which developed after further incubation. (For details 

 of the spray technique, see refs. 3, 4, 5.) 



On the adaptation hypothesis, the bacteria present 

 at the end of the initial growth-period would all be 

 phage-susceptible, and spreading would serve only 

 to redistribute the members of a homogeneous popu- 

 lation. No striking differences in colony count be- 

 tween spread and unspread plates would therefore be 

 expected. 



On the alternative (spontaneous mutation) hypo- 

 thesis, both susceptible and resistant cells would be 

 present at the end of the initial growth-period 

 wherever a sufficient end-population had been reached. 

 Further, mutations taking place a generation or more 

 before the cessation of growth would each be repre- 

 sented by a minute cluster of resistant cells all de- 

 scended from the one original mutant. Where the 

 arrangement of the bacteria is left undisturbed, a 

 cluster would give rise to a single resistant colony 

 after the application of phage ; but where the bacteria 

 are redistributed over the surface of the agar by 

 spreading, a colony would develop from each resistant 

 cell. Higher counts would thus be expected from 

 the spread than from the unspread plates. This is, 

 in fact, what was found (see Table 2), the difference 

 being as much as fifty-fold whore the end population 

 was highest. 



Table 2. Resistant colonies developing after spraying agar plate 

 cultures of B. coli with phage 21, showing the effect of redistribution 

 by spreading prior to spraying 



• Estimated by washing and assay ; averages of three independent 

 determinations, unap. — unspread; ap. — spread. 



It might be suggested that the bacteria are less 

 likely to become 'adapted' when crowded together 

 in the developing colonies. If this were true, tho 

 proportion of bacteria becoming 'adapted' should bo 

 least where microcolony size is greatest. Table 2 

 shows that the average number of bacteria per micro - 

 colony at the time of spraying rose with increasing 

 period of incubation from 33 up to 54,900, but that 

 there was no corresponding decline in the proportion 



