SUPPLYING THE ROMAN EMPIRE 115judged either much too low or a little too high. They are the former if cavalry
received twice as much wheat as infantry, as they did in the second century
BC according to Polybius (except that citizen cavalry received three times as
much as infantry) (6.39.13).^10
The fi gure of 100,000 (or 150,000) tonnes will be too high if the estimated
daily grain ration of one kg is too high. This fi gure for consumption (there
is no other from the imperial period) is derived from late Egyptian papyri
which show soldiers receiving bread rations at three Roman lbs or around
one kg per day. The same soldiers were given two Roman lbs of meat, two
pints of wine and^1 / 8 of a pint of oil. Jones’ word for this diet was ‘positively
gargantuan’. In fact, the fi gure for cereal consumption, taken in isolation, is
only marginally higher than Polybius’^2 / 3 medimnos of wheat per month,
equal to 4 modii, or about 27 kg, for Roman and allied infantry in the
second century BC. This is in turn approximately the same as what had been
termed a standard Greek military ration of one choenix of wheat per day in
the classical period. The chained slaves of Cato’s second- century- BC treatise
On Agriculture ate much more, 4 or 5 lbs of bread per day, but they
consumed little else ( de agr. 56).^11
One hundred thousand tonnes of grain will do as a rough estimate of the
cereal consumption of the army under Augustus, rising to 150,000 tonnes
under Septimius Severus.
State organization of military supply has consequences for the geo-
graphical range over which supplies were sought, the status of the goods
that found their way to the camp and the methods by which they were
brought.
Most supplies were local in origin. That is, the military provinces and the
areas adjacent to them took the lion’s share of the burden of army supply.
Commodities judged essential or in considerable demand were sought from
further afi eld only insofar as they could not be obtained, or not in suffi cient
quantity, near the army base. In this category should be placed the Spanish
olive oil that found its way to the Rhineland and Britain. In general, long-
distance transport was expensive and ineffi cient. Even if governments had
been prepared to pay the cost, they would have been intolerant of the
ineffi ciency involved, when keeping army regiments contented and on a war
footing was at issue.
Living off the zone or region of occupation was not always a practical
proposition. Strategic rather than economic considerations were sometimes
paramount in the choice of site for an army base, as is obvious from the way
in which the army was deployed in Syria, north Africa or Britain. Again, the
army demand for food, raw materials and a wide variety of manufactured
goods simply could not always be met locally, especially in the early days of
conquest and pacifi cation, even if the tribes and communities involved had
been hospitable to a sizeable occupying force – an argument more applicable
to the northern frontier regions than to the more developed East. As
pacifi cation gave way to peaceful coexistence with the local population, the