Abstract
INTRODUCTION The question of the diagnosis of branch lesions in clinical electrocardiography is still largely unanswered. There are at least two important reasons for this. In the first place, although an abundance of experimental data pertaining to complete bundle branch block in animals has been accumulated, the extent to which the results can be applied to man is very uncertain. A further difficulty is discovered in the recent papers of Fahr,1 who would completely reverse the previous conceptions of branch lesions and preponderating ventricular hypertrophy. In the second place, our knowledge of the anatomy of the atrioventricular bundle and the Purkinje system is far from complete. We have not been sufficiently critical of researches on their normal structure, more especially in relation to the methods adopted. We have fallen into the habit of easily visualizing the nodes and tracts of the atrioventricular conduction system in the human heart, overlooking the