Joseph Lucas offered an electric bicycle lamp in February 1888. It ran on a rechargeable wet Bunsen cell, and a Birmingham jeweller named Vaughton designed it. It weighed over two kilos and cost 55 shillings. It stayed in the catalogue for about a year. Lucas did not offer another electric bicycle lamp for forty years.

The problem was the bulb. Filaments did not survive the vibration of bad roads, and replacements were expensive. A wet cell also had to be charged somewhere.

The usual account says electric lighting went nowhere until 1898. The French press gives a less tidy picture.

In January 1896 a Paris paper ran a column on light accumulators. They had been shown at the Salon du Cycle for use in bicycle lanterns. The cells were charged from Callaud gravity elements and gave at least three hours of light. The writer thought that enough to take them out of the category of toys and into practical use. He recommended them to military and civilian cyclists, and preferred them to the flammable Venetian paper lantern. The same column described a Dorville accumulator giving a sewer worker eight hours of light from a kilo and a half of cells.

That is trade journalism. It describes what was exhibited and claimed, not what sold. Three hours was the figure for a battery that had to be charged at home from a set of wet cells.

Dynamos were tried from the 1880s, once carbon-filament bulbs made the idea workable. The first was built by Richard Weber, a mechanic in Leipzig. He took out two British patents, 5,078 of 1886 and 16,603 of 1887. His was a miniature dynamo in a rectangular box, with a friction wheel that a spring pressed sideways against the tyre of a high-wheeler or tricycle. He tried to sell it as the Excelsior Electric Lighting at the Leipzig exhibition of 1889.

A series of American patents followed from 1894, and nobody agreed on where the dynamo should sit. Ernest Tillmann, a Frenchman living in New York, put it on the seat tube, driven by a friction pulley and belt. Others mounted it over the front wheel, under the saddle, or clamped to a fork blade.

One of those patents shows how much trouble the idea still gave. Gustavos Heidel of St. Louis filed in January 1899 and assigned the patent to the Globe Electric Company of the same city. His dynamo clipped to the front fork. A rubber-rimmed wheel ran on the tyre, and a cone pressed against the bevelled side of that wheel to gear the armature up.

Heidel fitted a governor. Weights on the armature shaft swung outwards as the speed rose and lifted the shaft, which eased the cone away. A fast rider could then not push enough current through the bulb to burn it out.

Two springs held the whole frame in either of two positions. Swung away from the machine, it would not rattle or mark the paint. Swung forward, it pressed evenly on the tyre. The bulb sat in a hood on a ball joint and could be aimed.

The bottle dynamo is usually credited to James Wilson of Battle Creek, Michigan, in 1901, driven by a pulley running on the tyre sidewall. Weber was pressing a friction wheel against the tyre in 1886. Heidel was running a wheel on the tyre from the front fork two years before Wilson's date.

There is a good record of what was still wrong with the idea in 1917. In March of that year the Swiss maker Otto Pletscher filed a French patent for a bicycle lantern with the dynamo built into it. He named two complaints about the electric lamps then on the market.

They gave no light when the dynamo was not being driven, so pushing the bicycle by hand left the rider in the dark. And the wiring between dynamo and lamp was easily damaged. One broken wire put the whole lamp out of use.

His answer was to make dynamo and lamp into one unit on the front fork. The wires ran inside a hollow shaft, where nothing could get at them. A rubber pulley pressed sideways against the tyre and drove a cylinder of permanent magnets round a fixed armature. The gearing was high enough, by his claim, to give usable light at walking pace. A spring held the pulley against the tyre, and a lever with a catch lifted it off again, with two positions that could be locked.

A patent records what somebody applied for, not what was built.

Electric cycle lighting became ordinary in Europe in the 1920s, once electricity was normal in houses and workshops. In the American mail-order trade it arrived earlier.

Take one page of the Montgomery Ward catalogue of 1916. The acetylene lamps run from $1.68 to $2.40. The Presto electric outfit is 90 cents without a battery and $1.30 with lamp and battery. The Victor outfit is $1.10 and $1.50, the De Luxe $1.95 and $2.35. The cheapest electric outfit undercuts the cheapest gas lamp by more than two to one.

The catalogue sells it on the points the patents had been worrying about for thirty years. The light is even and strong and white. It will not blow out. You turn it on and off with a switch.

By 1925 the same formats were reaching Dutch and German wholesalers. A Joh. Lehmkuhl price list of that year lists Bosch, Berko, Radsonne and Riemann electric bicycle lighting, beside the cycle candles and the plain lantern glasses.

Electric lighting replaced three kinds of flame lamp. They are covered under bicycle oil lamps, bicycle candle lamps and acetylene bicycle lamps. Acetylene stayed in use alongside electric lighting until the Second World War. For brackets, date stamps and glass colours, see identifying and dating an antique bicycle lamp.

Primary: American patent 644,823 of 1900 (G. Heidel, electric lamp for bicycles, assigned to the Globe Electric Company); French patent 484,938 of 1917 (O. Pletscher, bicycle lantern with integral dynamo); British patents 5,078 of 1886 and 16,603 of 1887 (R. Weber); Le Petit Temps, 15 January 1896, consulted through RetroNews; Montgomery Ward catalogue, 1916; Joh. Lehmkuhl wholesale price list, 1925.

Secondary: Tony Hadland and Hans-Erhard Lessing, Bicycle Design: An Illustrated History, the stronger of the two standard works on patents and dynamos; Peter W. Card, Early Cycle Lighting 1868–1948.