236 THE ANTIGEN-ANTIBODY REACTIONS 



do not show the same strict specificity as the bacteriotropins — normal unheated 

 serum promotes the phagocytosis of a wide variety of antigenically unrelated 

 bacteria — but from the first there was evidence that suggested the presence of 

 specific factors of one kind or another. Thus, it has been the general experi- 

 ence that the serum of any one person varies in its opsonic effect on different 

 bacteria ; that the sera of different individuals may show striking differences 

 when tested against the same bacterium ; and that any one person may show a 

 variation in the opsonizing power of his serum for a particular bacterium, especially 

 as the result of infection with that organism, or of artificial immunization. Varia- 

 tions of this kind have been studied extensively by Wright and his colleagues 

 (see Wright 1909). Moreover, evidence pointing in the same direction was obtained 

 by in vitro experiments. Bulloch and Western (1906) succeeded in removing the 

 opsonic power of normal serum for particular bacteria by selective absorption. 

 Other workers were unable to confirm these results, but the careful studies of 

 Hektoen (1908) on the normal opsonins acting on the red blood corpuscles of 

 different animal species afforded strong support for the correctness of Bulloch 

 and Western's contention. 



The explanation of these anomalous findings would appear to lie in the fact 

 . that the opsonic effect of normal serum resembles its hsemolytic and bacteriolytic 

 effects in being dependent on both antibody and complement. Chapin and Cowie 

 (1907) showed that normal serum may have its opsonic power for a staphylococcus 

 removed by absorption with that organism in the cold, but that such absorbed 

 serum may still have the power of reactivating normal serum that has been in- 

 activated by heat. The cocci that have been used for absorption, when washed 

 and resuspended in saline, show little if any increased susceptibility to phagocytosis 

 in the absence of unheated serum. Later (Cowie and Chapin 1907) they showed 

 that normal serum loses almost all its opsonic power when diluted fifteen times 

 with saline. If, however, such diluted serum is added to another sample of serum 

 that has been inactivated by heating at 55°-60° C. for 10 minutes the mixture is 

 almost as active in promoting phagocytosis as was the original unheated, undiluted 

 serum. It would appear that normal serum contains specific sensitizing antibodies, 

 present in amounts so small that they are ineffective in a dilution of 1 : 15 or more. 

 Even in undiluted serum these antibodies are unable, by themselves, to alter 

 the bacterial surface sufficiently to promote phagocytosis ; but when complement 

 is adsorbed by the incompletely sensitized bacteria the necessary change in surface 

 conditions is produced. It would seem, also, that complement is not without effect 

 on the action of the bacteriotropins ; for G. Dean (1907) has shown that the action 

 of a heated antiserum is increased by the addition of a little unheated normal serum. 

 Sleeswijk (1908) has confirmed Dean's results, and concludes that sensitization, as 

 a preliminary to phagocytosis, is primarily dependent on a specific antibody. 

 When this antibody is present in adequate concentration it can produce its effect 

 in the absence of complement, though added complement may enhance it. When 

 the antibody is present in very small amount complement is necessary to induce 

 adequate sensitization. (See also Ward and Enders 1933.) 



It is of interest to note that the action of complement in opsonization appears to 

 differ in some way from its action in haemolysis ; since, as Gordon, Whitehead and Wormall 

 (19266) have shown, the addition of ammonia renders Qomplement inactive as a haemolytic 

 agent, but removes none of its opsonic activity. 



Gordon and Thompson (1935) inactivated complement by ammonia, Congo red, acid, 



