536 r. M. BURNET 



mixture may differ according to the host strain on which titrations are made 

 (Kahnanson and Bronfenbrenner, 1942). 



Partly neutrahzed phage will fail to pass a gradocol membrane tlirough 

 which the same phage will pass readily in the absence of sermn (Andrewes 

 and Elford, 1933). This may well be due to adsorption of antibody of Lanni's 

 second type but two other phenomena described by Burnet et al. (1937) 

 appear to be relevant. These are the production of very small diameter 

 plaques by partly neutralized virus and the loss of sensitivity to inactivation 

 by specific bacterial extracts. This jDhenomenon is almost the converse of that 

 described by Cami and Clark (1954) and mentioned above. 



5. Bacteria killed by formalin actively adsorb phage and are then agglutin- 

 able by antiphage sera and can absorb out neutrahzing antibody (Burnet, 

 1933c; Tolmach, 1957). In other words, attachment to bacteria does not block 

 completely the sites at which neutralizing antibody molecules can be ad- 

 sorbed. It can also be shown (Burnet et al., 1937; Hershey and Bronfen- 

 brenner, 1952; Nagano and Mutai, 1954) that neutralized virus is still at least 

 partially capable of attachment to bacteria. 



6. Bacterial viruses can be inactivated by material extracted from the host 

 cell. These contain polysaccharide (Gough and Burnet, 1934) and probably 

 represent fragments or derivatives of the Upomucoproteins which make up 

 the somatic antigen of the cell wall of intestinal bacteria (Goebel and Jesaitis, 

 1953). One such component active against T5 has been shown by Weidel and 

 KeUenberger (1955) to be in the form of small spherical particles which can 

 be seen in electron micrographs attached to the tip of the tail of inhibited 

 phage. There is a very close resemblance in the kinetics of inactivation by 

 phage by antiserum and by such bacterial products (Burnet et al., 1937). 



7. The process of attachment of phage to bacterium is complex. Morpho- 

 logically it appears that the distal half of the tail sheath unwinds or frays 

 out into fine fibrils exposing a core which appears to be the agent responsible 

 for penetration of the cell wall and irreversible bmding of the phage. It is 

 stated by Tohnach (1957) that neutralization by specific antibody uiliibits 

 irreversible union of T2, but hardly affects the initial, reversible phase of 

 union. Once irreversible union has been affected, subsequent addition of anti- 

 phage does not prevent mfection (Delbriick, 1945). 



From this summary, a relatively simple interpretation of the action of anti- 

 body on phage can be derived. Neutrahzation results when sufficient antibody 

 molecules are attached to the sheath (fibrillar) protein of the distal third of 

 the tail in such a fashion as to prevent the proper functioning of the process 

 leading from reversible to irreversible imion. One would expect (and there is 

 notliing reported incompatible with the exj^ectation) that a very high con- 

 centration of antibody will also block reversible union. One antibody mole- 

 cule appropriately attached is apparently enough to interfere sufficiently 



