In earlier papers (Gordon, 1930) it was shown that congo red has an inactivating effect on serum complement, both haemolytic and bactericidal, and that this effect can be reversed by treating the serum and congo red mixture with charcoal, the charcoal removing the congo red and leaving the complement active again. A similar reversal of inactivation is obtained by using instead of the charcoal, heated serum (55° C. for 30 min.) or protein solutions. Later (Gordon, 1931), it was shown that congo red had an inactivating effect on the haemolysins of Streptococcus haemolyticus and B. welchii. The reversibility of this effect was not so easy to demonstrate as with complement. Charcoal had a destructive effect on the haemolysins and so could not be used. It was found, however, that when the concentration of congo red was just sufficient to neutralize the streptococcal haemolysin, the addition of cuprammonium artificial silk adsorbed the congo red and liberated the haemolysin. In the case of B. welchii this method of reversal was not suitable, as the artificial silk had a destructive effect on the haemolysin. Instead, reversibility was demonstrated by adding ox serum to the mixture of congo red and haemolysin. This brought about a redistribution of the congo red between the ox serum and the haemolysin and if the amount of congo red used had been only just sufficient to neutralize the haemolysin of B. welchii, then the haemolytic activity could again be demonstrated. Gordon and Robson (1933) showed that congo red interfered with the anaphylactic reaction tested both in vivo and in vitro, the guinea-pig uterus being used in the in vitro experiments, in which the inhibitory action of the dye was shown to be reversible. It was suggested that the congo red interfered with the entrance of antigen into the cell.