Yolk medium. The yolk reaction was due to the action of phospholipinase produced by the organisms. As in the case of Cl. we&hii a-toxin, the yolk curd-forming activity of B. cereus and B. mycoides also was associated with haemolytic activity. The substances responsible for these activities appear to be sirhilar, and may be identical. The yolk reaction has proved useful for the rapid identification of B. cereus, as it is more specific than any of the other distinguishing tests hitherto employed. The three positive organisms, B. cereus, B. mycoides and B. anthracis, have been considered by previous workers on the grounds of morphology and antigenic structure to be closely related. Nagler (1989) described the effect of CZ. welchii cultures and toxic filtrates on human serum in producing an opalescence and curding (Nagler reaction) ; later Macfarlane, Oakley & Anderson (1941) found that egg-yolk saline (a lecitho- vitellin emulsion) is even more sensitive than human serum (the lecitho- vitellin (L.v.) reaction). Macfarlane & Knight (1941) then showed that the production of oialescence in egg-yolk emulsion was due to the action of a lecithinase, which split phosphocholine from lecithin, in the CZ. welchii toxic filtrates which was probably identical with the a-toxin. Since then, the Nagler reaction and the lecitho-vitellin (L.v.) reaction have been extensively used for the titration of a-toxin of CZ. welchii (Nagler, 1939; van Heyningen, 1941 a) and for the rapid identification of certain clostridia (Hayward, 1941, 1948). It is interesting that Nagler (1989) observed a non-pathogenic aerobic spore- bearing bacillus which caused an opalescence in human serum like that caused by CZ. welchii; but the effect was not inhibited by CZ. welchii antitoxin. Crook (1942) and Hayward (1943) also mentioned that some aerobic bacteria caused similar opalescence. In the light of these observations we have sought for similar enzymes in the aerobic spore-bearing bacilli. The work started with a survey of the Bacillus group to ascertain which of them produced an egg- yolk reaction. A test of 260 identified strains, comprising 24 species, showed that only B. cereus, B. mycoides and some strains of B. anthracis were positive. In comparison with other characteristics generally used for the differentiation of the BaeilZzcs group, the yolk reaction seems to be much more selective. One can hardly find any other characteristics which are confined to only two or three species of this very large group of organisms, and the test deserves the attention of taxonomists of the. Bacillus group. We have investigated the