ELECTROLYTES, CLOTTING IN BLOOD LOSS AND NORETHYNODREL 859 



calcium released from binding sites on sérum albumin (Carr, 1953) and micro- 

 somal membranes (Carvalho, 1963) over approximately the same pH ranges. 

 The in vitro data gave the quantity of calcium bound at each pH level under 

 equilibrium conditions; the reaction was complète in approximately 5 seconds 

 (Carr, 1953). We calculated the différence between calcium bound at a given pH 

 and the next lower pH and assumed that with an identical pH shift under transient 

 conditions, the same quantity of calcium would be freed. Since magnésium and 

 calcium have virtually identical binding affinities in various proteins and in micro- 

 somal and red cell membranes (Carr and Woods, 1955; Carvalho, Sanui and 

 Pace, 1963), we suggest that the rise of sérum magnésium during hemorrhage was 

 a resuit of the pH decrease. It seems likely that the freed magnésium originated 

 from intravascular binding sites, but further experiments are needed to estimate 

 the contributions from intracellular membranes, limiting membranes or proteins. 



Schrier (1966) and Hoffman (1962) showed that magnésium préserves mem- 

 brane cohésion and normal impermeability of the red blood cell. With loss of 

 bound magnésium during hemorrhage, intracellular thromboplastin may have 

 leaked out, thereby enhancing the extrinsic clotting process. The late increase in 

 sérum phosphate during hemorrhage, observed also by Beecher et al. (1947), may 

 be additional évidence of membrane leakiness (Schrier, 1966). The decrease in 

 coagulation time and increase in TEGa with hemorrhage may equally reflect 

 accélération of the fibrinogen-fibrin conversion in the présence of increasing 

 hydrogen ion concentration (Shulman and Ferry, 1950). Other mechanisms for 

 the increased coagulability cannot be excluded, but cell and platelet counts gave 

 no indication of splenic or pulmonary discharge between samples. An increase 

 in fibrin stabilizing factor (Sigg and Duckert, 1963) was not tested for. We suggest 

 that pH changes during rapid hemorrhage by repeated cardiac puncture may have 

 resulted in a loss of magnésium from membrane sites, increased membrane perme- 

 ability and, directly or indirectly, accélération of clotting. 



Ether anesthesia was not likely to have produced the increases in magnésium 

 and other sérum ion concentrations. In an unanesthetized rabbit, bled repeatedly 

 from a single cardiac puncture, AMg was 0.47 mEq/1, ACa 0.51 mEq/1, and 

 AK 1.12 mEq/1 (unpublished data). Beecher et al. (1946) observed increases in 

 sérum magnésium and phosphate concentrations in battle casualties after severe 

 blood loss. Veragut and Smith (1964) reported that sérum potassium increased 

 as pH decreased in dogs hemorrhaged after pentobarbital anesthesia. However, 

 prolongation of the experiment time, estimated as the elapsed time to 40 percent 

 hemorrhage, may resuit in further accumulation of hydrogen ion and greater 

 shifts in sérum ion concentration (compare expt. 3 with 4 and 5). 



To interpret the hemorrhage experiments in the injected animais, we assumed 

 that pH shifts were equal if etherization and withdrawal times and sample volumes 

 were équivalent between groups. In the Enovid-injected rats, particularly on the 



