652 
PHYSIOLOGY 
waves which are comparable to those in hu- 
mans in all respects. Another major difference 
between dogs and humans is that in the former 
the right ventricle is cranial to the left while 
in man it is at the same level or slightly caudal. 
Since the first experimental studies on bundle 
branch block were made in dogs and only the 
longitudinal lead was registered this has re- 
sulted in the famous confusion between right 
and left bundle branch block patterns which 
persisted for twenty years until F. N. Wilson 
ended it by introducing precordial leads. In pri- 
mates the bundle branch block patterns are quite 
comparable to those in humans.^i 
The electrocardiogram of the cat shows less 
variability of the T waves than that of the dog, 
and unipolar leads from the left regions of the 
chest facing the apex of the left ventricle have 
in our experience always shown upright T 
waves, while those facing the lateral wall of the 
left ventricle may show negative or diphasic T 
waves. We have not studied the intramyocardial 
temperature gradients in cats because of the 
relatively small size of their hearts compared 
with the thermo-couple junction, but we have 
studied in 9 normal cats and 24 cats, subjected 
to epinephrine infusions, the content of potas- 
sium, sodium, calcium and magnesium in the 
endocardial and epicardial half of the left 
ventricular wall.-*' There was almost a straight- 
line correlation between the polarity of the T 
wave in the two locations mentioned above and 
the difference in the K/Na ratio between the 
subendocardial and subepicardial halves of the 
left ventricular wall facing these locations. The 
correlation with the other electrolytes was less 
pronounced. The same correlation was present 
in 56 patients who died soon after an electro- 
cardiogram was taken and in whom we deter- 
mined the concentrations of these electrolytes 
in the subepicardial and subendocardial halves 
of the left ventricular wall."*^ Since the duration 
of QRS is much greater in humans than in cats 
it is necessary to compare the ventricular grad- 
ient (QRST area) rather than the absolute 
amplitude of the T wave with the transmural 
electrolyte gradients. Also, the location of the 
precordial leads did not exactly match that of 
the myocardial regions studied since we com- 
pared lead Vs with the findings at the left ven- 
tricular apex, lead V3 with those at the septum 
and lead aVL with those of the left ventricular 
base. Nevertheless, the agreement between the 
findings in cats and humans justifies compari- 
son between these species with respect to the 
T wave. 
A final problem concerns the comparison of 
the electrocardiographic limb leads and the so- 
called unipolar leads in humans and experi- 
mental animals. In humans the heart lies in 
the same plane as the origins of the arms and 
legs, and this is approximately true of the pri- 
mates. However, the baboons and the marmosets 
have a more elongated thorax, and in these mon- 
keys the origin of all four extremities is con- 
siderably posterior to the heart. The central 
terminal in these monkeys reflects the back 
potentials and the leads taken with it are not 
truly "unipolar". In an unpublished study with 
Dr. J. E. Craighead we found that the differ- 
ence between the findings in the limb leads and 
the more nearly orthogonal vectorcardiographic 
systems of Frank and McFee were considerable. 
In the Iris monkeys, whose thorax is more 
barrel-shaped, these differences were much less 
pronounced. Similarly, there should be a con- 
siderable difference between the limb and vec- 
torcardiographic leads in dogs with narrow 
chests (collies, terriers and German shepherds) 
than in dogs with square chests such as beagles, 
and in cats which also have a more rounded 
thorax cross-section. These differences should 
be even greater in cattle and in birds, where 
the thorax cross-section is very elongated and 
the cranial extremities originate near the pos- 
terior chest wall. 
SUMMARY 
In the choice of an experimental animal for 
the study of cardiac electrophysiology, the frog 
and turtle have the advantage of low cost and 
upkeep and of a low oxygen consumption which 
makes it easier to avoid hypoxia during in-vitro 
studies of basic properties such as transmem- 
brane potential, effects of injury and excitation- 
contraction coupling, which are fairly similar 
to those in the human heart. However, findings 
concerning excitability, activation and repolari- 
zation of the heart are not comparable to those 
