Could the Persian Empire logistically support an army several million strong?

[Erpingham]

Plenty of Zulu War images out there but, while they are keeping us thinking, here are two different images of columns at choke points.  Note in the first the column compresses, in the second it attenuates.  I suspect the first one they are coming out of a drift.  Note how bad the going is on a dirt track when everything has to use a narrow front, which may explain the spreading out in other pictures.

[Erpingham]

OK, let's assume an army of 480 000 men. About the best they could hope for in terms of roads is a track wide enough for a cart, allowing, say, a column 4 men abreast.

That means a column 120 000 men long.

Allow two yards depth per rank and you have a column 240 000 yards or 136,36 miles long.

Justin is channeling Delbruck here (didn't expect to say that ).  He reckoned that, using the German army staff estimates, that a 33,000 man corps took up 14 miles of road.  So he reckoned Herodotus' army would be 2,000 miles long.  Of course, he was ridiculing the literalists of his time.

[Erpingham]

From the figures Justin has been using - a daily march of about 20km - we can calculate we need 24,000 men per kilometre.  Each man needs four square metres of march space, by Justin's formula.  So a march column of 96,000 sq m per kilometre or 96 m wide. Obviously, its more complex than that because of the animals and baggage.  Given I think the army would be looser than this, because of the gaps between sub columns and between units, we'd need a cleared route of about 150 m width.  If we concluded 100 metres of cleared route was realistic, we could have an army of 320,000. 

If we use the Roman army density Justin quotes, I suspect a 150m cleared road would give us around a 250,000 army but I haven't calculated it out.

[Justin Swanton]

From the figures Justin has been using - a daily march of about 20km - we can calculate we need 24,000 men per kilometre.  Each man needs four square metres of march space, by Justin's formula.  So a march column of 96,000 sq m per kilometre or 96 m wide. Obviously, its more complex than that because of the animals and baggage.  Given I think the army would be looser than this, because of the gaps between sub columns and between units, we'd need a cleared route of about 150 m width.  If we concluded 100 metres of cleared route was realistic, we could have an army of 320,000. 

If we use the Roman army density Justin quotes, I suspect a 150m cleared road would give us around a 250,000 army but I haven't calculated it out.

So if the Persians can clear 150 metres of pathway, what's stopping them clearing 600 metres? They had 4 years to do it.

Napoleon's Grande Armée was divided into corps of 20 000 to 30 000 troops, which is the practical size for marching on a single road, with each corps taking a separate route to the battlefield. The road system in Europe at the end of the 18th century was far more developed than that in Thrace and Macedonia (and in the rest of the Persian Empire for that matter).

For a Persian army of 480 000 men, or just 200 000 men, to march to Greece, it would either have to do it in a string of separate corps a la Napoleon all following each other on the same road, or take separate routes along pathways broad enough for a column at least 4 men wide. Any evidence for either option?

[Prufrock]

For a Persian army of 480 000 men, or just 200 000 men, to march to Greece, it would either have to do it in a string of separate corps a la Napoleon all following each other on the same road, or take separate routes along pathways broad enough for a column at least 4 men wide. Any evidence for either option?

For the second, yes: H (7.121) says the army was split into three commands while moving through Thrace.

The order of the army's march, from Doriscus to Acanthus, had been such as I will show. Dividing his entire land army into three parts, Xerxes appointed one of them to march beside his fleet along the coast. [3] Mardonius and Masistes were the generals of this segment, while another third of the army marched, as appointed, further inland under Tritantaechmes and Gergis. The third part, with which Xerxes himself went, marched between the two, and its generals were Smerdomenes and Megabyzus.

http://www.perseus.tufts.edu/hopper/text?doc=Hdt.%207.121&lang=original

[Justin Swanton]

For a Persian army of 480 000 men, or just 200 000 men, to march to Greece, it would either have to do it in a string of separate corps a la Napoleon all following each other on the same road, or take separate routes along pathways broad enough for a column at least 4 men wide. Any evidence for either option?

For the second, yes: H (7.121) says the army was split into three commands while moving through Thrace.

The order of the army's march, from Doriscus to Acanthus, had been such as I will show. Dividing his entire land army into three parts, Xerxes appointed one of them to march beside his fleet along the coast. [3] Mardonius and Masistes were the generals of this segment, while another third of the army marched, as appointed, further inland under Tritantaechmes and Gergis. The third part, with which Xerxes himself went, marched between the two, and its generals were Smerdomenes and Megabyzus.

http://www.perseus.tufts.edu/hopper/text?doc=Hdt.%207.121&lang=original

If 30 000 men is the practical size for a force marching along a single road from one campsite to another in a single day, then 3 groups aren't enough for Xerxes' army. He'll need at least 7 or so (as did Napoleon) for an army 200 000 strong, and 16 for an army 480 000 strong. He'll also need to keep them separated for the entire route.

[Prufrock]

If 30 000 men is the practical size for a force marching along a single road from one campsite to another in a single day, then 3 groups aren't enough for Xerxes' army. He'll need at least 7 or so (as did Napoleon) for an army 200 000 strong, and 16 for an army 480 000 strong. He'll also need to keep them separated for the entire route.

Our old friend Maurice has each column about 60,000-80,000 strong, from memory.

[Mark G]

Napoleons columns always took a road.
Always.

Infantry to the ditch alongside, guns and horses on the road itself, with half width so that messengers could return along it, and a crash was not going to block the entire army.

There is no comparison to the magical manoeuvres you are proposing here.

It really is astounding that you two are still seeking to prove this.

Has anyone actually changed their opinion yet?

[Erpingham]

So if the Persians can clear 150 metres of pathway, what's stopping them clearing 600 metres? They had 4 years to do it.

An interesting question.  What does Herodotus actually say on the road building programme?  Where we have examples of armies building routes of advance, do they tend to clear huge areas or keep it fairly narrow and road like?  I've seen quoted distances for the overall distance between the boundary ditches of a Roman road in Britain of up to 100 m.  Edward I had a law that roads had to have a clear space 200 ft (60m) either side.

Incidentally, do we know work started on the road building programme immediately.  As I have often said to colleagues who'd say "It's Ok, we have a year to do it" - "Only if we start today".  Do we know they started immediately?

For a Persian army of 480 000 men, or just 200 000 men, to march to Greece, it would either have to do it in a string of separate corps a la Napoleon all following each other on the same road, or take separate routes along pathways broad enough for a column at least 4 men wide. Any evidence for either option?

I thought we had no roads at all (according to Patrick)?  The answers have to be somewhere in the geography - could the army advance along several parallel routes?  So far, we've assumed no.  Also, there has been an emphasis on the whole army moving between camps a day apart every day.  The idea that the same routes and camps were used over multiple days was rejected.  If the army is spread over several days in a long column it brings new march rate issues (will the disturbed and manured ground slow the march of the rear elements) logistical issues (fodder will be depleted at the camp site, stockpiles may no longer be adequate for the rear units) and certainly some hygiene issues of reuse of fouled ground.

[Erpingham]

Has anyone actually changed their opinion yet?

I didn't realise this was the idea   I've just been providing Justin with info to make more realistic calculations.

On further reflection : I think I did change my view a bit in the second major on this a few years ago.  I became more open to the idea that we could be talking an army of 250,000 or a bit more, including support troops.  I wouldn't reject 500,000 as impossible, just feel it is very unlikely.  I must admit, I find ancient (and especially medieval - no surprise) logistics interesting so I don't mind learning more.

[Flaminpig0]

The only Iranian source I have been able  to find http://www.iranchamber.com/history/achaemenids/achaemenid_army.phptalks about the  slow pace of marches due to 'the heavy baggage-train which often comprised litters for conveying the wives and concubines of the commanders'  which would tend to indicate that strategic off road travel for Persian forces wasnt a  realisitc option.

[Erpingham]

In the continuing spirit of opening up thinking on what a "road" might have meant, I found this by Historic England

Most new roads dating from the post-Roman period to the
18th century were merely heavily used trackways. They share
the characteristics of trackways and, apart from a few town
streets, bridge approaches and causeways, they were not
metalled. They follow the natural contours of the land and
avoid existing boundaries. When a road was obstructed or
impassable, travellers had the right (enshrined in the Statute
of Winchester of 1285) to diverge from its course, causing
multiple hollow-ways running alongside each other to be
created. Prior to enclosure, roads were often not restricted by
walls or hedges and so they tended to be wide, consisting of
large numbers of roughly parallel hollow-ways spreading out
across broad swathes of the countryside.

Greece may, of course , have been different but the idea of a trampled route rather than a narrow paved street in the Roman manner might help.

[RichT]

I thought we had no roads at all (according to Patrick)? 

Of course there were roads in Greece - albeit not nice metalled straight maintained roads like Roman ones (outside of cities). Probably more like the unsurfaced roads still to be found in rural areas today. Sometimes they followed river beds, which is fine in summer, or if you've already drunk the river dry 

Edited to comment on:

Prior to enclosure, roads were often not restricted by walls or hedges and so they tended to be wide, consisting of  large numbers of roughly parallel hollow-ways spreading out  across broad swathes of the countryside.

A good model for this type of thing for those of us in southern England is the Ridgeway.

[Jim Webster]

So if the Persians can clear 150 metres of pathway, what's stopping them clearing 600 metres? They had 4 years to do it.
[/quote]

The basic maths is that it would take four times as long. But actually it would take far longer because the area nearer the central line of the road/track would be easier to work on.
As an interesting side issue, because nobody was actually using this notional 600 meter frontage in the three years or four years before the army came through, unless you went to the trouble of physically uprooting stuff and burning it, you'd probably find that you'd have scrub and brush growing again so it would still need redoing.

[Erpingham]

Prior to enclosure, roads were often not restricted by walls or hedges and so they tended to be wide, consisting of  large numbers of roughly parallel hollow-ways spreading out  across broad swathes of the countryside.

A good model for this type of thing for those of us in southern England is the Ridgeway.

Example from Twyford Down