is not diminished by previous use, as it would be were it dependent upon their reserve of nutriment. These various considerations rule out a “chemical” explanation of the old-fashioned type. It is premature to do more than outline the “physical” theory which seems destined to take its place; and the reader will perhaps forgive if, for the sake of clearness, the case is put with unjustifiable definiteness and simplicity. Proteid substances are constituted of clusters of molecules. The form of the clusters depends upon the salts (or, more precisely, the ions) with which they are associated, and the associations depend upon the electric charges which the ions carry. In resting nerve-protoplasm the clusters are small, and, since the total surface-area of a number of small spheres is greater than the surface-area of the same weight of matter when condensed into large spheres, there is, so to speak, more surface for the ions to cling to. Conversely, when the ions leave the small clusters, the latter are not protected from the influence of mutual attraction. They fuse into larger clusters. Fusion is carried to its extreme limits when a protein coagulates. A nerve-impulse is a “wave” of partial coagulation. The positive electricity generated in a cell-body by the metabolism of its tigroids repels the positively charged ions which cling to the nearest protoplasm-clusters in the axon. Like acrobats swinging from trapeze to trapeze, each flight of ions dispossesses the ions from the clusters in front of it; and in this way the disturbance progresses down the axon as an electric wave.