252 THE ANTIGEN.ANTIBODY REACTIONS 



response, and of valency ; and variations in k would explain the effect of salts, lipins 

 and other non-specific substances on the progress of immune reactions. 



The analysis of specific precipitates in the region of antigen excess should 

 provide the best direct evidence for or against the multivalence of antibody. In 

 their review Hooker and Boyd (1942) point out that in only two systems have the 

 recorded estimates of the molecular ratio of antigen to antibody been more than 

 unity, namely, with diphtheria toxin (Pappenheimer, Lundgren and Williams 

 1940), and with Heidelberger and Kendall's (1934) azo-dye ovalbumin antigen — 

 and in both cases the value of 2 is in doubt. We have already noted that the 

 value of 2 accords well with Hershey's descriptive theory of the lattice hypothesis. 

 It is clear that precise analyses of a large number of antigen-antibody compounds 

 of this kind are required to settle this point. 



We may summarize the position with regard to the unity or diversity of anti- 

 bodies as follows. 



The antibodies in a monospecific antiserum prepared against a single antigen 

 may vary in physicochemical properties, such as particle size, electrophoretic 

 mobility, readiness to dissociate from union with antigen, and perhaps in valency ; 

 in " combining power " and " avidity " ; in resistance to destructive agents ; and 

 in capacity to form complexes with antigens that are susceptible to other substances 

 in the reacting system, such as electrolyte, complement, or a phagocyte. In no 

 case, however, is there good evidence of association between a given set of physico- 

 chemical properties, and a particular type of serological reactivity. The hetero- 

 geneity in a collection of antibody particles present in a certain preparation may 

 affect the different serological reactions in different degrees, but there is no reason 

 to suppose that opsonins, precipitins, antitoxins and so forth differ in kind, either 

 constantly or fundamentally. In this sense then, we may accept the hypothesis 

 of the unity of antibodies. 



The Nature and Properties of Antigens 



The two outstanding characters of antigens, their power to stimulate antibody 

 production and their specificity, may be to some extent explained in terms of 

 their chemical constitution. The basis of specificity is the more clearly understood, 

 and it will be convenient to discuss it first. 



The Basis of Specificity. 



Specificity is not absolute. There is no one natural antigen of which it would 

 be safe to predict that it would not react with antibody prepared against another 

 antigen. The variety of syntheses of which living tissues are capable, though 

 wide, is limited, and many tissue constituents in one organism are similar to those 

 in another organism. Where these constituents are antigenic, we often find the 

 similarity reflected in the serological reactions of the antigens with homologous 

 and heterologous antibodies. 



Thus the lens proteins of mammals are antigenically similar, though they differ 

 markedly from other proteins found in the animal body. This type of specificity 

 is referred to as " organ " specificity, and has been noted in antigens from brain, 

 testes and placenta, and in mammalian proteins like keratin, fibrinogen and thyro- 

 globulin (see Landsteiner 1936). 



On the other hand, antigens are often " species " specific. We can differentiate 

 by serological tests between the egg albumin of the duck and of the hen (Dakin 

 and Dale 1919), between the haemoglobins of different animal species (Higashi 



