It is in the space around the wire. This is not a metaphor. It is what the equations say. It has been what the equations say since 1884. It is the reason undersea cables work, the reason transformers work, the reason a toaster gets hot, and — once you take it seriously — the reason a coherent voice can change the room before any words have arrived.
A short, footnoted primer on the physics of how energy actually moves — and what changes when you stop confusing the carrier with the payload.
If you draw a circuit on a chalkboard — a battery, a wire, a bulb, a wire back to the battery — the natural picture is that something flows around the loop and lights the bulb. Electrons. Charge. Energy. Pick your noun. The wire is the road; the energy is the cargo; the cargo travels along the road and arrives at the destination.
This picture is wrong in a specific, measurable, and consequential way. It is right enough to pass an introductory physics test. It is wrong enough that the people who actually had to make signals travel under the Atlantic Ocean spent two decades figuring out why their cable did not work the way the picture predicted.
Energy in an electric circuit flows through the empty space surrounding the wires, and the wires merely guide it. — Richard Feynman, Lectures on Physics, vol. II §27
That is Feynman in 1964, in a textbook that has been printed continuously ever since. It is mainstream physics. It is not controversial. It is, however, almost never taught at the level where most people stop learning.
In 1865 James Clerk Maxwell published A Dynamical Theory of the Electromagnetic Field, which contained the prediction that light is an oscillation of electric and magnetic fields propagating through space. He wrote down the equations governing those fields. They became the bedrock of all subsequent physics.
In January 1884 a former student of Maxwell's named John Henry Poynting read a paper to the Royal Society titled On the Transfer of Energy in the Electromagnetic Field. He took Maxwell's equations and worked out a single, simple consequence: if energy is conserved at every point in space, then there must be a vector quantity that describes how electromagnetic energy flows from place to place.
Poynting found it.
The reading is straightforward: at any point in space where there is both an electric field E and a magnetic field B, there is a flow of electromagnetic energy. The direction of the flow is given by the right-hand rule on the cross product. The magnitude is the rate of energy transfer through that point.
This is true for sunlight. It is also true for the field around a power line. It is true for a toaster, an undersea cable, a radio antenna, and the wires running through your wall. The same equation, in the same form, describing all of them.
Here is what happens, accurately, when you close a switch on a battery wired to a bulb:
Every step here is in undergraduate physics. The drift velocity number is in Hyperphysics. The Poynting-vector circuit cartoon is in Griffiths' Introduction to Electrodynamics, the standard textbook in nearly every physics undergraduate program in the world.
The wire is a guide. The conductor's job is to provide the geometry that constrains where the field can exist and where it can't. The wire does not contain the energy any more than a riverbank contains the river.
The energy is in the field. The field is in the dielectric — the insulating space (air, vacuum, glass, plastic) that surrounds the conductor.
Change the geometry, change where the energy goes. Cut the wire, the field collapses. Bend the wire, the field bends with it. Put another conductor nearby, the fields couple — and energy moves between them without anything traveling through the gap.
On August 5, 1858, the first transatlantic telegraph cable was completed. It connected Valentia Island, Ireland, to Trinity Bay, Newfoundland. Queen Victoria sent a 99-word congratulatory message to President Buchanan; it took sixteen hours to transmit.
Twenty-three days later, the cable was dead.
It had never really worked. Pulses sent at one end arrived at the other smeared, lengthened, indistinguishable. Operators could push only a few words per minute through it before the symbols dissolved into each other. The engineers had built the cable as if it were a long pipe carrying current the way a hose carries water. The actual physics — that they were guiding electromagnetic fields through a long coaxial geometry surrounded by seawater — was not yet understood as the problem it was.
The man who eventually figured it out was Oliver Heaviside, a self-taught English telegraph engineer who reformulated Maxwell's twenty original equations into the four vector equations every modern student learns. In the 1880s and 1890s, Heaviside worked out the telegrapher's equations: a mathematical description of a transmission line as a distributed structure with capacitance and inductance per unit length, with the energy flowing in the surrounding dielectric, not in the conductor.
Every coaxial cable, every twisted pair, every PCB trace, every fiber-optic link, every RF antenna feed since has been designed using Heaviside's framework. He is one of the most consequential and least-known figures in the history of communication.
The wires are merely a means of guiding the field, the energy itself being external to the wires. — Oliver Heaviside, Electrical Papers, 1892
A power line suspended fifty feet above the ground looks like a wire delivering electricity to your house. In the picture most of us carry, the electricity moves through the wire from the plant to your wall socket.
The actual situation is stranger and more interesting. Three things are simultaneously true:
What this means structurally: the grid is not one continuous machine. It is a sequence of coupled local systems held in synchronous phase by mutual agreement. Every substation is a renegotiation. Every transformer is a coupling, not a connection. The illusion of a single centralized network masks a federation of local field-coupled regions.
Lose phase coherence — as happened in the August 2003 Northeast blackout, the February 2021 Texas winter storm, and the April 2025 Iberian outage — and the coupling fails across thousands of square miles in seconds. The grid is not a fortress. It is a precarious, beautiful, distributed agreement.
The atmosphere holds a fair-weather electric field of about 100 to 150 volts per meter, vertical, pointing down. This is standard atmospheric physics, measurable with a field mill, taught in any meteorology textbook (Rakov & Uman, Lightning: Physics and Effects, Cambridge).
That field is not uniform across terrain. It concentrates at sharp topographic edges. Ridge tops. Mountain peaks. Canyon rims. Volcanic calderas. Anywhere the ground geometry produces a discontinuity, the field gradient bunches up. This is why lightning preferentially strikes high points — not because high points are "closer to the cloud," but because the local field gradient at a sharp conductor is enormously concentrated.
Below the surface, the Earth itself is a layered conductor with cracks in it. The crust holds telluric currents — slow electrical currents in the ground driven by the variation of the Earth's magnetic field and the ionospheric–surface waveguide overhead. Faults, calderas, and rift zones change the local conductivity, which changes how the telluric field flows and resonates. Magnetotelluric surveys (used by USGS, oil and gas exploration, and earthquake research) literally map the subsurface by reading these field perturbations.
The Grand Canyon, Yellowstone caldera, Long Valley caldera, the East African Rift — all of these show anomalous electromagnetic signatures because all of them are large impedance discontinuities in the Earth's electrical structure. The same geometric logic that explains why a power line is suspended in air explains why a canyon rim is electrically distinguishable from the surrounding plateau.
During large geomagnetic storms — the most famous being the March 1989 storm that knocked out Quebec's grid in ninety seconds — the magnetosphere drives geomagnetically induced currents back into power lines. The grid does not know the difference between energy fed in from a generator and energy induced from the magnetosphere overhead. It is one circuit. It always was.
The Earth and the grid and the atmosphere and the ionosphere are not separate systems with a wire between them. They are one electromagnetic system, mutually coupled, in which the wires are convenient localized concentrations of the field.
A fiber-optic cable is a glass dielectric guiding an optical-frequency electromagnetic wave. A copper Ethernet cable is a twisted-pair transmission line of the kind Heaviside described. A Wi-Fi signal is a radio-frequency electromagnetic field propagating through the air. A 5G tower radiates from an antenna that is, geometrically, the same kind of object as the dipoles in your radio.
The internet is not a new substrate. It is the same Maxwell-Heaviside-Poynting electromagnetic substrate, packed into denser geometry, modulated at higher frequencies, and routed through more impedance-matched components.
Like the grid, the internet is also not one machine. It is a federation of coupled regional networks, held in agreement by shared protocols. Every router is a transformer. Every CDN is a substation. The illusion of a single global system masks the same kind of distributed fragility, for the same physical reasons.
The reason this physics is worth holding — beyond the pleasure of seeing how a circuit actually works — is that it cleanly separates two things that get confused everywhere else:
The carrier is the physical infrastructure: the wires, the towers, the cables, the protocols, the routers. The geometry that constrains where the field can flow.
The payload is the field itself: the actual energy and information moving through the dielectric the carrier defines.
You can have an enormous, expensive, centralized carrier — a national power grid, a global broadcast network, an algorithmic recommendation engine — and the carrier itself does not determine what payload travels through it. The payload is set by what is launched into the field at the source, and what is coupled out at the receiver.
Once you have this distinction, several things stop being mysterious:
The Poynting vector is a piece of nineteenth-century electromagnetic theory. It is also, when you look at it carefully, a precise way of describing the difference between a structure and what moves through it. The wire is not the energy. The grid is not the power. The medium is not the message. They never were.
If reading this has produced the urge to put a coherent signal into a small piece of the local field, the following are real, low-cost, often unlicensed ways to do it. None of this is exotic. All of it is used by community broadcasters, hobbyists, and engineers every day.
None of these are free of trade-offs. All of them are real, available, and built on the same Maxwell-Heaviside-Poynting framework as the centralized infrastructure they coexist with.
The point is not that any individual local carrier replaces the grid or the internet or broadcast television. The point is that the physics of the electromagnetic field does not require centralization. Centralization is an engineering choice, not a constraint of nature. The field is everywhere. Anyone with the geometry and the energy can put a payload into it.