Somers Roundabout

This is a bench model. A passenger-carrying wheel is regulated engineering requiring certified design and inspection — nothing here is a basis for building one.

Somers specifies a pair of parallel driving cables in grooves on the rims, explicitly so that one can fail without stopping the wheel. Designed redundancy, in 1893.

Glue a little bench rigidly to a disc and turn it. At the top it is inverted. This is why the carriages cannot be fixed.

Cut curved segments and overlap the joints so no two land in the same place. A wooden wheel is strong only if its joints are staggered.

Make a pair of identical wheels on a common axle. Carriages hang between them, and the pair resists the sideways loads a single rim cannot.

Cut a rounded groove around the outer face of each rim. The cable sits in the groove and drives by friction — there are no gear teeth anywhere.

Use radial members in tension where you can. A wheel loaded at the rim carries far more as a tension structure than as a set of compression struts.

Suspend cars on pins between the rims so they swing. Gravity levels them at every position with no mechanism at all — this is the elegant part.

The pivot must sit above the loaded car's centre of gravity or it will hang inverted. Check it loaded, not empty — passengers move the centre.

Place them at equal intervals. Uneven spacing makes the wheel unbalanced, so it accelerates through part of each turn and drags through the rest.

Run two cables at equal tension. If one carries everything, the redundancy is decorative — and that is a dangerous kind of decoration.

Release one cable while the model turns. The wheel should keep going on the other. Testing the failure case is the only way to know the redundancy is real.

Weight a single car and turn the wheel. It now wants to rest with that car at the bottom — which is why real wheels are loaded and unloaded in a set order.

Run it and watch for a car that keeps swinging. A pendulum near resonance with the rotation is deeply unpleasant — damp it or change the pivot height.

The patent. US 489,238, "Roundabout", granted 3 January 1893 to William Somers of Atlantic City, New Jersey. The specification describes a rotating vertical frame with suspended passenger carriages, driven by a pair of parallel cables running in grooves in the outer rims, and it makes a point of the dual-cable arrangement so that failure of one leaves the other driving. Somers built fifty-foot wooden wheels at Asbury Park, Atlantic City and Coney Island in 1892 — before the Chicago World's Columbian Exposition opened.

The dispute with Ferris. George Washington Gale Ferris Jr. rode Somers' Atlantic City wheel and went on to build the 264-foot steel wheel that was the centrepiece of the 1893 Exposition in Chicago. Somers sued for infringement. Ferris's defence was in essence that his machine was a different thing — a vastly larger steel structure, engineered as a bridge-builder would engineer it, rather than a scaled-up timber ride — and he prevailed. Both claims have some force: the suspended-carriage principle is Somers', while the structural engineering that made a 264-foot wheel possible is genuinely Ferris's. What is not in dispute is the order of events, and it is worth stating that the patent for the ride predates the wheel that gave it its name.

Gravity as the mechanism. Hanging each carriage from a pivot above its loaded centre of mass means the car is a pendulum, and a pendulum's rest position is directly below its pivot regardless of where that pivot has been carried. No linkage, no gearing, no control system, no power. It is one of the cleanest examples in mechanical design of using a physical law instead of a mechanism, and the same reasoning appears in gimballed compasses and ship's lamps. The failure mode is equally instructive: get the pivot BELOW the centre of mass and the stable position inverts.

Timber, and why it was replaced. Somers built in wood because it was cheap, workable and familiar, and he compensated for its weakness with laminated and staggered rim joints and generous section. Wood's problem in this application is not strength but consistency — it changes dimension with moisture, creeps under sustained load, and fails at knots and joints unpredictably. Ferris's steel allowed calculable strength and far larger spans, and every observation wheel since is steel. Somers' contribution survives as the arrangement, not the material.