RECOMBINATION IN BACTERIOPHAGE 



49 



the refrigerator. Agar for layer plating has the same nutrient composition, 

 but contains only 0.5 to 0.6 percent agar, the optimal concentration depending 

 on the age of the plates and on other variables. 



All platings of virus are made by the agar layer method, by adding an aliquot 

 of virus measuring 0.1 to 0.5 ml and about 0.1 ml of a day old unaerated broth 

 culture of bacteria, to 2 ml of melted soft agar at 45° C, and pouring the mix- 

 ture over the surface of an agar plate at room temperature. The plates are in- 

 cubated 18 to 24 hours at 37°C without inverting, during which time fluid 

 collects on the agar surface without undesirable effects if conditions are 

 optimal. 



Bacterial counts are made by spreading 0.1 ml aliquots on the surface of 

 dried agar plates. 



THE LINKAGE SYSTEM 



The mutants m and h {minute plaque and host-range modification, respec- 

 tively) have been crossed with wild type, with the mutants called rl, r7, and 

 rl3, and with each other. The crosses with wild type yield only the parental 

 types of virus, confirming that the mutants are unit modifications. The crosses 

 with r mutants, and the intercross, extend the linkage system previously de- 

 scribed (Hershey and Rotman 1948) as shown in fig. 4. The h locus is closely 

 linked to the locus rl3\ the m locus belongs in a third linkage group C. 



It should be understood that the diagram of fig. 4 is only a convenient 

 representation of linkage relations whose structural basis remains to be eluci- 

 dated (Hershey and Rotman 1948). The question of linear structure will be 

 returned to in the discussion of this paper. At this point we insert an experi- 

 ment which supports the general interpretation in terms of linkage. 



A 



ri m 

 — H — 



B 



1% 



h. P13 



-H— 



± 



M5% 



P 2, 4, 8, 9, 3, 6, 7, 5 

 I I I I I 1 1 1 



L 



T 

 7% 



J 



>15% 



Y\5% 



m 



Figure 4. 



-Linkage relations among mutants of T2H. The percentages indicate 

 yields of wild type in two factor crosses. 



The double mutants h r7 and h rl3 were isolated by making the respective 

 crosses hXr7 and hXrl3. The three crosses r7Xrl3, h r7Xrl3, and h rl3Xr7 

 were then compared with respect to the yield of r + virus, classified without 



243 



