Attacking the marching column


"Getting there is half the fun" would not describe the Roman army's experience. Army travel was brutally hard work. In the previous pages the march is described in some detail. As the daily march was described and illustrated in that section the daily march averaged about 10 miles, took some nine hours from the time the lead contingents left the first camp to the time the trailing rear guard units arrived in the second camp. The overall length of the column was almost fifteen miles.

The illustration below is taken from those pages to show the general arrangement of the various elements of the march.

This page will examine the march from a slightly different perspective.

Daily schedule

Moving the army was exhausting physical labor from dawn to dark. During the campaign season there would be roughly 16 to 17 hours of daylight. The day's schedule might look something like this:

4:00 Sunup, rise, prepare meal

5:00 Take care of animals, begin taking down tents

6:00 First contingents leave old camp

10:00 Lead contingents arrive at new camp

2:00 Fortifications for new camp complete

3:00 Last contingents arrive at new camp

4:00 Set up tents, unload pack animals, store baggage

4:00 Dispatch forager for food and fodder

6:00 Prepare evening meal

7:00 Camp chores, care for personal items

8:00 Ensure that all animals are secure for the night

9:00 Sundown

Bad roads

So far the description follows the general scenario worked out in the pages on the march. However, what was described there was an optimistic scenario. The illustration below shows the general marching order as previously described. The road is eighteen feet wide, width enough for six men abreast, three horsemen either mounted or leading their horses, three or six pack animals, and two wagons.

However there may be a problem with modeling the column's marching formation on this basis. Hans Delbruck, discussing the Persian army, describes the problems all pre-mechanized armies faced.

"The width of a marching troop column depends on the width of the road. If the road is too narrow for the column, even at only a few places, that still creates a march disruption that builds up progressively toward the rear and finally becomes completely intolerable. The troops who are marching farther toward the rear are forced to wait for hours and use up their strength in so doing, or, if they are not well disciplined, they fall out of formation. the foremost troops extend out in the same manner, and the column falls completely apart. Every good commander therefore considers it of the highest importance to avoid march jam-ups, or since with large masses that is hardly ever attained, to reduce them to a minimum. For this reason intervals are established between the various units, so that the smaller holdups can immediately be absorbed, and the higher leaders are constantly concerned with maintaining the intervals. . . . Modern troops also see to it very deliberately that half of the road remains as open as possible. In the case of every marching column it is absolutely necessary, especially in enemy territory, that movement and communication be possible alongside the column for high-ranking officers, liaison officers, messengers, and under certain circumstances also for quickly moving forward a special unit, such as cavalry. That cannot have been any different with the Persians. . . .there are numerous rivers to be crossed, mountainous land to pass through, passes to overcome. At many places the bridges, fords and mountain paths were undoubtedly not broader but narrower than those with which modern armies have to contend. The Persians must have marched with a column . . . certainly often not even 4 men wide, but only 2 men in width, using quite naturally at the same time, wherever possible, several parallel roads." He goes on to relate a Prussian experience in 1870 when a force tried to march with a wider than usual front. Their experience was that it was "very tiring because of frequent stops, holdups, and resumptions of the march . . . caused breaks to develop in the column, which showed up as a lack of good order."

Hans Delbruck, Warfare in Antiquity, p.118

Several points can be made from Delbruck's observations.

When moving into enemy territory the Romans would almost always have moved along water courses since camps always had to be located next to good water sources. Whenever possible they would have followed the local roads and paths which undoubtedly wound through the countryside, paralleling the streams and rivers. The most probable type of road that they would have encountered was unimproved: a path probably wide enough for a cart or wagon with the brush cleared from it and a set of wheel ruts worn into the earth. Bridges, if there were any, would have accommodated a wagon but probably not two abreast.

Certainly in some places the road would wind through fields or meadows which would allow many men to walk abreast. However, Delbruck's observation is that the army cannot march with a front that is wider than the narrowest part of the road before it. Therefore, if there is but one narrow place in the road, the army march formation cannot effectively be wider than the obstruction.

The Romans sent teams of engineers and workers ahead of the column to improve the road. We should not think of these men as paving the road or building bridges, there would be no time for that degree of labor. The real task would be to ensure that there were no unexpected narrowings of the road or other blockages. In this way a column of constant width could navigate the road without the jam-ups that Delbruck describes.

This is why the army would not attempt to move in several parallel columns on either side of the road. Surely in a ten-mile stretch there would be obstructions somewhere along the way -- a rock fall, swamp, a dense and impassible stand of brush -- which would block travel to the sides of the road.

A secondary observation is to underline the importance of local knowledge of the terrain ahead. When starting the march in the morning the commander would need to know the condition of the road in front. If it narrowed at some place, he would have to adjust his column width at the beginning of the march.

This points up the value of the Roman road system. They provides a wide paved level surface, but almost as importantly, they were of a reliable and known width so that the commander could fix the width of his column without concern.

Revised marching columns

Heeding the admonitions of Delbruck, the six man wide marching column illustrated above needs to be re-evaluated. Whenever the army left Roman territory it would likely have to use unimproved local roads, or even trails. As noted above, it is improbable that these roads would accommodate more than a single wagon. The marching column that would fit in this narrower road system is illustrated below.

The implications of a narrow road such as the one illustrated above are difficult to imagine. The fifteen-mile column stretches to at least twice, perhaps three times, its length. If a fifteen-mile column takes nine hours to move ten miles then a thirty-mile column cannot move much further than six or eight miles in a day, and that only if it has elements on the road from dawn to dark.

The problems are not solely with logistics. The long thin column is particularly vulnerable to attack.

Dangers to the marching column

It is easy to see why a column only two men wide and maybe thirty miles long would be vulnerable to attack. The soldiers are so strung out that it is almost impossible for them to protect the column. The baggage train alone might be twenty miles in length. No matter how and where the soldiers are placed, they will either be too few to be effective or too far away to respond in a timely manner.

But all marching columns are vulnerable, not just narrow ones. The following vivid description of an attack on a marching column is culled from correspondence with Ed Valerio, an American expert on Roman warfare.

A column that is loaded down with wagons, dependents, etc., was pretty much indefensible. In essence it could either move or fight, but not both.

The first thing to note is that the column is more-or-less bound to roads or at the very least improved tracks, and if the approximate point of entry is known, then all possible routes the column can thereafter take are also known. Secondly, seeing as the column is moving at a max. average rate of 3 mph, decent infantry (who can easily make 4 mph) or better yet cavalry (horsemen can average 8 mph in reasonable terrain), the opposition can intercept the column handily. Indeed, if the opposition is a cavalry force, it can literally run rings around the column. Once intercepted, the moving column is extremely vulnerable to any sort of attack

The column has a lot of noncombatants: dependents, servants, slaves, assorted lackeys, and so on, plus all their mules and wagons, en-column. All this gaggle would have been easy prey. The troops must have been strung out throughout the length of the column, it would have been very difficult for them to gather together, and even once they did- what to strike out at? The attackers always held the initiative -- any number of them could attack anywhere with any degree of severity at any given time.

The fact is, if one's troops are en-column, one simply can't deploy them rapidly except on one flank or the other. But what if one's column is several miles long and a message from the commander can only travel along the column's axis, and at the speed of a (at best) trotting horse as well? Let us assume that he wishes to mount an attack on the enemy which is harassing him from one side of the trail, and he wants his troops to deploy accordingly, and he sets messengers out to convey his orders. How much delay does a messenger accrue if he has to get round a felled tree, or a clump of wagons backed-up on one whose ox has been shot dead in it's traces, or he has to pass through a zone under attack? Assuming he eventually does find the intended recipient of his message, what does he say if that unit is already deployed, but is defending against an attack from the other side of the trail? Its already been an hour since he left the commander and heaven only knows what the situation is now back there, so what is he to do? Go back with his recipients objections to what is obviously an impossible order under his present circumstances? That'll take another hour.

A column can be nickeled-and-dimed to death by bands of the enemy ranging in size from a few tens to a few hundred attacking as opportunity permits. All the enemy commander had to tell his guys was to range themselves pretty-much as they pleased along the march-route and attack anything that passed their way. Remember, it would be automatic that the attackers would have a numerical superiority at any given point- at least until reinforcements arrived from up or down the column. If things got too hot, they simply ran back into the bush the way they came to regroup and catch their breath. The Romans certainly couldn't pursue them very far without risking being cutoff. From the attackers point of view the whole business might have been great fun -- run in and bash a few heads and maybe grab some stuff off a mule or wagon, them back into the bush for a snack. Certainly, the Romans weren't going anywhere in a hurry, and the next time the band attacked it might get real lucky and hit a load of laden wagons. Just think of this sort of thing going on episodically- perhaps tens of small encounters occurring along the column length simultaneously for miles -- from dawn to dusk. In the aggregate, Roman casualties would mount steadily all day long. Cohesion of the column would automatically come apart as portions of it were brought to a halt even for a short time, and folks ahead simply kept going. The Romans were totally confined to a well-defined single track. Even without parallel trails smallish bodies of lightly-armed attackers could easily pace the column. The key to understanding the mechanics of the battle is to accept that neither side could communicate amongst themselves. For the Romans this meant that nowhere could they gather together a meaningful response. For the attackers, however, it was basically a non-event -- as long as the various bands followed their standing instruction to pace the column and attack as opportunity permitted, they didn't need further communication.

Meanwhile, the road/trail rapidly deteriorates under an unaccustomed pounding by thousands of animal hooves. Your unit (century, cohort, whatever...) is sandwiched between a panicked gaggle of wagons and pack-mules piling up against your rear, and a panicked gaggle of the same ahead. You can hear screams and shouts from up ahead, but you can't see round the bend in the trail. The rumor is that the enemy are felling trees across the track, and anyway you don't know how far ahead or behind the next unit is, or indeed what shape it is in. You plod on. Do you pick up that wailing bunch of civilians- some hurt -- grabbing at your legs? or do you leave them be and continue to go forward picking your way round the wreckage of their wagons? Comes an 'open' area, and you can actually form up six-abreast- okay! It's spooky though, because all round you there are sounds coming from the woods. Guys out there crashing around and laughing and shouting, and sometimes you can actually get a glimpse of them, but so what? All of a sudden a whole bunch of them charge out of the woods chucking javelins at you and stabbing out with their swords -- but before you can form up to fight them they are gone again. The Centurion says 'forget it- no pursuit. Too great a chance of getting cut off out there...' Anyway, that's three more guys down- one seriously so, and one other with a leg wound which bodes not well for his prospects long-term. You figure it's only been about three hours since sun-up that means about eleven more hours of this to get through -- and the wild enemy warriors are still laughing and carrying-on out there in the bush ...good luck!

To Summarize

Marching columns were extraordinarily vulnerable to attack for a number of reasons:

  1. The attackers did not have to be well organized or in communication with each other. They could work as small groups acting independently along the route, attacking as the opportunity presented itself.
  2. The attackers could pick their time and which part of the long column they wished to attack. They could guarantee that they always had local numerical superiority.
  3. The attackers could use hit and run tactics.
  4. Even a small local attack had the potential of bringing the column to a stop. If any part of the column adopted a defensive formation it could no longer continue to move. Once this happened a gap would open between it and the leading elements. The integrity of the column would disintegrate.
  5. If large numbers of defenders left the column to deploy in a regular fighting formation or to chase the attackers then the column would be left undefended against other attacks. If small units of defenders were dispatched to chase the attackers they risked being cut off and isolated once they left the column.
  6. Killing or wounding baggage animals could be done from afar and easily since they were unarmored. Even a few dead and dying pack animals would disrupt or halt the column. The loss of large numbers of pack animals would cripple the army.
  7. It was easy to block unimproved roads and paths. A few trees or a rock slide into the middle of the column would bring it to a halt and isolate the various units from each other.

Something is missing in this picture

All of the above rather convincingly argues that an army column was so vulnerable that one wonders how they ever moved anywhere at all. But ancient armies did march through enemy territory all they time, and they did it rapidly and safely. In fact, we only know of a very few successful attacks against a column en-march (they will be discussed later). The reality is so different from what it would seem to have been that it should be cause for some amazement.

How can this be so? If columns are so easy to attack why wasn't that the primary military tactic?

Clearly all ancient armies were able to defend their lines of march. The picture painted above is not complete. Unfortunately, the ancients did not tell us much about the military tactics relating to marching. We are left with the fundamental fact that neither the Romans nor their many enemies actually attacked each others marching columns with any degree or regularity or success. We can only speculate about the reasons for this.

Defensive measures

The army could do some things to guard against attacks.

Information was the single most important thing. Good commanders sent mounted scouts far ahead and wide to the flanks to look for the enemy. Engineering and labor battalions were sent units ahead to clear the roads and paths. Large armies drew a great deal of attention to themselves. Whenever possible the commander kept himself well informed of the movements of the enemy army. He might acquire information from local natives, from captives, spies or scouts.

More active defensive measures included using highly mobile cavalry units to protect the flanks. The flying squadrons, in conjunction with the scouts, would be charged to keep enemy forces at a distance or to intercept and destroy them should they attempt to approach.

In hostile territory the soldiers themselves may be instructed to march in battle gear; that is, with arms in hand and without their normal personal baggage.

The most vulnerable parts of the column, the baggage train, would be broken up into smaller units and interspersed among the fighting me so that protection was never far away.

It was more difficult that it might appear to attack the enemy's column. In the situation where one army is pursuing the other, the two armies will move at very nearly the same speed. The leading army almost certainly chooses the best route, leaving only marginal secondary side routes available for any type of flanking maneuver.

For two armies closing toward each other, the situation is a bit different. It would be difficult to know with certitude the exact route the enemy might choose to use. Therefore, hard to know just where and when an ambush might work. This would make it nearly impossible to pre-position large numbers of men along the enemy's route of travel.

Intelligence from native informants or spies would very probably reveal the ambush. Large forces (armies) attracted attention and their presence would be known to everyone in the area. Scouts, spies or informants would keep both sides well informed about the location of the other army.

Small forces that might attack in a hit and run style could be detected, run off or annihilated by the roving cavalry flanking forces. And, even if a small scale attack were launched, the army could temporarily take defensive postures (shields up, for example) and signal for help, counting on the cavalry to arrive within minutes and kill the attackers. Attacks by small forces stood a good chance of being suicide missions; and, knowing this, most enemy fighters would be more than reluctant to attempt it.

The ambush could backfire if it were detected. A large attacking force would have to be hidden along the route in a strung-out formation and amid many natural obstacles for concealment. If detected the tables would be turned and this force would be vulnerable. The attackers could muster local numerical superiority and operate from favorable ground.

Examples of attacks

Attacks against the rear guard

There are many examples of attacks against the rear guards of armies. Typically these take the form of harassment by cavalry and light infantry troops which cause the rear guard to stop and take defensive positions. Usually the column itself continues forward. The rear guards were almost always comprised of fast moving troops so that they would not have a problem closing the gap to the column once they had driven back the attackers. Attacks of this nature do not seem to have been successful in significantly harming or delaying columns and would probably have been more of a nuisance than a real danger.

Attacks against the entire column

We only know of a few successful attacks against whole armies in column. Here are four examples.

  • The Allobroges & Hannibal: In crossing the Alps Hannibal's column was attacked by the Allobroges, a mountain tribe, who threw stones, rolled boulders and shot arrows down on his army from the heights above. Hannibal was ineffective at either protecting his column or in attacking the tribesmen. He was forced to overnight on the mountain and seize the high ground while it was undefended at night.
  • Hannibal & Flaminius: Hannibal, heading south into the heart of Italy, was being pursued by Flaminius who may have had an army of about 30,000. On the night of June 21 Hannibal hid his entire army in the hills bordering the shore of Lake Trasimene. Even though the precise site of the battle is still debated, the general outlines are fairly definite. The Roman army was contained on one side by the water, and hemmed in front and back by Hannibal's forces. The Romans marched in dense fog and were unaware of the presence of the enemy. The attack came along long sections of the Roman flank, catching the Roman column before it could properly form up. The Romans were pushed back into the marshy ground around the lake, perhaps even into the water itself. Bogged down as they were, they were not able to maneuver or to form an effective fighting front. Or to escape. This was the single most successful ambush of a column that we know about. An entire army was virtually annihilated in a single day.
  • Ambiorix & Sabinus: In 54 BCE during Caesar's conquest of Gaul he placed Q. Titurius Sabinus and Lucius Aurunculeius Cotta in command of 1 legion and 5 cohorts in a winter camp in the territory of the Eburones. Ambiorix led an uprising against the Romans and tricked the gullible Sabinus into believing that his position was hopeless. Under a promise of safe conduct Sabinus led his army out of camp, intending to join Cicero's forces at his camp. Ambiorix set an ambush a few miles away. The Roman column, encumbered by baggage, entered a ravine. The Gauls blocked both ends and attacked the column. Sabinus abandoned the baggage and formed his cohorts into a square. The Gauls sent showers of missiles into the formation and rapidly retreated whenever a cohort tried to attack. Eventually Sabinus tried to parley with Ambiorix under a flag of truce. He was murdered during the meeting and the Gauls then attacked and nearly annihilated the Roman force. (Caesar, Gallic Wars, V 32ff)
  • Arminius & Varus: In AD 9 the Roman governor Varus ventured 60 or 80 miles into pacified German territory to expand the reach of the provincial government. After a summer's activity he began the trek home. As he approached a heavily forested narrow pass, the Teutoburger Wald he learned that his trusted German commander, Arminius had fomented a rebellion among the tribes. The various forts left along the way to guard the route had been overwhelmed and the path ahead was blocked. His column of about 3 legions, family members and many civilians associated with the government was heavily encumbered with baggage. As Varus sought another route out of Germany his column was subjected to 4 days of continual attacks. By the end of the fourth day all but a very few soldiers were killed outright or captured and subsequently executed. Throughout the march, German tribesmen who had been secretly assembled harassed the column. Since the Romans were deep within German territory it would have been a simple matter for the Germans to converge on the column from every direction. This is different from the usual situation in which the enemy force was more concentrated in a single location and had difficulty getting forces in position to surround a column. Here the Germans already surrounded it by virtue of the circumstances, it was only a matter of calling the tribal groups to arms and setting them out along the one or two routes available to the Romans. It is likely that the Roman column was harassed from all sides almost continuously throughout its 4 day ordeal. Major battles, in which the Roman military system might be able to prevail, were not necessary. The Romans did not have provisions to enable them to fortify and hold a camp; they were forced to move, to try to return to the safe haven of Roman territory. And, when they moved, the column was vulnerable to attack by the Germans who were everywhere. Valerio's vivid description fills in the details.

Lessons from these examples

  • The Allobroges & Hannibal: The attack was possible because Hannibal was constrained to move along one known trail. The hillmen knew the terrain intimately well and could easily occupy the heights above the trail. There, they were safe from attack and had all of the advantages of looking down on Hannibal's army. However, this attack was not successful. I damaged Hannibal's army, delayed it, caused some serious losses, but did not halt it. The likely reason is that only a small part of the column came under attack -- the portion directly under the cliffs. The attack was shortened when Hannibal occupied the heights overnight.
  • Hannibal & Flaminius: This is the most spectacular and most completely successful ambush of a marching column in history. The keys to this battle were that Hannibal knew exactly the route the following army would take. He could reconnoiter the ground at leisure since he camped in the area the night before. He could block the Roman army both in front and back and attack all along the open flank. The lake hemmed the Romans in on one side, limiting their mobility. Escape from the trap, once the attack began, was virtually impossible. The dense fog, though not something Hannibal could have planned on, certainly helped as well.
  • Ambiorix & Sabinus: The deception achieved by Ambiorix places this battle into a unique category. Sabinus, believing he was under a pass of safe conduct, entered a prepared killing zone completely unaware of the danger.
  • Arminius & Varus: As far as major defeats go, this was one of the most serious that Rome suffered. It ranks lower than Hannibal's victory at Lake Trasimene simply because there the victory was complete in a single day, Arminius' victory was strung out over 4 days. The special circumstance of this battle was that it took place deep within German territory, enabling the Germans to come at the Romans from all sides and to be positioned along all escape routes. The unimproved paths and narrow roads that wound through the countryside made travel difficult for a large army and provided endless opportunities for attacks by large or small groups of German tribesmen. Varus' scouting, the key to avoiding ambush, may have been hindered by the terrain. However, it may also have been that allied German cavalry under Arminius were the very units tasked with scouting. The final key to understanding Arminius' success is that, as a trusted part of the Roman army, he knew in advance exactly when the column was departing and which route it would take.

From these examples some of the key elements necessary for a successful attack can be ferreted out:

  • Surprise: Except possibly for the attack on Hannibal in the Alps, each of the other attacks was a complete surprise. Ambiorix and Arminius achieved surprise through deception, Hannibal through misdirection -- the Romans were convinced he was racing for Rome and never suspected he would turn back toward them.
  • Secure base: In each of these examples the attackers were relatively safe from detection and secure from attack. The cliffs protected the Allobroges, the fog shielded Hannibal, the sides of the ravine sheltered the Gauls, and the dense forests hid the Germans.
  • Limited maneuverability: In all cases the army being attacked was not able to maneuver its units into defensive formations. Hannibal was on a narrow alpine trail, Flaminius hemmed in by the lake, Sabinus confined in a ravine (but even so he managed to form a defensive square), and Varus forced to march along narrow forest trails. In 3 of the examples the defeated army was hemmed in front and back and could not escape the attack. Varus, though not closely confined, was nevertheless, hemmed in by virtue of being some 30 miles inside hostile German territory. Flaminius and Sabinus were closely confined. The one example in which the army was not hemmed in front and back, Hannibal in the Alps, is the instance in which the attack was not successful in completely destroying the enemy.
  • Limited mobility: In all cases the escape routes were cut off. Hannibal had no place to go except to stay on the trail. Flaminius and Sabinus were both blocked in front and behind. Varus was caught in the middle of enemy territory and, even though he did manage to travel 20 miles, could not escape.
  • Advance knowledge of the route and time: The attackers all had advance knowledge of exactly when and where the column would be.
  • Ability to pre-position forces: In each of the examples, the attackers had considerable time to pre-position large forces in hidden positions well in advance of the actual battle. This is in high contrast to the more typical situation. When one army pursued another the only good road in the area would be occupied by the army in front, side routes, if there were any at all, would be longer and likely slower. This made it virtually impossible for the pursuer to move sizable forces around the flanks and get ahead of the army being chased.
  • Limited scouting and lack of good intelligence: In each case the army being attacked had limited scouting ability. Hannibal could not effectively scout the high ground above the trail. Flaminius simply failed to scout adequately (due perhaps to the fog or perhaps due to his own firm belief about Hannibal's intentions). Sabinus believed he was under a safe conduct agreement and scouts were probably not even sent out. Varus could not effectively scout the dense forests and, also, the attacks which blocked the escape routes happened well beyond the range of his scouts.


One final example may underscore the points above. In the Gallic wars Caesar was ambushed by the Nervii. But what is interesting is just where the ambush was located. The Nervii apparently knew both the general disposition of the forces in the marching column (or thought they did), the route Caesar was following, and the location of the new camp to which the column was marching. They were in front of the column and had time and opportunity to pre-position their forces wherever they wished. Since they were in their home territory, they would have known the terrain quite well. Yet they did not attempt an attack along the route of the march but, instead, laid in wait at the location of the new camp and tried to attack the legions while they were engaged in digging fortifications and dispersed in foraging parties. They obviously concluded that the column was less vulnerable to attack than the legions at the campsite.

Despite the apparent attractiveness of attacking a marching column, the fact that attacks were so infrequent is testimony to the care with which all ancient armies protected themselves. A fairly large number of factors had to combine to make an attack feasible. In the general course of events, too many of the critical factors were outside the control of a potential enemy, so attacks on columns remained the exception. The purpose of this page has been to point out the inherent vulnerability of columns, the defensive measures necessary for their protection and the special circumstances that were present in a few well-known successful attacks.

Leave a comment

Related Posts

Introduction to a New Model

Introduction to a New Model

The new infantry model At about the same time that I first developed The Roman Army web page Adrain Goldsworthy’s book,
Read More
A day’s march

A day’s march

This page is an attempt to reconstruct a ‘typical’ day’s march for the army. Before showing the marching column there ar
Read More
The order of march

The order of march

The order in which the various units moved is uncertain. Delbrück, who devoted over 600 pages of his 4-volume work on th
Read More
March: the whole army

March: the whole army

In the previous section the elements that would constitute a single legion were explored. When several legions were asse
Read More