136 C. A. KNIGHT 



relationship. Furthermore, it gains strength when coupled with the sero- 

 logical tests, since there seems to be only one case reported (Bawden and 

 Kassanis, 1945) in which serologically imrelated plant viruses showed 

 strong cross-protection. 



3. Similarity in Host Range 



Plant viruses, in common with viruses in general, show a specificity for 

 hosts in whose cells they can multiply. This specificity is narrow or broad, 

 depending upon the plant virus (Holmes, 1938; Price, 1940). However, 

 members of the same group of viruses would be expected to have similar 

 requirements for host conditions suitable for their multiplication. "While ex- 

 tensive systematic studies have not been made to test this assumption, ob- 

 servations in genera] attest to its vahdity. Hence, strains of a virus have 

 similar host ranges, although, as pointed out earher, they may not be 

 identical. Obviously, this test for strain relationship is less highly specific 

 than the serological or cross-protection tests; nevertheless, it can be quite 

 significant. 



4. Similarity in Method of Transmission 



Some plant viruses are transmitted mechanically (such as by rubbing juice 

 from an infected plant on the leaf of a healthy plant), others are transmitted 

 by grafting, by means of parasitic plants, or by insect vectors. Some viruses 

 may be passed by several or all of these means. In the case of insect trans- 

 mission, specificity is often shown by the type of insect vector and by 

 whether or not the virus undergoes a latent period in the vector and persists 

 for more than a Hmited time. Members of the same \arus group are commonly 

 transmitted by the same methods, but sometimes unrelated viruses are, too 

 (Bawden, 1950). Thus, similarity in method of transmission can help to show 

 strain relationship between viruses, although this would be considered in- 

 sufiicient if it were the sole evidence for strain relationship. 



5. Similarity of Response to Genetic Change in Host 



This test has not been widely used, largely because of the restrictions im- 

 posed by the test plants required. In brief, one must have two groups of test 

 plants differing specifically in genetic constitution, which in turn must result 

 in distinguishably different disease symptoms when a given virus is inoculated 

 to plants of the two sets. It is then found that viruses which belong to the 

 same virus group as the first one (that is, are strains) likewise cause, when 

 inoculated to the two sets of plants, two sharply distinguishable disease 

 types. Viruses unrelated to the one for which the test plants were established, 

 but which can infect these hosts, produce indistinguishable diseases in the 

 two sets of plants. 



