(a) In experiments on gastrocnemius muscles of the cat performing external work, the work was graded, either by altering the intensity of motor nerve stimulation or by changing the load. Only under the former conditions was the steady-state increase in blood flow conductance at all consistently related to the work performed. (b) In such experiments, efflux of inorganic phosphate and of potassium were closely correlated with the increase in vascular conductance, in the form of a typical dose-response curve. There was no significant relationship between changes in plasma osmolarity and conductance. 2. (a) Whereas most soleus muscles in the cat exhibit virtually no functional vasodilation, those with a relatively low resting flow tend to do so. (b) There was a relationship between the vasodilation, if any, and release of phosphate in fifteen out of seventeen experiments on soleus muscles. No relationship was found between any release of potassium or change of plasma osmolarity, and absence or extent of functional vasodilation. 3. (a) Terminal arterioles and collecting venules in the rat's spinotrapezius muscle were observed in vivo under low power magnification while the muscle was bathed in various test solutions, so that the vasodilator properties of hyperosmolar solutions, potassium and phosphate could be studied. (b) The dilator effect of hyperosmolar solutions was much the weakest: solutions of 340 m-osmole/kg elicited the largest responses, but these seldom exceeded 50% mM-K+ and 3.2--6.4 mM-Pi (as NaH2PO4) elicited 25--50% of maximum dilation, while 9--10 mM-K+ and 16--20 mM-Pi dilated arterial vessels fully. The latency to onset of dilation was shortest (5 sec) with Pi and longest (15 sec) with K+. 4. These findings, together with those already in the literature, lead to the conclusions that (a) hypersomolarity is unlikely to be in an important factor initiating or maintaining functional hyperaemia in skeletal muscle, (b) while K+ release may contribute in fast muscle it can hardly be the factor initiating the response, and (c) inorganic phosphate is the only substance tested so far whose efflux is consistently related to the vascular response in all muscles under a variety of experimental conditions, and whose efficacy as a vasodilator seems adequate.
