532 F. M. BURNET 



relatively prolonged course of immunization, they provide an antiserum that 

 will react with all determinants of tobacco mosaic virus, the x proteins 

 present in sap from infected plants and the A proteins resulting from mild 

 chemical treatment of the virus. The conditions of line production in gels are 

 probably too complex to allow any precise statement beyond the indication 

 that disaggi'egated or altered fragments of protein are present m x and A. 

 Whether these components are present as such in the virus itself and whether 

 they play a part in the synthesis of the virus is unknown. 



Electron micrographs indicate clumping of the virus of tobacco mosaic 

 by antiserum m a fashion resembling the agglutination of bacteria (Anderson 

 and Stanley, 1941; Malkiel and Stanley, 1947). Using two spherical viruses, 

 Southern bean and bushy stmit, Black et al. (1946) observed principally a 

 failure of the aggregated particles to form the regular arrays seen without 

 antibody, and some evidence of an increased separation between the particles. 

 Bawden and Pirie (1938) drew attention to the characteristically different 

 macroscopic appearance of the precipitates with a spherical virus (bushy 

 stimt) and a filamentary one (TMV); the former being compact and granular, 

 the latter loosely flocculent, resembling closely the and H types of agglut- 

 ination seen with bacteria. 



3. Animal Viruses 



In this field, aggregation reactions involving virus particles which had 

 been concentrated and at least partially purified have been reported for 

 vaccinia viruses, influenza viruses, and poHo viruses. There is httle doubt that 

 they ^\^U be equally demonstrable w4th any virus which can be obtained in 

 purified form and for which a high titer immune serum is available. Relatively 

 httle work has been reported and the results have not significantly influenced 

 the general trend of virological research. 



Immunological work with purified vaccinia virus was initiated by Craigie 

 and Tulloch (1931), and subsequently elaborated by Craigie (1932) and by the 

 group associated with Rivers at the Rockefeller Institute. Two soluble anti- 

 gens, L — S protein (Shedlovsky and Smadel, 1942) and NP, nucleoprotein, 

 were obtained from the elementary particles; both the corresponding antisera 

 were shown to agglutinate particle suspensions (Smadel et al., 1942). Neither 

 of these antisera, however, showed any power to neutralize the infectivity of 

 the virus. The power of a hyperimmune serum to inactivate the virus may 

 be related to a residual agglutinin x, which remains after absorption with the 

 two chemically characterized antigens (Smadel and Hoagland, 1942). 



Detailed study of the precipitin reactions of purified influenza virus by 

 Knight (1946) was prmcipaUy concerned to demonstrate the presence in or 

 on the virus surface of antigenic groups with host rather than viral specificity. 

 Virus prepared from infected allantoic fluid showed chick antigenic quahties. 



