24 Transactions of the Royal Society of South Africa. 
During stimulation the moving part of the instrument is deflected in 
the opposite direction to the demarcation current, and this constitutes the 
negative variation. 
A negative variation accompanies activity of hving nerve, and v^hen 
the leading off electrodes are applied to a nerve dissected out from the 
body of a newly killed frog, one on the uninjured longitudinal surface and 
the other on the cross-section, the instrument is deflected under the 
influence of the action current developed at the electrode on the uninjured 
spot. The changes giving rise to action current do not, in consequence of 
the injury, occur at the cross-section, or only to a slight extent. 
The positive after- variation was regarded by Hering and also by Head,* 
who later investigated the phenomenon, as occurring in the neighbourhood 
of the electrode upon the longitudinal surface, and was attributed to a 
process occurring in the nerve opposed to that process which may be 
termed excitation. 
According to Bering's conception, stimulation of a nerve produces 
phasic upward and downward changes, and at the close of a period of 
stimulation the positive variation which occurs evidences a process of 
building up or assimilation, a restitution process in the nerve which 
follows its period of activity. 
Head discussed the possibility that the occurrences giving rise to the 
positive change are located at the cross-section. The appearance of 
positive change would be presented if the cross-section of the nerve were 
to become more negative to the longitudinal surface than before stimula- 
tion, that is to say, if the demarcation current were increased by the 
tetanisation. Head came to the conclusion that it is more probable that 
the process is localised not at the cross-section but at the longitudinal 
surface, and agrees with Hering that it accompanies a process of 
restitution. 
Wallerf found that where the nerve has been preserved for several 
hours before stimulation, in place of the usual negative variation accom- 
panying activity, a positive variation may be obtained. In explanation of 
this reversal Waller suggests several possibilities. (1) It may be due to a 
reversal of action (or of greater action) in the parts of the nerve at and 
near the two leading-off electrodes. Since the normal negative deflection 
is due to action predominant near the electrode on the longitudinal 
surface of the nerve, the positive deflection now obtained might depend 
upon action becoming predominant at the cross section in consequence of 
some accidental and progressive injury near the longitudinal lead off. 
(2) The positive deflection may be a phenomenon of the same order as the 
positive after-variation described by Hering. (3) The reversal may be due 
* Pfliiger's Arch. Bd. 40 S. 207, 1887. 
t Phil. Trans. London, 1896. 
