x ELECTROMOTIVE ACTION IN NERVE 261 



speaking, small ; in spring frogs the single variations are short, 

 with comparatively large total effects. 



In the non-medullated nerves of Cephalopoda, Fuchs (I.e.} 

 estimates the average duration of the negative variation with 

 stronger stimuli at O'Ollo sec., with weaker excitations at 0'OOS2 

 sec., i.e. a value intermediate to those given by Bernstein and 

 Head. With a method corresponding to that of Hering and 

 Head, the duration of the negative variation would no doubt 

 be found even greater than Head's estimate of it in frog nerves. 



Fuchs further investigated the significance of the retarded 

 variation of non-medullated nerve, and indicated a possible 

 relation with the slow rate at which the excitatory process is 

 transmitted. If (as cannot be doubted) the transmission of 

 excitation does really depend upon some sort of propagation from 

 section to section, the prolonged duration of the process must be 

 advantageous, and we know experimentally that non-medullated 

 is far less sensitive to very brief stimuli than medullated 

 nerve. 



These facts show that the tetanic negative variation is 

 rhythmically discontinuous, and (notwithstanding its apparent 

 steadiness) of an oscillatory character in nerve as in muscle, 

 We have next to ask what magnitude can be reached by the 

 single negative swing on strengthening the stimulus ? Will the 

 demarcation current corresponding with the maximum of the 

 variation fall each time to zero, or even reverse itself (as 

 Bernstein found in medullated frog's nerve) ; or will there merely 

 be (as in muscle) greater or less diminution of the existing P.D., 

 in the rhythm of the stimulation ? The question in both cases 

 can be decided by the rheotome method, on making the galvano- 

 meter closure as short as possible, and then finding that position 

 of the slider which corresponds with the maximum of the varia- 

 tion. If compensation is then cut out. and that fraction of 

 current measured which is sent by the rotating rheotome into the 

 galvanometer circuit, and the tetanising key opened, it can be 

 seen at once whether the variation is less than, equal to, or 

 greater than the current of rest. By such experiments Bernstein 

 determined the negative variation with an augmented stimulus to 

 be much in excess of the current of rest, in frog's nerve. On 

 repeating the experiment Hermann (27, p. 385) at first found 

 the variation to be much less than the current of rest, and 



