That digitalis causes alterations in the heart resulting in changes in the form of the electrocardiogram is now well known. Robinson and Wilson have estimated in experiments on cats that the quantity of digitalis which can induce a change is about 30 per cent, of the calculated lethal dose. So far, however, no evidence has been presented to show that this amount of digitalis is beneficial—except in cases of fibrillation of the auricles in which block of auricular impulses, mainly through stimulation of the vagus nerves, takes place.A beneficial action must be based on the ability of the drug to increase the volume output of the heart, and it must be able to do this in therapeutic doses, that is to say, in doses which influence the T wave of the electrocardiogram or reduce the rate in auricular fibrillation. We have accordingly injected this amount into the veins of dogs, 11 of which received the tincture of digitalis and 19 of which received g-strophanthin; and into cats, 5 of which received g-strophanthi...