Two centuries separate a twitching compass needle from the Wi-Fi chip in your phone. This is who took each step, what they actually showed, and where the famous stories drift from the record.
Common mix-up: “Marconi invented radio,” or its rebuttal, “Tesla did, and the Supreme Court said so.” Neither holds. Radio was assembled by many hands: Maxwell’s theory, Hertz’s proof, Branly’s and Lodge’s detectors, Bose’s millimeter waves, Tesla’s tuned circuits and Marconi’s relentless engineering. The 1943 Supreme Court case invalidated parts of one Marconi patent, citing earlier work by Lodge, Tesla and John Stone Stone; it did not crown anyone.
In 1600 William Gilbert, physician to Elizabeth I, published De Magnete. He carved a sphere of lodestone, a terrella, walked a small compass over it, and showed the needle behaved just as it did on Earth: the planet itself is a magnet. He also catalogued materials that attract chaff when rubbed and named them electrica, from the Greek for amber. That is where the word “electricity” comes from.
For the next century and a half, electricity was a parlor trick of rubbed glass and sparks. Benjamin Franklin made it a science. He argued for a single electric “fluid” whose excess or lack he called positive and negative, and he proposed that lightning was the same thing as a laboratory spark. French experimenters led by Thomas-François Dalibard confirmed it with a tall iron rod at Marly-la-Ville in May 1752, following Franklin’s published plan. Franklin’s own kite experiment, in June 1752, is known mainly from his brief letter that October and Joseph Priestley’s 1767 account; there were no independent witnesses besides his son, and some historians doubt it happened quite as told. Franklin’s sign convention stuck, which is why the electron, discovered 145 years later, ended up “negative” and why conventional current flows the opposite way to the electrons.
In 1785 Charles-Augustin de Coulomb hung a needle from a fine wire, a torsion balance sensitive enough to measure the push between two small charged spheres. The force fell off as the square of the distance, the same law as Newton’s gravity. Then in 1800 Alessandro Volta, settling an argument with Luigi Galvani about twitching frog legs, stacked discs of zinc and copper separated by brine-soaked card. His “pile” gave the first steady electric current. Everything that follows needed it.
Coulomb’s balance measured forces of a few ten-millionths of a newton by the twist of a silver wire. Charge leaked away into humid air during each run, and his data covered only a small range of distances, so his inverse-square law was as much a confident inference as a proof. Henry Cavendish had shown it more precisely in the 1770s with nested spheres (no charge appears on an inner sphere if, and only if, the law is exactly inverse-square), but he never published; Maxwell edited his notebooks a century later.
Modern tests of the same kind, by Williams, Faller and Hill in 1971, showed the exponent is 2 to within about 3 parts in 10¹⁶.
In April 1820, in Copenhagen, Hans Christian Ørsted ran a current through a wire near a compass and saw the needle swing to point across the wire. He published that July, in Latin, in a four-page pamphlet sent across Europe. Electricity in motion makes magnetism.
André-Marie Ampère saw a demonstration in Paris on 11 September 1820 and within a week had shown that two parallel currents attract and opposite ones repel. Over the next six years he built the mathematical theory he called electrodynamics. The unit of current carries his name.
Michael Faraday, a bookbinder’s apprentice turned assistant at the Royal Institution, asked the reverse question: can magnetism make electricity? On 29 August 1831 he wound two coils on an iron ring and saw a galvanometer kick, but only at the instant he connected or disconnected the battery. That autumn he showed that pushing a magnet into a coil does the same thing, and built a spinning copper disc that made a continuous current, the first dynamo. The key was change: a changing magnetic field drives a current. Every generator and transformer on Earth runs on it (Chapter 2).
Faraday had almost no mathematics. Instead he drew “lines of force” filling the space around magnets and charges, and treated them as physically real. In 1845 he found that a magnetic field rotates the polarization of light passing through glass, the first hint that light and electromagnetism were related. In 1846 he even speculated, in a short paper, that light might be a vibration of the lines of force themselves.
Joseph Henry in Albany, New York, very likely observed induction in 1830 or 1831, around the same time as Faraday, but published after him. Henry did discover self-inductance, and the SI unit of inductance, the henry, honors him. Priority in science goes to publication, and Faraday’s 1831 papers were thorough and fast.
James Clerk Maxwell set out to put Faraday’s lines of force into mathematics. In “On Physical Lines of Force” (1861–62) he added one term to Ampère’s law, the displacement current: a changing electric field acts like a current and makes a magnetic field, just as a changing magnetic field makes an electric one. With that term, the equations allowed a self-sustaining ripple of electric and magnetic fields that travels through empty space.
Its speed came out of the equations as one over the square root of two constants measured on lab benches: one from forces between charges, one from forces between currents. Wilhelm Weber and Rudolf Kohlrausch had measured the ratio in 1856. Maxwell worked out 310,740 km/s. Hippolyte Fizeau had timed light with a spinning toothed wheel in 1849 at roughly 315,000 km/s. Maxwell wrote that “we can scarcely avoid the inference that light consists in the transverse undulations of the same medium which is the cause of electric and magnetic phenomena.” His 1865 paper made the theory self-contained, and his Treatise on Electricity and Magnetism of 1873 laid it all out. He died in 1879, aged 48, before anyone made such a wave on purpose (Chapter 3).
Heinrich Hertz did, in Karlsruhe between 1886 and 1888. His transmitter was two brass rods ending in small spheres, fed by an induction coil; when the gap sparked, current sloshed back and forth along the rods at tens to hundreds of megahertz. His receiver was a loop of wire with its own tiny gap. Meters away, faint sparks jumped across it, visible only in a darkened room. In 1888 he bounced the waves off a zinc sheet on the wall of his lecture hall. Incoming and reflected waves combined into a standing wave: places where the receiver sparked strongly, and nodes where it went dead. The spacing between nodes is half a wavelength, so he could measure the wavelength directly. With mirrors and prisms of pitch he showed the waves reflect, refract and polarize just like light.
Asked what use the waves were, Hertz reportedly said, “It’s of no use whatsoever… just an experiment that proves Maestro Maxwell was right.” The quote is everywhere, but it appears only in later retellings attributed to his students; the exact words, and whether he said them at all, can’t be traced to a primary source. Hertz died in 1894 at 36. In 1930 the unit of frequency was named after him.
Maxwell’s 1865 paper had twenty equations in twenty unknowns, written component by component, and leaned on potentials rather than fields. The compact four-line form taught today is largely the work of Oliver Heaviside, a self-taught former telegraph clerk, who recast them in vector notation in 1884–85; Josiah Willard Gibbs developed the same vector calculus independently. Hertz arrived at a similar streamlined form. For a while physicists called them the Hertz–Heaviside equations.
Hertz’s first wavelength figures, from waves on wires and in air, disagreed with each other, partly because of an error in his calculation of the oscillator’s frequency that Henri Poincaré pointed out. Later, better-controlled experiments by others confirmed that the waves travel at the speed of light.
In 1887 Hertz also noticed that ultraviolet light shining on his spark gap made sparks jump more easily. He reported it and moved on. It was the photoelectric effect, which Einstein would explain in 1905.
Drag the timeline sideways to pan and use the zoom slider to spread out crowded decades. Tap an event (or focus the timeline and use ← →) to read its card. Events marked ? have a disputed or legendary part.
The spark gap on the left launches waves toward a metal wall on the right; the reflection sets up a standing wave. Slide the receiver toward and away from the wall and watch its spark rise and vanish. Change the wavelength to move the nodes.
Simplified: the receiver responds to the electric field, which must be zero at a metal wall; the slow weakening of the wave with distance from the transmitter is ignored. Hertz’s own room-sized runs used wavelengths of several meters.
Hertz’s receiver was a spark you could barely see. Practical radio needed a better detector. In 1890 Édouard Branly in Paris found that a glass tube of loose metal filings, normally a poor conductor, suddenly conducted when a radio wave arrived. Oliver Lodge named it the coherer, added a tapper to shake the filings loose again, and in 1894, months after Hertz’s death, gave lectures in London and Oxford showing signals detected between rooms and buildings.
In Calcutta, Jagadish Chandra Bose built apparatus working at wavelengths of a few millimeters to centimeters, frequencies we now call millimeter wave and use for 5G. He invented waveguides, horn antennas and a galena-crystal detector, one of the first semiconductor devices, patented in the U.S. in 1904. A widely repeated story has him ringing a bell by radio through walls at a public demonstration in 1895; the details come from later accounts. He declined to patent most of his work.
Nikola Tesla’s contribution was resonance. His high-frequency coils and his 1893 lectures described tuned circuits at transmitter and receiver, the basis of selecting one station among many. His wireless patents were filed in 1897 and granted in 1900. Tesla’s grander goal, sending power through the earth from his Wardenclyffe tower, never worked.
Guglielmo Marconi was the one who made it a business. Starting in 1895 near Bologna, he improved range by raising tall antennas and grounding the transmitter, moved to England, and filed a British patent in June 1896. On 12 December 1901, at Signal Hill in Newfoundland, he and his assistant George Kemp reported hearing the Morse letter S, three dots, sent from Poldhu in Cornwall, about 3,400 km away. Only the two of them heard it, through an earpiece amid heavy static, and some historians think they mistook noise for signal. But by February 1902 ship’s officers aboard the SS Philadelphia were logging Poldhu messages over 2,000 miles out, and transatlantic service followed. Marconi shared the 1909 Nobel Prize with Karl Ferdinand Braun.
Spark transmitters could only make dots and dashes. Reginald Fessenden wanted a smooth, continuous wave he could shape with sound. He sent speech over about a mile in December 1900, and his later alternator transmitter at Brant Rock, Massachusetts, carried voice reliably. The famous Christmas Eve 1906 broadcast of a violin and a Bible reading to ships at sea rests on Fessenden’s own recollections decades later; no record from the time has been found, so treat it as likely but unproven.
Marconi Wireless Telegraph Co. of America v. United States began as a claim that the U.S. government had used Marconi’s patents during World War I without paying. The Court ruled in June 1943 that key claims of Marconi’s U.S. patent 763,772, on four-circuit tuning, were invalid because Lodge, Tesla and John Stone Stone had described the essentials first. That made a good story for Tesla, who had died that January, but the decision was about one patent’s claims, not about who invented radio.
While radio grew at the low-frequency end, the high end opened up. On 8 November 1895 Wilhelm Röntgen found that a covered cathode-ray tube made a screen across the room glow: X-rays (Chapter 12). In early 1896 Henri Becquerel found that uranium salts fogged sealed photographic plates without any light at all: radioactivity. Paul Villard spotted the most penetrating part in 1900, and Ernest Rutherford named it gamma rays in 1903.
The bigger shock came from heat. Classical physics predicted that a hot object should radiate infinite energy at short wavelengths, which obviously doesn’t happen. On 14 December 1900 Max Planck presented a formula that fit the measured glow perfectly, but only if energy was traded in lumps of size h × f. He later called it “an act of desperation” and didn’t believe the lumps were real.
In 1905 Albert Einstein took them seriously. Light shining on metal knocks out electrons, but whether it does so depends on the color, not the brightness: dim ultraviolet works, bright red doesn’t. Einstein explained it: light itself arrives in quanta, and each one either has enough energy to free an electron or doesn’t. Robert Millikan, who set out to disprove it, confirmed the prediction precisely by 1916. Einstein’s 1921 Nobel Prize was for this, not relativity (Chapter 6).
The classical Rayleigh–Jeans law says radiated power per unit frequency grows as f², without limit. Planck’s law adds a factor that cuts off exponentially once h f exceeds the thermal energy k T, because a mode can’t be excited by less than one whole quantum. At room temperature k T ≈ 0.025 eV, so modes above roughly 10 THz (thermal infrared) are frozen out. That is why warm things glow in the infrared and not in X-rays.
Edwin Howard Armstrong invented the regenerative receiver, the superheterodyne (still the architecture of most radios) and, with patents granted on 26 December 1933, wideband FM. By encoding sound in frequency rather than amplitude, FM ignored most static. RCA resisted it, and Armstrong spent his last years in patent litigation.
Radar. On 26 February 1935, near Daventry, Robert Watson-Watt and Arnold Wilkins showed that an RAF bomber flying through a BBC shortwave beam reflected enough signal to detect. Britain’s Chain Home network was running by the start of the war. In 1940 John Randall and Harry Boot built the cavity magnetron, a compact source of powerful microwaves, which made short-wave radar practical.
The microwave oven. Percy Spencer, a self-taught engineer at Raytheon who built magnetrons, is said to have noticed a candy bar melting in his pocket near a working radar set, then popped corn on purpose. The story’s details vary with the teller, but the patent is real: filed 8 October 1945. The first Radarange, in 1947, stood nearly six feet tall.
Information. In 1948 Claude Shannon published “A Mathematical Theory of Communication.” It defined the bit and proved that every channel has a maximum error-free data rate set by its bandwidth and signal-to-noise ratio. Every Wi-Fi and cellular standard since is an engineering race toward Shannon’s limit (Chapter 9).
Frequency hopping. Actress Hedy Lamarr and composer George Antheil patented a “Secret Communication System” in 1942: a radio-guided torpedo whose transmitter and receiver hop together among 88 frequencies, synchronized like player-piano rolls, so an enemy can’t jam it. The Navy shelved it. Frequency hopping had been described earlier, and modern spread-spectrum radio grew mainly from separate military work in the 1950s, so the line from their patent to Bluetooth, which hops among 79 channels 1,600 times a second, is inspiration rather than direct inheritance. Their recognition came late: an EFF Pioneer Award in 1997 and the National Inventors Hall of Fame in 2014.
The oldest light. In 1964 Arno Penzias and Robert Wilson, testing a horn antenna at Bell Labs in Holmdel, New Jersey, at 4,080 MHz, found an excess hiss of about 3.5 kelvin from every direction. They cleaned out pigeon droppings; it stayed. It was the cosmic microwave background, the cooled afterglow of the early universe, published in 1965 and rewarded with the 1978 Nobel Prize.
Wi-Fi. On 9 May 1985 the FCC, pushed by engineer Michael Marcus, allowed unlicensed spread-spectrum devices in three “junk” ISM bands shared with microwave ovens: 902–928 MHz, 2,400–2,483.5 MHz and 5,725–5,850 MHz. Anyone could build a radio there without a license. The IEEE 802.11 standard followed in 1997, at 1 and 2 Mbit/s; 802.11b reached 11 Mbit/s in 1999, the year the brand “Wi-Fi” was coined (a pun on hi-fi, not an abbreviation of anything). In Australia, John O’Sullivan’s team at CSIRO, drawing on radio-astronomy signal processing, patented techniques for beating indoor multipath echoes; the U.S. patent was granted in 1996. CSIRO enforced it against the industry and reportedly collected over A$400 million in settlements, a result admired in Australia and criticized by some as overreach elsewhere (Chapter 10).
Before 1985, nearly every radio transmitter needed a license tied to a specific use. The ISM bands were considered nearly worthless, polluted by ovens and industrial heaters. Allowing anyone to use them, provided they spread their signals and kept power low, created the first large commons for radio innovation. Cordless phones, Wi-Fi, Bluetooth, Zigbee and baby monitors all grew in that commons. Spectrum economists still argue about how much spectrum should be licensed versus shared; the 2020 opening of 6 GHz to Wi-Fi was the same argument decided the same way.
Each dot is a historical value for the speed of light. Tap a chip to highlight one; switch to the zoomed view to see how measurements closed in on 299,792.458 km/s.
| Idea | Theory or discovery | First proof or device | Everyday result |
|---|---|---|---|
| Induction | Faraday, 1831 | Faraday’s disc dynamo, 1831 | Power grids, transformers, wireless chargers |
| Radio waves | Maxwell, 1861–65 | Hertz, 1886–88 | Broadcasting, phones, Wi-Fi |
| X-rays | Röntgen, 1895 | Hand radiograph, Dec 1895 | Medical imaging, airport scanners |
| Light quanta | Planck 1900, Einstein 1905 | Millikan, 1914–16 | Solar cells, camera sensors |
| Radar | Reflection of radio waves (Hertz) | Daventry, 1935 | Air traffic control, weather radar, car radar |
| Information limit | Shannon, 1948 | Error-correcting codes, 1950s onward | Every digital link |
| Unlicensed radio | FCC, 1985 | 802.11, 1997 | Wi-Fi, Bluetooth |