SEROLOGICAL VARIATION 269 



antigenic constitution. The latter refer as R-S relation to the situation in 

 which two strains may be identical in one direction but differ markedly in 

 another. It is seen between mouse and egg lines of the same strain, lines of the 

 same strain from different laboratories, and in strains from the same epidemic. 

 It emphasizes the need for multiple cross reactions to obtain a reasonable view 

 of strain relationships, particularly when mouse lines may disclose hidden 

 antigens in the egg lines. 



Another phase of serological variation under active study at present 

 relates to antigenic and genetic recombination among strains. It is discussed 

 elsewhere in this volume. 



IV. Variation in Arthropod-Borne Viruses 



A. Demonstration of Strain Relationships 



Knowledge of these "arborviruses" has advanced progressively, with the 

 result that instead of a diffuse array of individual items of diversified origin, 

 they are congregating as groups of antigenically interrelated strains. The 

 differences may well be imprints made on a basic antigenic configuration by 

 the respective hosts and vectors in which they are maintained, just as 

 individual human hosts influence the strains of influenza virus they entertain. 

 This recalls Hammon's (1948) description of a grandfather or stem virus. 

 Although numerous investigators have contributed to the advance, the 

 extended observations of Smithburn (1952, 1954) with the neutralization 

 test in mice first demonstrated a series of no less than 12 interrelated agents 

 whose cross-reactions were usually nonreciprocal; each of them had, however, 

 an antigen in common with at least one other member of the group. The 

 development of hemagglutinating techniques for the encephalitis viruses 

 by Sabin and Buescher (1950), Chanock and Sabin (1953a,b,c, 1954a,b), 

 Sweet and Sabin (1954), led to systematic study of relationships by Casals 

 and Brown (1954), who formulated two sharply defined groups, A and B. 

 Later Casals (1957) classified 47 strains of arborviruses with the addition of a 

 group C of eight members, while eight others were as yet unclassified. 



B. Value of Successive Exposure in Demonstrating Relationships in Virus 



Groups 



Smithburn (1954) demonstrated that an animal immune to one virus of a 

 group may respond to a second virus, not only by producing antibody against 

 -the second, but by producing more antibody to the first and also antibody 

 capable of reacting with still other viruses. Diverse cross-reactions were 

 noted with serum of human subjects convalescent from infection with one or 



