244 ELECTRO-PHYSIOLOGY THAI'. 



nerve, which thus stands in the same relation to medullated 

 fibres as smooth to striated muscle. In agreement with this, 

 the single closure (eventually the opening also) of a constant 

 current produces as a rule a marked negative variation of the 

 demarcation current, an effect that occurs in medullated frog's 

 nerve only under special conditions (Biedermann, 3). 



Let the two connective nerves of Anodonta be prepared 

 together, with leading-off electrodes at the cross-section (after kill- 

 ing one end), and at a point of the longitudinal surface some 6 

 mm. higher up, while unpolarisable electrodes are simultaneously 

 applied to the other end of the pair of nerves (which are 

 moderately stretched between two supports), these electrodes being 

 connected by means of a reverser with 12 Dan. cells. There will, 

 then, at each closure of the exciting circuit, after compensation of 

 the demarcation current, be a more or less marked deflection of 

 the magnet in the direction of decrease, or negative variation, of 

 the nerve current, the magnitude of which is essentially dependent 

 upon the direction of the exciting current. If the latter flows 

 towards the end led off (which may be termed descending), the 

 action is invariably much stronger than with the opposite 

 direction. Careful investigation of this phenomenon leaves no 

 doubt that it arises from the consequences of exciting the nerve 

 by the constant current, i.e. is a negative variation in the true 

 sense of the word. This is shown not merely by the independ- 

 ence of the direction of the deflection from that of the current, 

 but also by the time-distribution of the effect, and the relations 

 exhibited between strength and direction of exciting current on 

 the one hand, and magnitude of effect on the galvanometer upon 

 the other. 



As regards, first, the distribution of the negative variation. 

 With descending direction of current this usually occurs so that 

 the deflection seems to begin at the moment of closure, or 

 imperceptibly later, reaches its maximum with tolerable rapidity, 

 and then, the current being still closed, declines rapidly at first 

 and afterwards more slowly. If the circuit is opened during this 

 period there is sometimes a marked delay in the backward pro- 

 cess, or even a reinforcement of the negative deflection. In most 

 cases, however, the opening has no effect, or there may even be 

 a positive after-variation, which with prolonged closure of the 

 exciting current may develop while the current is passing. This 



