The Hodgkin-Huxley model of the action potential is usually taught as four differential equations. Its authority comes from the preparation it was built on: the squid giant axon, a nerve fiber large enough to thread a wire through. The procedural choices made in Naples and Plymouth in the late 1930s and 1940s are the reason those equations still describe spike initiation with unusual accuracy.

The Squid Axon as Chosen Instrument

The squid giant axon is a peripheral nerve fiber that can reach roughly half a millimeter in diameter, hundreds of times thicker than a typical mammalian unmyelinated axon. It was first described by L. W. Williams in 1909, then forgotten until J. Z. Young demonstrated its function in the 1930s while working at the Stazione Zoologica in Naples, the Marine Biological Association in Plymouth, and the Marine Biological Laboratory in Woods Hole.

The axon drives part of the squid's water jet propulsion system, which the animal uses for brief, fast movements. Evolution built the fiber for escape behavior, not for the convenience of electrophysiologists. That accidental size is the whole reason the axon became a standard preparation: no one designed it as a model organism.

Size matters because it changes what an electrode measures. A fine glass pipette pressed against a thin axon records a blurry average of many cells. A wire pulled through a half-millimeter fiber records one membrane. The difference between an averaged signal and a single-unit recording is the difference between a noisy correlation and a mechanism.

Why Squid, Why Naples, Why 1939

The Stazione Zoologica supplied live squid on a daily basis, which is not a trivial logistical achievement. A preparation that must be dissected within hours of the animal's death imposes a schedule on the experimenter. Naples had the boats, the tanks, and the staff to make that schedule routine.

Woods Hole and Plymouth offered alternatives, and work moved among the three sites across the decade. Each location had its own species and its own seasonal abundance. The choice of station was partly a choice of which squid would be available in which month, a constraint that rarely appears in the published methods.

Dissecting a single axon by hand and mounting it across a partition let the experimenter control the internal and external solutions separately. That control is what turned a nerve fiber into a measurable electrical component. The preparation was not a proxy for a human neuron; it was a system simple enough to solve.

The Voltage Clamp as Procedural Innovation

Kenneth Cole and George Marmont developed the voltage clamp in the 1940s to hold a membrane at a chosen potential while measuring the current that flowed. Alan Hodgkin and Andrew Huxley adapted the technique in 1949. The feedback circuit compares the measured membrane potential to a command voltage and injects whatever current is needed to keep the difference near zero.

Clamping the voltage separates two variables that normally move together. In an unclamped axon, current changes the voltage, and the voltage change alters the current. Holding one constant lets the other be read directly. This site has argued in a related piece on quasiparticle mass measurements that controlling one variable at a time is often what settles a dispute.

The clamp also bought time resolution. A step from one holding potential to another produces a transient current that rises and falls within a few milliseconds. Recording that transient from a single axon, rather than from a population, keeps the fast components from being smeared by conduction delays along the fiber.

Separating Sodium and Potassium Currents

To isolate the currents, Hodgkin and Huxley changed the external solution. Replacing sodium with choline removed the early inward current and left a slower outward current, which they attributed to potassium. The substitution was crude by modern standards, but it separated two conductances that had previously been entangled.

Tetrodotoxin, purified later from pufferfish, confirmed the sodium pathway by blocking it selectively. That confirmation arrived after the equations, which is worth remembering when a model is described as having predicted a channel. The model described a conductance; the pharmacology came later and matched.

Five papers appeared in the Journal of Physiology in 1952. The division into five parts was itself a methodological choice: each paper handled one piece of the problem, from the sodium hypothesis to the quantitative description. A reader who wants the equations can find them in the fifth; the earlier papers carry the measurements.

From Equations to Action Potential Shape

The model is a set of nonlinear differential equations that approximate the electrical behavior of an excitable membrane. Four state variables describe the membrane potential and three gating variables, each with its own voltage-dependent rate constants. The parameters were fit to clamp records from squid, not derived from first principles.

Fit to those records, the equations reproduce the spike threshold, the overshoot, and the refractory period. They also predict that a brief suprathreshold stimulus produces an all-or-none response, which is what the axon does. The agreement is close enough that the model is still used as a default in computational work.

Later recordings from vertebrate nodes of Ranvier found the same qualitative conductances, with different kinetics. That is a real generalization, and it is narrower than the claim that the model describes every neuron. The equations describe a membrane with two voltage-gated conductances; cells with more channels need more terms.

One concrete test came from temperature. Hodgkin and Huxley reported their experiments at around 6–8 °C, and the rate constants they measured are steeply temperature-dependent. When later workers warmed squid axons to roughly 20 °C, the action potential shortened and the kinetics sped up, but the overall shape remained recognisable. The model could be rescaled by a temperature factor, not rebuilt. That robustness is a direct consequence of fitting to raw transients rather than to averaged summaries, and it is why the equations survived the move to other preparations.

Perfusion and the Squeeze-Film Problem

Threading a wire through the axon is one way to sample the interior; another is to perfuse it. By the 1960s, workers had learned to squeeze the axoplasm out of a giant fiber and replace it with a defined internal solution, turning the axon into a test tube with a membrane. That advance came after the original equations, but it grew from the same logic: if you can control the inside and the outside separately, you can ask which side a conductance depends on.

Perfusion also introduced new artifacts. The membrane can be damaged by the roller used to expel the axoplasm, and the seal at the ends can leak. Investigators learned to check the resting potential and the action potential amplitude before trusting a record. Those checks became standard practice, and they are the reason a modern report includes a statement about resting potential and input resistance. The squid axon taught the field what to control, and also what to watch out for.

What the Squid Method Still Teaches

Check the size of the preparation before trusting the signal. A fiber large enough for internal perfusion gives a cleaner measurement than a small process, and a method that averages across cells will hide the mechanism you are looking for. The squid axon was chosen because it was big, not because it was representative.

Use a clamp to separate voltage from current. If both move during the measurement, the record cannot distinguish cause from consequence. A related piece on carrying a turbulence model into a new domain makes a similar point about borrowed formalisms.

Substitute ions to isolate conductance pathways, and confirm the substitution with a selective blocker when one becomes available. Fit equations to raw traces rather than to published averages, because averaging across preparations can flatten the transients that constrain the parameters. State the species and the temperature in every report, since rate constants are steeply temperature-dependent.

• Measure the axon diameter before choosing an electrode.
• Hold voltage constant while recording current, and hold current constant while recording voltage.
• Change the external solution one ion at a time.
• Report the temperature alongside every rate constant.

The trade-off is real. The squid axon is not a human neuron, and the model's parameters do not transfer without refitting. A preparation chosen for accessibility will always raise the question of relevance. Hodgkin and Huxley answered it by fitting to the preparation they had and stating the assumptions, which is a narrower claim than the one often attributed to them.