THE ALL-OR-NONE PRINCIPLE 261 



hypothesis has been put forward that the initiation of an 

 impulse depends upon the rate of change of concentration 

 of ions at the point stimulated. More certain is our 

 knowledge concerning the nature of the impulse when it 

 is being propagated. There are two possibilities. Either 

 the impulse is launched with a certain quantity of energy 

 which carries it to its destination, or it is dependent for 

 its conduction upon a renewal of energy by molecular 

 changes at each successive point in its course. In the 

 first case we should expect to find that a nerve-fibre is 

 capable of carrying impulses of different strength accord- 

 ing to the intensity of the energy with which the impulse 

 is started. In the second case the intensity of the dis- 

 turbance would be independent of the strength of stimulus 

 and dependent Mily upon the nerve-fibre itself. The fibre, 

 in other words, would obey the all-or-none principle. Proof 

 that the latter supposition is true comes from Adrian's 

 experiment. When a given length of nerve is narcotised 

 for a certain time the impulse is extinguished as it traverses 

 the narcotised portion. Suppose that a length of nerve 

 D (Fig. 39, A) is narcotised in such a way that the 

 impulse started at III is just abolished at the distal 

 end of D. Now suppose this length to be divided into 

 two, d and d', separated by a length of healthy nerve I. 

 When the impulse emerges from d it will be reduced to 

 half its intensity. If it remains at this mtensity until 

 it enters d' it will again be completely extinguished at 

 the distal end of d'. Its intensity will be represented as 

 shown by the continuous line in Fig. 39, B. If, on the 

 other hand, the impulse arrives at the muscle with un- 

 diminished force, it means that every time it enters a healthy 

 part of the nerve it recovers its initial intensity as shown 

 in Fig. 39, C. Adrian found the latter to be the case, thus 

 proving that the nervous impulse rather resembles a train 

 of gunpowder. If a part of the train is shghtly damp 

 there is a delay in the rate of conduction, but provided 

 that the molecular changes arc able to pass through the 



