702 COMPARATIVE ELECTRO-PHYSIOLOGY 



surface. And it will generally be found true that while this 

 natural current of rest is from the less excitable A to the 

 more excitable B, the current of response, on the other hand, 

 which occurs on excitation, is from the potentially more 

 excitable, and therefore now more excited B, to the less 

 excitable and therefore less excited A. 



Such is the course of events in the normal or primary 

 condition. But under the excitation due to preparation, or 

 accidental disturbance, the more excitable surface becomes 

 the more excited, and relatively to the other, galvanometri- 

 cally negative. In consequence of this, the natural current 

 is reversed, and we have a resting-current due to the after- 

 effect of injury or accidental excitation, flowing from the 

 more to the less excitable. Thus, while the natural current, 

 in the primary condition, was from the less excitable A to 

 the more excitable B, that is to say, A -> B, this reversed 

 current of rest, due to accidental excitation or injury, is from 

 B-+A. Even now, however, B may be more excitable 

 than -A, hence fresh stimulation will induce a responsive 

 current from B to A. In the primary condition, such a 

 responsive current would have appeared as a negative varia- 

 tion of the natural current A -> B. But when the primary 

 condition has been so modified that the natural current is 

 reversed, and has become B^>A y the normal responsive 

 current B -> A will appear as a positive variation. Still 

 another variation is possible, when the normal response itself 

 undergoes reversal owing to fatigue, under which condition 

 this abnormal response, relatively to the reversed current of 

 rest, appears as if it were the normal negative variation 

 (fig. 119). It has, however, been shown that if we discard 

 this unreliable test of response, by the variation induced in 

 an antecedent current of rest the so-called negative varia- 

 tion it will be found that the responsive current always 

 flows from the more to the less excited. 



In order to determine which of two points in an anisotropic 

 tissue is the more excitable, it is necessary, as now under- 

 stood, to determine the direction of resultant response, under 



