1595 
of the diastole of the first systole of a group (vide Fig. 6 at 0). 
This restored normal rhythm continues now during 95 systoles, 
and passes then again into bigeminy. After about one minute | 
Fig. .6. 
convert this bigeminy again into the normal rhythm, as I did in 
Fig. 4 and 6. This normal rhythm continues now for over half an 
hour, whilst the dimension of the systoles gradually decreases. Then 
the experiment is stopped. 
Consequently with this frog’s heart after the poisoning with vera- 
trine’ we have converted: 
1. The halved rhythm in two ways into the normal one by an 
induction-stroke (Fig. 1 and Fig. 5). 
2. The halved rhythm into bigeminy (Fig. 3). 
3. Bigeminy into the halved rhythm (Fig. 2). 
4. Bigeminy into the normal rhythm (Fig. 4 and 6). 
Consequently we can discretionally with a heart the ventricle of 
which pulsates in the halved rhythm, convert this rhythm either 
into the normal twice as rapid rhythm or into bigeminusgroups. 
Therefore we have to regulate the dimensions of the systoles by inter- 
fering at special moments of the heartperiod with the stimulant. 
These dimensions depend on the duration of the preceding pause. 
Consequently if we obtain the normal rhythm, after having applied, 
at a special moment of the diastole, during the halved rhythm, an 
extra-stimulation, then we can obtain ventricle-bigeminy by applying 
the extra-stimulation a little later. 
-‘Fhe-experiments, in which 1 converted ventricle-bigeminy into 
the normal rhythm by an extra-stimulation in the diastole of the 
first curve of a group, which, consequently, produces an extra- 
systole, before the 2nd curve would have begun, are entirely in 
conformity with this view. 
The difference in dimension between the ventricle-systoles of the halved 
rhythm and those of the normal one is very great. The dimension of the 
first systole of the bigeminus-groups is a medium between these. It 
is always smaller than the yentricle-systoles of the latest halved 
