282 Transactions of the Boyal Society of South Africa. 
point. The point of stimulation and the position of the leads to the 
galvanometer have been varied. 
Since electrodes connected with the two ends of the galvanometer 
string are applied at separate points of the heart muscle, we may take it 
■that the curve obtained when the ventricle contracts, represents the 
algebraic sum of two similar electrical variations affecting the right- and 
left-hand electrodes respectively and therefore oppositely directed. 
In order to throw light on such questions as where activity commences 
in the ventricle when spontaneously beating, and what path the excitation 
takes through the musculature, it is necessary to deduce from the curve 
the form of the electrical change which takes place under each electrode. 
In a previous note" I have brought forward evidence derived from 
Fig. 4. 
cases of systolic alternation in the tortoise ventricle, which would indicate 
that the electrical disturbance which accompanies activity in the tissue 
under one electrode when the wave of excitation is propagated to it, 
•consists of a preliminary positive change, appearing as a descent of the 
•curve, followed by a negative change shown by an ascent of the curve. 
The negative change after attaining its maxiinum shows a decline, which 
is later checked. 
When large tortoise ventricles are available, and are stimulated 
artificially, it is possible to apply the electrodes at a considerable distance 
from each other, and as conduction of excitation is slow in the tortoise 
heart, the variation under one electrode may be distinctly seen to run its 
* These Transactions^ vol. iv., part i., p. 73, 1914. 
