1) The writers state that after the inactivation of fresh Guinea pig'serum at 54°C, for 30 minutes, this serum keeps the thermolable fractions in noticeable quantities. Such serum if mantained at 6°C. for 18 hours, reacquires part of the lost alexic activity. 2) The Ritz's, and Coca's 3rd component is formed at least by distinct elements, one sensible to tje action of formaline and the other to the action of sodium hydrosulphite. 4) Ammonia, formaline and sodium hydrosulphite are able to destroy the thermostable components of the serum inactivated by heat (56°C., 30 minutes). The emulsions of yeast, organs and gelose are not able to do so. 5) The yeats emulsions added to a mixture o equal parts of G. p. fresh serum and serum inactivated by heat are capable to destroy the 3rd component present not only in the fresh serum as also in the inactivated one. 6) The writers do not exclude the hypothesis of a flocculation in which the inactivated serum exerts the rôle of a protector colloid, but they believe that the inactivation of the complement by yeast emulsions is due to the action of thermolable substances of fresh serum after the adsorption of antitriptic substances by said emulsions. 7) The different components which form the alexin adhere to the sensitized blood cells in the following order: Blood cells-sensitizer-Thermostable fraction sensible to ammonia -Thermolable globulin fraction - Thermolable albumin fraction - Thermostable fraction sensible to formaline and thermostable fraction sensible to sodium hydrosulphite. 8) In the Bordet-Wassermann's reaction, stark positive, is fixated especially the thermolable globulin fraction and not only the 3rd component as it could be supposed. The thermolable albumin fraction remains free and active in the fluid. 9) The writers think that the hepatic origin of the alexin must be accepted. According to experiments on dogs intoxicated by chloroform one observes not only the diminution of the alexic activity but also the decrease of all components of the alexin, with, perhaps, the exception of the thermolable globulin fraction.