It is now commonly believed that in many patients in circulatory shock there is &dquo;inadequate perfusion of blood&dquo; through the tissues of the various organs of the body. The direct causes of inadequate perfusion are different in different types of shock, and several sets of causes can be operating in any one patient-' It has been shown2 that three types of circulatory shock can experimentally be separated from each other, i.e., (a) pure hemorrhagic shock with minimal trauma from (b) traumatic crushing shock with minimal hemorrhage, from (c) &dquo;reservoir retention&dquo; shock. Following hemorrhage with minimal trauma there are prolonged contractions of arteries, capillaries and veins. Following crushing trauma with minimal hemorrhage, there is a precipitation and agglutination of the blood into a circulating sludge. Following extensive burns, in man and other mammals there is a severe intravascular agglutination of all the circulating blood (reviewed by Knisely3) . In reservoir retention shock in several species including man,2 the outlet sphincter valves of the liver (reviewed by Knisely et a1.4) close, and in consequence the liver, portal vein bed and tributaries store so much blood that not enough is left to fill the rest of the circulatory system and maintain venous return and circulation. (For basic physiology, see Krogh and Lindhard5; Krogh6. 7; Knisely et al.8 ; Thomas and Essex9; and Knisely et al.4) The main purpose of this paper is to show that in two different types of shock, hypovolemic and septic, the inadequate perfusion of tissue is directly caused by two separate mechanisms and that, in consequence, separate (though perhaps in practice, overlapping) types of treatment are needed for each. The paper summarizes certain historical studies and then presents findings from our own animal laboratories and from our clinical Shock Trauma Unit.Experiynental bleeding of human beings: In 1947 Knisely et al. reported that healthy people (blood donors) responded to blood loss by contraction of arterioles, capillaries, and venules.l° In order to see if controlled hemorrhage does initiate visible spasms of conjunctival vessels, the blood and vessels of 34 blood donors, unanesthetized, unoperated relatives and friends of hospital patients who volunteered to donate blood, were studied before, during, and for 30 to 60 minutes immediately after each donated blood From the