134 LINUS PAULING 



to within less than o-8 A, a small fraction of an atomic diameter. It has also been 

 shown that a negatively charged group is present in the antibody at very nearly 

 the minimum distance of approach to a positively charged group in the haptenic 

 group (the trimethylphenylammonium cationic group) of an azoprotein used in 

 producing the antiserum. Many observed ratios of combining powers of different 

 haptens could be explained only in terms of the formation of specific hydrogen 

 bonds between antigen and antibody. 



The theory that the process of duplication of the gene is a two-step process 

 involving at each stage the formation of a structure complementary to that of 

 the molecule acting as a template has been given strong support through the 

 formulation of the Watson-Crick structure of deoxyribonucleic acid [8]. 

 Evidence that genes are molecules of nucleic acid rather than molecules of 

 protein has been gathered from several sources in recent years. The eflforts to 

 determine tlie molecular structure of fibres of deoxyribonucleic acid culminated 

 in the formulation by Watson & Crick of their structure, which involves two 

 complementary molecules twisted about one another to form a double helix. 

 At any level in one of the molecules there may be a nucleotide with any one of 

 the four nitrogen bases adenine, thymine, guanine, or cytosine. The sequence 

 of nitrogen bases in one of the polynucleotide chains may be, so far as we know 

 now, completely arbitrary. However, according to the Watson-Crick structure 

 the nature of the second polynucleotide chain is completely determined by that 

 of the first; the second chain must be complementary to the first, with thymine 

 wherever the first chain has adenine, adenine wherever the first chain has thy- 

 mine, c3rtosine wherever the first chain has guanine, and guanine wherever the 

 first chain has C3rtosine. This complementariness in structure is the result of the 

 ability of adenine and thymine to form two hydrogen bonds with one another, 

 and of guanine and cytosine to form three hydrogen bonds with one another. 

 (Watson & Crick proposed that guanine and cytosine form only two hydrogen 

 bonds with one another, of a nature, however, somewhat different from those 

 in the adenine-thymine pair; it was then pointed out by Pauhng & Corey [9] 

 that there are three hydrogen bonds in the guanine-cytosine complex.) With 

 Watson & Crick [10], we may accept as a reasonable molecular mechanism for 

 the duplication of the double polynucleotide chain of deoxyribonucleic acid a 

 process that involves the separation of the two chains, each of which then serves 

 as the template for the synthesis of the other. In the nomenclature used above, 

 we may describe the Watson-Crick double helix as AA~i ; its process of dupUca- 

 tion would involve the separation into A and A~i, each of which would then 

 serve for the synthesis of a complementary structure, giving AA~i and A^^A, 

 two double molecules identical with the original one, and each containing one- 

 half of the original double molecule. 



We may ask whether it might not be possible that the two chains A and A~i 

 of the Watson-Crick double helix are really a single chain, A and A~i being 

 attached to one another at one end. An argument has been presented to show that 

 this suggestion is to be rejected [11]. If A and A~^ were separate chains, the 

 process of duplication of AA"i would be as described above. If they were the 

 same chain, the process of duplication would consist in untwisting the double 



