The Unbroken Soliton

John Scott Russell and the Solitary Wave on the Union Canal

Linear nineteenth-century hydrodynamics, formalized by Astronomer Royal George Biddell Airy and Cambridge Lucasian Professor George Gabriel Stokes, proved mathematically that any isolated mound of water in a channel must flatten and disperse as it travels. Under linear wave theory, any localized water heap can be decomposed via Fourier analysis into a sum of sinusoidal waves of different wavelengths. Because the phase speed of a surface gravity wave in shallow or intermediate water depends on its wavelength—with long wavelengths outrunning short wavelengths—the component sine waves inevitably drift out of phase. Any isolated pulse must therefore spread out into a trailing packet of diminishing ripples.

In August 1834, Scottish engineer and naval architect John Scott Russell observed a single mass of water violate that theorem along the Union Canal at Hermiston, six miles west of Edinburgh. Engaged by the Union Canal Company to determine whether steam-powered navigation could replace horses, Russell was measuring the tractive resistance of a canal boat being drawn rapidly by a pair of horses when the hull struck an underwater obstruction and stopped dead. The volume of water that the boat's prow had been setting in motion did not subside into turbulent foam or break into an oscillating train of crests and troughs. Instead, it gathered ahead of the bow into a smooth, sharply defined rounded elevation some thirty feet long and eighteen inches high above the resting canal surface. Detaching from the stationary vessel, it rolled forward down the channel as a single solitary heap.

Russell spurred his horse along the canal towpath and galloped beside the solitary crest. Moving at roughly eight miles per hour, the smooth mound of water preserved its exact length and bell-shaped profile mile after mile until it finally wound out of sight in the curves of the channel. Convinced that he had witnessed a fundamental mode of fluid motion—which he named the great wave of translation—Russell built a thirty-foot glass-walled wave tank in his garden. There he conducted hundreds of controlled experiments by dropping calibrated weights at one end of the trough.

Russell's tank experiments uncovered three quantitative properties that contradicted linear wave theory. First, the solitary wave was purely an elevation above the undisturbed water level with zero depression trough behind it. Second, its speed $c$ depended directly on its crest height $A$ above the resting water depth $h$ according to $c = \sqrt{g(h + A)}$, meaning that taller waves traveled faster than shorter waves. Third, when Russell launched a tall solitary wave behind a shorter, slower solitary wave in his tank, the taller wave caught up to the smaller one, merged with it nonlinearly, and emerged ahead of it. Both waves retained their exact original heights and profiles as if they were solid particles. Despite Russell's reports to the British Association for the Advancement of Science in 1844, Airy dismissed the phenomenon in 1845 as an ordinary periodic wave misread in a restricted channel, and Stokes argued that a wave of permanent form was impossible within standard fluid approximations.

The mathematical resolution arrived only when Joseph Boussinesq in the 1870s and Dutch physicists Diederik Korteweg and Gustav de Vries in 1895 retained both small finite amplitude ($A/h$) and small channel dispersion ($(h/\lambda)^2$) at the same order of approximation, deriving the Korteweg-de Vries (KdV) equation for the surface elevation $\eta(x,t)$:

Airy had retained the nonlinear amplitude term while neglecting dispersion, predicting that every finite wave must steepen until its crest overtakes its base and breaks. Stokes had retained linear dispersion while neglecting finite amplitude, predicting that every localized pulse must flatten and spread apart. The Korteweg-de Vries equation revealed that in shallow channels, the nonlinear convective steepening term $\eta(\partial \eta / \partial x)$—which pushes the tall peak forward faster than the skirt—acts in exact opposition to the dispersive third-derivative term $\partial^3 \eta / \partial x^3$, which pulls steep, short-wavelength gradients backward. At the exact hyperbolic-secant-squared wave profile ($\eta(x,t) = A \, \text{sech}^2[\sqrt{3A / 4h^3}(x - ct)]$) that Russell chased on horseback along the Union Canal in August 1834, nonlinear amplitude steepening and third-order wavelength dispersion lock into perpetual balance. That equilibrium allows a self-reinforcing soliton to travel across water, plasma, and transoceanic fiber-optic cables without losing its form.