TECHNIQUES based on the anaphylactic response have been used in many investigations into possible differences between normal and neoplastic tissues. Maver and Barret (1943) used systemic anaphylaxis along with precipitin and complement fixation techniques in an attempt to distinguish between cathepsins of normal and neoplastic tissues. Zilber (1958) and his co-workers employed anaphylaxis extensively as an experimental tool. They reported differences betweeni normal and malignant tissue in rats, mice and rabbits and inferred that the method demonstrated the existence of tumour-specific antigens. Similar studies were done on rabbit tumours bv Artamonova (1959), on various human tumours by Gorodilova and Shershul'skaia (1959), on various tumours of rats and mice by Levina (1959), on experimental mouse hepatomata by Korosteleva (1959) and on the Guerin rat epithelioma by Ryzewska (1962).Makari (1955) It is evident that if Makari's claim were valid, this would be a test of considerable clinical importance. Even if the test gave positive results in a limited number of cases, as reported by Hackett and Gardonyi, it might still be a useful investigatory tool in tumour research. Accordingly it was decided to undertake an evaluation of the applicability of anaphylactic tests to this type of study. Two tests were investigated: the Dale-Schultz reaction and the measurement of histamine-release from chopped lung. In the first part of the project (Dale, 1965a and b) it was shown that some of the assumptions on which the use of the tests was based were invalid. Certain general procedural limits were outlined within which the tests might possibly be used to analyse antigen mixtures. Using these limits a mock tumour antigen study was then carried out to determine whether.if a cancer antigen were present in tissue and serum, it could be detected by the tests (Dale, 1965c). From this latter study it was concluded that the tests might possibly give positive results if a cancer antigen were either highly antigenic or else were present in fairly high concentration in both tumour extract and serum. In the present study two questions were posed