VIRUS GENETICS, MULTIPLICATION, AND PHYSICS 185 



NWS with no enzymatic or indicator activity, but neurotropic; 

 WSE with firm nonekiting adsorption to red cells, and WSM, a 

 very heat-resistant hemagglutinin , can recombine after multiple 

 infection to produce recombinant variants in much the same way 

 as found by Hershey and Rotman (1949). Thus, recombination 

 can be considered as a property of animal viruses also. 



Stated very boldly, the way the geneticist works is by dis- 

 covering a means of genetic combination, for example mating, 

 and then analyzing the progeny in terms of some clear property 

 to see what proportions of the progeny show the property. "One 

 white, one black and two khaki" is fundamentally what is 

 sought, although the whole beautiful and intricate pattern of 

 modern genetics contains the description of genotypes involving 

 thousands of separate genes. Phage genetics has only a small 

 number of phenotypic expressions to consider at the present mo- 

 ment, but is rendered a powerful and challenging subject by the 

 ability to observe large numbers of progeny and the fact that 

 triparental recombination is also observed, thus adding another 

 dimension to the range of possible experimentation. 



To illustrate the method of study, consider multiple infection 

 of E. coli cells with T-2 phage. These can be different strains of 

 T-2, for example T-2r, a rapidly lysing form, and T-2/?, a phage 

 which will multiply in J5/2, a different form of the host bac- 

 terium. It is possible, by plating on a mixture of B and B/^, to 

 tell the proportions of r and h and of recombinants between 

 them. The rapid lysis feature can be exhibited by 15 different 

 types of virus corresponding to 15 mutant forms. These are num- 

 bered Ti to ri5. Now, if Ti and h are used to produce a mixed in- 

 fection, a process represented by ri X h, there result, from many 

 bacteria, the following proportions: 30% ri, 30% h, 20% rih, and 

 20% with neither feature, the wild type. These figures are not 

 the standard proportions of genetics, but this is not a standard 

 process. The important fact is that relatively high proportions of 

 the recombinants rji and the wild type are produced. If, in place 

 of a mixed infection of h and ri, h and r^ had been used, a much 

 smaller recombinant proportion, only 0.8%, is found. Hershey 

 infers that h and ru are closely linked, so that they go together 



