Microelectrodes with an external diameter of less than 05p,. penetrate the surface membrane of skeletal muscle fibres without causing appreciable damage and can be used to record the restingpotential and action potential (Ling & Gerard, 1949; Nastuk & Hodgkin, 1950). This technique has rapidly been extended to other tissues such as frog heart (Woodbury, Woodbury & Hecht, 1950), myelinated nerve (Woodbury & Woodbury, 1950) and the giant axon of the cuttlefish (Weidmann, 1951a). The present paper contains an account of similar experiments with a preparation of mammalian cardiac muscle, and is a sequel to the preliminary reports by Coraboeuf & Weidmann (1949a, b) and Draper & Weidmann (1950). Most of the experiments were performed on the'false tendons' of the dog heart. These are slender bundles of the conductive system which run through the ventricular cavity of birds and mammals. Their muscular component is a syncytium of parallel fibres which in the dog have a diameter of about 30, u. In ungulates where the conductive elements are histologically well differentiated they are known as Purkinje threads (Purkinje strands, Purkinje fibres). In canine and human hearts the conductive fibres are said to differ in no essentialhistological feature from the fibres of typical cardiac muscle (Glomset & Glomset, 1940). We shall therefore employ the term false tendon to describe the dog preparation in preference to the better known one of Purkinje fibres. The principal results reported here are those concerned with the absolute magnitude of the action potential and resting potential. The effect of sodium concentration has also been examined, and the results support the view that electrical activity is due to a large and specific increase in the membrane permeability to sodium (Hodgkin & Katz, 1949; Hodgkin, Huxley & Katz, 1949).