IMMUNOLOGICAL METHODS IN THE STUDY OF VIRUSES 531 



progressively more important for diagnostic work in animal virology in recent 

 years, 



A. Precipitin Reactions 



1. Bacterial Viruses 



Macroscopic agglutination (or precipitation) of viruses of the CI 6 - T2 

 group was demonstrated by Schlesinger (1933) and Burnet (1933a). The latter 

 found that the range of viruses agglutinated corresponded to the group 

 defined on the basis of neutrahzation reactions. An indirect type of aggrega- 

 tion reaction with bacterial viruses is by the agglutination of formalin-killed 

 bacteria to which phage has been specifically adsorbed at high multiphcity 

 (Burnet, 1933c). This has been confirmed by Tolmach (1957), but little or no 

 use has been made of either method in recent years. 



2. Plant Viruses 



Precipitm reactions with rabbit antisera are used as a standard method in 

 aU laboratories working with plant viruses and the reaction has been exam- 

 ined in detail by Kleczkowski (1941), Schramm and Friedrich-Freksa (1941), 

 and Malldel and Stanley (1947). In general, the results correspond to those of 

 other typical antigens, the quantitative relations at the equivalence point 

 being what would be expected in relation to the size of the virus particle. 

 With tobacco mosaic virus, Malkiel and Stanley calculated that about 60 

 molecules of antibody were attached to each virus rod at the equivalence ratio. 



The specificity relationships between related but not identical virus types 

 conform to what has been found with bacteria or animal viruses. Precipita- 

 tion with a heterologous antiserum is less complete than with homologous. 

 Less antibody is precipitated with heterologous virus than with homologous. 

 Absorption with homologous virus removes all antibody while heterologous 

 absorption leaves most of the activity against the homologous virus. Such 

 relations hold irrespective of the host plant on which the virus was grown 

 (Mallviel, 1947). Complement fixation reactions can be show^n with tobacco 

 mosaic virus as antigen by standard technique (Chester, 1935). 



More recently, agf.r gel precipitation methods (Ouchterlony, 1949) have 

 been extensively used, the most recent account bemg that by Kleczkowski 

 (1957). Most workers have been concerned wdth tobacco mosaic virus (Sang 

 and Sobey, 1954), and, specifically, with the relationship of the antigenically 

 related proteins free of nucleic acid, which are either fomid in infected tissues 

 (Takahashi and Ishii, 1952; Commoner and Rodenburg, 1955; Commoner and 

 Yamada, 1955) or produced by the action of alkah on purified tobacco mosaic 

 virus (Schramm, 1943). The results reported by Kleczkowski, which are con- 

 cordant with those of Jeener et al. (1954) indicate that more than one line- 

 forming component is present in tobacco mosaic virus. If rabbits are given a 



