12 BRENT, MEDAWAR AND RUSZKIEWICZ 



or i-o per cent Triton loo or sodium deoxycholate no longer 

 sensitizes upon injection. However, it is difficult to attach much 

 weight to experiments in which the surface-active agent must 

 either be injected in company with the antigen exposed to its 

 action or be removed by organic solvents that may themselves 

 be damaging. 



For many biological purposes an antigen which has merely 

 passed through a soluble state is not good enough: the antigen 

 must be soluble in the form in which it is used. We have therefore 

 taken advantage of Castermans' (1961) finding (cf Mann, 

 Corson, and Dammin, i960) that sensitizing activity is not wholly 

 lost by antigenic sediments which have been exposed to high 

 pH's (in our experience, even as high as 11 -o). We have found it 

 convenient to bring an aqueous suspension of antigenic sediment 

 to pH 10' or II- o by the very slow addition of o-oiN-NaOH, 

 using an internal glass electrode system to monitor the pH, and 

 preventing high local concentrations of alkali by brisk stirring 

 with a plastic-coated magnetized iron bar. After removal of 

 insoluble matter by spimiing at 30,000^ for 30 min., the super- 

 natant was brought to pH 8-0 with o-oin-HC1, and NaCl was 

 added to a fmal concentration of o-I5M. Any precipitate so 

 formed was removed by a second centrifugation at 30,000 g 

 for 30 min. The fraction of residual activity in the supernatant 

 fluid has not yet been determined, but it is variable, and certainly 

 very small. 



The point of biological interest that has emerged from the use 

 of this solubilized preparation, and of the original "crude aqueous 

 preparation" (which contains some antigenic matter in a fmely 

 divided state), is that soluble or semi-soluble preparations do not 

 sensitize when administered intravenously. This point may be 

 taken as firmly established: the experiments summarized in 

 Table III are confmed to those that had strict internal controls. 

 In addition there are faint indications, now being investigated, 

 that an intravenous injection of soluble antigen may sometimes 



