This discovery was eventually explained by the fact that the DNA molecule consists of two, paired chains of linked nucleotides in which adenine (on one chain) is always paired with thymine (on the other chain) and guanine is always paired up with cytosine.

However, as the analysis of DNA from different sources continued it was quickly seen that DNA molecules could differ in the relative amounts of A+T and G+C; they were not always the same, in fact, they could found across a whole range of values.

Different organisms had DNA with characteristically different percentages of G+C and A+T.

1. Cells of the specimen under investigation are broken up mechanically in a liquid solution that prevents destructive enzymes from breaking down the valuable DNA molecules.

2. RNA molecules are removed from the mixture of cell contents by destroying them with special digestive enzymes. These enzymes catalyze the breakup of long RNA molecules by breaking them in to smaller and smaller pieces until they no longer interfere with the extraction of the DNA molecules.

3. All the DNA molecules in the broken cell mixture are tightly bonded to a special mixture of kieselguhr (or, diatomatious earth, a white powder like corn flour, made of the skeletons of diatoms) and poly-L-lysine (an artificial polymer of the amino acid lysine).

4. The DNA molecules in the broken cell mixture bind exclusively to the poly-lysine, which is in turn bound to the white kieselguhr powder, making an insoluble compound substance.

5. All the other molecules in the cell (proteins, carbohydrates, amino acids, sugars, etc.,) remain in the liquid solution.

6. The total mixture (of everything) is poured into a narrow, hollow glass tube which has a very small pore mesh filter at the bottom to trap the larger particles of kieselguhr. In this way the poly-lysine kieselguhr (and everything bound to it) is trapped inside the tube.

7. The tube is placed upright in a clamp and the contents of the tube allowed to settle. This forms a column of solid material at the bottom of the tube (trapped there by the filter) with the liquid on top.

8. The unwanted solution containing all the non-DNA molecules is washed away by passing a very dilute solution (0.4 molar) of common salt (sodium chloride) through the column of poly-lysine kieselguhr. This removes all the other cell-contents and leaves the DNA bound to the poly-lysine, which is in turn bound to the solid kieselguhr.

9. The DNA molecules are released from their complex with the poly-lysine kieselguhr (PLK) by passing stronger and stronger solutions of salt through the column of PLK and DNA. This breaks the connection between the DNA and the artificial lysine polymer.

10. Usually the DNA-PLK association is broken by salt in the concentration range 1.7 molar NaCl (sodium chloride) to 2.1 molar NaCl.

11. The concentration of sodium chloride in the solution of liquid being passed through the column of PLK is steadily increased between these two values. This is called a NaCl gradient, and changes gradually and in a regular fashion as the solution steadily gets more and more salty.

12. The released DNA is washed out of the bottom of the column and collected in a series of test tubes, called fractions.

13. The amount of DNA in each fraction can be measured and quantitated.

14. The concentration of salt it takes to release the DNA from the PLK depends on the amount of Guanine + Cytosine in the DNA molecules. The base composition (%G+C) of the DNA can therefore be determined by the fraction number at which the DNA leaves the column of PLK.

15. The relationship between the %G+C base composition of the DNA and the strength of the salt solution that releases it from the PLK can be determined by using standardized DNA molecules. Molecules of DNA with known base compositions can be bound and then eluted (washed off the column) from the PLK and the fraction number determined.

16. Comparing the fraction number at which known DNA elutes from the PLK column with that of the unknown DNA can be used to find the base composition of the unknown DNA.