Contemporary evidence has left unanswered the question as to whether or not during shock the small vessels are constricted or dilated. That is not to say that the subject has not been extensively investigated. But it has remained a vexed problem, perhaps because the methods employed for demonstrating caliber changes in vessels have usually been indirect ones and often highly unphysiological.Not only is there disagreement as to whether arteriolar constriction occurs in shock, but also as to whether such constriction, if it does occur, is of sufficient intensity to play an important part in the mechanism of shock.The clinical observation that during shock patients exhibit extreme pallor has usually been cited as evidence indicating vasoconstriction. Not only is the skin pale and cold but incision into it is almost bloodless. We have inspected the eyegrounds of such patients and found the arterioles markedly constricted. This is at least suggestive that arteriolar constriction is generalized if the analogy holds for shock as it does in hypertension. Both in animals (1, 2) and man (3) blood flow to the periphery decreases conspicuously during shock. This has been demonstrated by the finding of decreased cutaneous venous oxygen and increased arterio-venous oxygen difference (4--6), as well as by other methods.According to the work of Seelig and Lyon (7) vascular tone in shock seems to be maintained to some degree by nervous impulses, for severance of the sciatic nerve, when shock had occurred, resulted in marked increase in venous blood flow. Proportionately there was a greater increase in flow in shocked than in normal animals, suggesting vasoconstriction in the former. These measurements were only approximate, however, and Bartlett (8) believed he could show from calculation that if vasomotor tone had been markedly increased the change in outflow after section of the sciatic nerve should have been greater than it was. He injected salt solution under constant pressure into medium sized arteries partially isolated temporarily from the circulation and studied the rate of its inflow as the arterial pressure fell in consequence of induction of shock. The method is an artificial one whose results 215