58 D. A. L. DAVIES 



solutions have either been too dense (e.g. 6M-urea), or given 

 micelle formation (lysolecithin). There is therefore no evidence 

 for inhomogeneity, but only suggestive evidence for physical 

 homogeneity. The latter is, however, enhanced by the fmding 

 that the material sedimented in the ultracentrifuge, v^hen resedi- 

 mented in three fractions at successively increasing g values, 

 cannot be distinguished by chemical analysis. 



Serological properties 



Information on immunological homogeneity has been obtained 

 by agar diffusion, using rabbit anti-mouse sera. Mouse anti-mouse- 

 H-2 sera do not give precipitin lines with H-2 antigen prepara- 

 tions in agar presumably because H-2 antibody is "incomplete". 

 Since ascitic fluid has a composition similar to that of serum, 

 though with a smaller total concentration of solutes (35 mg./ml. 

 after dialysis), the extent of removal of serum components can be 

 checked. Globuhns compose the greater part of the 0*75 vol. 

 water dilution precipitate, but in Fig. 2a it will be seen that they 

 are also present in large amount in the DP fraction, although 

 some other components have been substantially removed. The 

 Spinco supernatants contain the bulk of the remaining serum 

 components; four sedimentations leave an SP fraction in which 

 two serum components are still detectable, one of these being 

 globuhn, but six sedimentations reduce these to a negligible level. 

 The fmal TM fractions are free from serum components as tested 

 at a level to show contamination down to i per cent. 



Rabbit anti-mouse-cell sera, after absorption with normal 

 mouse serum, show six main lines due to cell-bound antigens 

 (Haughton, Boyle and Davies, 1962). No single serum is available 

 which possesses all six antibodies in detectable amounts. In Fig. 26 

 a serum having several of these components is shown reacting 

 with the H-2 products and it can be seen that the fmal TM2 

 fraction contains one of these antigens only, but this at a high 

 level of activity. This antigen (antigen no. 3, Haughton, 1962) 



