M. N. LEVY 
661 
in the patient represented by Figs. 1 and 2) , the 
hypothesis would apply so long as the influence 
of the critical circulatory conditions on the ven- 
tricular pacemaker differed quantitatively from 
that on the S-A node. The difference in the tim- 
ing of the initiation of impulses by the two pace- 
makers is, by definition, the P-R interval. The 
atrial contribution to ventricular filling depends 
upon the P-R interval. This dependence has 
been defined in dogs with complete heart block 
by Brockman.^'^ In patients or experimental 
animals with a constant ventricular rate, the 
stroke volume and total peripheral resistance 
determine the level of the arterial blood pres- 
sure. Finally, the arterial blood pressure has a 
reflex influence on the rate of the S-A node, me- 
diated by the arterial baroreceptors. The time 
from the change in P-R interval till the sympa- 
thetic and vagal neurotransmitters affect the 
S-A node involves an appreciable delay. This 
time lag plays a role in the rhythmic oscillations 
of the P wave about the QRS complex, as ex- 
plained below. 
The feedback system illustrated in Figure 7 
facilitates the comprehension of the mechanism 
responsible for A-V synchronization in iso- 
rhythmic dissociation. At the left border of Fig- 
ure 4, for example, the P wave lies behind St. 
Hence, atrial contraction makes no contribution 
to ventricular filling. As the P moves in front of 
St (to the right of the first vertical deflection in 
the lower tracing), the atrium begins its con- 
traction before the ventricle, and the ventricu- 
lar stroke volume begins to increase. This is re- 
flected by a rise in the arterial blood pressure. 
As the pressure increases, the baroreceptor re- 
flex begins to decelerate the S-A node. The P be- 
gins to advance more and more slowly ahead of 
the St, accompanied by a progressive elevation 
of the blood pressure. Ultimately the S-A nodal 
frequency becomes equal to the ventricular pac- 
ing frequency momentarily. This is the maxi- 
mum P-St interval recorded while the P wave 
precedes St (between the first and second verti- 
cal deflections in the lower tracing). The S-A 
node continues to decelerate, so the P wave then 
begins to move back toward St; i.e., the P-St in- 
terval decreases from the maximum value. As 
the P-St interval diminishes, the blood pressure 
begins to fall. As the P crosses St (signalled by 
the second vertical deflection in the lower trac- 
ing) and then lies behind St, the atrial contrac- 
tion occurs after ventricular systole, and the 
blood pressure continues to fall. This reflexly 
accelerates the S-A node, with the result that 
the P moves behind St at a decreasing rate until 
a minimum P-St interval is reached (between 
the second and third vertical P-St deflections). 
At the minimum P-St deflection, the S-A nodal 
frequency again briefly equals the ventricular 
pacing frequency. With continued acceleration 
of the S-A node at the lowered arterial pres- 
sure, the P wave again begins to move back to- 
ward the St from behind. When it again passes 
in front of St, the blood pressure begins to rise, 
and the cycle is repeated over and over again. 
In the feedback loop depicted by Figure 7, if 
there were no delay in the system, the arterial 
pressure would stabilize at the value which re- 
flexly produced that S-A nodal frequency which 
equalled the fixed ventricular pacing frequency. 
This would be achieved at some specific P-St in- 
terval. Once this was reached, A-V synchroniza- 
tion would be sustained at a steady arterial 
pressure and P-St interval.'' Because of the 
delay, however, the level of cardiac efferent 
neural activity is appropriate not to the prevail- 
ing P-St interval, but to that existing at some 
previous point in time (equivalent to the delay 
in the system). Thus, the system "hunts" for 
the appropriate P-St interval, and in the proc- 
ess overshoots and undershoots the appropriate 
value repetitively. 
The role of the arterial blood pressure oscilla- 
tions in achieving A-V synchronization was 
demonstrated by the failure to produce iso- 
rhythmicity when the blood pressure oscillations 
were attenuated (Figures 4 and 5). This was a 
consistent observation. However, the presence 
of an intact baroreceptor reflex arc apparently 
is not essential in all cases, because isorhythmic 
dissociation could be produced after bilateral 
vagotomy and stellatectomy in many experi- 
ments. The range of ventricular pacing fre- 
quencies within which the S-A node would syn- 
chronize with the ventricular pacemaker was 
much smaller, however, indicating that when 
the reflex arc was intact, the baroreceptor reflex 
was the principal mechanism. It has been dem- 
onstrated that heart rate varies inversely with 
