Summary and Conclusions Over 3500 determinations of the hydrogen-ion concentration of the surrounding mediums were made by colorimetric methods in relation to various phases of biologic hemolysis. As a result of these observations the following conclusions are drawn: The cells have a strong power of altering the acidity, even in the presence of isotonic phosphate solutions, hence the pH determined in the final mixture is often quite different from that of the buffer solution used. In using solutions which have been saturated with carbon dioxid enough is carried in the solution so that the change in hydrogen-ion concentration, due to the cell, is much less marked.The effective range of acidity for biologic hemolysis is from about 6.0 to 8.3, with a rather broad optimum near neutrality. These limits are not sharply defined by the methods here employed.A study of the corpuscle in various buffers by means of the ultramicroscope shows no significant change in morphology except in extremely acid solutions.The effect of acid on amboceptor is similar to that of hemolysis as a whole. The hemolytic power cannot be restored entirely by changing the acidity; in other words, part of the amboceptor is destroyed, not merely inactivated.Similar results were obtained for the action of acid on complement alone, that is, most of the complement was merely inactivated but a little was destroyed under the conditions of the experiment.The buffered salt solution described for use in the Wassermann reaction by Mason and Sanford undoubtedly stabilizes a variable factor in the reaction but does not act by buffering the final mixture to neutrality since the buffer value of the cells is already sufficient to accomplish this, except in extreme cases.The experiments of Kolmer on the influence of hydrochloric acid and sodium hydroxid were repeated, determining also the final acidity in each tube. The results are somewhat variable, due in part to the difficulty of reading some of the tubes because of obscuring hemolysis.The work of Coulter on the union of the cell and amboceptor is confirmed by a different method. The point of maximal combination is well toward the acid side and the amount gradually decreases as neutrality is approached.Experiments on the union of complement to the sensitized cell would indicate that in an extremely acid or alkaline medium, the complement is not present, either in the cell or in the supernatant fluid, in a form in which it can be reactivated by neutralization.Complement can be split by hydrochloric acid over a range of at least pH 5.8 to 6.4. The addition of alkali in the standard method makes it very alkaline but keeps it within the range of optimal hemolysis. The complement can still be split however, if less alkali is used. Filtration of the split complement increases the final alkalinity, due to the amount of the acid mixture which remains on the filter paper.The mixture of mid piece and end piece which have been reunited is distinctly more alkaline than if the mixture was neutralized before the centrifuging and washing of the globulin. Evidently a considerable amount of acid has been lost somewhere in the process. Experiments show that the acid lost in the wash water will not account for the change but apparently some of the acid is bound by the precipitate so that it is not available for neutralization later. When extremely acid and alkaline distilled water is used, the wash water is brought nearer to neutrality than the original. If the wash water becomes more acid the precipitate becomes slightly more alkaline than it was at first, and vice versa.The results of complement splitting by carbon dioxid are variable and seem to depend more on the heaviness of the protein cloud than on the actual acidity. Failure is frequent at acidities which will cause splitting if attained by the use of hydrochloric acid.Splitting of complement by dialysis was only successful in one case. In all the attempts there was no great change in the acidity of the solutions.An attempt was made to split complement by acid phosphate solutions in the presence of an excess of amboceptor as was done by Michaelis and Skwirsky. The results were not very successful.A group of experiments based on those of Bronfenbrenner and Noguchi on complement splitting was carried out. With the use of hydrochloric acid, the acidity in the different tubes is very similar, regardless of whether or not the complement is split. In using carbon dioxid, the acidity is not increased by bubbling carbon dioxid through for more than two minutes. The reactivation of the hydrochloric acid end piece depends on specific factors, not on changes in acidity. The same is true of the end piece split by carbon dioxid.