JOSHUA LEDERBERG 



must be taken to minimize residual growth of auxotrophic inocula at the 

 expense of growth factors carried over with the cells or resident as impurities in 

 the test medium (114, 116). 



Serological variation is of key importance in medical bacteriology, but its 

 genetic study is barely under way. The work of Kauffmann and of Edwards 

 and Bruner on Salmonella illustrates the provocative information now at hand 

 (8, 9, 10, 71 ). A compendium of the serological variation of bacteria generally 

 would be inappropriate here, but the subject has been treated exhaustively in 

 excellent reviews and books (66, 36, 54, 32). 



Given two alternative forms of a gene — say a and A, each of which mutates 

 into the other at a definite rate — it is easy to show that an equilibrium will 

 eventually be established such that the ratio of a : A will be equal to the ratio 

 of the mutation rates to the respective conditions (42). Most mutation rates 

 are so low (of the order of one per million or billion cell divisions) that muta- 

 tional equilibria would take too long for human observation, even if the neces- 

 sary constancy of the environment were possible. Mutations involving the 

 antigenic structure of the flagella of Salmonella have, however, been found 

 to have unusually high rates (accounting for the readiness with which they 

 have been found), and Stocker (124) has described mutational equilibrium 

 as approached from inocula of either form. 



Induced mutations. — Since Muller's announcement in 1928 that X-rays 

 would induce mutations in fruitflies, an extensive segment of genetic research 

 has concerned the discovery of mutagenic agents and the conditions of their 

 effect. Higher organisms like Drosophila and maize are indispensable in the 

 finer analysis of the cytological basis of induced genetic alterations, but micro- 

 organisms are very useful tools in the screening of new agents for mutagenic 

 activity, and in the study of the gross quantitative aspects of such activity. 

 The same types of mutants already mentioned as best enumerated on a selective 

 basis are particularly useful here. Dose-response data have been published 

 for mutations induced by X-rays (113, 45) and by UV (49, 104) but their 

 interpretation, especially for UV, is far from simple. 



It has been thought that radiations induced mutations by direct photo- 

 chemical processes, i.e., that the gene itself might be activated by the absorption 

 of a quantum of UV, or by collision with a secondary electron following an 

 X-ray quantum absorption, (60, 80). Some revision of this concept is now 

 necessary on the basis of recent research. The effectiveness of X-rays is 

 potentiated by the presence of oxygen, and there may be a tenfold difference 

 between the doses required for a given effect in oxygen as against an inert atmos- 

 phere (47). This argues for a radiochemical intermediate, possibly some free 

 radical (peroxide?) which depends upon oxygen for its production under the 

 influence of X-radiation. 



Owing to the powerful penetrability of X-radiation, it has been used in 

 Drosophila studies more extensively than UV, which penetrates through living 

 material so poorly that there are serious experimental difficulties in its ap- 



