138 III. THE EVOLUTION OF VIRUSES 



It is most remarkable that this agent is reported to consist mostly 

 of nucleic acid, and according to Hotchkiss (17) the protein content of 

 the agent can be, at most, 0.2 per cent and is probably less. If this 

 is true, it must be admitted that the template action is sometimes 

 accomplished by nucleic acid only without being associated with any 

 protein, although this possibility is not entirely inconceivable as dis- 

 sussed previously (Part II, Chapter VIII). 



The nucleic acid concerning this reaction is confirmed to be 

 desoxyribonucleic acid, whereas this acid isolated in a pure form is 

 reported to exist in forming virus-like particles ; its molecular weight 

 about 8,000,000 and the length of the particle, 3000m// (18). This sug- 

 gests the possiblity that even the particles composed of nucleic acid 

 only can behave like a virus. 



From avirulent strain could be extracted similar nucleic acid 

 resembling, in its chemical composition, to the agent isolated from the 

 virulent strain, but it was wholly inactive in producing transformation 

 (19), a fact which should be naturally expected, since the agent from 

 the virulent strain can transform the avirulent only because the 

 former structure is more powerful than the avirulent, and accordingly 

 this avirulent strain itself or its particles should be unable to exert 

 any influence upon the virulent. 



The nucleic acid complex isolated from the avirulent strain (R- 

 type), however, can transform a variant (ER-type), which produces 

 very rough colonies, to R-type, indicating that ER-type is still weaker 

 than R-type. In this connection, it should be noted that transfor- 

 mation in the reverse direction, R->ER, can be carried out with a 

 transforming agent of the ER variant if anti-R serum is present 

 (20). This fact shows that despite its weak structure the agent isolated 

 from ER variant can overcome the R-type owing to the assistant 

 cooperation of the antiserum capable of disturbing the structure of 

 R-type. 



Similar transforming agents have been isolated from various other 

 bacteria such as E. colt (21), Shigella (22), and Haemcphillus (23). 

 Furthermore, it has long been known that certain strains of bacteria, 

 grown in the presence of culture filtrates or extracts of related orga- 

 nisms, acquire some of the properties of the latter. Thus, it was 

 possible to convert colourless and nonproteolytic culture of B. pyocya- 

 neus into virulent and pathogenic form, by two consecutive passages 

 in the culture filtrate of a virulent, pigmented and proteolytic strain 

 of the same species. There are also claims that cultures of strepto- 

 cocci or of Gram-negative bacilli exposed to the cells or products of 

 serologically different, but biologically related species, can develop the 

 specific agglutinability of the latter (24). 



