It has already been stated that nerve fibres are capable of conducting impulses in either direction - from or to the nervous centres. The position and character of the terminal organs determines the direction in which the nerve impulse usually produces results. In the ordinary peripheral nerves there are generally both kinds - efferent and afferent fibres, carrying impulses in different directions.

When we reflect that the passage of an impulse along a nerve is brought about by a molecular change in the axis cylinder, we are at once struck with the rapidity with which impressions are transmitted from one part of the body to another. This velocity is, however, only relatively great. When we compare it with the velocity of the electric current or of light, we at once see how much slower the rate of nerve impulse is, and that it may be compared with rates of motion commonly under our observation. To take every-day examples - viz., nine metres per second is about the rate at which a quick runner can accomplish his 1oo yards; race horses can gallop about 15 metres a second for a mile or so; a mail train at full speed travels about 30 metres a second, and the velocity of nerve force has been estimated to be in cold-blooded animals 27 metres per second; and in man about 33 metres per second. So that the intercommunications between man's brain and the various parts of his body only travel about the same rate as an express train, and about twice as fast as the quickest horse can gallop.

Different methods may be employed for the measurement of the rate of transmission of nerve force. The simplest is, with a good myograph, such as described in Chap, xxv, p. 462, to make a muscle draw two curves, one over the other, in one of which the stimulation is applied to the nerve close to the muscle, and in the other as far as possible away from it. The difference in duration of the latent period in the two curves, shown by the tuning-fork tracing, corresponds to the time taken by the impulse to travel along the part of the nerve between the two points of stimulation, the length of which can be directly measured; and hence the velocity of the impulse estimated.

Utilizing the fact that the extent of the deflection of the needle of a galvanometer is in proportion to the duration of a current of known strength passing through it for a short time, an accurate measurement of the difference in time of remote and near stimulation of a nerve may be made. By a special mechanism the time-measuring current is sent through the galvanometer at the same moment that the stimulating current goes through the nerve, and the instant the muscle begins to contract, it breaks the current passing through the galvanometer, so that this time-measuring current lasts only from the moment when the nerve is stimulated until the muscle begins to contract.