An action potential generated at one region of an axon can trigger action potentials in neighboring regions. This produces propagation without requiring the original patch of membrane to physically carry the signal down the axon.

When one segment depolarizes, positive charge spreads locally through the cytoplasm and extracellular fluid. This local current depolarizes adjacent membrane.

If the neighboring segment reaches threshold, its voltage-gated Na$^+$ channels open and generate a new action potential there.

active segment
    ↓ local current
adjacent segment depolarizes
    ↓ reaches threshold
new action potential

The signal is therefore regenerated repeatedly along the axon.

Why action-potential amplitude does not fade with distance

Passive voltage changes weaken as current leaks through the membrane and spreads through the cytoplasm. An action potential avoids progressive loss because each new axonal segment actively regenerates the full voltage-gated response.

This is different from simply conducting one fixed pulse of charge through a wire.

Why propagation is normally forward

Local current spreads in both directions from an active segment, but the membrane immediately behind the action potential is refractory because many Na$^+$ channels are inactivated and K$^+$ conductance remains elevated.

The region ahead is excitable, while the region behind is temporarily less able or unable to fire. This asymmetry makes propagation effectively unidirectional under normal conditions.

Conduction speed

Propagation is faster when local current can influence more distant membrane before dissipating. Larger axon diameter reduces resistance to longitudinal current flow, and electrical insulation by myelin can greatly increase effective conduction speed.

The key mechanism is therefore local passive current coupled to repeated regenerative action-potential generation.