arrd Departrirerrt of Rndiotherapy, Hairrrrrersrrrith Hospr'tnl"ALTHOUGH it now seems certain that haeniolysis irr vivo in patients with the cold-hacmagglutinin syndrome is determined by the presence of hgh-thermal amplitude high-titre cold auto-antibodies, the exact way the antibodies bring about the haemolysis has not been completely settled. 111 vitro, adsorption of complement, which with most sera occurs best if they arc acidified to pH 6.5-7.0 (Dacie, 1950(Dacie, , 1955, is most easily demonstrable if the erythrocyte-serum suspension is allowed to stand at 15-25' C. This adsorption of complement gives risc to a minor or major degree of haemolysis of normal erythrocytes according to the serum used, as well as causing those cells which remain unhaemolysed to be strongly agglutinable by antiglobdm serum (Dacie, Crookston and Christenson, 1957).Dacie (1954) summarized the then available evidence on the nicchanism of hacmolysis irr vivo and concluded that the two most important factors were: (I) the thermal range of the antibodiesthe nearer to 37" C. the antibodies acted, the greater the in-vivo haeniolysis;and (2) the hacniolytic as opposed to the agglutinating activity of thc antibodies, thosc antibodies which most readily brought about haemolysis irr uitro being associatcd apparently with more haemolysis irr vivo than antibodes which had little or no haemolytic potency.Recently, wc have had the opportunity of studying an elderly female paticnt suffering from the cold-haemagglutinin syndrome and assessing with the aid of 51Cr the effect on the survival ofher erythrocytes irr vivo of: (I) exposure of the cells irr v i m to thc patient's hcat-inactivatcd serum at 4" C., resulting in marked autoagglutination; (2) exposure of the cells irr vitro to the patient's frcsh unacidified serum at 4" C., resulting in marked autoagglutination, modcrate sensitization to agglutination by antiglobulin serum but no detectablc hacmolysis, and (3) cxposure of thc cclls irr vitro to the patient's frcsh acidified serum at 20° C. (PH 7.0 approximately), resulting in a minor degree of autoagglutination, marked Sensitization to antiglobulin serum (a marked adsorption of complement) and a moderate amount of hacmolysis. The expcriinental data provide further support for the concept of the importance of thc hacniolytic property of the antibody and of the relative unimportancc of agglutination. EXPERIMENTAL METHODS Expt. I . Exposirre in vitro of the Patient's Erythrocytes to Heat-Itractivnted Serirrrr ot 4 O C.Blood was collected from the patient in a warmed syringe. Five ml. were added to I ml. of acid-citrate-dextrose (ACD) anticoagulant solution previously warmed to 37" C. and 20 ml.were delivered into a warm 30-1~11. bottlc aiid allowed to clot undisturbed at 37" C. The cells from the citrated blood were labelled at 37" C. with 50 vc of 51Cr in the usual way (Lewis, Szur aiid Dacie, 1960) and then washed three times with warm salinc. While the labelling process was being carried out, 10 nil. of autologous serum which had bcen separat...