iv GENERAL PHYSIOLOGY OF NERVOUS SYSTEM 209 



near the first. This fact seems to us important, because it 

 corroborates Engelrnarm's theory that the strength of the current 

 corresponds with the intensity of the lesion in the injured nerve, 

 and that this process of injury is arrested at the next node of 

 Ranvier. It also gives support to Hermann's theory that the 

 uninjured cell elements are incapable of developing electromotive 

 phenomena, and that the critical points of demarcation between 

 the healthy tissue and that injured by the section are determined 

 by these nodes. 



Nerves that are wholly dead are incapable of giving currents. 

 Any lesion of a nerve along its course, by cauterising, crushing, 

 compression, etc., renders it negative to the normal parts. Local 

 changes of temperature in the nerve, if insufficient to produce 

 structural lesions (e.g. up to 30 C.), give rise to electromotive 

 phenomena, the heated tissue becoming positive to the normal 

 tissue, as occurs in muscle. 



It may be deduced from all these facts that the electrical 

 phenomena of resting nerve depend on the negativity (on the 

 galvanometer) of the altered or injured portions of the fibres, in 

 relation to the uninjured, which justifies the name of demarcation 

 currents given by Hermann. 



The strength of the demarcation currents does not appear to 

 be in strict relation with the area of cross-section. The frog's 

 sciatic gives a more vigorous current ( = O'02-O'Oo volt) than a 

 large nerve of the horse or monkey ( = O'OOS volt according to 

 Biedermaiin). This is probably due to the varying resistance and 

 susceptibility of the nerve to external agents. It may be affirmed 

 in general that every cause which decreases the functional capacity 

 of the nerve must also diminish the intensity of its demarcation 

 current. 



An important fact discovered by different observers, both for 

 vertebrates and invertebrates, is that non-niedullated fibres yield 

 greater differences of potential and therefore larger currents than 

 medullated fibres, independently of their sectional area. From 

 this it may be inferred that the seat of the electrical phenomena \ 

 is not the medullated sheath but the axis-cylinder of the nerve. 



Another remarkable fact is that while in a mixed nerve- 

 composed of afferent and efferent fibres the two transverse 

 sections, or two points equidistant from these upon the longi- 

 tudinal surface, are equipotential when connected with the 

 galvanometer, this is not the case for nerves composed of one 

 kind of fibre only afferent or efferent. The central cross-section 

 of an afferent nerve (e.g. a dorsal root of a frog's spinal nerve) is 

 negative to the peripheral cross-section, but the central cross- 

 section of an efferent nerve (e.g. the electrical nerve of Torpedo) is 

 positive to the peripheral cross-section (Fig. 135). In these cases 

 the equator is not equidistant from the two cross-sections, but 



VOL. Ill P 



