144 CHARLES A. THOMAS, JR. 



intervals, one of which was a pair of old polynucleotide chains, and the 

 other made of two new chains. Such a structure has been proposed 

 (Cavalicri and Rosenberg, 1961a,b). This structure, consisting of a 

 pair of duplexes, would be thought to arise from anotlicr unsi)ecified 

 kind of replication scheme. The essential feature of this proposal is that 

 it requires the biologically produced density hybrid to be of four i)oly- 

 nucleotide chains rather than two. These four chains are pictured as 

 being arranged in two conventional duplexes which are united l)y fre- 

 quent inter-duplex links. After one round of replication this structure 

 is pictured as one old and one new duplex, whereas the conventional 

 interpretation argues that a single duj)lex now consists of one old and 

 one new polynucleotide chain. To accept this hypothesis would demand 

 many adjustments in the interi)retation of a wide range of experiments, 

 and a full discussion would be ([uite lengthy, and largely hypothetical. 



If the biologically produced hybrid molecules are of this doublc- 

 du]ilex structure, then it seems clear that the bonds which unite the 

 dui)lex strands must have almost the same thermal stability as the 

 hydrogen bonds that unite the polynucleotide chains, for the experiments 

 of Schildkraut, Marmur, and Doty (1961) show that the biologically 

 produced hybrid is converted uniformly with time to heavy and light 

 subunits as the native structure of the molecule (as evidenced by 

 density) is destroyed. Since a 20-minute heating at 90.2°C fails to result 

 in the separation of subunits, while an equivalent heating at 93.8°C does, 

 we must conclude that the hypothetical inter-duplex linkages must be 

 as resistant to heat treatment as the hydrogen bonds, but at the same 

 time no more resistant. 



Another special condition must be placed upon these hypothetical 

 linkages between duplex molecules, namely, that of species specificity. 

 DNA samples from different species cannot be renatured together to 

 form molecules of hybrid density, whereas those from the same or a 

 closely related species can (Schildkraut et al., 1961). If the double 

 duplex structure did exist before denaturation, it must have recovered 

 during the renaturation process because transforming DNA has a similar 

 absorbance-temperature profile, and the same rate of thermal inactiva- 

 tion of biological markers as the original DNA, facts which strongly 

 indicate that the basic structure of the molecules is the same before 

 heat denaturation and after renaturation (Marmur and Doty, 1961). 



Thus, it is still possible to think of the biologically produced hybrid 

 in terms of a doubled-duplex molecule, provided that one is willing to 

 allow the inter-duplex linkages to have veiy special properties which are 

 almost the same as the hydrogen bonds which unite the polynucleotide 

 chains. On the other hand, all of the above experiments are easily under- 



