366 ELECTRO-PHYSIOLOGY CHAP. 



easy to prove that the secondary contraction is by no means 

 invariably due to negative variation of a pre-existing current, 

 as was admitted later by du Bois-Eeymond himself, when he in- 

 vestigated the negative variation of " parelectronomic " muscle. 



Before pursuing' this point any further we must, however, 

 attack another question, which was left in abeyance in an 

 earlier connection. It was stated that appearance of secondary 

 tetanus might be taken as a proof that the muscle current 

 undergoes no continuous diminution during contraction, but that 

 during that time it is constantly varying backwards and forwards, 

 although these movements are not followed by the magnet, on 

 account of its sluggish reaction. The rheoscopic limb of the frog, 

 however, leaves us in doubt as to how nearly the summits of the 

 single curves of variation approximate to the zero line (indicated 

 by dots in Fig. 113) whether they do reach it, so that the 

 current is nil at the moment of contraction or finally exceed it, 

 which corresponds with a reversal of current. 



Du Bois-Eeymond himself attempted to solve the first 

 question (Untersuchungen, ii. p. 120), and with this object con- 

 structed apparatus " by which the muscle could be submitted to a 

 rapid series of excitations through its nerve, moment by moment, 

 in rapid succession. After each moment of excitation, the muscle 

 current could be closed for a brief period, and this closure might 

 follow at a given time between any two stimuli. If the muscle 

 current therefore sinks between any two stimuli during tetanus 

 in a normal curve, and then rises again, its deepest point will be 

 reached so soon as the closure of the muscle current coincides in 

 position with this point." The problem is still better expressed 

 in the accompanying diagram (Fig. 114). 



Let the abscissa (0, T) represent the time, on which are 

 drawn the ordinates, i.e. height of the muscle current (h), so that 

 the line (m, m), etc., corresponds with the line of the current 

 during rest. In the equidistant moments of time (t, t 1 , t' 2 , t 3 \ 

 etc., there is always an excitation of the muscle, which re- 

 sults in a negative variation of the existing " current of rest," 

 and its course, as will be shown, represents the curve (m, o, m), 

 between each two stimuli. The true form of the latter is 

 easily determined if the galvanometer circuit is closed between 

 every two excitations, for a moment only (T), at periodically 

 repeated and uniform intervals (t 1 , t 2 , t 3 , etc.) The same 



