Fall-line cities

Why do so many old cities sit exactly at the point where a river stops being navigable?

Draw the eastern United States and mark Trenton, Philadelphia, Baltimore, Washington, Richmond, Petersburg, Raleigh, Columbia, Augusta, Macon, Columbus. They sit on different rivers, were founded by different people at different dates for different stated reasons — and they fall very nearly on a single line running southwest from New Jersey to Georgia.

Coincidence is not a serious answer to that. Something about that particular line made it the sensible place to stop, on river after river, for people who had never compared notes. What could a line in the landscape be doing?

Start underground, because the line is geological before it is anything else. Inland lies the Piedmont: old, hard, crystalline rock. Seaward lies the coastal plain: young, soft, unconsolidated sediment. The contact between them runs for well over a thousand kilometres, roughly parallel to the coast. A river crossing it is doing two different jobs on either side — cutting slowly through resistant rock above, cutting easily through loose sediment below — so at the contact the bed steepens into rapids or falls. Every river that crosses the boundary develops that break at about the same distance inland, because the boundary itself is a line. That already explains why the sites line up. It does not yet explain why anyone built there.

So put yourself in a boat. You are coming up from the sea with cargo, and the water is deep and tidal and easy — until the rapids. Above them your vessel cannot go. Ask the obvious question: where do you unload? Not earlier, because every mile of river carried is cheaper than a mile carted. Not later, because there is no later. The rapids fix the highest point that ocean-going or tidal craft can reach, and cargo must change vehicle exactly there.

Now notice what a compulsory change of vehicle brings with it. Someone has to lift the goods, so there are labourers. The goods wait, so there are warehouses. Two prices meet — what a cargo is worth on the water and what it is worth on the road — so there are merchants, and where merchants haggle there is a market. Upstream farmers bring produce down to the same point for the same reason in reverse. A place where every passing load must stop, be handled, and be traded is a place that accumulates people whether or not anyone intended a town. Geographers call the moment the break of bulk, and it is one of the most reliable town-founding conditions there is.

Then the second gift, from the same rapids. A river dropping steeply is usable power: dam it, run a race, turn a wheel. So the point that had to become a market is also the best mill site for miles — grain first, then textiles, iron, paper. The same feature that stopped the boats turned the machinery. That is why so many of these places became manufacturing towns rather than staying transfer depots.

One more thing the site tends to offer: rapids are shallow and often rocky, which makes them a good crossing. The lowest convenient bridging point on a river frequently sits at or just above the falls, so the inland road network converges there too.

Three refinements are worth making, because the tidy version overstates the case.

First, "the fall line" is a zone rather than a razor's edge, and not every city on the list sits precisely on the falls. Some sit at the head of tide — the furthest point reached by tidal water — which is close to the geological contact but not identical to it. Philadelphia is better described that way than as a falls town. The mechanism is the same; the marker is not.

Second, the two advantages are not simultaneous. Transhipment came first and founded the settlements; water power mattered much later, when there was industry to drive. A story that presents both as reasons for founding compresses two centuries.

Third, and most important: the mechanism explains where cities could form, not that they had to. A head of navigation with no productive hinterland behind it gets a landing and stays a landing. And once canals, then railways, then trucks made the rapids irrelevant as a barrier, the cities did not move — the reason for the location had expired while the location kept compounding. The line on the map is now a fossil of a constraint that no longer binds, which is why the pattern is easy to check and easy to misread as still operating.

The pattern generalises beyond one continent, incidentally: any head of navigation is a candidate. What makes the American case so legible is that here the geology arranged the barriers in a straight line.

• Why canals and railways, which erased the barrier, mostly reinforced the towns it had created rather than bypassing them.
• How heads of navigation elsewhere — inland ports on other continents — did and did not produce the same result.

Every term the collection defines is gathered in the glossary.