EVOLUTION AND VIRUSES 



long spell is needed for them to change their behaviour con- 

 siderably. Passage through a new host, or exposing an old 

 host to a new environment, is often all that is needed to 

 change some property strikingly. The apparent purposefulness 

 of the adaptation to a form better suited to the new host or 

 to the new environment may seem to suggest that the new 

 treatment is actually causing the change, but for this there is 

 no evidence. All the phenomena are adequately explained by 

 postulating selection operating on random-variants. Some of 

 the random variations probably reflect changes equivalent to 

 mutations, but not all need be. There is much evidence that 

 some of the offspring that are produced when two strains are 

 multiplying together in one cell combine characters that pre- 

 viously belonged to either one or other of the parent strains, 

 so that recombination of existing genetic factors could also be 

 a potent force in variability. However, with the immense 

 populations of virus particles that occur in infected cells, only 

 a small mutation rate would be needed to produce many 

 variants on which natural selection can begin to operate. And 

 selection is likely to operate quickly, for strains of one virus 

 compete strongly with one another. With plant viruses the 

 competition is such that when a cell is fully infected with one 

 strain, other strains are precluded from multiplying in it. 

 Hence, should a variant occur that multiplies and moves from 

 cell to cell a little faster than the parent type, it will not only 

 have this initial advantage, but by entering cells sooner it will 

 also prevent the parent type from coming later and interfering 

 with it. All that is needed, for example, to get strains better 

 adapted than the type strains of tobacco mosaic virus for 

 existence in plants at 37°, is to infect plants with the type 

 strains and put them at 37° for a day, after which the pre- 

 dominating strains at many infected centres are already 

 variants that multiply and move more readily at the higher 

 temperature. Most variants of viruses have, of course, been 

 found in the course of experiments designed specifically to 

 identify them, but there is abundant evidence for their 

 occurrence in the wild. Variants differing from one another 

 in almost every property for which tests have been devised 

 have been encountered. The most common type of variation 

 to be described, probably only because it is the easiest to 



22 



