Experimental evidence points to the following conclusions: (1) Resistance is offered at a synapse. This resistance must be overcome before an impulse can get through from neurone 1 to neurone 2. (2) The impulse does not, properly speaking, pass from 1, through 2. It infects 2, causing it to discharge a fresh impulse. (3) Time is of the essence of this process. Either the impulses head up at the synapse, or, passing through into the neurone, they produce a cumulative effect within it, which provokes it to discharge. (The latter hypothesis, which is the less likely of the two, transfers the resistance from the synapse to the neurone to be infected.) These conclusions are based upon experiments of the following kind: The minimal stimulus which will evoke a reflex action is determined. A stronger stimulus is then applied. The reflex occurs more promptly, and is more pronounced. But on further increasing the stimulus, it is found that the limit of effectiveness is soon reached. The proportional relation of response to stimulus is much less evident than it is when the experiment is tried with a nerve-muscle. Choosing a reflex action easily provoked, the afferent path is stimulated with an electric current interrupted fifty times a second. The impulses which flow down the efferent path to the muscle follow one another at the rate of about ten a second. A column of nerve-fibres within the spinal cord is stimulated fifty times a second. Again, the discharge into anterior roots has the natural rhythm of about ten. The cortex of the “motor area” of the great brain is stimulated with a rapidly interrupted current. The muscles which it governs contract with their natural rhythm. The cortex is sliced away, and the stimulus applied to the white matter beneath. A similar result is obtained. Evidence such as this points to an independence of action on the part of the neurones which one can express only in terms of resistance and explosion. But there is another line of thought which leads to the development of a picture of the working nervous system which seems at first sight incompatible with the one that we have sketched. The phenomenon of the knee-jerk (p. 274) reveals a nervous system so intimately linked together, so homogeneous, so mobile, that no event, however trivial, occurs in any part without sending a vibration throughout the rest. Instead of a multitude of batteries enveloped in a labyrinth of wires interrupted by myriads of switches which are crackling on and off, the image of a sheet of water better figures our conception—a material so frictionless that it is a-ripple from side to side and end to end, from the most distant rivulet which feeds it to the farthest trickle in which it drains away. It is a fluid in a state of infinite commotion, the movements of its particles varying in amplitude from tremulous quiverings which scarcely frost the silver of its surface to waves which, breaking on the muscular system, throw it up in heaps.