76 AN INTRODUCTION TO THE STUDY OF VIRUSES 
inoculation tests. The serological test can pick up this type of strain 
just as easily as any other. 
The facilities required for serological testing are not so extensive 
as for plant inoculation methods, the chief requirements being a 
suitable water bath to run at 50°C, a small power-driven centrifuge to 
take eight to twelve 15-ml tubes, and a supply of small tubes and 
pipettes. 
The carrying out of serological testing is not limited by the seasons 
as the growing of test plants may be. The result of the test is known 
within an hour or so, compared with two to three weeks for the 
inoculation methods. . 
The precipitin test therefore is a useful tool for the identification 
of a specific virus but is not sufficient to distinguish between virus 
strains since all strains of the same virus will react with an antiserum 
prepared against one strain. 
This merely indicates that the strains possess common antigens; 
it does not mean that they are antigenically identical. 
Strains of the same virus can, however, be differentiated anti- 
genically by the method of cross-absorption. This has been clearly 
explained by Bawden (1943) as follows— 
If each virus is not a simple unit antigen, but carries a number of different 
determinant groups, then the antiserum will also contain a number of 
different antibodies, each reacting specifically with its particular determinant 
group. The viruses which have one or more determinant groups in common 
will be precipitated equally by each other’s antiserum. But if, in addition 
to the common antigenic group, each strain contains specific groups, then 
their effects on their homologous and heterologous antisera will be different. | 
Each strain will react fully with its homologous antiserum, removing all 
the antibodies from it. It will remove from its heterologous antiserum, 
however, only those antibodies for which it has determinant groups, the 
others specific to the groups peculiar to the second strain being unaffected. 
The presence of such specific antibodies is shown by the formation of a 
precipitate when the virus strain used for producing the serum is added to 
a sample that has been allowed to react fully with a second strain. 
Chester (1936”) carried out a separation and analysis of ten tobacco 
mosaic virus strains by means of cross-absorption tests. His results 
are given in Table V. 
From this it may be seen that, after tobacco mosaic virus immune 
serum is allowed to react fully with aucuba mosaic virus extract, the 
serum is still capable of reacting with tobacco mosaic virus. The 
absorption of tobacco mosaic virus serum by an excess of aucuba 
