Observations on the mechanism of one type of so-called supernormal A-V conduction