The concept of vasodilator metabolites diffusing through the interstitial fluid to act directly on the vessel walls has never accounted adequately for the vasodilatation observed in a contracting muscle; for this dilatation is not confined to the capillaries and arterioles, but extends to the small intramuscular arteries and even to the larger arteries outside the muscle. In previous experiments, in which the vasodilatation was registered by the increase in venous outflow from the muscle, the post-contraction hyperaemia was thought to be produced in the main by an axon reflex in vasodilator nerve fibres, for it was grossly impaired by local anaesthetics, botulinum toxin or large, 'paralysing' doses of nicotine (Hilton, 1953(Hilton, , 1954. However, the nature of any nerve fibres involved remained obscure, since the post-contraction hyperaemia was not affected by chronic lumbar sympathectomy or posterior root ganglionectomy.It was shown over twenty years ago that the femoral artery itself dilates after contraction of the muscles of the lower leg, even when all the nerves to the limb are divided (Schretzenmayr, 1933;Fleisch, 1935). This reaction therefore had also been thought to be due to an axon reflex. It was abolished by local application to the artery, below the site of measurement, of phenol or cocaine, hence the responsible conducting elements appeared to run in the artery wall. Experiments now carried out on this dilator response of the femoral artery have confirmed these findings. Further, the femoral artery response has been found to be affected by drugs and surgical procedures in exactly the same way as the major part of the post-contraction hyperaemia. They both appear to be part of one and the same vascular reaction with a common conducting mechanism, and, since no known system of nerve fibres can be shown to be responsible, it is suggested that the conduction is effected by the smooth muscle of the artery wall.