Hi David!
That's an interesting problem! Here I have some information found in 'Der kleine
Pauly, Lexikon der Antike, 1979' (translation by me):
'The ancient or modern trials to calculate the population of ancient territories are by criterias of scientific statistics insufficient.
...
The problem of the population of
Rome first was cleared since 1940 by v.Gerkan by reference to the archaeologic-topographic basis claimed already by H.G.Wells. According to him
Rome could have in the time of the Late Republic never more than 624.000 inhabitants, the Imperial
Rome never more than 698.000 inhabitants. The hellenistic-roman towns of the East were smaller, with exception of
Alexandria perhaps. The population of the
Imperium Romanum is estimated to be 50 to 70 Millions. In the late
antiquity the population of
Rome goes down to 15.000 in the 6.century.
Constantinopolis rised at the same time to ca.600.000 under
Justinian I, reaching 1 Million never before 1453.'
I've read literature about calculating the population of
Rome. It starts with the building of the 'insulae' (the houses, in which most of the
Roman people live!), how high they are, howmany qm surface
area they have, how many persons could have inhabitant such 'insula', how great was the
area of the ancient
Rome, how big was the built-up
area of
Rome, how great was the building density, how was the distribution of the building density of this
area and so on. You have to take into account knowledge of the infrastructure, f.e. you have to calculate the quantity of water coming through the aquaducts to
Rome and so on and so on.
There are so many unknown quantities in this calculation that each quantifying should be seen cum grano salis.
Regards,
Jochen