VARIATION AND ITS CHEMICAL CORRELATES 135 



strains of tobacco mosaic virus and two cucumber viruses whose chemistry 

 will be developed in later sections. 



It is apparent from the data in Table I that all of the strains tested except 

 HR (Holmes' ribgrass) are closely related serologically to TMV. That the 

 antigens of two closely related strains, such as TMV and YA (yellow aucuba), 

 are not identical can often be brought out by cross-absorption tests, as de- 

 scribed by Bawden (1950). Another way of assessing the degree of relation- 

 ship is by use of quantitative precipitin tests (Malkiel, 1947; 1948), although 

 with certain strains of TMV tliis technique seems less able to reveal minor 

 serological differences than cross-absorption tests combined with semi- 

 quantitative precipitin reactions (Knight, 1942). 



A major point about the serological tests is that viruses which are almost 

 certainly strains by other criteria have thus far always shown positive 

 serological cross-reactions. This fact has led some workers to conclude that 

 the serological test is the most rehable evidence of strain relationship. In 

 general, this supposition appears valid, although in the cases of weak sero- 

 logical cross-reactions it would seem to be assuming more than present 

 knowledge of the chemical basis of serological specificity warrants. A case in 

 point is the relationship of CVS and CV4 (cucumber viruses) to the viruses of 

 the tobacco mosaic group. While these cucumber viruses show an immuno- 

 logical relationship to the TMV group (Table I; and Rochow, 1956), they 

 seem excluded on most other grounds (Knight, 1955a). 



2. Cross-Protection Tests 



Generally, a plant once infected with a virus remains infected throughout 

 its Hfe. Contact of the plant with other viruses capable of infecting that host 

 usually results then in one of two major responses: (1) the plant is suscept- 

 ible to infection by the second virus and new symptoms appear super- 

 imposed upon the first, or (2) the plant is resistant to infection by the second 

 virus and the latter is excluded. When it can be shown that this exclusion is 

 reciprocal, one has demonstrated cross-protection. (Synonyms include 

 "mutual antagonism," "cross-immunization," and "interference.") Often 

 protection can be shown conveniently in only one direction, and except for 

 the most rigorous comparisons this may suffice. 



Experience with plant viruses has shown, in general, that exclusion occurs 

 between related viruses, that is, between mutants or other variant strains of 

 the same virus group, whereas midtiple infections occur with imrelated 

 viruses. As Bennett (1953) has pointed out, there are exceptions to and 

 modifications of the rule, depending somewhat on methodology, type of 

 plant virus, and other circumstances. Despite these reservations the rule 

 holds weU enough to make it one of the most highly regarded criteria of strain 



