SCHIFF1 observed that cultures and culture filtrates of certain strains of Clostridium welchii possess the power to inactivate the blood-group A -substance contained in jpeptone and human saliva. The decompositiony -as considered specific for A -substance. We have examined a number of crude Cl. welchii filtrates fi r their capacity to destroy the specific blood-gr up substances, but have found them disappointingly weak. Through the kindness of Dr. W. EC van Heyningen, who has supplied a number of welchii (Type A) culture filtrate preparations partially purified with reference to collagenase (x-toxin)2, it has been possible to study the action of the enzymes contained in these preparations on the blood-group substances. The filtrates contained collagenase and hyaluronidase3, and in most specimens a- and 0-toxins were also present. The substrates used were preparations of -substance obtained from hog gastric mucin4 and A-, B- and O-substances which had been isolated from human pseudomucinous ovarian cyst fluids5. The human A- and 5-substances showed no significant O specificity. The A -substance isolated from hog mucin, although electrophoretically homogeneous at pH. 4.0 and 8.0, nevertheless is composed of two mucoids, one of which possesses A specificity and the other O specificity. The mixed material, which is usually referred to as hog mucin 4-substance, cannot be separated readily into its sero-logically specific A and O components by any of the simple chemical or physical techniques employed so far, but full details of this aspect of the dual specificity of hog mucin '.4-substance' and of preparations of A- and B-substances isolated from the saliva and gastric juice of persons belonging to groups A, B and AB will be discussed elsewhere.