iv ELECTROMOTIVE ACTION IN MUSCLE 385 



centre. Hence it is a great advantage to lead off, in indirect 

 excitation, from the actual seat of the excitatory process. If each 

 nerve-ending lay exactly in the middle of the corresponding fibre 

 of the parallel-fibred muscle, a negative wave of excitation, or con- 

 traction, would obviously be propagated from it in both directions 

 through the muscle at the moment of excitation. Then, on lead- 

 ing off from the middle and tendon end of such a muscle to the gal- 

 vanometer, while single shocks were sent into the nerve at equal 

 intervals by a Bernstein rheotome, a diphasic action current would 

 be demonstrable, consisting of a first " atterminal (abnerval)," and 

 a second " abterminal (adnerval) " phase. Such a response was 

 actually found by Hermann in his experiments with the sartorius 

 preparation. Both halves of the muscle at first indicated an 

 atterminal current directed from the centre to each tendon end, 

 while a little later, i.e. at the interval required by the excitatory 

 wave to traverse the muscle from centre to tendon end, an abter- 

 minal action current appeared, which, owing to the decrement of 

 the exciting wave, was always weaker than the first current. 

 On leading off from both tendon ends, we have at each moment 

 the algebraic sum of the effects in either half; this sum of 

 course = in a properly symmetrical muscle, in others its sign 

 varies with the time -interval. These experiments of Hermann 

 give physiological proof of the unclulatory course of excitation in 

 indirect stimulation, and we may now proceed to consider the 

 action current in its more complicated examples, with indirect 

 excitation of the gastrocnemius. S. Mayer (I.e.) found first a 

 descending and then an ascending action current, on leading 

 off from both tendinous ends of this muscle, after each single 

 excitation ; or, as it was then expressed because the first current 

 was identified with the negative variation of the muscle corroded 

 at the achilles expansion a variation first negative and then posi- 

 tive appeared, a fact confirmed later by du Bois-Eeymond with 

 Bernstein's rheotome, which S. Mayer had also employed, and by 

 Hermann with the (non-repeating) "fall-rheotome" described above. 

 If the tendo achilles was corroded, the ascending (positive) half of 

 the current was absent. Holmgren (29), moreover, by means 

 of a light magnet (without rheotome), had frequently observed, 

 before Mayer, a diphasic variation on the gastrocnemius, as well 

 as cases of simple positive, and negative variations. According to 

 Hermann the gastrocnemius may be regarded, diagrammatically, as 



2 C 



