CONSIDERABLE differences in experimeptal methods and wide variations in the purity of the antigens used have contributed to great confusion in the literature concerning the antigenicity of the serum proteins and especially of the albumin fraction. That serum albumin is an antigen giving rise to specific precipitin formation [Michaelis, 1902; Hektoen & Welker, 1924;Kimura, 1928] is directly contradicted by workers who find that albumin has no antigenic properties [Nolf, 1900;Landsteiner & Calvo, 1902;Linossier & Lemoine, 1902; Ruppel et al. 1923]. In view of the recent separation of albumin into crystalbumin and a new protein seroglycoid [Hewitt, 1936;1937] an attempt is made in the present paper to clarify the situation. EXPERIMENTAL Preparation of antigens The antigens employed were all derived from horse serum which was fractionated as follows. Serum (1 vol.) was mixed with water (1 vol.) and saturated ammonium sulphate solution (1 vol.). The precipitated euglobulin was filtered off, redissolved, reprecipitated with one-third saturated ammonium sulphate solution, filtered off and dialysed free from salts. The dialysate was diluted with 4 vol. of distilled water and the reaction was adjusted to pH 5-9. It is interesting to note that in adjusting the reaction hydrochloric acid is much more effective in precipitating euglobulin than is acetic acid. The precipitated euglobulin was centrifuged down, washed with water, dissolved in 4 % saline and dialysed. This isoelectric precipitation of the euglobulin was repeated three times. Finally the euglobulin solution was neutralized and sterilized by filtration through a Seitz filter.To the one-third saturated ammonium sulphate filtrate from the euglobulin precipitate was added more ammonium sulphate solution, raising the ammonium sulphate concentration to one-half saturation, thus precipitating the pseudoglobulin. The precipitated pseudoglobulin was filtered off and dialysed, and the dialysate was diluted with 4 vol. of distilled water. Euglobulin was removed by adjusting the reaction to pH 5-9 and centrifuging. The pseudoglobulin was then separated into two fractions, A being precipitated by 40 % saturated ammonium sulphate solution, and B not precipitated by 40 % but by 50 % saturated ammonium sulphate solution. After several reprecipitations the protein precipitates were filtered, dialysed etc., as described in the case of euglobulin.After removal of the globulins crystalbumin was obtained from the filtrate by addition of acetic acid and repeated recrystallization and seroglycoid was separated from the mother-liquors as previously described [Hewitt, 1936;1937].