Marconi Wireless Telegraph

A spark oscillator radiates across the whole spectrum and interferes with radio services. Build it only shielded, at minimum power and with no external aerial, as a bench demonstration. High voltage: discharge everything before touching it.

Marconi claims the transmission of signals through air, earth or water by means of high-frequency oscillations. Look in the text for the elevated aerial and the earth connection: they are the real contribution.

Two small loops with spark gaps: a spark in one produces a spark in the other a few centimetres away. It works — and at that distance it is useless.

Wind a primary of a few thick turns and a secondary of many fine turns on a soft iron core. The vibrating interrupter generates the high-voltage pulses.

Two metal spheres facing each other across a small gap. Every spark is a damped pulse containing a wide band of frequencies — not a pure wave.

Connect one side of the spark gap to a vertical wire, as high as possible. Range grows with aerial height: this is the empirical discovery Marconi made by trial.

Earth the second terminal through a buried plate. Try it with and without: the difference in range is enormous, and this is the central point of the patent.

The conducting ground acts as the second element of the aerial, like a mirror image of the vertical wire. You get a dipole twice as tall without having to build it.

A small glass tube of metal filings between two electrodes. At rest its resistance is very high; struck by a wave the particles weld together and it conducts.

The coherer stays conducting after the signal: it has to be shaken. Without the automatic decoherer the receiver takes one pulse and is then silent for ever.

The coherer switches very little current. A relay amplifies mechanically and drives the bell or the Morse register.

The receiver wants an elevated aerial and an earth too. Symmetry between the two ends is what multiplies the useful range.

Move the receiver away and note the maximum distance for three aerial heights. The graph you get is the one Marconi plotted in 1895.

Run two at once: the receiver cannot separate them. The spark occupies the whole spectrum, and tuning — British patent no. 7777 of 1900 — was born of this.

The patent. US 586,193, “Transmitting Electrical Signals”, filed 7 December 1896 and granted 13 July 1897 to Guglielmo Marconi; the corresponding British patent is no. 12,039 of 1896. Marconi began experimenting in 1894 in the attic of Villa Griffone near Bologna. The Italian post office showed no interest, and he moved to England, where he found funding and customers — maritime ones above all.

What he actually added. Hertz had produced and detected electromagnetic waves in 1887, treating them as a verification of Maxwell's equations with no practical use; Édouard Branly invented the coherer in 1890 and Oliver Lodge had used it to receive signals. Marconi invented none of these parts. His contribution was engineering and experimental: the raised vertical aerial and the earthing of transmitter and receiver, which together turned a range of metres into one of kilometres. It is a clear case where the invention is not a new component but a new configuration of known ones.

Why the earth matters so much. Conducting ground behaves as a reflecting plane: it creates an electrical image of the vertical wire, so an aerial of height h behaves roughly like a dipole of twice the length. You get half the aerial free, and low-angle radiation — the useful kind for covering distance over the earth's surface — improves markedly. Vertical aerials with a ground plane still work for this reason.

The limit of the spark, and how it was overcome. A spark oscillator emits a damped pulse occupying an enormous bandwidth: two nearby stations interfere hopelessly and there is no selectivity at all. Marconi attacked this with the tuned circuits of British patent no. 7777 of 1900, which introduce tuning. The definitive answer, though, came with continuous waves generated by thermionic valves, after De Forest's Audion: only then did voice, music and broadcasting become possible. The transatlantic transmission of 12 December 1901 was announced as reception of the letter S in Morse; the experiment was not rigorously documented and some historians still dispute it, which is worth reporting honestly.