Take 2,750 centimetres of copper and measure how fast a wave crosses it. Laid out as coaxial cable: 81% of the speed of light. Wound into a Tesla extra coil: 126% — past the limit, on a bench, with an ordinary frequency counter. Eric Dollard has spent fifty years arguing that this is not an error but the second kind of electricity, the longitudinal wave that engineering walked away from in 1919. This document reconstructs the whole system — the physics, the machines, the arithmetic he had to reinvent to describe it — so you can follow it, argue with it, and see exactly where it stops being testable.
How to read this
01 · The core distinction
Two waves, one wire
02 · The two fields
Magnetism and dielectricity
03 · The strange arithmetic
Counterspace, per-henries and the algebra he had to invent
04 · The machines
Extra coils, magnifying transmitters, virtual grounds
05 · What was abandoned
Tesla, Marconi, and the 1919 cut-off
06 · The wider worldview
Ether, organic electricity, and the rest of it
07 · At the bench
How you would actually build and measure this
08 · The grid
Adversary and instrument
09 · Experiments with a falsifiable claim
What you could actually go and test
10 · Where the mathematics comes from
Heaviside, Steinmetz, Macfarlane, versor algebra
11 · Where it leaves physics
The parts no experiment can reach
12 · The lineage
Who he is standing on
13 · The claims register
198 claims, sorted honestly
testable difference unfalsifiable / re-description agrees with standard physics
14 · In his own words
Forty quotes, unfiltered
15 · The cheat sheet
Explain it in one breath
Every load-bearing concept, compressed to the sentence you'd actually say to another human. Fifty-three ideas, in the order they build on each other.
16 · The verdict
What's actually here
Genuinely interesting
The velocity-factor measurement. The same 2,750 cm of wire, measured twice: 81% of light speed as RG-62 coax, 126% as a Tesla extra coil. Both numbers are real, both are explicable — but almost no working engineer has ever seen the second one on a bench, and his whole argument hinges on it.
The recovered literature. Heaviside's operational calculus, Steinmetz's transient theory, Macfarlane's versor algebra, Bewley on travelling waves. A real body of pre-1930 engineering mathematics that genuinely did get pruned from the curriculum — and he reads it in the original.
The telluric channel. Ground-coupled propagation is respectable physics (Zenneck, Sommerfeld) and commercially near-abandoned. His near-field ratio claims are cheap to test with two receivers and a car.
89 testable differences, 51 with a number attached. Most will probably fail. But they are claims with predicted magnitudes and settling experiments, not vibes.
The textbook in different words
Half the register — 100 of 198 claims — predicts exactly what standard theory predicts once you translate the vocabulary. Different story, same meter reading.
"Electricity is not electronics" is a naming decision, not a measurement. Field energy versus conduction loss is ordinary transmission-line theory with the labels swapped.
The 126% coil result is classified here as agreeing with standard physics: it is a real measurement of a slow-wave structure's phase velocity, not of energy outrunning light. Four further claims were relabelled that way on audit, and four the other way.
"Wires are translators, not conductors" restates the Poynting picture — which every RF engineer already uses to design cable.
Beyond what any experiment can reach
Only nine claims are unfalsifiable — but they are load-bearing ones: counterspace as the destination of missing energy, space having exactly one true dimension, the ether as a substance rather than a bookkeeping device.
The suppression material is history-shaped but not settleable as stated. Notably, he throws out the popular versions of it himself — no scalar weapons, no atmospheric Tesla car, and RCA did not bury a working system.
The golden-ratio discharge, formative forces and "organic electricity" leave physics entirely and become a philosophy of nature. Worth reading as that, not as a prediction.
What you could actually go and test
The paired-receiver telluric ratio — buried versus elevated antenna at 2, 8, 15 and 30 miles. He predicts 5:1 near in and a crossover by 30 miles; standard theory predicts one mode with one distance law.
Quarter-wave resonance on a scale-model extra coil against a light-speed baseline. Does 126% reproduce, and does it move when you change the winding pitch?
The no-antenna 160-metre station, built against a low-impedance ground.
The toroid transformer and the homopolar disc — two bench anomalies that need no exotic apparatus and no belief in anything.
Start with a garden hose. A ham radio club in Santa Barbara, and a man named Chris Carson has just taken the microphone from Eric Dollard — a hard act to follow, he warns the room, because Dollard can talk like this for five hours and have you convinced by the fifth. So instead of talking, he hands one end of an ordinary garden hose to Dollard, walks off with the other, and shakes his end sideways. You watch the ripple travel. It takes time to cross the room. That, Carson says, is the transverse wave, and that finite travel time is the speed of light — in classical theory, the speed limit. Then he does it again, differently.
This time he jerks the hose endwise — a straight shove along its own length. And the far end moves at the same instant. No ripple crossing the room. If you move one end, he says, the other end moves instantaneously; there is no time delay, and that is the one we completely ignored. Feel two things at once here, and keep both. One: something genuinely different just happened in the same hose. Two: a hose is not a wire, and a demonstration that lands in a room is not a measurement. We will come back to this hose.
There is also a measurement, and it is stranger. January nineteen eighty-eight, a small lab in northern California, a camera running. Dollard takes twenty-seven and a half metres of wire — ninety feet — and asks it one question: how fast does a disturbance travel down you? Laid out as ordinary coaxial cable — a centre wire running inside a metal tube that shields it, the two sharing one axis — the answer is sensible. Eighty-one percent of the speed of light. And that figure is not a discovery: a signal in a cable is dragged below light speed by the insulating material packed between the wire and its shield, and eighty-one percent is, near enough, what the manufacturer prints on the datasheet for that cable. So the ruler here is thoroughly ordinary, which means everything rests on what he does next. He winds that same wire into a Tesla coil — a tall skinny coil wound with many turns of fine wire, dozens to hundreds depending on the build, one end of it left free, driven until it rings at its own natural frequency — and asks again. This time: one hundred and twenty-six percent of the speed of light. Same wire, same room, same afternoon — twenty-six percent over the universal speed limit. Or, as he puts it on camera, we might get a ticket from the relativity police.
The man doing this is Eric Dollard, and he is worth placing properly, because everything written about him online is worship or dismissal, both lazy. As a young man he had the run of a dying wireless station at Bolinas on the California coast — the Marconi station RCA took over, half-abandoned above the Pacific. By his own account he ran a small research laboratory in the Alexanderson room of the Marconi building, repairing and installing cables to earn his stay. Not a job. Not a title. A kid with keys to a cathedral. And he has spent the forty years since arguing one thing: when the industry chose which half of electricity to build the world out of, it picked the wrong half.
Now the part you would normally hear first. The internet will tell you Bell Labs called Dollard their Golden Boy and their Angel of Electricity, that he holds a doctorate, that the Navy trained him. None of that is his. It traces to one anonymous fan article, by a man Dollard's own organisation later denounced as a fabricator, plus a line of publisher's marketing copy. What he says about himself is far smaller and has not changed in decades. The other side is just as plain: no peer-reviewed papers, no patents, five stapled booklets with no editors or referees. Which is why no mainstream rebuttal exists — there is nothing in the literature to rebut. Neither fact decides whether the physics is right. That is what the rest of this is for.
So, the deal, before we go further — this subject is a minefield of bad faith in both directions. This document keeps three registers separate and never blends them. One: what Dollard's system claims, in his own terms, as strongly as he would — because a worldview you have only met in caricature is one you do not understand. Two: what standard physics and electrical engineering actually say. Not what a forum thinks they say: the named effect, the equation, the measurement, with its source. Three: what an experiment could settle — the bench test, both predictions, how to read the result. When they agree, you will be told. When they collide, you will see both and where the evidence points. What you will never get here is register one dressed up as register two.
One more thing, because people get this wrong. Understanding a worldview and endorsing it are different acts. You can follow Dollard's reasoning all the way down, see exactly why it convinces working engineers, and still come out certain he is wrong about the central claim. That is not fence-sitting. It is the only honest way to read anybody. What you get for the hour is the real engineering with the numbers left in — not Tesla had secrets, but actual impedances, frequencies, coil dimensions, and where each came from. And a verdict at the end: what is right, what is wrong, and what nobody has bothered to check.
Fair warning: the answer is genuinely mixed, and if you came for a clean win in either direction you will be let down twice. Of the two hundred claims here, about a hundred turn out to be textbook physics in unfamiliar clothes — real effects most engineering degrees simply skip. Eighty-nine are checkable differences, where his prediction and the standard one come apart and an afternoon on a bench settles it. Nine are unfalsifiable, built so no measurement could ever disagree. And the one hundred and twenty-six percent? A real measurement. He took it, he published a table showing the number slides with the shape of the coil, and two years before that film he wrote in print that it does appear to be a phase velocity, not a group velocity — the speed at which the crests of a pattern seem to slide past you, rather than the speed at which the energy and the signal actually get anywhere. Only the second of those is capped at light speed. Nothing went wrong with the measurement. Something went wrong with the reading of it. Let us go and find out what.
So here's the ledger. Everything Dollard says across the whole corpus, deduplicated down to two hundred distinct claims, each one dropped into one of three piles. The sorting rule isn't "do I believe this" — nobody needs my opinion. The rule is: what would it actually take to find out? That question has exactly three answers, and they're what the tabs above this list do. One hundred and one claims agree with standard physics. Ninety make a testable difference. Nine are unfalsifiable.
Pile one is the biggest, and its flagship is the claim that sounds most like heresy. Dollard says the energy in a circuit doesn't travel through the copper — it travels in the space around the wires, and the wires are only boundaries that steer it. That is not fringe. That is the Poynting vector, the electric field crossed with the magnetic field, standard since eighteen eighty-four, taught in Feynman's Lectures, Volume Two, chapter twenty-seven — including the provocative part where energy enters a resistor sideways, through its skin, not along the leads.
"Agrees" doesn't mean trivial and it doesn't mean wrong. It means both accounts predict the same needle on the same meter. And this pile has a startling member — the centrepiece measurement of his career. In nineteen eighty-eight he measured a Tesla extra coil ringing far higher than its wire length says it should, and read that as a wave moving at one hundred and twenty-six percent of light speed. The measurement is real. It's also exactly what ordinary helical-resonator theory predicts for that coil. Two stories, one reading.
Pile two is where the document earns its keep: eighty-nine claims where a bench experiment gives different numbers depending on who's right. Here's one. Dollard says his longitudinal wave crosses the earth at two hundred and ninety-one thousand miles a second — pi over two times the speed of light. So key a ground-coupled transmitter against a GPS-locked clock, put a receiver two hundred miles away, read the arrival time. He predicts two hundred and ninety-one thousand; standard physics says nothing arrives before one hundred and eighty-six thousand. Nowhere to hide.
Another, cheaper. He claims several receivers tuned to the same transmitter can each draw full power without dividing it between them — the pie doesn't get cut. So put a wattmeter on one receiver, switch on a second, then a third, and watch what the first reads. Standard theory says the number falls. He says it holds. Same shape at Landers: two aerials, one lightning strike, and he predicts the ground channel hears it before the radio one.
Pile three is tiny — nine claims — and needs the most care, because "unfalsifiable" gets used as an insult. Here it describes the shape of a claim, not the man. It means no instrument reading comes out differently depending on whether it's true. Counterspace is the clean case: the energy missing from his generator went into an inverse space that is, by construction, unreachable by measurement. And some may well be true. "Morgan cut Tesla off once he realised he could never meter the power" could be right — it's an archive question, not a bench question.
Now the part a document like this normally skips quietly. Once the piles were built, a critic pass went back and argued with every label. Eight moved — and they moved in both directions. Four dropped out of testable into agrees; four climbed out of unfalsifiable into testable. That two-way traffic is the point. A sorting that only ever moves one way isn't a sorting, it's a verdict in a costume.
The demotion worth telling is the headline comparison in the nineteen eighty-eight film — the longitudinal network lighting a fluorescent tube while the transverse network manages a dim neon bulb. Same power going in, he says, within fifteen to twenty percent. Except in the same breath, on camera, he says why he only knows that roughly: we don't have a high frequency watt meter to measure this right now. So "same power" is an estimate, not a measurement — and the two networks also differ in impedance, resonance and sharpness of tuning, so a brighter lamp is what plain filter theory expects. That comparison is what the whole demonstration rests on.
The four promotions cut the other way, and they were a failure of attention rather than of charity. Each time the vague version had been kept while the specific one got thrown away — and the specific one was testable. "The plasma forms looked alive" is poetry. "They never went to the same place twice" is spatial statistics: film the discharges, pull the strike coordinates, test them against a random spread. Same with the golden-ratio claim, whose discarded version named thirty-six and seventy-two degree angles and a frequency condition you settle with a Fourier transform.
So use the tabs, filter it, argue with it. But here's the number to carry out. One hundred and thirty-three of the two hundred claims carry a number, a named device, or a recipe you could build from — a figure with units on it, not a mood. Two thirds of the corpus is checkable, in principle, by one person with a signal generator, an oscilloscope and a free weekend. Whatever you conclude about Eric Dollard, he mostly said things specific enough to be wrong. That's rarer than it sounds.
Forty quotes, no filter. One instrument first, the most useful thing on this page: his register shifts with the kind of claim he's making. Solid ground makes him specific. Contested ground makes him funny. Off the map, he goes grand. Same steady delivery throughout — the sound of the sentence tells you where you're standing.
Start with the ones where he's simply right, and better said than your textbook. The birds on the power line. Electricity trapped between two mirrors. Heaviside's perfect conductor that's really a perfect obstructor. All three are Poynting's theorem, eighteen eighty-four: the energy really does travel in the space between the wires. Nothing heretical. Just vivid.
The electron ones are that move pushed a step too far. He's right that electrons don't carry your energy — they crawl about ten inches an hour; one leaving your socket takes a day to reach the lamp. Textbook. "Electrons are the resistance" is the overreach; they are the charge carriers, measured a century ago. Listen for where the true sentence becomes the extra one.
Then the inversions — counterspace, per-centimeters, the space between the lines on the ruler, two points already one. Finest teaching in the corpus. The ruler line especially: a hard idea in a sentence you could repeat to a child. Whether counterspace is real is a separate argument. As explanation, it's superb.
Now the famous one, with the driest joke in physics attached — a hundred and twenty-six percent of the speed of light, and a ticket from the relativity police. Two years earlier he'd written in print that this does appear to be a phase velocity. And his own table slides that ratio from eighty percent to two hundred and eighty-seven, purely with coil shape. He is more careful than his fans.
The angry ones aren't all bluster. "There never really were any Maxwell's equations" is flatly true — the four you were taught are Heaviside's compression of Maxwell's twenty. Kidnapped and murdered is just the temperature of the grievance. With Heaviside: the fingernails and the block of stone are real biography. The "because of Einstein" is his own joinery.
One cluster isn't argument at all. It's grief. Tesla, victim of his own poetry. The pigeons and the library. The ark he's loading with dead engineers. The frog they'd rather have in formaldehyde than alive. That one's about a historical society blocking a tower replica — the most portable sentence he's ever said.
The funny ones are load-bearing too. Quarks out of my condensers is an epistemology on a bumper sticker. The cycles-per-second pedantry says naming a unit after a man forgets what the thing does. Thomson's toast — a Nobel laureate hoping his own electron stayed useless — is his favourite object.
Then it goes wide. A sun that's an anode inside a vacuum tube. Light we understand no better than Pythagoras did. Faraday's line that matter cannot act where it is not — the seed of the ether worldview. God, electricity and music in one breath. Notice these arrive in the same calm voice as the coil geometry. That's the tell.
And then, without warning, back to the bench. The tank-circuit Q he demands, against the ham handbook. Ground rods that are mostly for show. Your hand on the washing machine. Dielectric discharge as one of the most unpleasant things an engineer meets — gorgeous old prose for "this will hurt you." All checkable.
Two that should move your prior. He takes an axe to scalar waves — the thing half his audience believes in — with one clean point: the scalar is the part that doesn't wave. And he says outright there is no free energy. A man running a con doesn't take those two things from his own crowd.
Last thing, and it changes how you hear all forty. The inflated résumé online — the Bell Labs golden boy, the doctorate — is not his. It traces to a fan he later called a fabricator, and to marketing copy. What he claims is smaller and hasn't moved in forty years: Farnsworth telling him to use mathematics when you have no equipment, and ten years of this work done on park benches.
Right — this is the one that matters. Everything so far has been Dollard's system on its own terms. This is the honest verdict: not a debunking, not a defence. Four movements — what he gets right that nobody teaches, the measurement everything rests on and why it fails, one mistake repeating across a hundred and fifty-four years, and what's still open.
Start with what's right. "Electricity flows in the space between the wires, not inside them" sounds like heresy. It's Poynting's theorem, eighteen eighty-four, and it's in the Feynman Lectures — volume two, chapter twenty-seven — energy arriving through a resistor's side surface, not down the copper. Not fringe. Textbook. Just so badly taught that engineers hear it and assume they're being sold something.
Second: the circuit theory every engineer learns genuinely does break down for these coils. And the loudest people saying so aren't his fans — they're the Corum brothers, IEEE-published Tesla-coil researchers, who wrote that lumped-element theory "does not, and cannot" describe a structure where the finite speed of light matters inside it.
Third: the units. When he says a capacitance is so many centimetres, that isn't eccentricity — in the CGS system Heaviside and Steinmetz worked in, capacitance is literally a length, and a metal sphere's capacitance equals its radius in centimetres. His complaint that the reciprocal vocabulary got binned is documented too: in nineteen eleven a Cornell professor, Vladimir Karapetoff, named the inverse henry the yrneh — henry backwards, the same joke as mho for the inverse ohm. Published, never used, gone.
And one more people assume can't exist. A refereed IEEE paper — a group at the University of Split, two thousand and eight — modelled Tesla's transmitter as an antenna over lossy ground and found no significant radiation in the Hertzian sense. Peer review landing on the structural claim: the thing is not a radio transmitter. It doesn't say longitudinal waves exist. But it's real, and it's on his side.
Now the centrepiece. It's the spine of everything, so it deserves a fair beating. Nineteen eighty-eight, on film: an extra coil wound with twenty-seven and a half metres of wire. At light speed along that wire, quarter-wave resonance lands at two thousand seven hundred and twenty-seven kilocycles. He measures three thousand four hundred. Divide: one point two five. A hundred and twenty-six percent of the speed of light. The measurement is real; the arithmetic is correct.
Here's why it fails anyway. A coil's resonance isn't wire length divided by light speed. It's set by axial height, diameter and pitch — how far apart the turns sit — because the wave climbs the axis, hopping turn to turn, not winding round the copper. The Corums solved that properly, and their model, from people sympathetic to Tesla, predicts that exact measurement with a wave that isn't fast at all. It's crawling: two to four percent of light speed.
And this should land hardest. Run the two standard textbook models on an ordinary Tesla secondary, then do Dollard's division on their predictions. One gives a hundred and sixty-five percent of light speed, the other two hundred and seventy-five. Mainstream physics routinely predicts a bigger excess than the one he measured and called anomalous.
Then three things close it, all from inside his own camp. One: his own table. Theory of Wireless Power, nineteen eighty-six — the ratio running from about eighty percent up to two hundred and eighty-seven, purely by changing the shape of the winding. A velocity is a property of the medium. It doesn't change because you wound the same wire on a fatter former. That's geometry wearing a velocity's clothes.
Two: the control. On that same film he measures a coax cable as the comparison and gets eighty-one percent of light speed — the neat matched opposite of the coil's hundred and twenty-six. But eighty-one percent is, within error, the published velocity factor of RG-62. It's on the datasheet. Not a matched pair of anomalies — one anomaly and one datasheet.
Three — the most disarming line in the corpus, and it's his. Nineteen eighty-six, in print, two years before the film: "It is interesting to note that the velocity measured on the Tesla coil is also pi over two greater than the velocity of light, but this does appear to be a phase velocity rather than a group velocity." Phase velocity above light speed is standard physics; it carries no energy and no information. He conceded the distinction the argument turns on. His followers hardened it. He didn't.
Which brings us to the finding that ties this document together. The same mistake happens three times, a hundred and fifty-four years apart — and each generation cites the previous one as independent proof.
Eighteen thirty-four. Charles Wheatstone runs half a mile of wire through the vaults under King's College, sparks it, and watches the flashes in a spinning mirror. The number that comes down to us is two hundred and eighty-eight thousand miles a second — faster than light, and the origin of this whole legend. It isn't a measurement. He resolved no displacement at all; the deviation, he says, could not have exceeded half a degree — a lower bound, not a result. "I am not prepared to state the results with numerical accuracy." Biographers hardened a detection limit into a fact.
He also timed the spinning mirror by ear — matched its pitch to G-sharp, inferred eight hundred revolutions a second. His own later footnote admits a mechanical counter gave not more than six hundred. Twenty-five percent off, on the experiment's most important constant.
Nineteen oh-five. Tesla files US patent seven eight seven, four one two, states a propagation velocity of four hundred and seventy-one thousand two hundred and forty kilometres a second, and derives his earth-resonance frequencies from it. That number is pi over two times the speed of light, exactly. Not measured, not inferred — assumed, as an input, on the way in. Tesla's earth-resonance arithmetic is superluminal by construction.
Nineteen eighty-eight. Dollard measures a coil, divides by wire length, gets one point two five, and cites Tesla's pi over two as independent corroboration. Nobody in that chain ever measured pi over two. And the diagnosis was there all along: Weber, in the eighteen-fifties, called Wheatstone's result meaningless, because the propagation velocity came out depending on the length of the circuit — so it "was not well defined."
So what's settled. Three things, refuted cleanly. The sun emits no transverse energy — no. Hale measured plane-polarised sunlight in nineteen oh-eight, and a longitudinal wave has no polarisation to measure; every solar magnetograph built since is a polarimeter. Net energy from a fusor — no. Rider's nineteen ninety-five analysis shows the loss is structural, not an engineering problem. It genuinely fuses, amateurs do it in garages — it just never comes out ahead.
And wireless power at planetary scale — the Navy built exactly that architecture and published the numbers. Project ELF: forty-five kilometres of buried ground dipole, two point six megawatts in, eight watts radiated. Not eight kilowatts. Eight. About half a bit a minute. Not incompetence — the ratio of antenna size to wavelength. Tesla's error was never propagation — it was coupling. The launch, not the path.
Now the part it would be dishonest to skip. The Zenneck wave — a signal bound to the ground's surface rather than radiating away — is a real, exact solution of Maxwell's equations, and has been since nineteen oh-seven. Whether you can launch one efficiently over real earth has been argued by top antenna theorists for a hundred and eighteen years and is still not closed. The disagreement is about aperture size, not physics. No peer-reviewed, replicated demonstration exists either way.
Which leaves the man. What's documented is real and modest: Science Director, then Vice-President, of the Borderland Sciences Research Foundation, nineteen eighty-six through eighty-seven, with the title "Wireless Engineer" printed beside his name on the organisation's own masthead. Five self-published booklets, one conference paper, no refereed journal article ever, and no patents.
And here's the part that cuts in his favour — precisely what most people use against him. The inflated résumé, Bell Labs calling him their "Golden Boy," trained by RCA and the Navy, "Doctor" Eric Dollard — is not his. It traces to two documents: an anonymous fan article from twenty twelve, by someone his own organisation later called a fabricator, and a line of publisher's marketing copy. It appears in no source he controls. What he says about himself is smaller, and hasn't changed since nineteen eighty-seven.
He never published in a refereed venue, which is the dull reason mainstream engineering never engaged him — there was nothing in the literature to engage. But the people who build and model these coils did look. Paul Nicholson, who maintains their reference simulation, first concedes the physics — yes, the coil's normal oscillation is a longitudinal mode, we model them routinely — then rejects the vocabulary. The effects are real and already modelled; the exotic language isn't needed.
So — what feeds the ledger at slash evidence. One live bet: efficient Zenneck launch over real earth. One bench experiment, absurdly cheap — wind two coils with identical wire length and different diameters, sweep them, and watch the so-called velocity slide from sixty-eight percent to a hundred and sixty-four percent of light speed on shape alone. A spool of magnet wire and an afternoon settles the question that generated the whole system. Three closed cases with receipts: the polarised sun, the fusor, and eight watts. And one thing that isn't a verdict at all — a detection limit in eighteen thirty-four became a constant, the constant became an assumption inside a patent, and the assumption came back a century later as corroboration. Nobody lied. Everybody cited. The chain broke anyway.
Part two: two waves, one wire. Everything here is Dollard sorting the world into two piles, and before he can do that he has to take the word "electricity" away from you and hand it back meaning something else entirely. Two terms will keep coming up, so let's plant them now: the magnetic field, which wraps around a wire in rings, and the dielectric field, which runs straight in and out of it like spokes on a wheel. Those two fields are the cast; the waves are what they do together.
We've just met two kinds of wave: the ordinary one that wiggles sideways, and Dollard's longitudinal one that pushes along the direction it travels. Now we go a level deeper and split electricity itself, because in his system there is no such thing as a single electric field — there are two of them, and they are opposites in almost every respect. This track is that pair: magnetism, the one that circles a wire, and dielectricity, the one that runs straight out of it. Then, at the end, the thing you get when you run both at once — which is the only thing Dollard will agree to call electricity at all. Get this straight and the strange arithmetic waiting in the next section stops being strange.
Right — this is the stretch where the arithmetic goes strange, and it's the theoretical spine of everything that comes after it. You've just met two fields that behave in opposite ways: one that grows by getting bigger, one that grows by getting closer. This section is what happens when Dollard takes that seriously enough to rewrite the units, then the algebra, then eventually the conservation of energy itself. Seven ideas: an inside-out space, upside-down units, a fourth way to store energy, a field that breathes, a clock-face algebra, speed that stops being a thing, and energy that's allowed to go missing.
Everything so far has been Dollard's argument — two waves, two fields, and a strange arithmetic to write them in. This part is where he stops arguing and starts building, so let's call it what it is: the machines. Nine of them, in order — the single coil he says does all the work, the transmitter that uses the planet as a component, the ground system that decides whether any of it works at all, and finally the thing in his workshop that makes a galaxy appear inside a dead lightbulb. Watch for the change in register here: this is the part of the story with a shovel in it.
Everything so far has been how the thing works and what he's built with it. Which leaves the question you're probably already holding: if any of this is real, where is it? This track is Dollard's answer, and it's history rather than physics — what was abandoned. Two concepts: a fork in the road that radio took a century ago, and the year he says the science of electricity died.
Everything up to here has been machines and history — coils, grounds, and the moment in 1919 when Dollard says the whole science got thrown out. This next stretch is the wider worldview: what he thinks electricity actually is, once you stop treating it as engineering. Three ideas, and they get harder to test as we go — electricity as a mass-free, almost living thing; sparks that come out shaped like seashells; and his own explanation for why none of this is in use. Worth saying out loud before we start: this is the part where you get to decide how much rope to give him.
Part seven: at the bench. Everything up to now has been theory — two waves, two fields, a strange arithmetic, a shelf of machines — and this is where Dollard stops philosophising and tells you what to build, how to drive it, and how to measure it without fooling yourself. Two bits of vocabulary run through the whole track: when he says kilocycles, read kilohertz, thousands of cycles a second; and decibels, dB, are just a compact way of writing enormous ratios, where every ten dB is another factor of ten in power. It is also, for my money, the most honest stretch of the entire corpus — because almost all of it can be checked by someone with a bench and a weekend.
We've spent the last stretch inside his lab — coils, shielding, the receiver at the end of the chain. Now walk outside and look up at the wires over the street, because this track is about the grid as adversary and instrument. It's the thing drowning out every faint signal he's trying to hear, and it's also the largest piece of electrical apparatus humans have ever built. Two concepts: a wiring decision that he says turned the entire network into an antenna, and a fact about power almost nobody notices — plus the much larger claim he hangs on it.
Everything up to here has been Dollard telling you how electricity works. This part is different — it is the part of his system that can lose. Seven experiments, most of them cheap, each with a stated result and a real way of coming back and saying no. If you keep one section of this whole thing, keep this one, because a claim you can check is worth more than a claim you can admire.
We have just spent a long stretch on things you could actually go and build. Now the opposite move — the paper trail. This track is where the mathematics actually comes from: which books he is standing on, which term he says got quietly deleted from the equations, and exactly which slot in the standard accounting he thinks is sitting empty. And it belongs right here, immediately before he leaves physics behind altogether, because this is the last section where every claim has a page number attached to it.
Last stop. Everything up to now has had something at the end of it you could actually check — a meter to read, a coil to wind, a paper from the eighteen-eighties to pull off a shelf. This track is where Dollard's system walks off the edge of physics and keeps right on walking, and that's the theme: where it leaves physics. Two concepts — a Sun with nothing inside it, and a gravity that pushes you instead of pulling you. The thing worth watching here isn't whether he's right; it's that he says all of this in exactly the same calm engineer's voice he used for coil geometry.