x ELECTROMOTIVE ACTION IN NERVE 243 



measured by the resulting swing of the galvanometer magnet, is pro- 

 portional at all points of the nerve to the strength of the original 

 demarcation current, and is therefore maximal on leading off from 

 the cross-section and most positive point of the longitudinal 

 surface, nil on leading off from two isoelectric points. Even with 

 a maximal negative variation, it can be seen directly with 

 an aperiodic galvanometer that the diminution of the nerve 

 current during tetanic excitation never amounts to complete 

 abolition, so that a greater or less fraction of E.M.F. remains con- 

 stant. The negative variation of the demarcation current is, as 

 we should a priori anticipate, the same in non-medullated as in 

 medullated nerve. 



In the olfactory nerve of the pike it was found by Kiihne and 

 Steiner to be very vigorous, which agrees with the high E.M.F. 

 of the " current of rest." Since non-medullated nerve, like 

 muscle, reacts better to stimuli of prolonged duration than to 

 short induction shocks, the negative variation is perceptibly 

 stronger when the tetanising excitation is effected by the rapid 

 make and break of a constant current. This is more especially 

 the case (as tested on molluscan nerve Biedermann, 3) when 

 the unfavourable effects of a current flowing in one direction are 

 eliminated by introducing a rotating reverser, or by quickly turning 

 a Pohl's commutator. At the close of the rhythmical excita- 

 tion the magnet usually returns with decreasing rapidity to its 

 position of rest, or there may in stale preparations be a negative 

 remainder of the deflection. On attempting to tetanise molluscan 

 nerve in the ordinary way by means of a du Bois' sliding inductorium, 

 and thus to obtain a negative variation of the demarcation current, 

 there is usually complete absence of effect, under the most favour- 

 able relations of excitability, even when the coils are pushed close 

 together. This minimal activity of brief currents is also very 

 marked with interrupted constant currents, where the magnitude 

 of the negative variation does not .usually increase (as in 

 medullated nerve under similar conditions) with increasing 

 frequency of stimulation, but on the contrary undergoes a 

 diminution, which is the more considerable in proportion as the 

 interruptions of the current are more rapid, i.e. as each single 

 stimulus is shorter. Since a similar reaction may be observed in 

 the electrical excitation of the non-medullated claw-nerves of 

 crayfish, this must be a widespread property of non-medullated 



