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

[Jim Webster]

The surplus grain was exported to areas outside the Persian Empire, so the hungry citizens remain fed with the domestic harvest. The buyers from outside Persian borders have to find their grain from elsewhere, it's not Xerxes' problem.

Xerxes forbids the export of the surplus grain and buys it himself or takes it as a tax, along with the extra grain he orders to be grown for the campaign.

that's the problem with having a really big empire. Apart from a handful of Greeks, some of whom may have Medized to guarantee their grain supplies, the grain is bound for within the Empire.
If hungry citizens could be fed with the domestic harvest, then they wouldn't buy grain from Egypt in the first place.
So there's still no surplus grain. He's got to order more and find a way of increasing the manpower and farmable land of Egypt by a considerable percent

I'm working on the assumption that most of Egyptian surplus grain was not destined for the Persian Empire but for the Greek cities and other customers on the Mediterranean littoral. But do we have source material on this?

Even presuming Egypt didn't export a single bushel beyond the Empire's borders, at most it is being asked to supply something like 7,5 percent its annual harvest of a single year. Is it asking too much from Egyptian farmers to farm 1,9% more land over 4 years?

Actually if the population of Egypt is 5 million, (7.5 million is Ptolemaic, at least 300 years later than our date and after considerable immigration and agricultural improvement) and you're asking them to feed six million for 6 months. So I suggest my figure of 50% increase is more accurate.
So you have the four years.
In year one I suppose Xerxes could have the Nile flogged to teach it a lesson and make sure the next three Niles are higher and allow more land to be cultivated. But other than that year 1 is going to consist of planning the major irrigation works you need to put a minimum of 15% more land into cultivation, and finding the labour to work it.
You're looking at a project on the same sort of scale as the Ptolemaic work on the Faiyum

[Jim Webster]

no, it's not a 100kg. It starts off at 90 and gets down to 80 by the time you deduct saddle and kit. That's why I ended up with 28 days before you run out

Fine, so 2/3 of the load reaches its destination. Bear in mind that the handlers would have eaten half that anyway.

Work also on the assumption that the grain would be stored in amphorae but carried in sacks.

you mean it'd be put in all these amphorae apparently necessary for sea transport when it got to the docks? 

[Justin Swanton]

The surplus grain was exported to areas outside the Persian Empire, so the hungry citizens remain fed with the domestic harvest. The buyers from outside Persian borders have to find their grain from elsewhere, it's not Xerxes' problem.

Xerxes forbids the export of the surplus grain and buys it himself or takes it as a tax, along with the extra grain he orders to be grown for the campaign.

that's the problem with having a really big empire. Apart from a handful of Greeks, some of whom may have Medized to guarantee their grain supplies, the grain is bound for within the Empire.
If hungry citizens could be fed with the domestic harvest, then they wouldn't buy grain from Egypt in the first place.
So there's still no surplus grain. He's got to order more and find a way of increasing the manpower and farmable land of Egypt by a considerable percent

I'm working on the assumption that most of Egyptian surplus grain was not destined for the Persian Empire but for the Greek cities and other customers on the Mediterranean littoral. But do we have source material on this?

Even presuming Egypt didn't export a single bushel beyond the Empire's borders, at most it is being asked to supply something like 7,5 percent its annual harvest of a single year. Is it asking too much from Egyptian farmers to farm 1,9% more land over 4 years?

Actually if the population of Egypt is 5 million, (7.5 million is Ptolemaic, at least 300 years later than our date and after considerable immigration and agricultural improvement) and you're asking them to feed six million for 6 months. So I suggest my figure of 50% increase is more accurate.
So you have the four years.
In year one I suppose Xerxes could have the Nile flogged to teach it a lesson and make sure the next three Niles are higher and allow more land to be cultivated. But other than that year 1 is going to consist of planning the major irrigation works you need to put a minimum of 15% more land into cultivation, and finding the labour to work it.
You're looking at a project on the same sort of scale as the Ptolemaic work on the Faiyum

Fine, let's take 5 million as the population of Egypt and 6 million tops as the size of Xerxes' army and allow 6 months for the campaign, working on the fair assumption that Xerxes intended to strip Greece of grain to help supplement supplies for the return journey. We also assume Egypt supplied half the required 88% grain - generous estimate - needed the feed the army (the 12% comes from the Greek hosts on the army's route).

So Egypt needs to come up with 44% of the grain. That's the size of army (6 million) ÷ size of Egyptian population (5 million) ÷ 2 (half a year's supplies) ÷ 4 (4 year's preparation) x 44% (percentage Egypt must supply) = 6,6% of the annual harvest. This is the absolute worst case scenario.

I suggest that Xerxes doesn't spend year one flogging the Nile but pays for extra labourers to dig irrigation channels and plant the grain without wasting time - if he can afford to dig a canal for his fleet he can afford this. The labourers stay on the job for the next 4 years and gather in the extra harvest. They can then stay and make Egypt a larger exporter of grain than it originally was or they can go home.

Bearing in mind you need to cultivate at most another 6,6% of the land. Assuming 90% of the population of 5 million work the land that's 297 000 extra labourers committed to the job. Not inconceivable.

[Justin Swanton]

no, it's not a 100kg. It starts off at 90 and gets down to 80 by the time you deduct saddle and kit. That's why I ended up with 28 days before you run out

Fine, so 2/3 of the load reaches its destination. Bear in mind that the handlers would have eaten half that anyway.

Work also on the assumption that the grain would be stored in amphorae but carried in sacks.

you mean it'd be put in all these amphorae apparently necessary for sea transport when it got to the docks? 

Only keep the grain in amphorae when in storage or when shipping during the 4 years' preparation (when speed of transport is not of the essence). When it is being carried by mules or shipped to the army on the march keep it in sacks.

[Justin Swanton]

As some sort of indicator of the kind of money the Persian Empire could have disposed of, the Seleucid Empire was required to pay Rome and indemnity 3000 talents on the nail and 2000 talents a year for 12 years thereafter. This had to come from the surplus of Seleucid economic production.

Presuming the Persian army cost 400 talents a day to feed (based on the Thasian incident), 4 months on campaign less 12% from local Greek cities would cost 105 days x 400 talents = 42 000 talents in all, or 10 500 a year presuming 4 years' stockpiling. A comparable figure to the Seleucids keeping mind all this money was destined for internal use, not to be given away without any recompense.

[Erpingham]

I note that we still consider the grain was carried by in amphorae.  I continue to be troubled by this, as there seems very little evidence to support this.  Most studies seem to distinguish between amphora cargo and bulk grain cargo.  For example, David Gibbins writes this in a paper called Shipwrecks and Hellenistic Trade

As in the mediaeval period, grain was most efficiently carried in large, dedicated sitegoi, the grain being laden in sacks or a rifuso directly into the hold

(sitegoi are grain ships, a rifuso is the process of overloading the contracted amount, to allow for spoilage)  Hellenistic and classical sitegoi seem to have been smaller than their Roman descendents - 100-400 tonnes seem to crop up.

[Justin Swanton]

I note that we still consider the grain was carried by in amphorae.  I continue to be troubled by this, as there seems very little evidence to support this.  Most studies seem to distinguish between amphora cargo and bulk grain cargo.  For example, David Gibbins writes this in a paper called Shipwrecks and Hellenistic Trade

As in the mediaeval period, grain was most efficiently carried in large, dedicated sitegoi, the grain being laden in sacks or a rifuso directly into the hold

(sitegoi are grain ships, a rifuso is the process of overloading the contracted amount, to allow for spoilage)  Hellenistic and classical sitegoi seem to have been smaller than their Roman descendents - 100-400 tonnes seem to crop up.

Very good. The grain would spend only a few days in the hold of a ship so amphorae would probably not be required. But I suggest that for long-term storage, at least on the Aegean coastline, sealed amphorae (as used for olive oil or wine) would be necessary.

[Erpingham]

Someone (Justin?) was asking about Byzantine figures

The approximate maximum weight a horse or mule can transport over reasonably long distances is about 250 1b (114kg) and a little more over short stretches, although the optimum has generally been set at about 200lb in modern and immediately pre-modern pack-trains. In the late third-century Edict of Diocletian (14.11) a load of 200 Roman pounds (65.49 kg/ 144 1b) is prescribed; a sixth-century source gives mules a total burden of 156—66 Roman pounds (110—16 1b/ 50—3 kg). Similar limits are established by the imperial legislation on the public post. A mid-tenth-century Byzantine text gives somewhat higher values ...... : three categories of load are specified: (a) saddle—horses carrying a man (presumably not armoured and carrying military panoply) and their own barley were loaded with four modioi each -106 Roman pounds=75 1b (34 kg); (b) unridden saddle-horses carried eight modioi—212 Roman pounds—I50 1b (68 kg); and (c) pack-animals loaded with barley carried ten modioi—265 Roman pounds= 187 1b (85 kg). Thus the maximum permitted load for an animal in the imperial baggage train in the ninth and tenth centuries was set at 10 modioi without the pack- saddle (stigma) and harness which, according to the legislation of the fourth sixth centuries, weighed approx. 50—60 Roman pounds (35—42 1b/ 16—19 kg, equivalent to 51—62 Byzantine pounds).


From Appendix 1 in Warfare, State and Society in the Byzantine World, 565-1204 by John F. Haldon.  Most of the Appendix can be viewed via Google. Appendix 2 is on using grain to feed armies but alas most of it is inaccessible.  I suspect some members of the forum own the book and may be able to help.

[Jim Webster]

Fine, let's take 5 million as the population of Egypt and 6 million tops as the size of Xerxes' army and allow 6 months for the campaign, working on the fair assumption that Xerxes intended to strip Greece of grain to help supplement supplies for the return journey. We also assume Egypt supplied half the required 88% grain - generous estimate - needed the feed the army (the 12% comes from the Greek hosts on the army's route).

So Egypt needs to come up with 44% of the grain. That's the size of army (6 million) ÷ size of Egyptian population (5 million) ÷ 2 (half a year's supplies) ÷ 4 (4 year's preparation) x 44% (percentage Egypt must supply) = 6,6% of the annual harvest. This is the absolute worst case scenario.

I suggest that Xerxes doesn't spend year one flogging the Nile but pays for extra labourers to dig irrigation channels and plant the grain without wasting time - if he can afford to dig a canal for his fleet he can afford this. The labourers stay on the job for the next 4 years and gather in the extra harvest. They can then stay and make Egypt a larger exporter of grain than it originally was or they can go home.

Bearing in mind you need to cultivate at most another 6,6% of the land. Assuming 90% of the population of 5 million work the land that's 297 000 extra labourers committed to the job. Not inconceivable.

1) Why would Xerxes intend to strip an already grain deficient area of grain? He couldn't guarantee it, it could go in the flames of the burning cities. Anything could happen to it. But still I said 6 months for easy reckoning. But given that Greece had a population of 12.5 million or thereabouts, and by the time he'd conquered it they could have eaten half their harvest anyway, so even if they handed over everything, there might not be enough grain to feed them. After all with 6 million people coming in, he's increasing the population of Greece by 50%

2a) So what other countries are supplying the extra grain? Remember they have to be coastal or riverine. There aren't really a lot of areas that satisfy these conditions. He can hardly haul it from Babylonia and Asia Minor isn't a major grain exporting area. Neither are the Phoenician cities and Northern Syria only really took off under the Seleucids. So where did it come from?

3) The first year of the four years isn't going to produce anything extra. The size of the crop is fixed by the time Xerxes sends his messengers. The first year is spent in creating new water basins, digging new channels and that sort of thing. It's also spent frantically trying to find the extra labour force. The Egyptian labour force for building was available in the high water time because there was damn all else they could do. Unfortunately for Xerxes he needs it in the low water time when they're at their busiest doing agricultural stuff. So the year he actually does all this digging and suchlike, he still doesn't get extra yield because his farm workers are not doing farm work, they're doing engineering.

4) this 6.6% extra land of course depends on you finding somewhere else which is just as good at producing grain for export as Egypt. When you've come up with this place, then we can go with 6.6, otherwise it's over 12%

5) Your 300,000 (or 600,000 if we're on 12%) labourers. On the assumption that they'll all kept pretty busy just doing agricultural things, what do you want them to stop doing so they can start breaking land into production?

6) This is an absolute worst case scenario. No it isn't, you might get a low or a high Nile. Probably one in five is too low or too high, one being as bad as the other. So in a bad year, you're feeding Egyptians out of stocks.

[Jim Webster]

As some sort of indicator of the kind of money the Persian Empire could have disposed of, the Seleucid Empire was required to pay Rome and indemnity 3000 talents on the nail and 2000 talents a year for 12 years thereafter. This had to come from the surplus of Seleucid economic production.

Presuming the Persian army cost 400 talents a day to feed (based on the Thasian incident), 4 months on campaign less 12% from local Greek cities would cost 105 days x 400 talents = 42 000 talents in all, or 10 500 a year presuming 4 years' stockpiling. A comparable figure to the Seleucids keeping mind all this money was destined for internal use, not to be given away without any recompense.

You walk into a grain market with money and try to buy when there's no grain to sell. Waving the cheque book around is no good. It was Alexander the Great who poured cash out in front of horses to point out to the Satrap who'd sent it that he needed the grain not the money.
The difficult bit isn't producing the money, it's producing the grain

[Jim Webster]

I note that we still consider the grain was carried by in amphorae.  I continue to be troubled by this, as there seems very little evidence to support this.  Most studies seem to distinguish between amphora cargo and bulk grain cargo.  For example, David Gibbins writes this in a paper called Shipwrecks and Hellenistic Trade

As in the mediaeval period, grain was most efficiently carried in large, dedicated sitegoi, the grain being laden in sacks or a rifuso directly into the hold

(sitegoi are grain ships, a rifuso is the process of overloading the contracted amount, to allow for spoilage)  Hellenistic and classical sitegoi seem to have been smaller than their Roman descendents - 100-400 tonnes seem to crop up.

Very good. The grain would spend only a few days in the hold of a ship so amphorae would probably not be required. But I suggest that for long-term storage, at least on the Aegean coastline, sealed amphorae (as used for olive oil or wine) would be necessary.

I think I calculated that something like 88 million of them would be needed. We should have found the traces of breakages archaeologically by now
Anyway, storing grain or any other crop in anaerobic conditions doesn't necessarily leave it fit for human consumption. You can end up with anaerobic digestion where anaerobic bacteria attack it, they produce lactic acid as a byproduct and the grain ends up pickled in lactic acid. It's called silage, I make it every year 

[Flaminpig0]

I note that we still consider the grain was carried by in amphorae.  I continue to be troubled by this, as there seems very little evidence to support this.  Most studies seem to distinguish between amphora cargo and bulk grain cargo.  For example, David Gibbins writes this in a paper called Shipwrecks and Hellenistic Trade

As in the mediaeval period, grain was most efficiently carried in large, dedicated sitegoi, the grain being laden in sacks or a rifuso directly into the hold

(sitegoi are grain ships, a rifuso is the process of overloading the contracted amount, to allow for spoilage)  Hellenistic and classical sitegoi seem to have been smaller than their Roman descendents - 100-400 tonnes seem to crop up.

Very good. The grain would spend only a few days in the hold of a ship so amphorae would probably not be required. But I suggest that for long-term storage, at least on the Aegean coastline, sealed amphorae (as used for olive oil or wine) would be necessary.

I think I calculated that something like 88 million of them would be needed. We should have found the traces of breakages archaeologically by now
Anyway, storing grain or any other crop in anaerobic conditions doesn't necessarily leave it fit for human consumption. You can end up with anaerobic digestion where anaerobic bacteria attack it, they produce lactic acid as a byproduct and the grain ends up pickled in lactic acid. It's called silage, I make it every year 

Another point is would the Persian Empire have the ability to radically increase the production of amphorae to the level that Justin's argument is now dependent on- I am assuming that no one is going to argue that the things were produced in Henry Ford style factories as part of a centralised command economy?

[Justin Swanton]

I note that we still consider the grain was carried by in amphorae.  I continue to be troubled by this, as there seems very little evidence to support this.  Most studies seem to distinguish between amphora cargo and bulk grain cargo.  For example, David Gibbins writes this in a paper called Shipwrecks and Hellenistic Trade

As in the mediaeval period, grain was most efficiently carried in large, dedicated sitegoi, the grain being laden in sacks or a rifuso directly into the hold

(sitegoi are grain ships, a rifuso is the process of overloading the contracted amount, to allow for spoilage)  Hellenistic and classical sitegoi seem to have been smaller than their Roman descendents - 100-400 tonnes seem to crop up.

Very good. The grain would spend only a few days in the hold of a ship so amphorae would probably not be required. But I suggest that for long-term storage, at least on the Aegean coastline, sealed amphorae (as used for olive oil or wine) would be necessary.

I think I calculated that something like 88 million of them would be needed. We should have found the traces of breakages archaeologically by now
Anyway, storing grain or any other crop in anaerobic conditions doesn't necessarily leave it fit for human consumption. You can end up with anaerobic digestion where anaerobic bacteria attack it, they produce lactic acid as a byproduct and the grain ends up pickled in lactic acid. It's called silage, I make it every year 

Keeping in mind that everyone, everywhere, has to be able to store grain in such a way that it keeps for up to 12 months. This applies in a wet or dry climate. I suggest that grain that is meant to last several years (which is not most of the grain) care must be taken to keep it dry. This may require sealed amphorae or it may be enough to keep it in rooms or pits specially constructed to keep humidity at a minimum, I don't know, I'm not a farmer. I suspect though that if the BC folks could preserve their grain for 12 months they could preserve it for considerably longer.

[Jim Webster]

Keeping in mind that everyone, everywhere, has to be able to store grain in such a way that it keeps for up to 12 months. This applies in a wet or dry climate. I suggest that grain that is meant to last several years (which is not most of the grain) care must be taken to keep it dry. This may require sealed amphorae or it may be enough to keep it in rooms or pits specially constructed to keep humidity at a minimum, I don't know, I'm not a farmer. I suspect though that if the BC folks could preserve their grain for 12 months they could preserve it for considerably longer.

The only way to do it for multiple years (assuming you cannot dig pits six or more meters down into seriously cold ground) is to have active storage. Firstly you get rid of the oldest first anyway. But secondly you'll have bins and inspect each bin regularly.
Inspection a bin means going into it and looking for infestations. The minute you find one you've got to be prepared to empty the bin (on the grounds that these people cannot flush it with C02 or similar) and use it immediately.
Secondly your inspector will doubtless have a couple of slaves with in him the bin and they'll dig down and that way you check if it's overheating. If it is, empty it immediately, spread it out thin on a hard surface and re-dry it. Then put it back in a different, dry and fumigated bin while the one it was in is cleaned and fumigated.

The alternative is opening the door after four years and going "Oh look, a bit full of uneatable rubbish"

[Erpingham]

I think I calculated that something like 88 million of them would be needed.

Amphora studies are an entire sub-set of archaeology - I'm not going there   However, a quick google turned up the single reference to amphora production rates in Egyptian papyri - two men could make 30 amphorae in four days.  Say four per man per day.  So 22 million man days work.  I'll let Justin create a scheme for organisaing the labour, recalling that the work was often seasonal (the potters were also farmers) and that it would have to be fitted in alongside regular production for wine, olive oil and perhaps fish industries (I don't know at what point salt fish became a big export industry).