was, leads to the obvious inference that he was not a good commander.

The almost complete lack of preparation for the march does suggest another possibility: did Valens and his generals have accurate information as to the horde’s location? Did they know the best route to get there?

There is a possibility they may not have. Valens certainly had the horde’s location, both from his procursatores and Fritigern’s letter. But did he know how to get there? Part of the exploratores’ job was to know the local terrain. Part of the procursatores’ job was to find the best route for the army to march to battle. Both had the job of providing intelligence as to the terrain.

Eight hours is an extremely long march to battle, especially on a forced march without adequate water and food; it is also twice as long as it took the Goths to march from their encampment to Adrianople after the battle. [129] The issues with the left wing cavalry suggest that the road the praesental army took to the battlefield was inadequate. The cavalry scattered on the road may have been looking for something, likely another route to the front, but maybe they were returning after looking for the laager. The aggressiveness of Valens during the consilium had disappeared – mysteriously, some would say – by the time he and his army reached the battlefield.

And according to Ammianus, the procursatores had already gotten part of the report on the location and state of the Gothic horde wrong in their underestimate of the troop strength.

Did they get the route to the battlefield wrong as well? Did they underestimate the distance? Pick a bad route?

Unfortunately, with Ammianus making no mention of such an issue (not that he necessarily would, given his hostility to Valens), the hard evidence for such a problem is weak. Nevertheless, it is intriguing to note that maps as the Romans understood them were not the same as maps today. Most Roman maps were simple strips based on individual roads, much like the “Trip-Tiks” put out by the American Automobile Association. It would show towns, natural features, shortcuts and waypoints like Nice along that road. Some of these “maps,” called “itineraries,” were simply lists of such waypoints. [130]

But these maps usually did not show off-road distances and features, and most of the praesental army’s journey from their camp outside Adrianople to the Gothic wagon laager was largely off-road. The procursatores would have had to fill in that information gap.

Be that as it may, the deterioration in the condition of the troops was obvious. Their rush to fight the Goths had left them shackled for that same fight. While the fear and adrenalin of combat would rule initially, the thirst, hunger and fatigue would eventually take its toll, especially on the troops’ endurance.

This was a problem with no immediate solution. The food and water had been left with the baggage outside the walls of Adrianople. The Goths had their own food and water and were not likely to share it with Roman troops preparing to attack them. Even if other water sources in the area could be found, drinking from them so close to the enemy was too dangerous, and the Gothic troops were unlikely to allow the Romans time to refresh themselves before combat.

With few options for taking care of his troops in the interim, it should come as no surprise that when Fritigern again asked for peace talks, Valens would agree. Under the circumstances, it was a logical, rational decision.

And, like many of Valens decisions, it would turn out badly.

Very badly.

Another consilium

Fritigern had sent another ambassador forward and asked that men of noble rank be sent to the Gothic camp to serve as hostages while peace talks were conducted. The use of hostages to guarantee the safety of prominent figures or compel adherence to treaty agreements was a very common practice in the ancient world, but the nature and context of Fritigern’s offer is unclear due to ambiguities in Ammianus’ text.

The general inference has been that once the hostages were received, Fritigern himself would go to the Roman lines and negotiate with Valens in person. It must be noted that this assumption is by no means an unreasonable one. If true, it would suggest that the Gothic chieftain did not view his earlier treatment at Marcianople as particularly perfidious.

The ambiguities in Ammianus’ text do not end there. Some translations hint that Fritigern was prepared to use his political capital to sell any peace agreement to his recalcitrant coalition. [131] Other translations claim that the Tervingian chief offered to provide the Romans with water, food and other supplies, again, perhaps over the objections of his people. [132] This offer would represent a very astute observation on Fritigern’s part of the condition of the Roman troops, and it would be an offer that Valens would be almost compelled to accept. And he did.

But whom could the Romans send?

Austin and Rankov say that Fritigern’s offer to negotiate “successfully distracted the emperor and his staff for a while from devoting their full attention to the battle.” [133]

Except it seems to have distracted far more than Valens’ “staff;” it distracted the senior generals. For while Ammianus does not say it, he alludes to what must have been a second consilium, this time between Valens and his generals. The only general he mentions by name is Richomeres, the ambassador from Gratian’s court whose command responsibility in this engagement, if any, is unknown. It can be deduced that the Sarmatian general Victor, the Master of Cavalry, was in attendance as well; Victor was present with the emperor at the end of the battle long after the cavalry were gone, and having negotiated the truce with the Sassanids, he had diplomatic experience that would be useful in the negotiations that appeared to be imminent.

Beyond those two, determining the attendees involves speculation and conjecture, but if Richomeres and Victor were there, it is hard to see a reason why the other senior officers would not have been. This is true especially in the case of Trajanus, who, Ammianus hints, was now back in the fickle emperor’s good graces after his trashing of Sebastianus; Trajanus also had diplomatic experience, though part of that experience consisted of arranging a banquet for the assassination of the mercurial king of Armenia. It is likely that Sebastianus, Saturninus and other generals were there as well.

Many of these men would have been summoned from their duties on the front line, so screening, lining up and deploying the troops into battle line would have been left to lower level military tribunes. Those tribunes had trained for this, and the troops had practiced screening and deployment maneuvers countless times. They should have been able to handle it by themselves.

Ammianus describes the discussions that took place:

The proposal of the dreaded leader was welcome and approved, and the tribune Aequitius, then marshal of the court and a relative of Valens, with the general consent was chosen to go speedily as a surety. When he objected, on the ground that he had once been captured by the enemy but had escaped from Dibaltum, and therefore feared their unreasonable anger, Richomeres voluntarily offered his own services and gladly promised to go, thinking this also to be a fine act and worthy of a brave man. And soon he was on his way [bringing] proofs of his rank and birth. [134]

There are very curious elements to these discussions. For starters, the Romans seem to have tried to “low ball” the Goths on the hostage. Aequitius was only a military tribune; it was probably thought that he would be acceptable because he had a post with the court and was a relative of Valens. Apparently, Aequitius was not present for these discussions, probably helping direct the troops’ deployment in the absence of the more senior officers at this consilium. When summoned and told of the decision, Aequitius objected. It seems he had been captured by the Goths at the battle near Dibaltum, had managed to escape, and now feared the Goths would retaliate against him. To his credit, Valens did not pull rank on the issue. The problem was none of Valens’ generals was willing to go.

A more intriguing element is the behavior of the ubiquitous Richomeres. Ammianus says that he “gladly promised to go, thinking this also to be a fine act and worthy of a brave man.” That may be true, but it sounds somewhat like spin. That Richomeres, having less “skin” in this battle than Valens’ people, volunteered to serve as the hostage is curious by itself. In fairness to the eastern generals, they were deploying for a battle and may not have wanted to disrupt the troops any more than the consilium itself was doing by depriving them of their senior officers. But openly saying this was “a fine act and worthy of a brave man” – in front of the emperor, no less – can be read as a dig at the relative cowardice of Valens’ generals. Richomeres may have volunteered simply out of exasperation at the dysfunctionality of Valens’ army that he had witnessed in the last 24 hours; they had refused to wait for nearby reinforcements from Richomeres’ boss, had botched the march to the battlefield and now could not even conduct negotiations properly. [135]

Whatever the truth of the matter, Valens remained with his generals, and Richomeres went back to his escort and got his proof of rank and noble birth, whatever that was. He then set out for the Gothic line, apparently with the Gothic ambassadors as escort.

Richomeres would have seen the Batavians, the second line of Roman troops, still getting themselves lined up. Not that this was considered a problem because they were expected to fight much later, if at all.

He would have passed the left flank cavalry, still trying by various routes to get to the front line.

He would have passed the first Roman line. The right side of the line would have been lined up and deployed – shields up, spears leveled – by now. The center was likely close to being lined up, but not deployed. The left side was still probably getting lined up, though by now it would have been well along. All under the direction of the military tribunes, the lower level officers who, like their troops, would have seen the ambassadors traveling between the battle lines and wondered what it all meant, but who still had their orders to prepare for battle and were doing their best to carry out those orders.

As he and his little party left the relative safety of the Roman lines, Richomeres would have seen the scholae, part of the right flank cavalry, still performing their screening duties for the infantry deployment, still …

“What in Gratian’s name are they DOING?

The scholae attack

It is perhaps appropriate in this battle about which so much information is missing that at this critical juncture in Ammianus’ account there is what is called a lacuna, which is a fancy way of saying “gap,” of some 20 characters. [136] Like a recorded episode of CSI with five critical minutes missing. What happened in that gap is anyone’s guess.

Ammianus then describes the tragic events that followed:

As [Richomeres] was on his way to the enemy’s rampart, the archers and the targeteers, then under the command of one Bacurius of Iberia and Cassio, had rushed forward too eagerly in hot attack, and were already engaged with their adversaries; and as their charge had been untimely, so their retreat was cowardly; and thus they gave an unfavorable omen to the beginning of the battle. [137]

This paragraph by itself is so ambiguous as to be subject to varying inferences. Ammianus’ next paragraph does little to clear things up:

This unseasonable proceeding not only thwarted the prompt action of Richomeres, who was not allowed to go at all, but also the Gothic cavalry, returning with Alatheus and Saphrax, combined with a band of the Alani, dashed out as a thunderbolt does near high mountains, and threw into confusion all those whom they could find in the way of their swift onslaught, and quickly slew them. [138]

These two paragraphs can be and are interpreted differently by different commentators. Any of these interpretations may be true; at this point there is no way to prove or disprove any of them, especially given the lack of positive identification of the battlefield. But going back to Rome’s historical military practices and my own limited knowledge of Latin, I will attempt to present a coherent account of what happened here, hopefully using a minimum of speculation to fill in the gaps and making what speculation there is to at least be of the informed variety.

While the importance of cavalry had increased since earlier times, its role had changed very little. The cavalry’s role was to scout, to screen and if possible to deliver the decisive blow in any combat engagement. The last role in that list is what had increased most. But before that could happen, the cavalry still had to protect the infantry while the infantry was deploying for battle. [139]

That screening function, at least in the tactical sense, could include a number of activities. On at least one occasion, it involved the cavalry just standing there in its own line of battle. Mostly it involved the cavalry running around, hampering the view of the deploying infantry with both its mass and the dust it kicked up. The cavalry could also poke and prod the enemy line, finding out what it could about them and trying to disrupt the enemy’s own deployment. It was all about keeping the enemy from disrupting the deployment process, whether by blinding them, discouraging them, or giving them something else to worry about.

In most instances, the enemy would have cavalry of its own, who would move out and skirmish with the Roman cavalry. This was definitely a tradeoff, with the enemy infantry being protected, but at the cost of further blocking its view of the ongoing infantry deployment. It should be noted that this was skirmishing; it was not supposed to be or even lead to heavy combat. Not yet. After the infantry had finished its deployment (sometimes significantly afterwards) the cavalry would move to the infantry’s wings to protect the flanks and perhaps land a decisive blow on the enemy flanks.

At the beginning of the deployment, Ammianus has the cavalry in its proper position screening the infantry, or at least the right flank cavalry. Having led the march to the battlefield, these were the units designated scola or schola, the all-cavalry personal escort of the emperor, technically not part of the praesental army, but, theoretically at any rate, under his personal command (though in any major combat action they would have likely been placed under the command of the general in tactical command if that individual was not the emperor, which seems likely here). [140] Ammianus gives some information as to the units involved: archers (sagittarii) and “targeteers’” (the scutarii). The archers here would have undoubtedly been horse archers; while light infantry could perform a screening and skirmishing function, foot archers could not because they could not stand up in any kind of close combat. The scutarii (without the sagittarii designation) were the heavy cavalry, though not the really heavy cavalry like the cataphracts (catafractarii and clibinarii), armored lancers that would form the basis for the medieval armored knight. [141] Any cataphracts in the right flank cavalry would not have been involved in skirmishing, being too heavy for that purpose, and instead would have taken position on the right wing of the infantry. [142]

The Notitia Dignitatum lists the possible units involved as the scola scutariorum prima, the scola scutariorum secunda and the scola scutariorum sagittariorum. [143] That scola scutariorum sagittariorum listed is the same unit discussed by Ammianus at Adrianople. [144] It has been persuasively argued by historian David Woods that the scola scutariorum sagittariorum, were commanded by Bacurius, a prince from Iberia. [145] After this day, Bacurius would go on to have a long and distinguished career in the Roman military, full of heroic deeds. This day would not be one of them. Assuming Ammianus’ reference to Bacurius and Cassio was because of their positions as unit commanders, the accompanying scola scutarii would have been commanded by Cassio, about whom little is known. The man who would be commanding all the cavalry, Victor, was not with the scholae, but was back with Valens dealing with the attempted peace negotiations, about which Bacurius and Cassio likely had little if any knowledge.

The sagittarii and the scutarii would have had different roles in the screen. Despite their designation as scutarii, which indicates they may have been armored, the sagittarii would not have been expected to hold up in close combat. They would have been used to bombard the Gothic lines with arrows, inflict casualties and keep them off balance. The scutarii would have probably kept position between the Gothic line and the sagittarii, and also done their own harassment of the Gothic infantry. Included in the mix of these harassing tactics would have been mock charges, in which the cavalry would charge the enemy infantry, then turn away at the last minute, which would keep the infantry tense and their spears leveled when they did not need to be, thus helping to tire them out. As Ammianus makes clear, they had no orders to attack.

It should be noted that, though the actions of the scholae in harassing the Goths have been called “provocations” (to which the Goths did not respond) and inconsistent with peace negotiations, absent orders to the contrary, harassing the enemy is exactly what the screening cavalry was supposed to do. [146]

Many historians have interpreted the scholae attack at Adrianople as something akin to an impromptu Roman banzai charge; exasperated and impatient at the peace negotiations, these horsemen decided to take matters into their own hands and out of the blue charged and initiated the battle. [147] While this is certainly a possibility, it seems highly unlikely. It is not hard to find instances of Roman units attacking without orders or even against orders. “[S]oldiers’ entering battle against orders was hardly uncommon in the Roman army.” [148] In 168 BC, Aemelius Paulus was only able to restrain his troops from attacking at Pydna for a limited time. [149] Julius Caesar planned a limited attack at Gergovia in 52 BC as something of a rearguard operation. The attack was successful – so successful that his troops got carried away and tried to take the entire town, against orders, leading to a defeat with significant casualties and political repercussions. [150] At Gamala in AD 67, Vespasian’s troops took it upon themselves to attack an opening in a wall and were badly beaten by Jewish defenders. [151] In an incident during the siege of Jerusalem, some of Titus’ troops were tricked into charging the gates without orders – charging into an ambush. [152] In a second incident during that same campaign, Titus himself was surprised to hear a Roman trumpet blast emanating from atop the Antonia Fortress, taken by Roman legionaries in an action that was a complete surprise – to the Jewish defenders and to him. [153] But it is very difficult to find instances of such attacks in circumstances like Adrianople, with troops attacking even before their own line is deployed and ready for battle.

We will likely never know what exactly happened to start the Battle of Adrianople. But in applying these puzzle pieces to Ammianus’ account, an outline of the accidental onset of battle begins to form.

Since the Goths had no cavalry on the field yet, the scholae would have had full rein of the field between the two lines. They would have felt little fear from the Gothic infantry, and would have likely taken liberties with their screening duties. They may also have felt a need to make up for the lack of left wing cavalry in the screen, though they probably kept relatively close to their deployment area on the right. The rain of arrows would have gotten heavier and heavier, but, paradoxically, less and less organized. The mock charges would have gotten closer and closer before turning away. [154]

Until one of the scutarii got too close and could not turn away.

Some of the scutarii would have gotten tangled up with the Gothic infantry, armed with spears. In an oversimplified “rock-paper-scissors” comparison of ancient troop types, the order would have been “swordsmen-spearmen-cavalry.” [155] Since horses have historically shown a strong aversion to charging walls of shields bristling with forests of long, pointy things, spearmen usually trump cavalry. Cavalry must attack spear-armed infantry from the flank and rear to avoid the spears.

But here, some of the scutarii would have been charging from the front – the worst possible position for cavalry to attack spearmen. When they got too close, the Gothic spears took their toll. Horses were killed and riders injured or unhorsed.

The rest of the scutarii then joined their engaged comrades, trying both to extricate them from their entanglement with the infantry and to inflict casualties on the Goths –once again, attacking the spear-armed Gothic infantry from the front.

The sagittarii focused their archery on the engaged part of the Gothic line. And the Gothic infantry near the fighting turned to help their brethren. It was becoming a maelstrom, sucking in everything and everyone near it.

Richomeres, says Ammianus, was not permitted to continue his mission. His Gothic escorts likely shouted angry charges of Roman perfidy – again – at him before heading back to their own lines. In all probability, Richomeres himself was puzzled as to what was going on.

Valens and his generals, still preparing for those peace talks that would now never come (at least while Valens was alive), were also bewildered at what they were seeing. That confusion would have been followed by horror – combat had just started and his senior officers were not with the troops, but with him

Sebastianus, Saturninus, Trajanus, Victor and others got back to the front as quickly as they could. But events were already beyond their control.

Many Roman legionaries on the front line, more worried about their own deployment, their own upcoming combat and what was going on with these ambassadors coming and going, may not have even seen the initial “attack” by the scutarii, but they saw the remaining scutarii and sagittarii charging. And then the whirlpool sucking in the nearby Gothic troops. Their first reaction was likely bewilderment. For the tribunes, that confusion would have included far more serious questions: Why are they attacking? Did I miss the order to attack? Shouldn’t we be helping them?

With a yell, the first line of Roman infantry advanced. It was probably on orders from Valens, given through a blast of trumpets, though now he was more trying to catch up events than manage them. However, there is a possibility that there was no formal order for the first line to advance, but that some of the tribunes, thinking the scutarii attack meant they all were supposed to be attacking and had somehow missed the order, led the infantry in somewhat of a haphazard advance, maybe “zipping up” from the Roman right to the left in closing with the Gothic line. The legionaries themselves stuck to their training: double shield wall in front, brisling with spear points, the back ranks hurling missiles at their Gothic rivals. Then the lines closed and the traditional toe-to-toe slugging match began.

The left flank cavalry, still trying to deploy, joined in the initial infantry charge, probably at the tail end of it. Cavalry were running straight from the road into the enemy line.

So the battle for the fate of the world – not that anyone actually knew that – had started. Badly, for the Romans. The Army in the Emperor’s Presence – hungry, tired and desperately thirsty – had “attacked” by accident. Valens and his generals had lost control of the action and were struggling to even keep up with it. And the fighting had now sucked in all the Roman first line infantry and cavalry.

But not the Gothic cavalry.

Until now.

The attack of Alatheus and Saphrax

No one knows where they came from. No one knows where they went. But the contingent of Greuthungian, Alan and Hunnic cavalry under the Greuthungian regents Alatheus and Saphrax struck “like a thunderbolt,” changing the course of the battle and, indeed, of world history. [156]

Traditionally, the assumption has been that up until this point, the Gothic cavalry had been on a foraging expedition when, facing the Roman army, Fritigern recalled them. Recently, that assumption has been questioned. [157] The cavalry may have been hiding in ambush all along, though if this were the case it may have been prematurely executed, not that it mattered here. [158] There are other possibilities as well. After the horde’s presence for several months, the forage in this area would have been largely gone. The horde’s extremely slow movement suggests a lack of scouting and screening cavalry. The Roman procursatores and exploratores appear to have had no problem keeping the horde under continuous surveillance. And there is the matter of that Gothic raiding party that had attempted to seize Nice. There is a significant possibility that the cavalry under Alatheus and Saphrax were, in fact, the raiding party and were now just returning from that failed expedition. If this is the case, the fortuitous timing of their return would be such an incredible coincidence as to be worthy of a bad Tom Cruise movie. But the possibility cannot be discounted.

Estimates by historians as to precisely where on the battlefield the cavalry attack came from are all over the place – the Roman left, the Roman right, or both. Again, absent additional probative evidence, especially conclusive identification of the battlefield, any of these scenarios could be true. The scenario I present below for consideration is based on my read of Ammianus’ text and my study of military tactics.

Reset for a moment with the translation of Ammianus passage:

This unseasonable proceeding not only thwarted the prompt action of Richomeres, who was not allowed to go at all, but also the Gothic cavalry, returning with Alatheus and Saphrax, combined with a band of the Alani, dashed out as a thunderbolt does near high mountains, and threw into confusion all those whom they could find in the way of their swift onslaught, and quickly slew them. [159]

The “attack” by the Roman right-wing cavalry prompted the response by the Gothic cavalry, though Ammianus likely condensed the reaction time. Ammianus’ use of the singular in “thunderbolt” (fulmen) strongly suggests the decisive attack came in one wave. [160]

And the attack was devastating. The first target of the Gothic cavalry was the scholae who had started it all, in particular Cassio’s scutarii. Already engaged with the Gothic infantry, under a shower of missiles from troops behind the front line and on top of the wagons of the laager, the scutarii were now taken in the right flank by Alatheus and Saphrax. They had no chance. Likely cut to pieces by Gothic spears, the scutarii fled—a flight which quickly snowballed. With their bodyguards gone and unable to hold up in close combat, the sagittarii now beat a “cowardly” retreat as well. Caught up in the flight would have been any remaining cataphracts on the Roman right, better at shock than at defense, too small in numbers to stem the tide.

Ammianus’ description suggests that the charge by Alatheus and Saphrax was like an unstoppable wave. They “threw into confusion all those whom they could find in the way of their swift onslaught, and quickly slew them.” [161] “[A]ll those whom they could find in the way of their swift onslaught” sounds like a lot more than one wing of cavalry. And it probably was.

The Gothic cavalry would have chased the fleeing Roman cavalry deep into the Roman rear, where they came across the Batavian infantry. The Batavians, Germans whom the Romans had long ago resettled in what is now the Netherlands, were sort of Rome’s version of the famous Gurkhas of the British Army. [162] The Batavians were considered crack units, though their performance of late had been underwhelming. Now they would not get the chance to overwhelm or even just plain whelm. After their entanglement with the left wing cavalry, they were still lining themselves up when the Gothic cavalry appeared in their midst. While the Batavians were armed with spears, without being even lined up, let alone deployed to present a shield wall bristling with spearpoints, they could not stand up to the Gothic onslaught. The Batavians were “thrown into confusion” and scattered, with much of the fleeing infantry being run down.

The last Romans unfortunate enough to be caught in this Run of Terror were stragglers from the left-wing cavalry. Still trying to make their way to the front, still trying to figure out what was going on, they ultimately ended up staring into the face of the Gothic charge. No time to get organized, no way to fight back, they were all dispersed and driven into flight.

In one ludicrously well-timed charge, Alatheus and Saphrax had completely undermined the Roman battle line. The right wing cavalry was gone, much of the left wing cavalry was gone, the second line of infantry was gone. And the Gothic cavalry was now behind the Roman main battle line of infantry.

They reformed and prepared for another charge, this time at the rear ranks of the Roman first line infantry.

The noose was ready around the Roman neck. All that remained was to draw it tight.

The infantry clash

Yet another one of the mysteries of Adrianople has been the question of why the Roman infantry collapsed as it did. The scenario presented above is at variance with some popular (and not unreasonable) interpretations of Ammianus, in which the disintegration of the right wing cavalry and the charge of Alatheus and Saphrax take place before the infantry engagement. To be sure, Ammianus discusses the collapse of the scholae before the arrival of the Gothic cavalry, and does not mention the early dispatch of the Batavians at all. [163] Under this scenario, the Romans should have been able to survive and possibly even eke out a victory, albeit a costly one.

What happened?

The key for surviving this scenario would have been the praesental army forming a circular line of its own, forming the shield wall completely round the circle with the spears facing outwards. The spears and shields would have kept the cavalry at bay, while the troops facing outwards would have had a chance to resist the physical pressure they would have faced to squeeze the circle, thus further reducing its combat power. While so doing would not guarantee survival, let alone victory, failure to do so would almost guarantee disaster, as the Romans had found out at Cannae some six centuries before.

In fact, the Romans had long ago developed tactics for dealing with the potential encirclement they were facing now. A line could retreat upon itself into an oval-type formation known as an orbis, then expand outward. [164] Another tactic was to have every second cohort in the battle line turn about 180 degrees, in full view of the enemy, then have the entire line advance -- but because of the turnabout, the line would advance in opposite directions, pushing the encircling troops outward. [165] In other instances, the rear ranks of the Roman line reversed themselves for a similar effect. That could have been done here – but only if the Roman infantry had not already been engaged with the Gothic infantry. Once the Roman legionaries were engaged, there was no possibility of redeployment.

Similarly, the Batavians manning the second line were in an even better position to respond to the encirclement, except they had already been destroyed.

This also goes to the issue of Roman command and control, disrupted by the consilium. After the botched attack by the scholae, the senior commanders likely rushed back to the front line. It was all they could do just to get control of their own units, never mind responding to the situation.

For their part, the Roman legionaries, in close combat with their Tervingian counterparts, were giving a good accounting of themselves in spite of their thirst, hunger and fatigue. As Ammianus states:

On every side armour and weapons clashed, and Bellona, raging with more than usual madness for the destruction of the Romans, blew her lamentable war-trumpets; our soldiers who were giving way rallied, exchanging many encouraging shouts, but the battle, spreading like flames, filled their hearts with terror, as numbers of them were pierced by strokes of whirling spears and arrows. [166]

The initial conflagration caused by the scholae “attack” had spread all along the line. The initial advance by the Roman infantry seems to have been pushed back momentarily, maybe a symptom of a not-completely organized advance. But the soldiers quickly recovered and were holding their own in the slugging match. Unfortunately for them in their role as “anvil”, most of their “hammer” in the form of the Roman cavalry was already gone.

Most, but not all.

Part of the Roman left wing cavalry had managed to escape the apparent tangle on the inbound march. Though perhaps somewhat disorganized, they had made an effective attack. While the Roman legionaries had kept the Tervingi infantry engaged in the front, the Roman left-wing cavalry hammered at the extreme Gothic right. Fritigern’s troops buckled under the onslaught, and were gradually driven back, almost all the way back to the wagons.

The Roman legionaries and the left wing cavalry fighting so heroically under such difficult circumstances likely had no idea of the disaster that had taken place on their right and in their rear. It seems that Valens and his generals did not either.

But Fritigern did. And he knew exactly what to do.

From inside the laager, apparently through prepared sally ports, came a flood of Tervingian infantry reinforcements. Ammianus references “the superior numbers of the enemy” and says they “pour[ed] forth in huge hordes [and] trampled down horse and man.” [167] Maybe the troops within the laager did indeed outnumber the Romans, as Ammianus suggests. While Alatheus and Saphrax and their cavalry hammered away at the unprotected rear of the Roman battle line, archers and missile troops behind the Gothic battle line and on top of the wagons bombarded the legionaries with missiles, and these new infantry reinforcements worked their way around the Roman flanks.

Ammianus sets it up:

And because the left wing, which had made its way as far as the very wagons, and would have gone farther if it had had any support, being deserted by the rest of the cavalry, was hard pressed by the enemy’s numbers, it was crushed, and overwhelmed, as if by the downfall of a mighty rampart. The foot-soldiers thus stood unprotected, and their companies were so crowded together that hardly anyone could pull out his sword or draw back his arm. Because of clouds of dust the heavens could no longer be seen, and echoed with frightful cries. Hence the arrows whirling death from every side always found their mark with fatal effect, since they could not be seen beforehand nor guarded against.
But when the barbarians, pouring forth in huge hordes, trampled down horse and man, and in the press of ranks no room for retreat could be gained anywhere, and the increased crowding left no opportunity for escape, our soldiers also, showing extreme contempt of falling in the fight, received their death-blows, yet struck down their assailants; and on both sides the strokes of axes split helmet and breastplate. [168]

First priority for Fritigern was the hard-pressed Gothic right. The Roman cavalry, attacking ferociously, were almost at the wagon laager, but their charge was spent and they lacked support to exploit their success; the other left wing cavalry now gone. [169] The fresh Gothic infantry crushed the remaining Roman cavalry – literally, according to Ammianus.

The Romans, with their flanks now completely exposed, were forced to draw them back in on themselves. Left and right eventually met in the Roman rear, re-establishing some semblance of a shield wall in a circular formation. Inside that circle was the all of the Army in the Emperor’s Presence, living up to its name with Valens still in their midst.

Now the Goths surrounding the Romans put the squeeze on. Like a fist balling up a piece of paper, the Gothic infantry and cavalry pressed the Roman circle on all sides into a smaller and smaller area. As they did so, the praesental army lost combat power; with the circle getting smaller and smaller, the number of troops the Romans could put on the front line got fewer and fewer; the remaining combat power of the army diluted outwards while the Goths concentrated their combat power inwards at the Romans.

Even worse for the Romans, as the circle got smaller and smaller, as the Goths squeezed harder and harder, the legionaries got more and more bunched together. Worse than the most crowded college bar on a Saturday night, the praesental army became just one mass of humanity, troops pressed so close together they could not even move. Command and control became impossible. A spear could no longer be withdrawn after a thrust, nor could a sword. A soldier could be killed but the press of flesh would keep his corpse upright.

This ever-shrinking circle of squeezed Roman troops looked to the Gothic archers and missile troops not like a noose, but more like a target. They poured a shower of arrows and darts into the Roman body, with a near certainty that they would hit a target. The dust kicked up by the battle blocked out the sky, so the unfortunate legionaries could not even see the incoming missiles. The praesental troops may have tried to set up a fulcum, the late Roman version of the famous testudo formation of overlapping shields to protect against such missile barrages. [170] But they were so pressed together they could not even raise their shields.

Still, the Romans fought on.

In this great tumult and confusion the infantry, exhausted by their efforts and the danger, when in turn strength and mind for planning anything were lacking, their lances for the most part broken by constant clashing, content to fight with drawn swords, plunged into the dense masses of the foe, regardless of their lives, seeing all around that every loophole of escape was lost. [171]

Knowing they were completely encircled, weakened by thirst, hunger and fatigue; barely able to move; barely even able to stand due to the slippery blood, sweat and, yes, human waste accumulating at their feet; their combat power withering away; the legionaries of the Army in the Emperor’s Presence fought on. With courage borne of desperation, they gave no thought to strategy but let their training take over. They used their spears as much as space would let them until the spears were broken. Then they slashed and hacked with their swords, almost indiscriminately. Romans as well as Goths felt the blade of the praesental army’s spaethae.

But courage borne of desperation, adrenalin produced by desperation, inevitably runs out:

Now the sun had risen higher, and when it had finished its course through Leo, and was passing into the house of the heavenly Virgo, scorched the Romans, who were more and more exhausted by hunger and worn out by thirst, as well as distressed by the heavy burden of their armor. Finally our line was broken by the onrushing weight of the barbarians, and since that was the only resort in their last extremity, they took to their heels in disorder as best they could. [172]

Now the slaughter would really begin.

The end

It may seem counterintuitive but in ancient warfare, fleeing from battle was usually a good way to get oneself killed. Whether a Spartan hoplite, a Macedonian phalangist or Roman legionary, in battle the best hope for safety lay in the shield, spear, sword and the presence of one’s comrades.

But none of that could be used when running from the battlefield. Presenting one’s backside to the enemy meant the spear and sword could not be used to strike or drive off foes, the shield could not be used to absorb or redirect incoming blows, and the isolation meant vulnerability from all sides. Furthermore, while wearing a full suit of armor, running was difficult, swimming even more so. [173]

So the vast majority of casualties in battle were usually inflicted when one side had broken and was fleeing. A fleeing infantryman could be trampled by a horse, or feel a spear thrust in the back, or a sword slash across the hamstring.

Against this reality, which every soldier was taught, was the incredible, inhuman, barely endurable stress of battle, both physical and mental. Every battle was a question of which side would break first. The result was an often extremely large disparity of casualties on the ancient battlefield.

Ammianus says the barbarians broke the Roman line. That is true – mentally. Believing they had no escape, the surviving Roman infantry actually hacked their way through the Gothic troops and their own comrades until a glimmer of a possibility of escape could be seen. One legionary took saw the chance and took it, then more and more, until most of the army broke.

The rout of the praesental army was like so many others in ancient warfare:

And so the barbarians, their eyes blazing with frenzy, were pursuing our men, in whose veins the blood was chilled with numb horror: some fell without knowing who struck them down, others were buried beneath the mere weight of their assailants; some were slain by the sword of a comrade; for though they often rallied, there was no ground given, nor did anyone spare those who retreated. [174]

The picture is one of complete chaos of a routed army – panicked soldiers running in every direction, no semblance of organization or command, everyone just trying to get away.

Well, not everyone.

In the chaos created by running troops, the screams of the wounded and dying, the haze created by the dust that was continuing to be kicked up, and the gathering darkness of the approaching sundown, two legions of palatini did not retreat. One Ammianus identifies as the lancearii; the Notitia Dignitatum calls them lanciarii. These were spear-carrying infantry; though almost all the infantry carried spears, the lanciarii were the most senior, most elite infantry unit in the army. [175] They likely started out the battle deployed in the traditional spot for the best unit in the army: on the right side of the line. When the Roman infantry was encircled the lanciarii likely had been on the perimeter and remained so when most of the infantry behind them eventually fled.

The other legion Ammianus identifies as the mattiarii; the Notitia Dignitatum calls them matiarii. A rather mysterious unit, this palatine legion may have been armed with lead-weighted darts (though, to be sure, by this time most units were armed with at least some darts) or even maces. [176] Where this legion was in the original battle line is unknown, but if they remained standing with the lanciarii at the end, it seems likely they had been deployed very close to them.

These proud legionaries were joined in their stand by one more individual of some note: Flavius Julius Valens Augustus:

While all scattered in flight over unknown paths, the emperor, hedged about by dire terrors, and slowly treading over heaps of corpses, took refuge with the lancers and the mattiarii, who, so long as the vast numbers of the enemy could be sustained, had stood unshaken with bodies firmly planted. On seeing him Trajanus cried that all hope was gone, unless the emperor, abandoned by his body-guard, should at least be protected by his foreign auxiliaries.

On hearing this the general called Victor hastened to bring quickly to the emperor’s aid the Batavi, who had been posted not far off as a reserve force; but when he could find none of them, he retired and went away. And in the same way Richomeres and Saturninus made their escape from danger. [177]

This incident and, indeed, its mere description are more curious than they appear at first glance.

Valens had gone into battle with a minimum of insignia. Among those insignia had been the purple imperial robe signifying his rank, but by now the robe was gone; whether he had thrown it aside or just lost it in the middle of combat is unknown. [178] The loss of the robe may explain why he was able to go across the battlefield alone without being recognized by the Goths. The loss of his bodyguards may also help explain it; they may have abandoned the emperor, which, given Valens’ propensity for alienating people, seems likely, but they could also have just become separated during the battle, or simply all been killed.

The picture Ammianus paints of the eastern emperor is a pathetic one – amidst all the chaos of the fighting, the most powerful man in the world is slowly wandering alone among the dead and dying, in a horrified daze, in shock at what has just happened and wondering if – no, hoping – it is all just a very bad dream.

All very plausible, and it might very well be true – except Ammaniaus has no way of knowing what was going through Valens’ mind at this moment. For the emperor had no one with him, and, as far as anyone knows, from this point forward no one he would meet would survive the day. Witnesses would see Valens, but they did not talk to him.

Some translations of Ammianus have Valens seeking “refuge” with the lanciarii and the matiarii; others say he “fled” to the two remaining legions. Except his slow wandering across the battlefield and not away from it is not consistent with someone looking to flee. It is consistent with someone who cannot mentally handle the catastrophe that has befallen him. It is also consistent with someone who doesn’t know what to do, except that he will not leave the battle. Whatever Valens’ faults – and they were legion [179] – he would not abandon his troops on the battlefield. Nevertheless, his ending up with the lanciarii and the mattiari doesn’t sound like the proverbial Custer’s Last Stand.

But that does not mean it definitely was not. The lanciarii and the mattiari were holding their ground against the Gothic tide. Unfortunately, these two legions were not the larger legions of the Republic and the Principate but the bite-sized legions of the late Empire, and so were in danger of being overwhelmed by the far superior numbers of Goths.

The emperor’s plight was witnessed by Trajanus. He loudly called attention to the situation, saying words to the effect of “all hope was gone, unless the emperor … should at least be protected by his foreign auxiliaries.” The magister equitum Victor, ubiquitous except when and where he was needed most, took this as the cue to bring up those “foreign auxiliaries” – the Batavi seniores, the Batavians – to help Valens.

This raises a number of questions. Why, if the situation was so dire, did Trajanus not go and help the emperor himself, at least at first? Why, if Trajanus could see Valens, the two legions and their attackers, could neither he nor Victor see the Batavians? And why did Trajanus and Victor not find their inability to see the Batavians of some concern? These questions show that this incident was not so much an interesting story as a rather shallow stream of information that nevertheless can be panned for considerable gold.

When taken together, Trajanus’ statements, his initial hesitance to help Valens himself, and the effort to bring up the Batavian auxiliaries would seem to indicate that he was trying to form a new line, at the very least to try to punch through to the trapped emperor. The presence of the crack Batavians would help rally fleeing troops and create a cadre around which to form a new position. Calling attention to the emperor surrounded by enemy troops was one time-tested way of motivating troops; many Roman generals, most famously Scipio, would throw themselves into the middle of enemy troops to motivate their own soldiers to “rescue” them. Of course, that technique depended in part on the commander’s popularity with his own troops, and when it came to popularity with the troops, to put it mildly, Valens was no Scipio. And Trajanus could not help the emperor without enough troops.

There were more issues as well. By this time there were certainly issues with visibility. Dusk was approaching and the kicked-up dust continued to leave a sort of haze. The obvious reason no one could see the Batavians is that they had long since fled, scattered by the early charge of Alatheus and Saphrax. But it does not explain why Trajanus and Victor were not alarmed by their inability to see the Batavians when they could see the emperor.

The inference that can be drawn from this is that they did not expect to see the Batavians from where they were, that the Batavians had not been deployed where they could be seen from main battle, which was in front of where the Roman first line of infantry had initially attempted to deploy. This goes to support the belief, held by many historians, that the Romans had initially deployed the first line on a low ridge, with the Batavian second line deployed behind and thus screened by the ridge.

That Trajanus and Victor (not to mention Richomeres and Saturninus) could be so sanguine about the Batavians when they could not see them also suggests that they firmly believed these auxiliaries were where they were supposed to be. As all of these generals had been trapped with the Roman infantry by the Goths, this would also mean that none of these generals had seen any Batavian troops within the pocket. While not completely conclusive, this would seem to indicate that the Batavians had not been pushed into the first line infantry and trapped with them, but had fled separately, likely early in the battle on the charge of Alatheus and Saphrax.

With the Batavians gone, there was no hope of snatching a draw from the jaws of defeat, no hope of saving the emperor and the stalwart lanciarii and mattiarii from being overwhelmed by the vastly superior numbers of Goths. It was over and the generals knew it. Well, some of them did. Once again unable to help when he was needed most, though through no fault of his own, Victor fled. On seeing this, Richomeres and the veteran Saturninus joined him in flight, leaving the emperor and his few remaining troops to their fate. All three generals survived.

Trajanus did not. What happened to him is unknown, but Ammianus had a so-far-unexplained change of heart regarding this officer. When Trajanus was initially sent with Profuturus to try and stabilize the situation—an attempt that ended badly with the Battle of Ad Salices—the historian said both Profuturus and Trajanus “more rank and ambition than ability.” Now, with Trajanus dead on the battlefield, Ammianus says that of all the “distinguished men” who died at Adrianople, two stood out. One was Trajanus. It gives a hint, but no more than a hint, that maybe he died in an effort to rescue Valens.

The other death that stood out in Ammianus’ eyes was that of Sebastianus. The conduct of the best Roman general on the field is not otherwise mentioned, suggesting he may have been killed relatively early on while the Roman infantry was encircled by the Goths.

The Roman Emperor Flavius Julius Valens was never seen again. As told by Ammianus:

At the first coming of darkness the emperor, amid the common soldiers as was supposed (for no one asserted that he had seen him or been with him), fell mortally wounded by an arrow, and presently breathed his last breath; and he was never afterwards found anywhere. For since a few of the foe were active for long in the neighborhood for the purpose of robbing the dead, no one of the fugitives or of the natives ventured to approach the spot. [180]

This story indicates that while Valens was with the lanciarii and mattiarii he was hit by an arrow and died shortly thereafter. His body was never recovered. How anyone knew the emperor had been killed by an arrow if no one was with Valens and no one saw him is not explained.

Ammianus then relates a far more famous story of the end of Valens:

Others say that Valens did not give up the ghost at once, but with his bodyguard and a few eunuchs was taken to a peasant’s cottage near by, well fortified in its second storey; and while he was being treated by unskillful hands, he was surrounded by the enemy, who did not know who he was, but was saved from the shame of captivity.

For while the pursuers were trying to break open the bolted doors, they were assailed with arrows from a balcony of the house; and fearing through the inevitable delay to lose the opportunity for pillage, they piled bundles of straw and firewood about the house, set fire to them, and burned it men and all.

From it one of the bodyguards leaped through a window, but was taken by the enemy; when he told them what had happened, he filled them with sorrow at being cheated of great glory, in not having taken the ruler of the Roman empire alive. This same young man, having later escaped and returned secretly to our army, gave this account of what had occurred. [181]

Which of these versions is true, if either or if both, is anyone’s guess. [182] The apparent inability of the Goths to break down a wooden door, even though they are known to have had axes, seems rather odd, which has prompted discussion that the “cottage,” “cabin” or however it is translated was actually a fortified farmhouse, part of the web of Roman in-depth border fortifications, hard points and supply posts. [183] Many of the non-Arian Christians saw the immolation of Valens as a sort of divine retribution for his crimes against non-Arians, specifically the burning deaths of those 80 monks on that ship. [184] Afterwards it would come out that many of these people had predicted Valens would die by burning and his body would never be found. [185] None of which goes to support the accuracy of the account, but it certainly cannot be ruled out.

Either way, Valens never left the field of Adrianople. Two thirds of his army would remain in the Emperor’s Presence on that same field, including 35 military tribunes and those generals, Sebastianus and Trajanus. As Ammianus states, only the Roman disaster at Cannae was worse. [186]

The Siege of Adrianople

Only the coming of night – at around 7 p.m local time – ended the slaughter on August 9, AD 378. [187] This night would be darker than usual – there was a new moon on August 8, by the Julian calendar. [188] The almost complete darkness was fitting. When the sun set on Adrianople, the sun also set on the Roman Empire. The world would not emerge from the darkness for some 800 years. When it did, the Roman Empire would not be with it.

That night Goths apparently returned to their laager, the scattered Roman troops fled in all directions – fumbling, stumbling, staggering in the dark – a number of them somehow finding their way to Adrianople, to where they carried the news of the abject catastrophe. For the Roman people, August 9, AD 378 was every bit as traumatic as December 7, 1941, or September 11, 2001, was for Americans, with all the psychological scarring of the Vietnam War added on. [189] Emotionally, they would never recover.

In the midst of this shock and despair, with little remaining military or even civilian leadership on hand, the Romans somehow managed to form a semblance of an organized response, and even continued the use of exploratores to bring in intelligence information. On August 10, the Romans were well aware that the Gothic troops, hoping to seize the imperial treasury Valens had left behind within the walls of Adrianople, were on their way to the city.

A large number of survivors of the battle, apparently both infantry and cavalry, had made their way to Hadrian’s City and were begging for entry, but the civilian authorities, fearful of a ruse – with justification, as events would show – would not open the gates for them. These survivors gathered under the walls of Adrianople, apparently around the legions left to guard the praesental field army’s baggage and their fortifications. [190] While the returning survivors would have swelled the number of defenders, the organization was lacking and, with most of the senior officers dead, missing or not available, putting together a defense would be in the hands of lower-level tribunes.

Under the circumstances, these tribunes did an incredible job, much better than the Gothic troops who forgot Fritigern’s admonition to “make peace with walls.” By the “fourth hour” – between 10:00 am and 11:00 am – the Goths had surrounded Adrianople, in the process somehow managing to cover the distance between their laager and the city in half the time it took the Romans a day earlier. Ammianus gives few details as to the fighting. The Goths appear to have been in a complete frenzy to break into the city, and seem to have abandoned all pretense of planning and, indeed, rationality; the historian gives an image of the Goths literally hurling themselves at the Romans. [191]

Combat continued for an unspecified period. The most notable event took place during the “ninth hour” – between 3:00 pm and 4:00 pm – when some 300 Roman infantry deserted to the Goths. Or, more accurately, tried to. For when they got to the Gothic lines, in full view of the legionaries, they were immediately slaughtered by their new friends. After that, it seems that there were no more attempted desertions. [192]

After a further period of fighting of indeterminate length, a thunderstorm put a damper on things and the Goths sullenly returned to their camp. They appear to have taken heavy losses. The Romans had held their own.

Their August 9 victory would leave the Goths arrogant and flushed with power, but it would be the high point of their revolt; in the immediate aftermath of their success, their efforts moved away from the wisdom of Fritigern and closer to the wisdom of Wile E. Coyote. They sent a threatening letter demanding the surrender of Adrianople with – you can’t make this stuff up – a Christian priest; it was laughed off. [193] They had two Roman deserters pretend to have escaped from the Goths and successfully plead entry into the city, with the idea that they would ultimately distract the garrison enough to allow the Goths to storm the walls, but the Adrianopolitans had expected just such an attempt – this is why they had denied the survivors of the battle entry into the city – and the traitors were found out, tortured and beheaded. [194]

The Goths then tried to attack the city again, but in the interim the walls had been further fortified, and onto the battlements had been moved archers and the famous Roman artillery such as scorpions and ballistae. [195] Once again, the Goths seem to have been in a complete frenzy to get inside the walls, but their vast numbers seem to have only created a target-rich environment for the Roman missile troops and especially the Roman artillery, which made a deep impression on the barbarians. Once again, they were repulsed with heavy losses, and took to finger-pointing as to who did not listen to Fritigern’s warning about not attacking walls. [196]

Why they did so is anybody’s guess. For after attacking Adrianople and being frustrated by its walls – twice – they decided they would have better luck attacking an even bigger city with even bigger walls: Constantinople. The news the horde was coming sent the eastern capital into a panic, with Valens’ widow Domnica scraping together what troops she could to organize a defense. [197]

She needn’t have bothered. When Valens had left with the praesental army, the Arab cavalry had remained in the city; now those same cavalry were ready to defend Constantinople. They sortied out to meet the Goths and fought them to a standstill. But this was not Ad Salices, for the Gothic troops found the Arabs to be a bit too bloodthirsty – literally – for their liking. [198] The madness that had driven the Gothic efforts after August 9 seemed to disappear after that, and they ultimately retreated from before Constantinople.

The worst of the crisis was over; but while the high point of the Gothic revolt had passed, the Romans would never recover.


As soon as the exploratores reported that the Goths had left the area around Adrianople, messengers were sent out to Macedonia, Philippopolis and Serdica to find the Emperor Valens. [199] They had no idea he was gone.

Nor did Gratian. For some reason Valens’ magister equitum Victor, and not Gratian’s own comes domesticorum Richomeres, was given the task of informing the western emperor of the death of Valens and the destruction of the praesental army. The chaos created by the Goths in Thrace was such that Victor had to take a very circuitous route, through Macedonia and even through Thessaly in Greece, coming back through Moesia and Pannonia, to reach Gratian near Sirmium. [200] The news of his uncle’s death did not exactly move Gratian to tears. With his troops too few and too light to reverse the situation, Gratian turned around and returned to Gaul.

The division between the eastern and western empires was now more practical than theoretical, with the giant gash in the Balkans now separating them. Land communication between the two was now almost impossible. [201] The Goths continued their raiding and pillaging. Roman cities and towns in the Balkans were now alone, left to defend themselves with whatever limitanei they could find. [202]

The Roman government was powerless to do anything about the Gothic predations. For the civilian population, this resulted in fear and anger – fear at what might happen to them, anger that for all the Empire took in taxation it could not protect them. For the Roman military, the impotence was far more infuriating. That fury led one Julius, magister militum per Orientis (master of troops in the Orient, which was the far eastern part of the Empire), to strike out in a manner that was characteristically ruthless and uncharacteristically irrational. Before Adrianople, young Tervingian males had been sent east as hostages for military training to become Roman soldiers. Now Julius had them all slaughtered. [203] This brutal move could be justified because these Goths could potentially switch sides as Sueridas, Colidas and their troops had done, but a closer examination shows it was merely revenge for the disaster at Adrianople and a pointless waste of potential troops. [204]

Ultimately, Gratian would find a very capable emperor for the East: Theodosius. Theodosius would have to use all his skill and resourcefulness to end the Gothic conflict – 4 years later. In so doing, he would be forced to give the Gothic confederation virtually their own state in Thrace, in exchange for peace. And troops. Foederati. In this way Rome would make up its losses at Adrianople. In theory.

Foederati. Paying warlords to fight for Rome, with troops who had no loyalty to Rome, by giving them land that would no longer be ruled by Rome.

This sounds like no empire. But after Adrianople, this was the Roman Empire.

< Go to Part 2 of 3

* * *

Show Footnotes and Bibliography

* * *

Copyright © 2012 Jeffrey R. Cox.

Written by Jeffrey R. Cox. If you have questions or comments on this article, please contact Jeffrey R. Cox at:

About the author:
Jeffrey R. Cox is a litigation attorney in Indianapolis, IN, and an independent military historian specializing in World War II, ancient Greece and ancient Rome, which he has studied for decades. He holds a bachelor’s degree in National Security Policy Studies from The Ohio State University. He can be reached at

Published online: 04/01/2012.

* Views expressed by contributors are their own and do not necessarily represent those of MHO.
© 2018, LLC Contact Brian Williams at: