01Executive summary
Four milestones in moving goods overland — and a reminder that for bulk freight before the railway, water beat road every time.
James Brindley's Bridgewater Canal of 1761 showed that inland bulk water transport could be built and paid for commercially. From 1764 Tresaguet in France, and then Telford and McAdam in Britain, began treating a road as a designed layered structure rather than a hard surface laid on the ground. The Erie Canal opened in 1825 and immediately outcompeted the road it paralleled.
02Canals: why water wins for bulk
The reason inland navigation dominated bulk transport before the railway is physical. Rolling resistance on a poor road is high, a horse can draw only a limited load, and the load damages the road it travels on. Floating that same load removes almost all the resistance: a single horse on a towpath can move a barge carrying many times what it could pull in a wagon. Everything about canal economics follows from that ratio.
Canal engineering, though, is unforgiving in a way roads are not. A road can climb; water cannot. A canal must hold a level over long distances, change level only at controlled points, and be supplied with enough water to replace what every lock passage sends downhill.
Hold a level
Long pounds at constant elevation demand accurate survey and substantial earthworks, embankments and cuttings. The alignment problem is the same one the Roman aqueduct engineers solved, applied to a much larger cross-section.
Change level under control
The pound lock — a chamber with gates at each end — developed through the medieval Italian canals. Earlier Chinese practice used inclined planes with rollers and windlasses, a different solution to the same problem.
Supply the water
Every lock cycle discharges a chamber-full downhill. Summit levels need reservoirs, feeders or pumping, and water supply frequently governs the capacity of the whole navigation.
Stay watertight
A canal in permeable ground leaks away its own supply. Puddled clay linings were the practical answer, and maintaining them was a permanent operating cost.
Brindley's work for the Duke of Bridgewater in the Midlands is the point at which this became a commercial proposition rather than a princely undertaking. The Duke had travelled in Europe and been struck by the Languedoc Canal, whose locks, aqueducts and tunnel carried water transport through mountainous country. The influence on subsequent canal building, including in America and on the Erie, was preponderantly British and stemmed largely from Brindley.
The Erie Canal, opened in 1825, ran some 580 kilometres and opened a continental interior. Its effect on the competing road was immediate: stages, carriages and trains of Conestoga wagons had thronged the National Road for a decade, with around five thousand wagons unloaded at Wheeling in 1822 at a cost of roughly $400,000 in carriage. The canal competed with it altogether successfully from the year it opened.
03The road as a designed structure
Roman road building had understood that a pavement is a layered structure that must be kept drained. That understanding was substantially lost in Europe for a thousand years, and the recovery begins with Pierre-Marie-Jérôme Tresaguet, who started formulating his technique in 1764 — the year McAdam and Telford were boys of eight and seven.
Broken-stone roads existed before all three of them; Hubert Gautier's writings, first published in 1693, indicate that small pieces were being used. What Tresaguet, Telford and McAdam contributed was system: a specified structure, built to a method, for a stated purpose.
| Element | Purpose | Failure if omitted |
|---|---|---|
| Drained formation | Keeps water out of the subgrade so it stays stiff | The subgrade softens and everything above it fails regardless of quality |
| Large stone foundation course | Distributes wheel load over a wide area of subgrade | Concentrated load punches into soft ground |
| Graded broken stone | Angular fragments interlock under traffic into a dense mass | Rounded or oversized material never locks up and remains loose |
| Fine surface | Takes abrasion, sheds water, binds under traffic | Water enters the structure and the layers below lose strength |
| Camber | Sheds water laterally instead of letting it stand or soak | Ponding, saturation and rapid deterioration |
| Side drainage | Removes water clear of the structure | The whole structure sits in a trough of its own water |
Telford built a heavy foundation of carefully placed large stones and graded material above. McAdam argued that a well-drained subgrade could carry the load itself, and that what mattered was a relatively thin layer of angular broken stone, closely sized, that would interlock under traffic into a coherent mass — cheaper, faster and adequate. Neither is wrong. Telford's approach suits weak subgrades and heavy loads; McAdam's suits sound ground and lighter traffic. The genuinely important shared insight is McAdam's: the strength is in the drainage and in the interlock of angular particles, not in the mass of material. That is still how a granular pavement layer works, and it is still the thing most often got wrong on site.
04Choosing between modes
The period covered here contains a genuine, contested transport-mode competition, decided on economics rather than on preference, and it is instructive because the same reasoning recurs.
The lesson is not that one mode is better. It is that the competition was settled by cost per tonne over distance against capital committed, and that a new mode with a different position on that trade-off displaced the incumbent quickly and completely. The canal companies were not badly run; they were holding an asset whose economic basis had been removed.
05Takeaways for current practice
- Reduce the resistance rather than increasing the effort. Floating a load beats pulling it, and that single ratio explains the canal age.
- The strength of a granular pavement is drainage and interlock. Angular, closely graded, well drained — not thickness for its own sake.
- Match the structure to the subgrade and the traffic. Telford and McAdam were both right about different situations.
- Identify the consumable that limits capacity. For a canal it is water at the summit, and it is easy to design a navigation that cannot be supplied.
- A new mode does not beat the incumbent; it removes its economic basis. Canal operators were displaced by arithmetic, not by mismanagement.
The modern descendants sit in Austroads pavement design guidance and AS 1348 for road engineering terminology. Cited by number for orientation only — verify currency.