In a recent post about the theory of agricultural location, I missed out the definition of 'P' in Dunn's equation...thanks Mi for pointing this out. P is the price of the commodity (e.g. wheat) at the market. I've corrected the omission. Mi's remark made me think about the variation in wheat prices on a locality basis. If the wheat price varies from point to point, then Dunn's equation gets more complicated. I've found a remarkable set of wheat prices by county which last from 1770 1820. I have been looking at these to see what regional difference there might be. On the graph I've plotted the yearly mean price for our four counties of interest. Interesting things: the wheat prices converge when the variation is small. But when there are peaks (for example a huge jump in 1800) then there is considerable divergence. And it is always Somerset that is most expensive. Cornwall is always cheapest, with the two counties between them geographically, also between them with regard to price. I'd guess that this effect is caused by proximity to London. London was a major market, with about 14% of the population living there. This is really interesting (isn't it?).