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

[Jim Webster]

Secondly, of course with settled agriculture your famine level is close to your abundance level. Areas which have settled agriculture tend to have a population which can be supported under normal circumstances.

Even in the event of failure of your principal water supply?  I think you may need to cast your evidential net more widely.

Actually check on the figures, you'll find that a 10% drop in production causes a perfectly adequate famine. When combined with the inevitable private stockpiling to play the market
Again I refer you to The Grain Market in the Roman Empire: A Social, Political and Economic Study as a good starting point

[Jim Webster]

I mentioned amphorae in connection with grain transport.  Whether they were used for storage is conjectural, but it would be interesting to consider what effect this might have on the keeping qualities of their contents.

Evidence would be nice, sources and stuff
Have you any evidence for amphorae being used to transport grain?

[Jim Webster]

Which is what I am doing.  I agree about assessing the practicality; what I do not agree with is beginning with the idee fixe that it is necessarily impossible.  It might be possible or it might not, and we need to understand what the Achaemenid Empire could and, based on its other known and recorded activities, would, do.  This seems to be somewhat different to what people tend, on the basis of a 20th century mentality, to believe.

I don't have an idee fixe, I just looked at the evidence and decided that on the basis of very basic factors such as the time taken to get people through choke points, the logistics of supplying that number of men on the move etc, that it was never going to happen.

I merely looked at the evidence, should anybody provide me with evidence to show how the Persians moved forces numbering five or six million men about on a regular basis I'd be interested to see it.
With the Persian campaigns against Egypt, where we have much more information, Greeks being present in the armies, the numbers are far more reasonable

[Duncan Head]

Here's something: Caligula's ship-bridge across the Straits of Messena. That's a distance of 3 miles or about 5280 yards. A typical Roman merchant ship was about 16 yards wide. That gives you a ball-park figure of 330 ships. According to Lacius Curtius requisitioning the ships caused a severe famine in Italy, notably Rome, which implies that Caligula had expropriated the Carthage-to-Rome supply fleet, say 300-odd ships, having built the remainder.

If one supposes from this that the entire Roman Mediterranean merchant fleet numbered somewhere between 500 and 1000 ships of all sizes, it would be unreasonable to conclude that the regular Persian merchant fleet numbered more than a couple of hundred vessels. But with 4 years to prepare Xerxes had enough time to construct the extra ships he would need to supply a massive army.

Herodotos VII.1.2 does say that Xerxes requested the cities to provide more of, among other things, transport ships than in previous years.

Actually the Roman incident is a nice comparison. But on checking Herodotos VII.97, the 3,000 is the total of warships and horse-transports, not grain-ships - I'm not sure if Herodotos ever gives us a figures for those. So this may be a bit of a red herring, sorry.

Incidentally the most often-cited numerical analysis of Xerxes' supply requirements these days, more recent than Maurice, is probably T Cuyler Young, "480/479 BC - A Persian Perspective" (Iranica Antiqua XV, 1980). Can't find it online, and I only have a few old handwritten notes. But Patrick, and probably Justin, would hate him. He argues that even Maurice's 200,000+ for the army is much too large to be supplied. At one point he calculates that if Xerxes' army were the size that Maurice calculated, the army would require 85 grain-ships in continuous shuttle duty. And he thinks that's unrealistically too many ships! Even I, instinctively sceptical of Herodotos' propaganda figures, think this may be harsh.

(He's talking about the period before Salamis, when the Persian army and fleet is in Attica. Assuming 134,000 men in the fleet, 210,000 in the army and 75,000 horses - which IIRC is Maurice's figures - he reckons they'd need 892 tons of grain a day. For the three weeks between occupying Athens and the battle of Salamis, that's 18,732 tons. Assuming grain is being brought from the nearest depot is at Therma, five days' sail from Athens, assume (he admits uncertainty here) an average capacity of 130 tons per ship, the Persians must offload 7 ships per day at Athens, so they'd need 7x5 ships travelling each way on any given day, 7 loading at Therma, and 7 unloading at Athens - total 84 ships. Which he thinks is too many, though he never says why it's too many.)

[Flaminpig0]

One thing that isn't clear to me is why the Persians would have felt the need for such an extraordinarily large army.

Think like an Achaemenid.  Read Herodotus Book VII (especially 103):

When he heard this, Xerxes smiled and said, "What a strange thing to say, Demaratus, that a thousand men would fight with so great an army! Come now, tell me this: you say that you were king of these men. Are you willing right now to fight with ten men? Yet if your state is entirely as you define it, you as their king should by right encounter twice as many according to your laws. If each of them is a match for ten men of my army, then it is plain to me that you must be a match for twenty; in this way you would prove that what you say is true. But if you Greeks who so exalt yourselves are just like you and the others who come to speak with me, and are also the same size, then beware lest the words you have spoken be only idle boasting. Let us look at it with all reasonableness: how could a thousand, or ten thousand, or even fifty thousand men, if they are all equally free and not under the rule of one man, withstand so great an army as mine? If you Greeks are five thousand, we still would be more than a thousand to one. If they were under the rule of one man according to our custom, they might out of fear of him become better than they naturally are, and under compulsion of the lash they might go against greater numbers of inferior men; but if they are allowed to go free they would do neither. I myself think that even if they were equal in numbers it would be hard for the Greeks to fight just against the Persians. What you are talking about is found among us alone, and even then it is not common but rare; there are some among my Persian spearmen who will gladly fight with three Greeks at once. You have no knowledge of this and are spouting a lot of nonsense."

Observe carefully the mentality, perspective and thinking contained within that assertion.  That pretty much answers the question.

The Achaemenids, broadly speaking, felt that success in battle depended upon two things: bravery and numbers.  Persians had bravery.  Persian armies had numbers.  Put the two together and success was guaranteed.  So they thought.

Presumably that would  apply to every other invasion they launched including the one that led to  Marathon.?

[Flaminpig0]

One thing that isn't clear to me is why the Persians would have felt the need for such an extraordinarily large army.

Think like an Achaemenid.  Read Herodotus Book VII (especially 103):

When he heard this, Xerxes smiled and said, "What a strange thing to say, Demaratus, that a thousand men would fight with so great an army! Come now, tell me this: you say that you were king of these men. Are you willing right now to fight with ten men? Yet if your state is entirely as you define it, you as their king should by right encounter twice as many according to your laws. If each of them is a match for ten men of my army, then it is plain to me that you must be a match for twenty; in this way you would prove that what you say is true. But if you Greeks who so exalt yourselves are just like you and the others who come to speak with me, and are also the same size, then beware lest the words you have spoken be only idle boasting. Let us look at it with all reasonableness: how could a thousand, or ten thousand, or even fifty thousand men, if they are all equally free and not under the rule of one man, withstand so great an army as mine? If you Greeks are five thousand, we still would be more than a thousand to one. If they were under the rule of one man according to our custom, they might out of fear of him become better than they naturally are, and under compulsion of the lash they might go against greater numbers of inferior men; but if they are allowed to go free they would do neither. I myself think that even if they were equal in numbers it would be hard for the Greeks to fight just against the Persians. What you are talking about is found among us alone, and even then it is not common but rare; there are some among my Persian spearmen who will gladly fight with three Greeks at once. You have no knowledge of this and are spouting a lot of nonsense."

Just a thought but what would be Herodotus' source for this quote from Xerxes?

[Justin Swanton]

One thing that isn't clear to me is why the Persians would have felt the need for such an extraordinarily large army.

Think like an Achaemenid.  Read Herodotus Book VII (especially 103):

When he heard this, Xerxes smiled and said, "What a strange thing to say, Demaratus, that a thousand men would fight with so great an army! Come now, tell me this: you say that you were king of these men. Are you willing right now to fight with ten men? Yet if your state is entirely as you define it, you as their king should by right encounter twice as many according to your laws. If each of them is a match for ten men of my army, then it is plain to me that you must be a match for twenty; in this way you would prove that what you say is true. But if you Greeks who so exalt yourselves are just like you and the others who come to speak with me, and are also the same size, then beware lest the words you have spoken be only idle boasting. Let us look at it with all reasonableness: how could a thousand, or ten thousand, or even fifty thousand men, if they are all equally free and not under the rule of one man, withstand so great an army as mine? If you Greeks are five thousand, we still would be more than a thousand to one. If they were under the rule of one man according to our custom, they might out of fear of him become better than they naturally are, and under compulsion of the lash they might go against greater numbers of inferior men; but if they are allowed to go free they would do neither. I myself think that even if they were equal in numbers it would be hard for the Greeks to fight just against the Persians. What you are talking about is found among us alone, and even then it is not common but rare; there are some among my Persian spearmen who will gladly fight with three Greeks at once. You have no knowledge of this and are spouting a lot of nonsense."

Observe carefully the mentality, perspective and thinking contained within that assertion.  That pretty much answers the question.

The Achaemenids, broadly speaking, felt that success in battle depended upon two things: bravery and numbers.  Persians had bravery.  Persian armies had numbers.  Put the two together and success was guaranteed.  So they thought.

Presumably that would  apply to every other invasion they launched including the one that led to  Marathon.?

What is interesting is that the primary sources give huge figures for all Persian armies, and not just Persian armies but also armies of other nations of the Fertile Crescent during that era. So it's not just one figure by one writer that is being discounted. If the sources are so unreliable as to make an army five or ten times larger than it really was, then they can't be trusted in anything else they assert. And that's the end of history.

Thus far in this thread I haven't seen any killer argument - backed up with proof - that demonstrates the impossibility of such an enormous army marching, camping and being fed and watered from the Hellespont to Greece.

Take imperial Rome. The city needed 420 000 tonnes of grain brought by ship each year. If the average Roman merchant ship had a carrying capacity of 100 tons that means 4200 ship voyages to feed the city. It takes two days by sea from Carthage to Rome with a favourable wind and four days with an unfavourable one, so a round trip including loading time, repairs and so on probably took about 10 days. The Mediterranean was navigable for 8 months of the year so a ship could make the trip about 24 times at best which means at least 175 ships dedicated just to bringing grain to Rome, year after year. But the Roman Empire did it, free of charge. What is the big obstacle for the Persian Empire doing it as a once-off, with years of preparation beforehand?

[Jim Webster]

Take imperial Rome. The city needed 420 000 tonnes of grain brought by ship each year. If the average Roman merchant ship had a carrying capacity of 100 tons that means 4200 ship voyages to feed the city. It takes two days by sea from Carthage to Rome with a favourable wind and four days with an unfavourable one, so a round trip including loading time, repairs and so on probably took about 10 days. The Mediterranean was navigable for 8 months of the year so a ship could make the trip about 24 times at best which means at least 175 ships dedicated just to bringing grain to Rome, year after year. But the Roman Empire did it, free of charge. What is the big obstacle for the Persian Empire doing it as a once-off, with years of preparation beforehand?

strangely enough, the city of Rome didn't move about. It had taken centuries to get the infrastructure right

[Justin Swanton]

Take imperial Rome. The city needed 420 000 tonnes of grain brought by ship each year. If the average Roman merchant ship had a carrying capacity of 100 tons that means 4200 ship voyages to feed the city. It takes two days by sea from Carthage to Rome with a favourable wind and four days with an unfavourable one, so a round trip including loading time, repairs and so on probably took about 10 days. The Mediterranean was navigable for 8 months of the year so a ship could make the trip about 24 times at best which means at least 175 ships dedicated just to bringing grain to Rome, year after year. But the Roman Empire did it, free of charge. What is the big obstacle for the Persian Empire doing it as a once-off, with years of preparation beforehand?

strangely enough, the city of Rome didn't move about. It had taken centuries to get the infrastructure right

Are beaches an insuperable problem?

[Justin Swanton]

I mentioned amphorae in connection with grain transport.  Whether they were used for storage is conjectural, but it would be interesting to consider what effect this might have on the keeping qualities of their contents.

Evidence would be nice, sources and stuff
Have you any evidence for amphorae being used to transport grain?

I'm not sure amphorae would actually be necessary. One needs to keep the grain dry below decks for a few days during the ship's passage across the Aegean. At the print shop where I work, the floor is regularly flooded after a heavy rain. Water and paper make a bad combination, so we keep all paper on palettes. Problem solved. It's enough to ensure the sacks don't touch the bottom or sides of the ship's hull. The Romans solved the problem putting an inner layer of planks over the ship's ribbing, on which the grain sacks were stored. One imagines the Persians had a similar solution.

[Prufrock]

What is interesting is that the primary sources give huge figures for all Persian armies, and not just Persian armies but also armies of other nations of the Fertile Crescent during that era. So it's not just one figure by one writer that is being discounted. If the sources are so unreliable as to make an army five or ten times larger than it really was, then they can't be trusted in anything else they assert. And that's the end of history.

Not quite – there are numerous () reasons why numbers can be wrong without meaning that the rest needs to be chucked out.

Peter Green (the historian, not the guitarist) in his Greco-Persian Wars page 62 gives some of them for this particular instance (basically, Xerxes had every reason to exaggerate the numbers, while the Greeks after they won had no reason to revise them downwards, and that the numbers were for before the entrance into Europe).

Thus far in this thread I haven't seen any killer argument - backed up with proof - that demonstrates the impossibility of such an enormous army marching, camping and being fed and watered from the Hellespont to Greece.

You probably have, but choose not to see it as such

Take imperial Rome. The city needed 420 000 tonnes of grain brought by ship each year. If the average Roman merchant ship had a carrying capacity of 100 tons that means 4200 ship voyages to feed the city. It takes two days by sea from Carthage to Rome with a favourable wind and four days with an unfavourable one, so a round trip including loading time, repairs and so on probably took about 10 days. The Mediterranean was navigable for 8 months of the year so a ship could make the trip about 24 times at best which means at least 175 ships dedicated just to bringing grain to Rome, year after year. But the Roman Empire did it, free of charge. What is the big obstacle for the Persian Empire doing it as a once-off, with years of preparation beforehand?

Notes to the Landmark Herodotus (p.577) translating Herodotus' grain figures into modern quantities indicate the army required 4,700 tons per day (470,000 for 100 days; 1,715,500 for a year). These supplies are coming through beaches, in potentially unfriendly territory, and without dedicated unloading crews at the drop off point, and without troops in place to secure depots in advance (if advance depots is part of the argument) so it doesn't really compare to Imperial Rome's 420,000 tons coming through a purpose-built port with organized shipping and regular crews at both ends.

There is also the issue of choke points as mentioned by pretty much everyone. H 7.176 mentions the route narrowing to a space the width of a wagon in two places. Try getting 1,700,000 fighting men let alone the other 3.5M plus  followers, cavalry horses, baggage animals, food on the hoof, and baggage carts through spaces that narrow in timely fashion!

There are of course other objections too, but one that I don't think has been noted in this thread so far is the 'shock and awe factor' that 10 white horses, a chariot, 10 more white horses, 1000 picked cavalrymen, 1000 picked foot, and the 10,000 Immortals elicits in 7.40-41 & 7.55. If the army was really 1,700,000 strong, would so much be made of these comparatively small elite contingents?

[aligern]

No no Justin, you just cannot go equating all statements in an ancient source as having the same value.
If we took accounts of WW2 a contemporary souce of the Battle of Britain could give excellent information on the types and capabilities of the aircraft, worse qualuty of information on the numbers taking part, poor information on the effects of the bombing and even worse on the numbers shot down. The source of the information might be someone in the propaganda dept of one of the combatants, German, British or Italian and have an ulterior motive to increase or decrease numbers.
If 2000 years later someone was looking at the memoirs of an Italian pilot who took part, writing a report to his bosses and that was our only source then  should  the historians of the future just accept the numbers or that the CR42 was a good fighter because it had great visibility and a tight turning circle and coukd outfight a Gloster Gladiator. Of course someone could construct a logic that the Gladiator fought in the Battle of Britain because a few did, because the RAF clearly had them, because one was dug up in Iraq or Egypt and dated to 1941.
Ancient sources need to be looked at in their own context and understood for their own motivations. If later armies all seem to be smaller, far smaller and yet be the product of states that we know to have the economic power to support an army equal to any Persian then it is extremely unlikely that Persia deployed forces that were much greater in size.  In theiry Rome in the Late Empure could provide 650, 000 troops ( I think that's Agathias) so why do they not put out armies of halh a million? Well they have frontiers to defend, internal security to attend to and the ligisics would have been a nightmare. Please don't give comparative numbers from the Bible....its just not that sort of historical document.
The Greeks had a very good motive for consistently exaggerating the size of Persian armies. As had Caesar for exaggerating the numbers of Gauls or Tacitus for bigging up the numbers of Ancient Brits. The Greeks  were beating the Persians, but at the same time had a poor opinion of Persian organisation, motivation, equipment, morale...the lot. How could an army of such people be terrifying enough to be an heroic opponent...why, by giving them huge numbers of slave soldiers, formidable by their mere mass, despicable in equipment and freedom.
Such a huge army would just starve when the Greeks took control of the seas. It would be effectively immobilised by the problems of food distribution. The continuation of this argument relies, as Justin points out, on the difficulty of proving a negative......its more like religion than history. Can we prove that Moses didn't see a burning bush?? That pharaoh's was not drowned under the returning waves?


Ancient sources should be

[Erpingham]

Incidentally the most often-cited numerical analysis of Xerxes' supply requirements these days, more recent than Maurice, is probably T Cuyler Young, "480/479 BC - A Persian Perspective" (Iranica Antiqua XV, 1980).

Tuplin's paper, which I linked earlier, is ruthless in his critique of Young.  He notes, for instance, that Young forgets the fodder requirements of his animals.  Even so, he believes Young is too deterministic and believes Maurice's estimate better.

Are beaches an insuperable problem?

No, as long as the force is of a manageable size. However, feeding a army several times the population of Rome across an improvised beach landing on a hand-to-mouth basis while moving the entire supply operation every few days is a non-starter.  Then there is the weather.  You don't need a full on storm to disrupt a beach supply, just a heavy swell.  Your ships can't be sure they beach and get off safely, so they will have to anchor and wait.  Not all the beaches will be sheltered and, if the weather gets worse they may have to run for cover. Unlike Rome, you have no big granaries to buffer you against losing even a few days supply. 

[Justin Swanton]

What is interesting is that the primary sources give huge figures for all Persian armies, and not just Persian armies but also armies of other nations of the Fertile Crescent during that era. So it's not just one figure by one writer that is being discounted. If the sources are so unreliable as to make an army five or ten times larger than it really was, then they can't be trusted in anything else they assert. And that's the end of history.

Not quite – there are numerous () reasons why numbers can be wrong without meaning that the rest needs to be chucked out.

Peter Green (the historian, not the guitarist) in his Greco-Persian Wars page 62 gives some of them for this particular instance (basically, Xerxes had every reason to exaggerate the numbers, while the Greeks after they won had no reason to revise them downwards, and that the numbers were for before the entrance into Europe).

Thinking about the propaganda angle, wouldn't the Greek writers have just as much reason to boast about beating a 200 000 man Persian army as beating a 3 400 000 man one? Propaganda works if it is believable to those for whom it is destined. People in that era would have known that the Persians could not possibly field armies of several million men if in fact they didn't. If 200 000 was in fact the upper limit for a Persian army then Herodotus' contemporaries would have laughed at his figures. Propaganda exercise flops.

As an example, if I told you that 80 Rhodesian soldiers utterly defeated an enemy camp of 5000 guerrillas all armed to the teeth with the latest in Russian and Chinese military hardware, killing over 1000 of them, would you think that a propaganda exercise a la Herodotus?

Thus far in this thread I haven't seen any killer argument - backed up with proof - that demonstrates the impossibility of such an enormous army marching, camping and being fed and watered from the Hellespont to Greece.

You probably have, but choose not to see it as such

I'm all ears (or eyes in this case). 

Take imperial Rome. The city needed 420 000 tonnes of grain brought by ship each year. If the average Roman merchant ship had a carrying capacity of 100 tons that means 4200 ship voyages to feed the city. It takes two days by sea from Carthage to Rome with a favourable wind and four days with an unfavourable one, so a round trip including loading time, repairs and so on probably took about 10 days. The Mediterranean was navigable for 8 months of the year so a ship could make the trip about 24 times at best which means at least 175 ships dedicated just to bringing grain to Rome, year after year. But the Roman Empire did it, free of charge. What is the big obstacle for the Persian Empire doing it as a once-off, with years of preparation beforehand?

Notes to the Landmark Herodotus (p.577) translating Herodotus' grain figures into modern quantities indicate the army required 4,700 tons per day (470,000 for 100 days; 1,715,500 for a year). These supplies are coming through beaches, in potentially unfriendly territory, and without dedicated unloading crews at the drop off point, and without troops in place to secure depots in advance (if advance depots is part of the argument) so it doesn't really compare to Imperial Rome's 420,000 tons coming through a purpose-built port with organized shipping and regular crews at both ends.

One needs to look at this in terms of ships. 4700 tons per day means 47 ships with a carrying capacity of 100 tons or 16 ships with a carrying capacity of 300 tons. Choose a middle figure and say 30 ships that must offload each day or 3 ships an hour. Not actually such a big deal.

There is also the issue of choke points as mentioned by pretty much everyone. H 7.176 mentions the route narrowing to a space the width of a wagon in two places. Try getting 1,700,000 fighting men let alone the other 3.5M plus  followers, cavalry horses, baggage animals, food on the hoof, and baggage carts through spaces that narrow in timely fashion!

Herodotus mentions the chokepoint being at Trachis, which is just before Thermopolae. It is there precisely that Xerxes' problems began, not during the trip from the Hellespont to Greece.

There are of course other objections too, but one that I don't think has been noted in this thread so far is the 'shock and awe factor' that 10 white horses, a chariot, 10 more white horses, 1000 picked cavalrymen, 1000 picked foot, and the 10,000 Immortals elicits in 7.40-41 & 7.55. If the army was really 1,700,000 strong, would so much be made of these comparatively small elite contingents?

Why mention 10 white horses for an army of 200 000 for that matter? The context of the passage is important. Pythius the Lydian had asked Xerxes to release one of his five sons from the army to stay with him at home. Xerxes, furious, has the son executed and "set one half of his body on the right side of the road and the other on the left, so that the army would pass between them." It is not stated how wide the 'road' is, nor if it is the entire army or just a contingent of it that passes by. The army is in three sections: hoi-polloi come first, then a gap, then the king with his elite troops, then a gap, then more hoi-polloi. The mention of the white horses etc. is clearly meant to underscore the magnificence of Xerxes.

[Patrick Waterson]

I mentioned amphorae in connection with grain transport.  Whether they were used for storage is conjectural, but it would be interesting to consider what effect this might have on the keeping qualities of their contents.

Evidence would be nice, sources and stuff
Have you any evidence for amphorae being used to transport grain?

Justin has turned up something; I did a search for wrecks but it seems search engines have recently been reconfigured to avoid historical information in favour of contemporary advertising; Lucian in The Ship, or The Wishes describes a Roman period grain ship (the Isis) but not her cargo.

I say, though, what a size that ship was! 180 feet long, the man said, and something over a quarter of that in width; and from deck to keel, the maximum depth, through the hold, 44 feet. And then the height of the mast, with its huge yard; and what a forestay it takes to hold it! And the lofty stern with its gradual curve, and its gilded beak, balanced at the other end by the long rising sweep of the prow, and the figures of her name-goddess, Isis, on either side. As to the other ornamental details, the paintings and the scarlet topsail, I was more struck by the anchors, and the capstans and windlasses, and the stern cabins. The crew was like a small army. And they were saying she carried as much corn as would feed every soul in Attica for a year. And all depends for its safety on one little old atomy of a man, who controls that great rudder with a mere broomstick of a tiller!

So for the present I am left with: if you were moving grain during the Persian-Greek-Roman period, how would you get it on board, keep it dry during transit and unload it cleanly and expeditiously with a minimum of overall wastage?  And I think: amphorae.  They are waterproof (if you seal them), they are easy to load and stow, they should not shift around from waves lapping at the ship (you do have sand and shingle ballast in the hold, right?), they are rat-proof, and at the terminus you can unload them easily enough with just your crew and tackle if help is not forthcoming.  If for some reason you have multiple destinations you can unload exactly as much as is needed at each destination.  When it arrives, it is all date-stamped and certified for origin.

In fact, thinking about it, how can you not transport it in amphorae?

This is also the conclusion of Wikipedia, sourcing Adkins, L.; Adkins, R.A. (1994). Handbook to life in Ancient Rome.  To quote:

"In the Bronze and Iron Ages amphorae spread around the ancient Mediterranean world, being used by the ancient Greeks and Romans as the principal means for transporting and storing grapes, olive oil, wine, oil, olives, grain, fish, and other commodities. They were produced on an industrial scale until approximately the 7th century AD. Wooden and skin containers seem to have supplanted amphorae thereafter."

I guess these gentlemen know their stuff.  That is pretty much all I have been able to turn up so far.

Storage is the next question: Herodotus in III.6 mentions the Egyptians collecting all the wine-jars (amphorae) which arrive in their country and using them as the basis of a desert water supply scheme.  He mentions no such arrangement for grain jars (amphorae), implying they continued in circulation.  If one were stocking up for a campaign, it would make sense to keep the grain in its original amphorae, where you not only have the date and origin information on the seal but also need not bother about rodent infestation and can shift precise quantities without a lot of fuss and shovelling.  This would mean that at the end of the campaign you are looking at a clean, empty floor in the storehouse and any bits of unlucky amphorae are probably on a windswept beach somewhere.