342 ELECTRO-PHYSIOLOGY CHAP. 



pressure, which can only, be avoided by complete familiarity with 

 the deleterious matters on the one hand, and the extraordinary 

 sensibility of the muscle -substance on the other. Above all, 

 contact with any wound in the muscle, or the fluid by which 

 this is moistened, must be carefully avoided. For, as was pointed 

 out by du Bois-Eeymond, the exposed fibres, dying or dead, e.g. in 

 an artificial cross-section, are extremely active in developing cur- 

 rent. These facts are very striking in regard to du Bois' dictum 

 that only such matters as attack the muscle-substance chemically, 

 and thereby, as he said, destroy the parelectronomic layer, develop 

 electromotive action, since we are justified in assuming that the 

 muscle -substance itself does not undergo chemical alteration. 

 But it must be remembered that the exposed fibres rapidly set 

 up rigor, and undergo chemical changes, which of course develop 

 acids. Since, on the other hand, it is known that even very 

 dilute acids are highly injurious to the vital properties of 

 muscle, it is natural to refer the current-developing property of 

 the artificial cross-section to the acidifying of the muscle-substance. 

 How far this assumption is really justifiable must be decided later. 

 The theory of pre-existence of electromotive action meets 

 with special difficulties in the uninjured, or apparently uninjured, 

 adductor group of the frog. In the majority of cases du Bois 

 found a descending current between the two ends of tendon, 

 but there were also cases of complete absence of current, as 

 well as of a reversed current. The current between upper 

 end of tendon and equator (du Bois' " upper current ") was 

 greater, as a rule, than that between equator and lower end of 

 tendon (" lower current "). Yet du Bois also found the opposite, 

 there being even cases in which both ends of tendon were positive 

 to the equator. These discrepancies and confusions were, as 

 Hermann points out, sufficient in themselves to shake the par- 

 electronomic theory ; but such was not the case. On the con- 

 trary, on the ground of certain results with the adductor muscles 

 it obtained a wider extension from the presumption of a 

 parelectronomic strip, developed in many cases in place of 

 the parelectronomic layer (11). By this du Bois-Eeymond 

 designated the (rare) case in which an artificial section in the 

 neighbourhood of the end of the tendon is positive, and not, as 

 usual, negative, to the longitudinal section. It will be shown 

 later that all these irregularities admit of a simple explanation ; 



