340 General Discussion 



in which a spontaneous mutation of the usual type might be 

 expected. It was found that, at the end of a period of several weeks, 

 the entire population showed a high degree of resistance for strep- 

 tomycin. This is a very important experiment and appeared 

 convincing. It was initially confirmed by Szybalski ; but he has now 

 recanted. Szybalski has recently shown some of the reasons why 

 the experiment was fallacious (Szybalski, W. (1956), Microbial 

 Genetics Bull, 14, 6). One reason is that in the depths of the 

 medium, under relatively anaerobic conditions, sensitive cells are 

 able to grow in the presence of streptomycin. Another reason is 

 that the experiment cannot be reproduced when purified strepto- 

 mycin is used; it seems that impurities present in streptomycin 

 preparations can be utihzed for the growth of either sensitive or 

 resistant bacteria. In the case of sensitive cells, some are killed off 

 but others grow in the depths of the medium, where they are 

 unaffected by the streptomycin. There is thus a turnover of cells, 

 and among these a streptomycin-resistant mutant can arise in the 

 normal way. 



I would suggest that in this kind of experiment there is one in- 

 ternal control which would help clearly to distinguish between an 

 adaptation phenomenon and one due to a selection of resistant 

 mutants. This is to employ a bacterial population which is not 

 homogeneous but made up of several genetically marked types of 

 the same wild-type organism, mixed in different proportions. If 

 mutant selection operates, then the resistant population issuing 

 from the experiment is likely to be homogeneous for a single type. 

 If, on the other hand, the evolution of resistance is due to adapta- 

 tion of all, or of the great majority, of the cells, in the population, 

 then all the types should be represented in the resistant population 

 in their initial proportions. 



Dean: I am very interested in Dr. Hayes's comments on strepto- 

 mycin, because from time to time we have had indications that 

 " reduced " streptomycin is sometimes not adsorbed by the cells and 

 is not very toxic. The experiments which we described at this 

 symposium were mainly done in sugar media. They were of two 

 types. The first type was in a minimal medium in which the cells 

 eventually grow. The second was of the Szybalski type, where one 

 had omitted a nitrogen source or else had used phosphate buffer 

 only. In neither case did we see any evidence of lysis, even after 

 41 days, when investigated by total bacterial mass (turbidity). In 

 these experiments we dilute the samples and plate out less than 100 

 cells. They all form colonies, and if selection was taking place we 

 should have seen a differentiation of the colonies on the plates. If 

 we were simply selecting mutants during the lag phase we should 



