RECOMBINATION IN BACTERIOPHAGE 47 



little less. Many of the stocks, particularly of the r mutants, contain unknown 

 substances inhibiting adsorption of the virus if the dilution into bacterial 

 suspension is less than 1 : 100 or so. There is no difference in the adsorption of 

 different stocks at high dilutions of the virus. If a mixed infection is attempted 

 with an r stock which contains inhibitor, and an h or wild type stock which 

 does not, the adsorption of both viruses in the mixture is prevented equally. 

 No undesirable disturbance of the relative multiplicity of infection in mixtures 

 is therefore encountered. To obtain satisfactory levels of adsorption, all r 

 mutant stocks are sedimented in a refrigerated International centrifuge with 

 Multispeed head, and resuspended in a solution containing 1 percent Bacto- 

 peptone and 0.5 percent NaCl. Very little loss of virus occurs during sedimen- 

 tation, and probably no permanent aggregation, since sedimented mixtures of 

 r and wild type virus do not yield mixed plaques. The resuspended virus is 

 stable for at least several weeks. 



To make a cross, a two hour culture of E. coli strain H in nutrient broth, 

 containing 2X10^ bacteria per ml, is infected at 37°C in an aerated culture 

 tube (the "adsorption tube") with yV volume of a mixture of diluted viral 

 stocks containing 2X10^ plaque forming particles of each kind per ml. After 

 five minutes, during which equal numbers (about 50 percent) of each virus 

 are adsorbed, a 10* dilution is made into broth (the "growth tube") for further 

 incubation, and a second diluted sample is spun for the assay of unadsorbed 

 virus. Sixty minutes after infection, an assay from the growth tube gives the 

 average yield of virus from about 40,000 mixedly infected bacteria. 



For the single burst experiments, an additional dilution from the adsorption 

 tube into antiserum to neutralize the unadsorbed virus (Delbruck 1945c), 

 is made at the end of the adsorption period. Five minutes later, a further 

 dilution from the antiserum tube is made into broth to contain about one 

 infected bacterium per three ml. Before the 20th minute after infection, sam- 

 ples measuring one ml are distributed into a series of small tubes. The virus 

 yields in these tubes are assayed by plating 0.3 ml 60 minutes or more after 

 infection. The remainder of each sample, excepting those containing no virus 

 or unmixed yields, is plated on two additional plates the next day. About 10 

 percent of each sample is mechanically lost. An important feature of these 

 experiments is the guard against contamination of materials provided by the 

 fact that about f of the samples contain no virus, whereas the remainder 

 contain more than 100 particles. 



Viral yields from the growth tube are plated on sensitive bacteria (strain 

 S), on the indicator strain (No. 2 B/2H, 2K), and on a mixture containing 

 one volume S and two volumes indicator (day-old broth cultures). On the 

 mixed indicator all four types of virus can be recognized (fig. 3), and their 

 sum equals the count on S. Mixed indicators plates always show a few doubt- 

 ful plaques which can be identified only by sampling and retesting, but their 

 number is too small to be of importance, and mixtures of pure stocks can be 

 counted with satisfactory accuracy. The counts on the single indicator, giving 

 only the h mutants, are also satisfactory with mixtures of pure stocks. These 

 counts tend to be low, however, for mixed yields of h and A+ virus. The cause 



154 



