Abstract
The question of how quickly energy propagates in an electrical circuit has been the subject of renewed debate following a widely discussed thought experiment. This article examines the physical principles underlying the propagation of electromagnetic energy in circuits, challenging the conventional view that electrons carry energy from source to load. Through an analysis of surface charge distributions, the Poynting vector, and transmission line theory, we demonstrate that energy is transported by electromagnetic fields external to the conductors, and that a load can receive appreciable power on timescales determined by the speed of light rather than by electron drift velocity. Experimental validation using a 10-metre scale model confirms theoretical predictions, revealing that a small but measurable current flows through a load within nanoseconds of switch closure, well before signals from the distant parts of the circuit could have returned. These findings underscore the pedagogical and practical importance of field-based descriptions of circuit behaviour.
1. Introduction
In a thought experiment that has stimulated considerable discussion, a battery, a switch, and a light bulb are connected by superconducting wires of length half a light-second (approximately 150,000 km), with the bulb positioned one metre from the battery and switch. The question posed is: after the switch is closed, how long does it take for the bulb to emit light? The counterintuitive answer — that the bulb begins to glow in approximately 1/c seconds, the time for light to traverse the one-metre gap — has been met with scepticism, with critics asserting that it would violate causality or that the initial power would be negligible. This article provides a rigorous physical analysis of the problem, drawing on established electromagnetic theory, and presents experimental evidence that confirms the field-based prediction. We argue that the apparent paradox arises from common misconceptions about the nature of energy transport in circuits, and that a proper understanding requires moving beyond the lumped-element model to consider the distributed electromagnetic fields that are the true carriers of energy.
2. The Role of Surface Charges in Establishing the Electric Field
A fundamental misconception is that electrons carry energy from the battery to the load, or that they push each other through the wire like marbles in a tube. In reality, the average drift velocity of electrons in a conductor is extremely small — on the order of millimetres per second — while their thermal speeds are of order 10⁶ m s⁻¹. The energy dissipated in a resistive load is not carried by the electrons themselves; rather, it is extracted from the local electric field that accelerates electrons between collisions with the lattice.
The electric field that drives current is not simply the field of the battery. Instead, it arises from a combination of the battery's field and charges that accumulate on the *surface* of the conductors. When a battery is connected into a circuit, even with the switch open, charges rearrange themselves on the wire surfaces: an excess of electrons on the negative side, a deficit on the positive side. This surface charge distribution adjusts dynamically until the internal electric field is zero everywhere inside the conductor — the condition for electrostatic equilibrium in the open circuit.
For a closed circuit carrying steady current, the surface charge gradient is not uniform. The largest gradient occurs across resistive elements, where the electric field is strongest. As Chabay and Sherwood have shown in their influential textbook *Matter and Interactions*, this surface charge description provides a unified treatment of electrostatics and circuits, resolving phenomena that are otherwise mysterious. They write: *“By emphasizing the crucial role played by charges on the surfaces of circuit elements, it is possible to describe circuit behaviour directly in terms of charge and electric field”*.
When the switch is closed, the surface charges on either side of the gap neutralise each other on contact. The resulting change in the surface charge distribution propagates outward at the speed of light, modifying the electric field in the space around the wires. When this electromagnetic disturbance reaches the load, the electric field inside it becomes non-zero, and current begins to flow. This is the physical basis for the 1/c response: the bulb is affected by the change in the local field, not by the completion of a global circuit loop.
3. Energy Transport and the Poynting Vector
If electrons do not carry energy, how is energy transferred from the battery to the load? The answer is given by the Poynting vector S = E × H, which describes the flux of electromagnetic energy. For a DC circuit, the Poynting vector points outward from the battery through the space surrounding the wires, then inward toward the load. Energy flows in the fields *external* to the conductors, not through the metal itself.
This is not merely a mathematical convenience: it is a physical reality that has been confirmed by numerous analyses. In a simple circuit with a battery and a resistor, the vast majority of the energy flows through the space immediately adjacent to the wires, guided by the fields that are shaped by the surface charges. The wires serve to channel these fields, much as a waveguide channels electromagnetic radiation, rather than acting as a pipe for electrons.
In the Veritasium thought experiment, after the switch is closed, the Poynting vector initially points from the battery across the one-metre gap toward the load, regardless of whether the far ends of the wires are connected. This is because the fields that carry energy are established locally, not by the global topology of the circuit. As one commentator notes, *“electrical energy is not transported by electrons, it is transported by fields”*. This explains why a disconnected wire placed nearby can experience virtually identical fields to the connected one, at least until reflections return from the far end.
4. Transmission Line Theory and the Distributed Element Model
The lumped-element model — in which resistors, capacitors, and inductors are treated as discrete components — is a powerful approximation for circuits whose physical dimensions are small compared with the wavelength of the signals involved. However, for circuits with very long wires, this approximation fails. The distributed-element model, or transmission-line model, treats resistance, capacitance, and inductance as continuously distributed along the length of the line.
In the Veritasium circuit, the parallel wires form a transmission line with a characteristic impedance Z₀ = √(L/C), where L and C are the inductance and capacitance per unit length. When the switch is closed, the source initially "sees" this characteristic impedance, and a voltage wave propagates down the line at the speed of light. The initial current through the load is determined by the voltage divider formed by the load resistance and the characteristic impedance of the transmission lines.
For the experimental setup with wires 10 metres long and spaced one metre apart, the measured characteristic impedance was approximately 550 Ω. With a load resistor of 1.1 kΩ, the initial voltage across the load was about 4 volts from a source of approximately 18 volts, corresponding to a current of about 4 mA and a power of approximately 14 mW. This is many orders of magnitude greater than any leakage current, and sufficient to produce visible light from an LED. The experimental results, obtained using high-speed oscilloscopes at Caltech, confirm that a signal arrives at the load within nanoseconds, precisely as predicted by field theory.
5. Causality and the Speed of Information
A common objection to the 1/c answer is that it appears to violate causality: it seems to imply that information about the state of the far ends of the circuit (which may be broken) can reach the bulb faster than light. This objection is based on a misunderstanding. The initial current through the load does not depend on whether the far ends of the wires are connected or broken.
As the transmission line analysis shows, the load initially responds to the local electromagnetic field, and only later — after a time of order L/c, where L is the length of the wires — does it receive reflections from the far ends. In the words of one analysis, *“a bulb is not designed to simply pull passing EM radiation energy like an antenna. It is designed specifically to receive electrical energy through the non-radiating fields that accompany the wires it is attached to”*. The bulb responds to the fields that are established in its immediate vicinity; these fields are not dependent on the global connectivity of the circuit on timescales shorter than the light travel time to the far ends. Thus, causality is preserved: no information travels faster than light.
6. Experimental Validation and Pedagogical Implications
The theoretical predictions have been validated by multiple independent experimental efforts. In a scaled-down model with 10-metre wires, the initial voltage across the load was clearly visible above the noise floor, confirming that a non-negligible current flows within the first few nanoseconds. Similar results were obtained by other researchers using kilometre-scale wires. These experiments demonstrate that the field-based description is not merely a theoretical abstraction but a physically observable reality.
The pedagogical implications are significant. Most introductory textbooks omit the surface charge description of circuits, leading to persistent misconceptions about the nature of current and energy flow. As Chabay and Sherwood have argued, a unified treatment of electrostatics and circuits based on surface charges and fields can greatly enhance conceptual understanding. The Veritasium thought experiment, despite its initial controversy, has served to highlight these issues and stimulate renewed interest in the fundamental physics of circuits.
7. Conclusion
The question of how quickly a light bulb lights up in a very long circuit is not a trick question, but a window into the deep physics of electromagnetic fields. The answer — approximately 1/c seconds — follows directly from Maxwell's equations and is confirmed by experiment. The key insights are that energy is carried by fields, not electrons; that surface charges are essential for establishing the electric field in a circuit; and that the lumped-element model is only an approximation that fails when circuit dimensions are large compared with signal wavelengths. As one veteran circuit designer put it: *“I used to think in terms of voltage and current. And I used to think that the energy in the circuit was in the voltage and current, but it's not. The energy in the circuit is in the fields.”* This perspective, though counterintuitive, is essential for a correct understanding of electrical circuits and for the design of high-speed electronic systems.
References
- Chabay, R. W., & Sherwood, B. A. *Matter and Interactions*, 4th ed. Wiley.
- Chabay, R., & Sherwood, B. "A Unified Treatment of Electrostatics and Circuits." (1999). Available at matterandinteractions.org.
- Chabay, R., & Sherwood, B. "Polarization in electrostatics and circuits: Computing and visualizing surface charge distributions." *American Journal of Physics* 87, 341 (2019).
- "Surface charges on current carrying conductors." Physlab, LUMS.
- "Energy transfer in electrical circuits: A qualitative account." *American Journal of Physics* (2005).
- "DC Power Transported by Two Infinite Parallel Wires." arXiv preprint.
- "Transmission line theory." Electronics Specifier.
- "Lumped-Element Model." LibreTexts Physics.
- "The Cable to the Moon: Veritasium's Light Bulb Experiment in Low-Cost Miniature Form." arXiv:2502.xxxxx (2025).
- Physics Stack Exchange discussion on energy flow in circuits.