342 General Discussion 



at first, but the Lac~ (from untrained cells) will not produce large 

 colonies for a long while unless there are some fully trained cells 

 giving rise to colonies on the same plate. 



Dean: I don't agree with that. If you keep the Lac~ cells on the 

 plate long enough they will eventually reach Lac"^ size, even if you 

 have only four or five colonies on the plate. 



Pollock: Yes, but it is the speed of growth that is affected by the 

 presence of the fully trained cells. 



Dean: Surely if you have fully trained it, and if you have Lac"^ 

 and Lac~ cells on the plates, one type is going to come up quicker 

 than the other. 



Pollock: The lag is surprisingly short, in my experience. 



Davis: I would like to make a suggestion that might apply to the 

 Akiba and Szybalski type of experiment, in which it was originally 

 alleged that mutations occurred in a small population without a 

 background of growth. Szybalski now believes that there is indeed 

 an invisible background of growth that produces enough population 

 to give rise to these hereditarily changed cells as spontaneous 

 mutants. In addition to the elegant method that Dr. Hayes 

 suggested for seeing whether there is indeed selection in this appar- 

 ently completely barren medium, a simple way of detecting whether 

 there was background growth in such a medium would be to have 

 some penicillin present. It is known that penicillin will sterilize 

 cells only under conditions that permit them to grow. If you had a 

 background of cells that did not grow during this long period, they 

 should survive the penicillin, but if they went through some genera- 

 tions while maintaining a constant population density, they should 

 die out. 



Dean: In these experiments on Bad. lactis aerogenes and d- 

 arabinose, penicillin has very little action. 



Davis: You have to increase the concentration to affect aerogenes 

 but the mode of action is still the same. We can isolate auxotrophs 

 of aerogenes by the selective action of penicillin, which indicates 

 that it sterilizes only those cells that are growing; but you need 

 three or four times as much penicillin as you need for Esch. coli. 



Hotchkiss: If there is growth occurring in the Akiba-Szybalski 

 experiment, it is a very special kind of growth. I rather object to 

 the induction by which we generalize and say that penicillin must 

 kill any type of growing cell. Szybalski did attempt genetic labelling 

 by putting some streptomycin-resistant cells into the digest, and 

 he found that these with their particular genetic markers were not 

 among those recovered later. His recent conclusion is that these 

 could not have survived. They were not related sufficiently to the 

 others and did not have a selective advantage in the mixture. 



