The function of this bundle is to conduct impulses from the auricle to the ventricle, so that normally the ventricular beat follows that of the auricle in practically one-fifth of a second. The effect of digitalis on this bundle may be the retardation or prevention of conduction. This is usually a result of vagus stimulation, and it may be prevented by atropine. But in some cases, as demonstrated by Cushny, the effect of digitalis on conduction is not prevented by atropine, and in these digitalis presumably has a direct action upon the junctional tissues, either the auriculoventricular bundle proper, or the junctions of its ramifications with the proper muscles of the ventricles.

In therapeutics a prolongation of the auriculoventricular interval, e. g., to three-tenths or three-fifths of a second (incipient heart-block), is not uncommon from digitalis. It is an effect that can be ascertained only by tracings, but it is a toxic manifestation and calls for stoppage of the drug. More rarely seen from digitalis, but much more serious, is a degree of interference with conduction which results in occasional or frequent failure of the ventricle to beat in response to the auricle, i. e., a state of partial heart-block. In this the auricle beats faster than the ventricle. In mild degrees the auriculoventricular interval gradually lengthens, or suddenly lengthens, so that the ventricle intermits at regular intervals, i. e., skips every tenth, seventh, third, etc., beat, the tracings showing an independent auricular beat during the ventricular intermission; and the stethoscope no ventricular contraction. In marked stages the ventricle beats only in response to every second or third auricular beat, i. e., in 2:1 or 3:1 rhythm, the pulse being slow and regular. In these last states fainting spells are not uncommon.

Still less frequent from digitalis is complete heart-block, in which the ventricle receives no adequate stimulus from the auricle, and consequently beats at its own intrinsic rate, with entire disregard of the auricular beat. In the complete block of disease the rate of the ventricle is in the neighborhood of 30, and this is the normal intrinsic rate of the human ventricle. But in the complete block from digitalis, owing to the increase in muscular irritability, the rate tends to be faster, and may even exceed that of the auricle (Hewlett and Barringer). In this last type, in the absence of a careful study of tracings, the block may remain undetected. In ordinary cases, however, bradycardia should suggest the possibility of block; and in any heart a block should always be suspected when there is a sudden slowing of the ventricular rate with regularity. In auricular fibrillation a complete block is shown by the striking change from rapidity and irregularity in the action of the ventricle to slowing and regularity. Slowing from digitalis may, therefore, be due to auriculoventricular heart-block, as well as to an effect upon the sinus node. Indeed it is to this cause that the slowing obtained in auricular fibrillation or flutter is due.

When a partial block is already established by disease, digitalis is very prone to increase its severity or to change it to complete block. Some of the deaths from the intravenous use of strophanthin, the digitalis ally, have probably been produced in this way.

The following is an interesting case of permanently complete heart-block, in which the digitalis had the effect of bringing on short spells of doubling of the intrinsic rate of the ventricle with retrograde rhythm. It was a case on Dr. Nome's service at St. Luke's Hospital. In one of my tracings from this case the ventricular rate shows a sudden jump from 26 to 54, a drop of the auricular rate from 62 to 54, and a change of the rhythm to "reversed" or retrograde, i. e., the auricular systole followed that of the ventricular, instead of preceding it, both having the same rate. At the end of each such paroxysm there was a long pause of the ventricle, lasting some seconds, during which the patient had a passing attack of faintness or light-headedness, though lying flat in bed.

Fig. 17.

Fig. 17. - Chart comparing the effect of digitalis on the rate in cases having auricular fibrillation with those having a normal rhythm. The black dots represent the rate with auricular fibrillation and the white with the normal rhythm. The side figures represent pulse-beats. The top figures represent days (James Mackenzie in "Heart," vol. ii, No. 4, 1911).

Such a pause, sometimes following the doubling of a slow ventricular rate, is prone to occur in partial or complete heart-block, and may be accompanied by feelings of faintness, loss of consciousness, or an epileptiform convulsion, the typical Stokes-Adams attack. These effects are due to a momentary anemia of the medullary centers, the result of the ventricular stoppage. They are likely to be more serious if the patient is in the upright position.

Fig. 18.

Fig. 18.

Fig. 19.

Fig. 19.

Figs. 18 and 19. - Complete heart-block developing in a case with auricular fibrillation. On admission (tracing 18) the ventricle was very irregular, rate 146 to 200, with a countable radial pulse of 80 to 94. Infusion of digitalis, 4 drams thrice daily, was given for eleven days, then stopped. At this time the pulse was nearly regular, rate about 72. Four days later tracing 19 was taken, the pulse being quite regular, rate 54. Three days later, i. e., one week after the stoppage of the drug, the complete block was still present, the ventricular rate remaining between 50 and 60.