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

[Justin Swanton]

At the beginning of his Histories Herodotus mentions 'Persians best informed in history', 'Persian learned men' - λόγιοι having the sense of a) versed in tales or stories; b) learned, erudite (most common usage); or c) skilled in words, eloquent.

Its possible. A long list of things are possible.
But historiography, public libraries, scientific method - all that good stuff - were a work in progress.
So when Herodotus writes, "Persian's say," is it a sailor's tale over a glass of akraton at his local pub?

Or for example, Herodotus writes "Corinthians say, and the Lesbians agree" - so he's double-checked it right? Two sources? But they agree that Arion rode to Taenarsus on the back of a dolphin. So is that exemplary of his incredulity?

While very helpfully Herodotus sometimes refers to other authors' histories, does that mean he is going without in the vast majority of times he does not? Who knows.

What is so incredible about a dolphin saving the life of a man? They do it all the time. Here's a story of dolphin pushing a non-swimmer to safety. And another similar story. And another. And another.

Here's a video of a humpback whale pushing a diver to safety.

[Justin Swanton]

Maurice is coming apart for me. I'm willing to read Young - any chance of sending me the relevant pages as scans?

I don't have a copy of the Young article, but Mark K kindly quoted some of it on a yahoo group, which I will requote here (and do note the size of the force he is considering):

Though I can well imagine that many will have been glad to see no mention of this topic lately, I cannot restrain myself from mentioning another article: T. Cuyler Young, Jr., "480/479 B.C – A Persian Perspective" (1980), 15 Iranica Antiqua 213.


Young calculates some logistical and transport requirements for a hypothetical Persian force of 210,000 men and 75,000 horses, based on the assumption that each soldier required approximately 3 pounds of grain per day, and each pack and cavalry animal 10 pounds of grain and 10 pounds of straw or other fodder.  At pp. 225-227, for example, he considers the advance from Therma to Thermopylae :

Hide message history
"Once having departed from Therma we may reasonably assume that the Persians were marching through hostile country, and had left territories which they had been able to supply with stores before the campaign began. [Footnote 31: Persian control before the start of the campaign could hardly have extended much south of Therma. Certainly it reached no further south than Tempe, for early in the campaign season of 480 B.C. Greek troops, under Spartan command, had been active that far north. For details see Burn 1962: 339-345.] In other words, from Therma on they had either to carry their grain supplies with them (by pack-horse, or by sea), or to live off the Greek harvest, or both. There would have been no prepared grain supplies along the route of march.



"The chronology of this march is subject to various interpretations, but the figures offered by Maurice are reasonable: the march itself took thirteen days, and the Persians were some six days at Thermopylae . [Footnote 32: Maurice 1930: 233.] While in the latter position, supply by sea would have been both possible and relatively easy, since the navy could control the approaches to the bay of Lamia . The situation was much different along most of the route of march to Thermopylae , however, and it is on this stretch that the Persian Quartermaster Corps would have been severely tested. The army divided and marched south from Therma by at least two, if not three, different roads (following perfectly the Napoleonic maxim to march divided, fight combined). [Footnote 33: For an excellent map of the Persian routes of march south from Therma v. Burn 1962: 340.] Only one of these routes was along the coast, and even then contact with the sea was easy only for about the first thirty miles. This left a total of one hundred and ten miles for that column to march on an inland road before reaching Thermopylae, and the other two columns were inland the whole distance from Terma [sic] to Thermopylae . Thus, the coastal column was out of touch with the sea for about ten days, and the rest of the army could not be supplied by ship for the whole thirteen days of the march south to the pass where Leonidas was to win his niche in history.



"Let us return to some plain facts and figures of military reality with this march in mind. An army of 210,000 men and 75,000 horses required a total of 1,380,000 pounds of grain a day (fodder and water needs may be left aside to simplify the discussion). A single pack-horse can carry about two hundred and fifty pounds of grain, but we must remember that the pack-horse itself eats ten pounds of grain a day, so its effective carrying capacity beyond its own needs is two hundred and forty pounds [Footnote 34: Engels 1978: 19 and Table 1.] A simple calculation reveals that 5,750 pack-horses would be needed to carry south from Therma the grain necessary to feed for one day an army which ate 1,380,000 pounds of cereal a day (1,380,000 divided by 24). For a two day march one would need 12,000 pack-horses to carry 2,760,000 pounds of grain, a two day supply for the army with each horse now able to carry only 230 pounds because that horse would itself eat twenty pounds of grain in two days. If one continues with this straightforward calculation to cover a ten day march, the minimum amount of time the one coastal column would have been out of touch with the sea and hence unable to be supplied by ship, on that tenth day the army would require 4,710,000 pack-horses to provide a total of 706,560,000 pounds of grain which is what the troops and animals would have eaten in the ten days of marching.



"Even if: (1) we assume that at least half of the army's grain needs were met from the Greek countryside, and (2) we ignore the transportation which would have been required to move that foraged grain to the troops, one is still left with an impossible demand on the Persian Quartermaster General to move grain forward from Therma. The only reasonable conclusion from these calculations is that the Persian army which marched to Thermopylae did not even come close to numbering 210,000 men and 75,000 animals."

Young goes on to assert that similar considerations would apply to limit the size of the Persian force that could have moved from Thermopylae to Athens , and the size of the Persian force at Plataea .

Two things come to mind:

1. I don't see why the Persian army needed to be inland for longer than five days at any time. It starts out from Therma and follows the coast to Phila. It then moves inland along the valley of the Tempe river. That part is brutal, moving through a hilly pass for about 10km to the plain beyond but from what I can see from Google maps and street views, it's doable. It's then four days to the coast again at Pagasae. After that the army can then split into two halves, one half following the coast the other half going inland and both meeting five days later on the broad valley of the Spercheus river before Thermopylae.

2. Grain is harvested in Greece in June, just in time for the arrival of the Persians. Over the Tempe river pass there is a broad fertile plain and they can take all they want: food for the locals' needs for twelve months. Foraging parties would strip the area, giving the army more than enough for its requirements during its passage.

Is that several million men moving along the Valley of the Tempe River and through the hilly pass?

It would probably need to be over the hilly ground north of the Tempe valley which itself is a steep and narrow gorge. The army must be able to march in a column/columns several hundred yards wide at all times.

[Andreas Johansson]

He did use the work of earlier historians - at least, he quotes what "Hecataeus the son of Hegesandrus declares in his history".  (It's commonly suggested that he got his list of Achaemenid satrapies and their revenues from Hecataeus, but he doesn't actually say so.)  So he was not averse to using written sources. But given that we have no knowledge that he read Aramaic, his source would presumably have to have been interview of someone who could.

If the "History" of Hecataeus he refered to was the work now known by that name, it wasn't a history in the modern sense, but a book criticizing myths. And even if he did use earlier historians, would those have consulted a hypothetical Persian document in an archive or Kuriosenkabinett somewhere? Merely using written sources isn't the same as going to the archives for primary documentation.

[Dangun]

What is so incredible about a dolphin saving the life of a man?

As I suggested earlier, there are different places to sit on the incredulity continuum.

After being captured by pirates, and before being tossed from the ship, Arion played his harp and sang to Apollo. The music attracted dolphins who duly rescued him after he walked the plank. The dolphins carried him to a temple of Poseidon... and Hyginus writes that the dolphin was eventually placed in the heavens as the castellation Delphinus. (In my earlier post I didn't mean to leave this bit out - my apologies - I was just quoting from Herodotus.)

Myth and history got a little blurry.

[Prufrock]

As I suggested earlier, there are different places to sit on the incredulity continuum.

Yes indeed.

So far we've had practical evidence presented against the idea of a 5 million men plus expedition dismissed on what might be termed 'technicalities', while various theories with nothing to support them other than 'my impression is...' or 'I once did that...' or why wouldn't they just do this?' are accepted without question.

Wind and weather conditions are not always ideal for sailing - dismissed.
It takes quite a lot of time and effort to unload vessels - only if they're modern vessels.
That chokepoints make timely passage impossible for such a large number of people - well, they'd just go over the mountains.
That what we know of supply considerations would make feeding and watering a travelling band of 5 plus million folk impossible, and at particular points very difficult even for forces an order of magnitude lower - these are the Persians we're talking about! Normal considerations don't apply!
That men with experience moving large bodies of troops and who had walked the ground felt that it would not be possible - I dispute a calculation, and besides, he's from a different era - he must be wrong.
Herodotus doesn't always tell the truth - you can rely on me to know the truth when I see it.

And that's about where we're at!

[Patrick Waterson]

"Let us return to some plain facts and figures of military reality with this march in mind. An army of 210,000 men and 75,000 horses required a total of 1,380,000 pounds of grain a day (fodder and water needs may be left aside to simplify the discussion). A single pack-horse can carry about two hundred and fifty pounds of grain, but we must remember that the pack-horse itself eats ten pounds of grain a day, so its effective carrying capacity beyond its own needs is two hundred and forty pounds [Footnote 34: Engels 1978: 19 and Table 1.] A simple calculation reveals that 5,750 pack-horses would be needed to carry south from Therma the grain necessary to feed for one day an army which ate 1,380,000 pounds of cereal a day (1,380,000 divided by 24). For a two day march one would need 12,000 pack-horses to carry 2,760,000 pounds of grain, a two day supply for the army with each horse now able to carry only 230 pounds because that horse would itself eat twenty pounds of grain in two days. If one continues with this straightforward calculation to cover a ten day march, the minimum amount of time the one coastal column would have been out of touch with the sea and hence unable to be supplied by ship, on that tenth day the army would require 4,710,000 pack-horses to provide a total of 706,560,000 pounds of grain which is what the troops and animals would have eaten in the ten days of marching.

Er ... should that be 1,380,000 divided by 240?  (I suspect a typo because the result, 5,750 horses, works out correctly).

There are a few notable assumptions here:
1) Pack horses - mules and camels can take double the 250 lbs Young assigns to a pack horse.  Young may have up to twice as many animals as Xerxes actually used.
2) Young assumes no fodder anywhere along the route, despite the Persians having marched 'at the best season of the year' (Herodotus VII.50).
3) Curious geometric rather than arithmetic increase in animals.
1 day's supply = 5,750 animals
10 days' supply = 4,710,000 animals (should be 57,500)
If the army is eating 5,750 animal loads' worth on day 1, it will eat the same on each of days 2-10.  Ergo, the number of animals needed to carry supplies for ten days = 10x the number of animals needed to carry supplies for one day.  Young's arithmetic is wildly out.
4) Curious rationale for loading ("a two day supply for the army with each horse now able to carry only 230 pounds because that horse would itself eat twenty pounds of grain in two days") - the horse still carries 250 lbs; the consumption fo grain is effectively being levied twice.
5) The root of the problem: "carry south from Therma the grain necessary" - Young is assuming an 18th century depot system rather then the actual practice of having the baggage train move with the army.

Because of this fundamental mistaken assumption, Young's calculations are entirely worthless.  Unless you want to move an 18th century army between fortresses, and even then they somehow managed 10-day journeys without all this fuss (thinking of Frederick the Great's marches in the Seven Years' War).

[Duncan Head]

3) Curious geometric rather than arithmetic increase in animals.
1 day's supply = 5,750 animals
10 days' supply = 4,710,000 animals (should be 57,500)
If the army is eating 5,750 animal loads' worth on day 1, it will eat the same on each of days 2-10.  Ergo, the number of animals needed to carry supplies for ten days = 10x the number of animals needed to carry supplies for one day.  Young's arithmetic is wildly out.

It's not quite that simple - as Engels points out with detailed calculations, IIRC - because all the extra pack animals have to eat from the supplies they carry, as well. An army with x men, y horses, and z pack animals will need a lot less than an army with x men, y horses, and 10z pack-animals. (Unless they can graze, and of course the more animals you have, the harder it is to find enough grazing.) Doesn't Engels claim that no army can carry more than about eight days' worth of provisions, because beyond that the extra pack animals are just eating their own loads? There are ways round the problem (like redistributing loads and then eating the spare animals, feeding them to Thracian lions, or even just sending them back) but you can't just use an arithmetic progression.

[Patrick Waterson]

As I suggested earlier, there are different places to sit on the incredulity continuum.

Sitting on the analytical continuum might be a better bet.

Myth and history got a little blurry.

If this blurring occurs through trying to discredit Herodotus because he reports a seemingly miraculous rescue, I suggest it might unblur if one were simply to read the source without preconceptions.

We are coming up against a key factor in the readnig/understanding/interpretation of history.  I have noticed a tendency to smack down a Procrustean bed of 20th/21st century concepts and assumptions followed by surgery on anything protruding past those assumptions.  This does not result in history, merely in self-confirmation of self-delusion (given that any period's outlook and ethos will generally contain more delusion than fact).

So how do we get to history?  My recomendation is to treat it as an intelligence-gathering exercise, not a debating society exercise.  To do that, one must pay more attention to one's sources than one's preconceptions.  Cross-check the information in sources wherever possible, and consider the implications of any source statement, but consider it as accurately rather than as preconceptually as possible and be aware of the operating systems, ethos and outlook of the time.

So far we've had practical evidence presented against the idea of a 5 million men plus expedition dismissed on what might be termed 'technicalities', while various theories with nothing to support them other than 'my impression is...' or 'I once did that...' or why wouldn't they just do this?' are accepted without question.

We have not had practical evidence, only assessments by 20th/21st century people of how things would be done in the 20th/21st century.  This is about as valuable as a modern tactician explaining how a modern infantry company would take an objective and concluding this is how it was for a Roman century.

Wind and weather conditions are not always ideal for sailing - dismissed.

Why not come up with examples from classical sources of unloading delayed by weather?  If you are right about iffy conditions, there ought to be at least some instances where the best-laid unloading plans gang aft agley on account of playful Mother Nature.

It takes quite a lot of time and effort to unload vessels - only if they're modern vessels.

Modern vessels actually unload pretty quickly because of containerisation, which is essentially the same principle as using amphorae but on a larger scale.

That chokepoints make timely passage impossible for such a large number of people - well, they'd just go over the mountains.

Logical enough - the Assyrians did it, the Spartans did it.

That what we know of supply considerations would make feeding and watering a travelling band of 5 plus million folk impossible, and at particular points very difficult even for forces an order of magnitude lower - these are the Persians we're talking about! Normal considerations don't apply!

Maybe what we think we know is wrong.  See my post on Young.

That men with experience moving large bodies of troops and who had walked the ground felt that it would not be possible - I dispute a calculation, and besides, he's from a different era - he must be wrong.

Well, he is applying the calculations of an entirely different type of army with entirely different practices to a situation beyond his ken and experience.

Herodotus doesn't always tell the truth - you can rely on me to know the truth when I see it.

There is a bit more to it that that, Mr Bell.  Be fair!

[Patrick Waterson]

3) Curious geometric rather than arithmetic increase in animals.
1 day's supply = 5,750 animals
10 days' supply = 4,710,000 animals (should be 57,500)
If the army is eating 5,750 animal loads' worth on day 1, it will eat the same on each of days 2-10.  Ergo, the number of animals needed to carry supplies for ten days = 10x the number of animals needed to carry supplies for one day.  Young's arithmetic is wildly out.

It's not quite that simple - as Engels points out with detailed calculations, IIRC - because all the extra pack animals have to eat from the supplies they carry, as well. An army with x men, y horses, and z pack animals will need a lot less than an army with x men, y horses, and 10z pack-animals. (Unless they can graze, and of course the more animals you have, the harder it is to find enough grazing.) Doesn't Engels claim that no army can carry more than about eight days' worth of provisions, because beyond that the extra pack animals are just eating their own loads? There are ways round the problem (like redistributing loads and then eating the spare animals, feeding them to Thracian lions, or even just sending them back) but you can't just use an arithmetic progression.

A valid point, and an illuminating one in view of the classical period habit of carrying about one week's worth of provisions in the baggage train, although once one has allowed for the additional animals to carry fodder for the other animals the increase is still more arithmetic than geometric.  I calculated (roughly) 135,000 animals to feed the army for a week rising to around 400,000 if they have to carry fodder to feed themselves.  This is still a whole order of magnitude lower than Young's 4 million.

[Justin Swanton]

What is so incredible about a dolphin saving the life of a man?

As I suggested earlier, there are different places to sit on the incredulity continuum.

After being captured by pirates, and before being tossed from the ship, Arion played his harp and sang to Apollo. The music attracted dolphins who duly rescued him after he walked the plank. The dolphins carried him to a temple of Poseidon... and Hyginus writes that the dolphin was eventually placed in the heavens as the castellation Delphinus. (In my earlier post I didn't mean to leave this bit out, I was just quoting from Herodotus.)

Myth and history got a little blurry.

Let's summarise what happened to Arion in more contemporary terms. He is captured by Pirates off the coast of southern Greece and they intend to kill him. He decides to invoke the help of Apollo (not Poseidon notice) and disguises his prayer as a song so the pirates don't catch on and shut him up. The last thought in his head are the dolphins who, as dolphins do, are following the ship. Apollo doesn't answer his plea and he gets tossed into the sea by the impatient pirates. The dolphins, to his complete surprise, decide to help him out - they nudge him towards land and he catches hold of the fin of them. The dolphin gets the idea and takes him to the shore. By coincidence there is a temple to Poseidon nearby where they land, but I imagine there would be quite a few temples to Poseidon along the coastline to Greece where fishermen depend on the good graces of the god in order to ply their trade.

For some odd reason the dolphin doesn't just leave him on the beach with a wave of its fin and head back to sea, but gets stranded on the sand. Something obviously happened here to trap the dolphin, the details of which we are not told, which would be strange if this was a myth but sounds believable as an incomplete factual account.

How does that sound?

[Erpingham]

I'm not sure whether continuing with factual inputs is worth it after the dolphin episode but here goes.

From wikipedia Mule :

The median weight range for a mule is between about 370 and 460 kg (820 and 1,000 lb). While a few mules can carry live weight up to 160 kg (353 lb), the superiority of the mule becomes apparent in their additional endurance.

In general, a mule can be packed with dead weight of up to 20% of its body weight, or approximately 90 kg (198 lb).[6] Although it depends on the individual animal, it has been reported that mules trained by the Army of Pakistan can carry up to 72 kilograms (159 lb) and walk 26 kilometres (16.2 mi) without resting. The average equine in general can carry up to approximately 30% of its body weight in live weight, such as a rider.

From wikipedia pack animal

Loads for equids are disputed. The US Army specifies a maximum of 20 percent of body weight for mules walking up to 20 miles a day in mountains, giving a load of up to about 150 kg. However an 1867 text mentioned a load of up to 800 pounds (about 360 kg). In India, the prevention of cruelty rules (1965) limit mules to 200 kg and ponies to 70 kg.


Of course, we could use the idea that things were different in the past and ancient mules were super beasts capable of carrying vastly more or they were gigantically big, so 20% of their body weight was more but I'd suggest we aim at mule loads 72-150 kg for long range, difficult terrain work.  So Young is closer than Patrick here.

According to the Commonwealth Scientific and Industrial Research Organisation (CSIRO), maximum safe load for pack camels is 300kg, so Patrick is correct to say they can carry twice the load of mules.

[Patrick Waterson]

Thank you for taking the time and effort to write a response

In order .
1. Can you produce any evidence for this Cheka style organisation? for example  is it mentioned by any writers contemporary to the Persian empire

References to the 'King's Eyes' are found throughout Greek historical literature.

2. What would this organisation capable of replicating the work of a 'regular logistics corps' and able to organise the feeding of several million men be.

The same that kept provinces administered and armies raised in normal circumstances: the satrapal government system.

3. The Bronze tablet would it be accurate enough and be of  the right scale to lead people to a supply dump bearing in mind they are travelling across country in a land they have never  seen before and do not have the benefit of GPS. Are such Bronze tablets common and is map reading a common skill within the Persian Empire

These maps were highly functional (cf. London Underground maps) but were not used for this purpose.  The army marched - with guides - towards known locations (cities) where supplies had been gathered by the locals (under advance planning from above, see Herodotus VII.119) or obvious unloading areas ("There's the sea, there are the ships, here come the boats").

4. Which simply goes to prove that the invasion army was of a  reasonable  size as opposed to the HHAH that is to say there is no evidence for the creation unusual number of sailors and the like

This is a non sequitur.

For me the problem is that you and Justin start from the assumption that Herodotus must be right about the size of Xerxes army

Well, he is either right or wrong.  So we try checking out a few parameters.  Was the Achaemenid Empire capable of creating those conditions, operating with 5th century BC (not 20th century AD) methods?  Do the dimensions of the bridge across the Hellespont and the crossing time for his army match up with the claimed figures?  For what approximate size of army would the loss of sea supply following Salamis cause a more or less instant logistical collapse, given Mardonius' ability to maintain 300,000 over winter in Thessaly without maritime supply?  And so on.

and then work backwards to create the conditions that allow for its existence, it is not falsifiable.

Do you mean 'verifiable' rather than 'falsifiable'?  I would be saddened to think that falsification was one's aim when studying history.

Without wishing to be rude the Persian Empire that you need to depict is a totalitarian society with a magical ability to collect, analyse and disseminate  information- it has more in common with the worlds of M A R Barker or Clarke Ashton Smith  than than an actual ancient society.

I suggest getting to know Biblical period societies a bit better - in particular, the extent and depth of detail in their archives, e.g. Amarna archives, Pylos archives, Hattusas archives and the many Babylonian archives, tablets and hides under various regimes including the Achaemenids.  And yes, they were what we would consider totalitarian - in accordance with their laws and customs, and without continually outpouring a 'party line' - but, to someone thinking in 20th/21st century terms, totalitarian.

[Erpingham]

How does that sound?

Like something out of one of those American TV shows on minor channels that everyone watches for a laugh?

I suggest we stick to the area of Herodotus' historical writings.  Fortunately, we don't need to assume that, because he repeated fantastical stories, he is a fantasist. 

[Patrick Waterson]

From wikipedia Mule :

The median weight range for a mule is between about 370 and 460 kg (820 and 1,000 lb). While a few mules can carry live weight up to 160 kg (353 lb), the superiority of the mule becomes apparent in their additional endurance.

In general, a mule can be packed with dead weight of up to 20% of its body weight, or approximately 90 kg (198 lb).[6] Although it depends on the individual animal, it has been reported that mules trained by the Army of Pakistan can carry up to 72 kilograms (159 lb) and walk 26 kilometres (16.2 mi) without resting. The average equine in general can carry up to approximately 30% of its body weight in live weight, such as a rider.

From wikipedia pack animal

Loads for equids are disputed. The US Army specifies a maximum of 20 percent of body weight for mules walking up to 20 miles a day in mountains, giving a load of up to about 150 kg. However an 1867 text mentioned a load of up to 800 pounds (about 360 kg). In India, the prevention of cruelty rules (1965) limit mules to 200 kg and ponies to 70 kg.


Of course, we could use the idea that things were different in the past and ancient mules were super beasts capable of carrying vastly more or they were gigantically big, so 20% of their body weight was more but I'd suggest we aim at mule loads 72-150 kg for long range, difficult terrain work.  So Young is closer than Patrick here.

It really depends upon how kind the Achaemenids were to their animals.  Spanish in the Napoleonic Wars, for example, appear always to have loaded their mules to something like the 500 lbs mark.  US Army practice varied (as per the Wikipedia pack animals article) and Near Eastern societies on the whole are not noted for their concern for animal welfare.  So the average Achamenid mule load could be 250lbs, 500 lbs or anywhere in between (I am assuming that unlike the US Army they had not bred anything capable of standing up under 800 lbs).  Given that their masters would aim to carry the most material upon the fewest animals, I would go for around 500 lbs, noting that the load will get lighter every day so it is not quite as bad for the animals as it looks.

How does that sound?

Like something out of one of those American TV shows on minor channels that everyone watches for a laugh?

Actually I find Justin's suggestion quite reasonable.  We get unlikelier things happening in this day and age, and something noteworthy must have happened to Arion for such a tale (which is well within the known general behaviour patterns of dolphins) to be recorded.  What I dislike is the 'cultural racism' approach which seems to take for granted that any tale told in classical Greek sources must be the inventive product of an inferior mind.

[Erpingham]

Do you mean 'verifiable' rather than 'falsifiable'?  I would be saddened to think that falsification was one's aim when studying history.

Falsifiability is a criterion for something to be considered testable in scientific method.  The test must be able to show that the hypothesis is wrong.  Justin, who I know has studied the Philosophy of Science, will doubtless recognise this.

Whether the study of history is scientific and subject to the falsification test is another matter