88 NUCLEIC ACIDS AND NUCLEOPROTEINS 



must remember that the minimum estimate^ of the number of 

 possible permutational isomers in which a substance of the com- 

 position of calf thymus DNA could exist is phantastically high, 

 namely 10^^^^. Actually, this number will be even much larger, 

 as a molecular weight of 750,000 for DNA, used in these com- 

 putations, is too low. How then expect two different organisms 

 that cannot make identical proteins to be able to dupUcate this 

 enormous throw of dice? 



If we take it for granted that there exist a very large number 

 of different nucleic acids — a number presumably much larger 

 than can be shown merely by analysis — we may go on to ask: 

 Are there any principles that unify what appears so disparate? 

 No simple or recognizable repeating unit can be made out in 

 these giant polynucleotide chains; they do not seem to possess a 

 perceptible periodicity. But we became aware very early in our 

 work^ of several remarkable regularities apparently characteristic 

 of all deoxypentose nucleic acids. They are as follows: (a) The 

 sum of the purine nucleotides equals that of the pyrimidine 

 nucleotides, (b) The molar proportion of adenine equals that of 

 thymine, (c) The molar proportion of guanine equals that of 

 cytosine (+ methylcytosine). (d) The number of 6-amino groups 

 (adenine + cytosine + methylcytosine) equals that of 6-keto 

 groups (guanine + thymine). These regularities may be consid- 

 ered as well established. (See Table XVII in Ref. 13.) Polynu- 

 cleotide fragments produced by the enzymic degradation of DNA 

 do not show these regularities^^, a finding that will have to be 

 taken into account should well substantiated deviations from 

 these pairing principles be discovered. 



b. Fractionation 



Having plunged into a whirlpool, though decorated with har- 

 monious surface ripples, we proceeded to make it even more 

 chaotic. We asked a question. Granted, we said, that the DNA 

 of a given species represents a completely arrhythmic, highly 

 polymeric nucleotide chain, should we consider it as strictly 

 homogeneous or as consisting of a bundle of different chains, dif- 



