130 CHARLKS A. THOMAS, JR. 



ability to drtect nicks l)y |)r()(luciii<r a still liroaWcf l)aii(l (Thomas and 

 Pinkcrton, 1902 1. 



Two control cxpci'inicnts should l)i' niciitionciL If a tew interruptions 

 in the polynucleotide chains are jjroduced l)y the limited reaction with 

 DNase, then the shorter chains are easily observed by the disproi)or- 

 tionately broad band produced by the single polynucleotide chains, 

 whereas the undenatured duplex has about the same band wirlth as 

 before. In order to ride out the i)ossibiIity that formaldehyde is forming 

 cross-links between polynucleotide chains, N"-N^^ hybrid molecules 

 were produced by reannealing. On subse({uent dcnaturation in the 

 presence of CH^-O, the chains of different density were cleanly separable, 

 which meant that the formaldehyde was not causing them to be linked 

 together (Thomas and Berns, 19621. 



When the continuity of the molecule is maintained by a single chain, 

 one would expect that its tensile strength would be one-half that of the 

 duplex. Thus, interrupted molecules might lie nioi-c fragile to shear 

 than uniform dujilexes. The major conse(iuence of the decay of an 

 incorporated P'*- is the destruction of the phosphodiester linkage and 

 thus the interruption of the polynucleotide chain. Therefore, it is possible 

 to produce artificially interrupted molecules in this way. This expecta- 

 tion was confirmed l)y Levinthal and Davison (1961), who found that 

 T2 molecules which had suffered a substantial decay of P^- were broken 

 at lower rates of shear than was required to break undamaged molecules. 

 Since interruptions apparently do increase the shear fragility of the 

 molecule, this indicates that there were no interruptions in the orig- 

 inal molecule in those regions which can be broken by shear. Since 

 the maximum tension is developed in the middle, these experiments 

 suggest that ther(> ai'e no natui'al interi'uptions near the middle ])art of 

 the molecule. 



The sedimentation \-elocity analysis of the chromatographic fi-actions 

 of the first breakage product of whole molecules (Burgi and Hershey, 

 1961 1 shows that there is a single, ajiproximately Gaussian distribution 

 of molecular sizes produced by stirring. This distribution is not altered 

 by prolonged stirring at the critical s])eed, and would only be produced 

 if the most probable point of breakage were in the center of the molecule, 

 and if the probability of fracture decreased in an approximately normal 

 manner on each side of the middle. This would indicate that the proba- 

 bility of the breakage is governed by the hydrodynamic factors in the 

 region of the stirring blade, rather than weak points located at a special 

 location. The same kind of analysis was applied to the breakage of 

 half-molecules by stirring at a higher (critical I s])eed. This produced a 

 normal distribution of molecular lengths about otie-quarter the original 



