J. D. WATSON AND F. H. C. CRICK 



close to the amount of cytosine + 5-methyl cytosine, although the ratio of ade- 

 nine to guanine can vary from one source to another (Chargaff, 1951; Wyatt, 

 1952). In fact as the techniques for estimation of the bases improve, the ratios of 

 adenine to thymine, and guanine to cytosine + 5-methyl cytosine appear to grow 

 very close to unity. This is a most striking result, especially as the sequence of 

 bases on a given chain is likely to be irregular, and suggests a structure involving 

 paired bases. In fact, we believe the analytical data offer the most important 

 evidence so far available in support of our model, since they specifically support 

 the biologically interesting feature, the presence of complementary chains. 

 We thus believe that the present experimental evidence justifies the working 

 hypothesis that the essential features of our model are correct and allows us to 

 consider its genetic possibilities. 



V. Genetical Implications of the Complementary Model 



As a preliminary we should state that the DNA fibers from which the X-ray 

 diffraction patterns were obtained are not artifacts arising in the method of pre- 

 paration. In the first place, Wilkins and his co-workers (see Wilkins et al., 1953) 

 have shown that X-ray patterns similar to those from the isolated fibers can be 

 obtained from certain intact biological materials such as sperm head and bacterio- 

 phage particles. Secondly, our postulated model is so extremely specific that we 

 find it impossible to believe that it could be formed during the isolation from 

 living cells. 



A genetic material must in some way fulfil two functions. It must duplicate 

 itself, and it must exert a highly specific influence on the cell. Our model for 

 DNA suggests a simple mechanism for the first process, but at the moment we 

 cannot see how it carries out the second one. We believe, however, that its speci- 

 ficity is expressed by the precise sequence of the pairs of bases. The backbone of 

 our model is highly regular, and the sequence is the only feature which can carry 

 the genetical information. It should not be thought that because in our structure 

 the bases are on the "inside," they would be unable to come into contact with 

 other molecules. Owing to the open nature of our structure they are in fact fairly 

 accessible. 



A Mechanism for DNA Replication 



The complementary nature of our structure suggests how it duplicates itself. 

 It is difficult to imagine how like attracts like, and it has been suggested (see 

 Pauling and Delbruck, 1940; Friedrich-Freksa, 1940; and Muller, 1947) that self 

 duplication may involve the union of each part with an opposite or comple- 

 mentary part. In these discussions it has generally been suggested that protein 

 and nucleic acid are complementary to each other and that self replication involves 

 the alternate syntheses of these two components. We should like to propose 

 instead that the specificity of DNA self replication is accomplished without 

 recourse to specific protein synthesis and that each of our complementary DNA 

 chains serves as a template or mould for the formation onto itself of a new com- 

 panion chain. 



202 



