A note regarding the Population Detail Spreadsheet (see below). From my long-past classes in social geography, and many game supplements, cities arise as trading points at the center of several smaller entities. Since the Perilous lands does not have a notation of "small or large" cities -- just "cities" -- I am using towns to fill this void. From what I have seen of the cultures so far, the urban ration appears to range from .15 to .18 -- that is, urban population is abut 15% to 18% of the total population. This means that the remaining 85% to 82% is involved in agriculture, herding or fishing. I Figure towns to be about 1/2 services and 1/2 support, with villages nearly 100% support.
How does this come into game play? Well, when attempting to purchase "city goods" at a town cut all availability by 50% (thus 80% becomes 30%). The purchase of "city goods" in a village has availability cut by 80%. Likewise, the purchase of "Town" goods in the city should be cut by 25% and "village" goods by 50%.
What are "city goods"? Basically, what the GM considers to be luxury items or those items requiring a large selling base, like heavy metal armors, ships larger than a small boat (unless a shipyard is nearby). "Village" goods are those that are usually raw materials -- foodstuffs, raw hides, ore, wood, etc. "Town" goods are those that involve one step of production above raw goods (i.e. Cloth, from raw wool or cotton, leather goods, from raw hides, etc), or where the product is too large for a single village (i.e. raising horses).
All of these levels are subject to GM use, of course. If the players stumble out of the wilderness with battered armor and attempt to find it in the town of "Backwater", they may be out of luck unless they are willing to settle for leather armor.
The following concerns an Excel sheet I made.
Basically, you supply the city populations and number of forts and it chews through a bunch of assumptions to determine the number of towns and villages. I have tried to tweak it based on Ma'helas (yea, one data point :) so it may produce strange numbers for other cultures (i.e. Katai). There are four "magic" numbers that may be tweaked to fine-tune the sheet: "Magic T" affects the number of support towns per city; "Magic Tp" affects the average population of each town; "Magic Vp" affects the average population of each village; and "Magic CTS" affects the "City/Town Scale" which adjusts the average town size based on the smallest city size.
Feel free to use this tool and play with it. If you find magic numbers that make a more "realistic" fit, please let me know.