Let us imagine the substance of the fibril to be composed of, or at least to contain, the substances _a_ + _b_ which dissociate reversibly into the substances _c_ + _d_. At any moment, and in any particular physical state, as much of _a_ and _b_ pass into _c_ and _d_ as _c_ and _d_ pass into _a_ and _b_. There will be equilibrium. But now let a stimulus alter the physical conditions: prior to the stimulus the phase was _a_↓{m} + _b_↓{n} = _c_↓{p} + _d_↓{r}--the suffixes _m_, _n_, _p_, _r_, denoting the concentrations of _a_, _b_, _c_, and _d_--but after the stimulus the phase may be _a_↓{m1} + _b_↓{n1} = _c_↓{p1} + _d_↓{r1}. Now the element of the nerve substance (1) forms a system with the element (2). The condition in (2) is _a_↓{m} + _b_↓{n} = _c_↓{p} + _d_↓{r}, and that of (1) _a_↓{m1} + _b_↓{n1} = _c_↓{p1} + _d_↓{r1}, but these two together now fall into a new state of equilibrium and this is transmitted along the whole nerve-fibril with a velocity which belongs to the order of magnitude of that of chemical changes. If the stimulus remains constant (a constant electric current for instance), the new condition of equilibrium will be established throughout the whole length of the fibril and the nervous impulse will be a momentary one (as it is in this case). But if the stimulus is an intermittent one (an interrupted electric current, light-vibration, sound-vibrations), then in the intervals the former condition of equilibrium will become re-established and the nervous impulse will be intermittent (as it is). There would be no work done on the whole in the changes, except that done by the transmission of the changed state of equilibrium to the substance of the effector organ in which the nerve-fibril terminates--the substance of a muscle fibre, or the cell of a secretory gland, for instances. There would, probably, be a certain dissipation of energy as in the case of the propagation of an electric impulse through a poor conductor, but all our knowledge of the chemistry of the nerve fibre points to this amount of dissipation as tending to vanish.