The axonal membrane is not the seat of nerve conduction: it is the boundary between two osmotic reservoirs whose asymmetry is the thermodynamic engine of the action potential. Voltage-gated ion channels are not the generators of the nerve signal: they are its osmotic amplifiers, and their spatial distribution along the axon is a geometric necessity, not an arbitrary anatomical feature. The Ionic-Mechano-Hydraulic (IMH) model formalises this principle: intracellular K+ adsorbed on the cytoplasmic polyelectrolyte gel triggers an ionic phase transition; extracellular Na+ amplifies the resulting hydraulic wave through Nav channels; Kv channels close the osmotic cycle and enforce the refractory period. The conduction velocity is predicted from the elastic modulus of myelin, not from the density of the sodium channel. The model resolves a 75-year-old anomaly that Huxley and Stämpfli themselves described as impossible in a purely electrical system: a positive current enters a node before the membrane potential at the preceding node has reached its maximum. Ten falsifi-able predictions are presented that cover myelin mechanics, mechanoreceptor adaptation, terminal arborisation geometry, velocity-diameter scaling, and axon diameter limits derived from first physical principles. The Hodgkin-Huxley model is not discarded: it is explained.