According to the conception advanced by Starling,1 edema may be produced in at least two ways. The blood's capillary pressure (filtration force) may be increased while the plasma colloid osmotic pressure is kept constant, or the plasma colloid osmotic pressure may be lowered while the capillary pressure is kept constant. The rôle played by each of these two factors in fluid distribution has been the object of much investigation. The problem was first approached in our laboratory from the point of view of the capillary pressure change. It was found that acute rises in the arterial pressure (and hence in the capillary pressure) of cats and dogs were accompanied by increased concentration of their blood, the transudate being recovered in the tissues by analyses for water content.2 Furthermore, direct measurements, in other laboratories, of capillary blood pressure and colloid osmotic pressure in human subjects have given quantitative meaning