"It might be thought that this localizing power depended directly on nerve-distribution; that each touch-nerve had connection with a special brain-centre at one end (the excitation of which caused a sensation with a characteristic local sign), and at the other end was distributed over a certain skin-area, and that the larger this area the farther apart might two points be and still give rise to only one sensation. If this were so, however, the peripheral tactile areas (each being determined by the anatomical distribution of a nerve-fibre) must have definite unchangeable limits, which experiment shows that they do not possess. Suppose the small areas in Fig. 25 to each represent a peripheral area of nerve-distribution. If any two points in _c_ were touched we should according to the theory get but a single sensation; but if, while the compass-points remained the same distance apart, or were even approximated, one were placed in _c_ and the other on a contiguous area, two fibres would be stimulated and we ought to get two sensations; but such is not the case; on the same skin-region the points must be always the same distance apart, no matter how they be shifted, in order to give rise to two just distinguishable sensations. "It is probable that the nerve-areas are much smaller than the tactile; and that several unstimulated must intervene between the excited, in order to produce sensations which shall be distinct. If we suppose twelve unexcited nerve-areas must intervene, then, in Fig. 25, _a_ and _b_ will be just on the limits of a single tactile area; and no matter how the points are moved, so long as eleven, or fewer, unexcited areas come between, we would get a single tactile sensation; in this way we can explain the fact that tactile areas have no fixed boundaries in the skin, although the nerve-distribution in any part must be constant. We also see why the back of a knife laid on the surface causes a continuous linear sensation, although it touches many distinct nerve-areas. If we could discriminate the excitations of each of these from that of its immediate neighbors we should get the sensation of a series of points touching us, one for each nerve-region excited; but in the absence of intervening unexcited nerve-areas the sensations are fused together.