ANTIGENS, HAPTENS, ANTIBODIES, ETC. 425 



Depending on the source of the antibody, from horse or 

 rabbit senim, for example, on the intensity and length 

 of the course of immunisation, and on the exact method 

 of fractionation, the antibody may be found to be 

 associated with one or other of the globulins or divided 

 between them. This is not surprising when it is remem- 

 bered that the proportions of albumin, pseudoglobulin 

 and euglobulin which can be separated even from normal 

 serum are very variable. 



(b) Precipitation ivith Alcohol. — By this method, also, 

 the antibodies are separated with the globulin fractions. 

 Denaturation of the proteins by alcohol is avoided by 

 working at 4° C. or lower, or by bringing the alcohol 

 concentration rapidly above 90 per cent. By pre- 

 cipitating the antibody from an anti-pneumococcus 

 serum in the cold with 10 per cent, alcohol in the 

 presence of N/200-sodium chloride at pH 6-7, Felton 

 succeeded in removing about 90 per cent, of the inactive 

 protein. 



(c) Adsorption Methods. — Methods similar to those 

 introduced by Willstiitter for separating enzymes have 

 also been found effective in separating antibodies. For 

 instance the antibody to the typhoid bacillus can be 

 adsorbed on alumina and then eluted with dilute (N/100) 

 alkali. Diphtheria antitoxin, the flagellar typhoid 

 antibody and the -antibody of Salmonella enteritidis are 

 adsorbed by kaolin from which they can be eluted with a 

 solution of 2 per cent, glycine in 2 per cent, sodium 

 chloride. The eluted typhoid antibody solution gave 

 negative reactions for proteins but contained 0-6 mg. 

 of nitrogen per ml. ; it was not affected by proteolytic 

 enzymes. Only 15 to 20 per cent, of the antibody could 

 be recovered in this way, but that which was obtained 

 was about six times as concentrated as the original 

 serum . 



((/) Electrophoresis. — Antibodies have an isoelectric 

 puint at about /)H 5-5 and accordingly move in an 



