216 



HOROWITZ AND LEUPOLD 



6000— 



Abscissa : Colonies/plate 

 Ordinate ^ • = Secondary colonies/lo^ survivors 

 o = Temperoture mutcnts/lo* survivors 



4ooo- 



2ooo- 



Fig. 1. Frequency of secondary colonies (solid circles) and of temperature mutants 

 (open circles) per 10*' survivors as functions of population density. 



chemical mutants which have been fed 

 "syntrophicallv" by the wild types. 

 Figure 2 shows that the composition 

 of the population of secondary colo- 

 nies is also influenced by the popula- 

 tion density. It is seen that the yield 

 of temperature mutants per hundred 

 secondary colonies decreases with in- 

 creasing densities, indicating that the 

 yield of temperature mutants decreases 

 even more rapidly than does the yield 

 of secondary colonies in general. 



At least three selective forces are 

 at work in these populations: ( 1 ) com- 

 petition for food, (2) probably more 

 important, mutual inhibition by-prod- 

 ucts of metabolism, and (3) superim- 

 posed on these but acting in the 

 opposite direction, cross-feeding, or 

 syntrophism, the mutual exchange of 



essential growth factors. The first two 

 mechanisms are probably mainly re- 

 sponsible for the rapid decrease in the 

 yield of secondary colonies with in- 

 creasing plate densities. Their intensity 

 is evidently quite remarkable. They 

 are unspecific forces, however, and 

 cannot be expected to influence sys- 

 tematically the relative frequencies of 

 the two types of temperature mutants 

 which we set out to find. The third 

 influence, cross-feeding, is much more 

 dangerous in this respect, although its 

 direction and intensity are difficult to 

 predict. The effect of cross-feeding on 

 the selective isolation of the two kinds 

 of temperature mutants depends on 

 both the quantity and quality of the 

 output of growth factors bv^ wild type 

 E. coli. For example, let us assume that 



