Numerous investigators have observed certain toxic effects of digitalis on the hearts of animals, and clinicians have assumed that the results thus recorded are generally transferable to patients. These experimental findings are not entirely constant with different species. Their importance would obviously be enhanced if it could be shown just how far the effects produced by the drug on a certain animal parallel its effects on the human heart; and it would be of especial clinical value to know in what animal the toxic effects are most like those that are produced in man. For there is no exact knowledge of the quantitative effects of toxic doses of digitalis on the human heart, because it is only when a patient accidentally receives an excessive amount of the drug that clinical records showing its toxic action are obtained. So far as knowledge acquired in this way confirms the results of pharmacologie