1.

1. A systolic bifid apex impulse can be recorded in some, but not in all, cases of bundle-branch block. It may also occur in cases without bundle-branch block. These facts should be considered in assessing the diagnostic value of the systolic bifid apex impulse and indicate that it cannot be regarded as reliable evidence of the presence of bundle-branch block.

2.

2. Comparison of the interval between the beginning of the QRS complex and the beginning of the carotid pulse wave in cases with bundle-branch block, in normal controls, and in cases with cardiac decompensation, indicates that ejection from the left ventricle is significantly delayed in the common type of bundle-branch block. No comparable delay is observed in cases of cardiac decompensation without intraventricular conduction defect. Evidence is presented to establish the fact that the delay is associated with the particular type of aberrant spread of the excitatory process which occurs in the common type of bundle-branch block.

3.

3. Comparison of the intervals between the beginning of the QRS complex and the beginning of the C-wave recorded from the veins immediately above the right clavicle in cases of the common type of bundle-branch block and normal controls shows no significant delay in bundle-branch block. This finding suggests, although it does not definitely prove, that in the common type of bundle-branch block there is no significant delay in the contraction of the right ventricle.

4.

4. Roentgenkymographic tracings of the aorta and pulmonary artery were made in 6 cases with the common type of bundle-branch block, 2 cases with other types of intraventricular conduction defect and split first sounds, 8 normal controls with single first sounds, and 6 cases with split first heart sound, but without intraventricular conduction defect. The time relations of the aortic and pulmonary artery pulses were compared. The results were as follows: (a) In the common type of bundle-branch block, the interval between the beginning of the QRS complex and the beginning of the aortic pulse was significantly prolonged in each case. (b) In the normal controls with single first sounds the differences in time between the beginning of the aortic and pulmonary artery pulses fell within the limits of error of the method. (c) In the cases with split first sounds, both with and without intraventricular conduction defects, the differences in time between the beginning of the aortic and pulmonary artery pulses tended to be greater than in the control cases with single first sounds. In 2 cases the differences were as great as those observed in bundle-branch block. However, in one of these two cases the aortic pulse preceded the pulmonic; whereas in all cases of the common type of bundle-branch block, the pulmonary artery pulse preceded the aortic.

5.

5. All the findings support the hypothesis that in the common type of bundle-branch block there is asynchronism in the beginning of ejection from the two ventricles due to delay on the left side. The evidence also indicates that in some, possibly all, cases with split first sounds but without intraventricular conduction defect, there is asynchronism in the beginning of ejection from the two ventricles.

6.

6. The association between asynchronism in the beginning of ejection from the two ventricles and splitting of the first heart sound; the correspondence in the intervals between the two components of the sounds and the degree of asynchronism; and the relationship of the arterial pulse to one or the other component indicate that (1) the split first sound has a right ventricular and a left ventricular component, and that (2) separation of these components is due to asynchronism in certain of the early phases of cardiac contraction in the two ventricles.

7.

7. Splitting of the second heart sound can be heard on auscultation and recorded in most cases with bundle-branch block. It is also frequently present in cases without evidence of cardiovascular damage.

8.

8. It can be proved by the following methods for identifying aortic and pulmonic closure that splitting of the second heart sounds is due to asynchronism in closure of the two semilunar valves. (a) When the second heart sound is split, the component associated with aortic closure can be identified by its time relation to the incisura of the carotid artery pulse. (b) When the sound is widely split, the component associated with pulmonic closure can be identified by its time relations to the beginning of the descent of the V-wave recorded from the veins above the clavicle. (c) When either the diastolic murmur of aortic insufficiency or the opening snap of mitral stenosis is present, additional methods for confirming the identity of the components of the second sound may be available.

9.

9. In the common type of bundle-branch block, pulmonic closure usually precedes aortic closure. Occasionally they may be practically synchronous.

10.

10. In cases with split second sounds but without intraventricular conduction defect, either aortic or pulmonic closure may come first.

11.

11. All the evidence obtained in this study is in accord with the hypothesis that the common type of bundle-branch block is due to delay in spread of the impulse on the left side. The view held by Eppinger and Rothberger, and Lewis is no longer tenable. No support was obtained for the statement that it is unjustifiable to attempt to localize the side of bundle-branch block from the direction of the major initial complex in the three limb leads of the electrocardiogram.