208 ERNST FREESE 



a coiiuiion ()i'i<!;in. A new species was thought to conic about by the 

 gradual change of an old species. Laniaick (1809) atti'ibutcd this change 

 to both the crossing of existing fofnis and the inheritance of acciuircd 

 ciiaracters, to which tiie organism had achipted during its Ufetinie. Dar- 

 win (1859), however, explained evolution by natural selection and 

 reasoned that new forms arise by chance irresjiective of whether they 

 are useful, useless, or deleterious to the organism. But convincing evi- 

 dence, showing that ac(iuired (adapted) characters are not indefinitely 

 inherited, accumulated only slowly; for example, the spontaneous nature 

 of mutations in bacteria has been proved only recently (Luria and Dcl- 

 briick, 1943). Phenotypic variations of similar organisms could then be 

 divided into those which are heritable and those which are non-heritable 

 and for some time this proved the only experimentally verifiable distinc- 

 tion. During this period, the w^ord "mutation" was synonymous with all 

 "heritable variations." 



A further classification of heritable variations became feasible only 

 when Mendel's laws were rediscovered at the beginning of the twentieth 

 centuiy and when many crosses w^ere examined genetically and cyto- 

 logically (see scheme in Fig. 1). One learned to distinguish nuclear from 

 extranuclear (cytoplasmic) inheritance and their variations. The herit- 

 able nuclear variations were found to involve the alteration of the 

 chromosomes, either their number in a nucleus (polyploids, aneuploids, 

 etc.) or their information content. Alterations of the information content 

 could be further divided into those that occur through a recombination 

 of information already present in two pieces of genetic material and 

 those that occur by a mutation of the information within one chromo- 

 some and do not require the interaction of genetic material. This present 

 meaning of the word "mutation" is much more restricted than the 

 original one. 



The bipartition into nuclear and extranuclear variations, shown in 

 Fig. 1, is reasonable for organisms having a defined nucleus, and it is 

 useful, because the nucleus and the cytoplasm can often be separated by 

 natural or artificial means. But as is common to all classifications one 

 encounters difficulties. For certain microorganisms, e.g., bacteria, a 

 nuclear membrane may not exist and thus a division into nuclear and 

 extranuclear variations may be without meaning. Nevertheless, both 

 bacteria and viruses contain most of their hereditaiy information in one- 

 dimensional genetic structures (genomes) which must correspond to 

 physical entities. The existence of such a physical entity containing 



cytcsines; dCMP, dcoxycytidinc monophosphate; AP, 2-aniinopiirinp; BU, 5- 

 bromouracil; BUdR, 5-bromodcoxyuridine; EES, ethyl ethanesulfonatp; HA, hy- 

 droxylamine; NA, nitrous acid. 



