THE REVERSIBILITY OF ANTIGEN-ANTIBODY UNION 217 



a consideration of the later stages will necessitate assumptions whose validity 

 may not be independently demonstrable. 



Before discussing the evidence for the two views of the reaction, we must 

 first understand how far the two mechanisms — adsorption and chemical combina- 

 tion — can be distinguished. 



In the formation of chemical compounds of the type involving intramolecular 

 rearrangement of atoms, whether these compounds are freely dissociable or not, 

 the forces that hold the constituent atoms of the compound together are, according 

 to the electronic theory of valency, derived from a change in the distribution of 

 the electrons (see Marrack 1938). In some cases one atom may simply give up one 

 electron to another. Two oppositely charged ions are produced. In this case, 

 no true compound is fornied ; for in solution, the ions exist independently ; and 

 in the crystal, although the ions occur in a ratio corresponding to the conception 

 of a compound in definite proportion, there is no continuous pairing of oppositely 

 charged ions to form molecules of the compound, but an interpenetrating lattice 

 of the different kinds of ions, in which each ion is in an electrostatic relation with 

 all its immediate neighbours. In other cases electrons are shared between the 

 atoms concerned. The shared electrons may be contributed by one of the atoms 

 concerned (co-valent bond) or by both atoms (co-ordinated bond). Combination 

 of this kind must, of necessity, obey the law of simple multiple proportions. 



There is no clear distinction between the formation of compounds by co-valent 

 or co-ordinate bonds, and the formation of absorption complexes. In the latter, 

 two molecules are held together by the attraction of the nuclei of certain atoms 

 in one molecule for the electrons of atoms in the other, and vice versa, according 

 to the distribution of electrically active polar groups on the molecular surfaces. 

 With small molecules the distinction can be made because it is technically possible 

 to separate and identify the constituent molecules, and to determine whether or 

 not they were united in constant proportions. Large molecules are more difiicult 

 to separate and identify, because absolute differences, such as the presence or 

 absence of a methyl group, are proportionately small ; and by the same token, 

 the law of simple multiple proportions is meaningless as soon as the molecular 

 weights of the reacting substances are so large that the removal or insertion of 

 an atom is no longer detectable by available analytical methods. 



We may assume, if we wish, that adsorption depends primarily on intermolecular 

 forces as opposed to intramolecular forces, and that combination depending on 

 the intermolecular attractions will take place in varying proportions ; that is, 

 combination will not be subject to the law of simple multiple proportions. Never- 

 theless, the existence of compounds showing varying proportions of the two con- 

 stituents will not be an absolute criterion for the differentiation of adsorption 

 compounds from those that are formed by the intramolecular union of atoms. 

 If a molecule is very large, -and contains numerous active groups, it may be very 

 difficult to obtain a satisfactory differentiation on this basis. The nitro-compounds 

 of cellulose, for instance, have a composition that varies over a wide range. More- 

 over, a compound that is formed as the result of intermolecular forces may undergo 

 secondary changes in which intramolecular forces are involved. We may, however, 

 take it as a safe guiding rule that if the combination of two substances is shown 

 to be subject to the law of simple multiple jjroportions, it is an indication that 

 intramolecular forces are involved, while an apparent failure to obey this law 

 suggests, though it does not prove, that we are dealing with an adsorption compound. 



