As it is known by this time that DNA is a double-helix and a macromolecule, the next question - actually posed in the Watson, Crick & Wilkins paper - is how does this mega-molecule replicate itself. Of the many different conceptual notions, two seem to be the most probable:

    In the first case the parent double helix merely acts as a template and a completely new copy is made. Thus the parent molecule remains intact to remain in one daughter cell, and the other progeny cell acquires a completely new molecule, which has been formed into a helix presumably while the two strands were being lengthened.

    The second case requires that the strands unwind so that complementary strands can be made from each and the semi-new helices, which are then partitioned between the two daughter cells.

    The lore of the science behind this goes that Meselson and Stahl were taking a coffee break and wondering if they could make E. coli make DNAs having different densities such as that lead is denser than aluminum. If so, then you have a chance at solving this problem, if they had a way of discriminating between molecules of different densities. They had heard of isopycnography, which employed centrifuge tubes of salt solutions filled in a special way - such that the deeper one went into a tube the more concentrated (thus denser) the solution became. Upon centrifuging, molecules would be inclined to seek and remain at levels that were identical to their own densities. (Greek: iso-pycnos means equal density.)

    If the first alternative were true, after one generation, half of the DNA should be heavy and half light. If the second alternative were true, then after one generation, all of the DNA should be at mid-density, and then from there on more and more light-density DNA should appear with a residual band of mid-density DNA.

    Believing isopycnography was the tool they should use, they had to know how dense their gradients needed to be. As they sat in their coffeeshop with cups in front of them, they guesstimated that DNA probably had a density similar to that of finger nails. They put some clippings into coffee. The clippings sank. Then they started adding sugar, but the clippings never floated no matter how much sugar was added. Then they started adding salt. But that didn't work either. So they went to visit with their stockroom person to find a substance that was so soluble that it would float the clippings. The stockroom person suggested CsCl (cesium chloride), and that worked even though it cost an expensive $1/gram at that time, and they needed a lot of grams. So they were the ones who began the CsCl craze.

    Thus they thought they had a method to discern between these alternatives. Again they thought of tagging the molecules using isotopes - but this time they considered non-radioactive ones: heavy and light nitrogen atoms.

    Next are shown these researchers' controls: the first is showing how to set up a gradient with DNA; then come three tubes with different DNAs in them. Thus showing that although heavy and light nitrogen atoms differ rather little it is still enough to affect the overall densities of the DNAs that contain them.

    They started their experiment. First they knew that they needed some detectible DNA standards in their tubes, so they added traces of radioactive heavy and light DNAs - for instance, one might be labelled with C-14 and the other with P-32. Then they started their real experiment with cells that were grown for many generations in heavy, but non-radioactive nitrogen so that all the nitrogen atoms in the cells - including those in their DNAs were homogeneously labelled. With that done, they shifted their bacteria over to a light-nitrogen medium and followed the growth. After every cell division cycle they took a sample; isolated DNA; and subjected it to isopycnography. They saw how the DNA started out heavy, and then ALL OF IT became half-heavy; and then part became light and the rest remained half-heavy. (Notice that the light bands have a trace of red in them indicating that identifier (tracer), and the heavy bands have a trace of blue.)

    Thus the second laternative was supported, and so we now say that DNA-replication is semi-conservative.