The first scientists to determine atomic weights were John Dalton between 1803 and 1805 and Jöns Jakob Berzelius between 1808 and 1826. Atomic weight was originally defined relative to that of the lightest element hydrogen taken as 1.00, and in the 1820s Prout's hypothesisstated that atomic masses of all elements would prove via a whole number rule to be exact multiples of this hydrogen weight. Berzelius, however, soon proved that this hypothesis did not always hold even approximately, and in some elements, such as chlorine, atomic weight falls almost exactly between two multiples of the hydrogen weight. Still later, as noted, this was shown to be an isotope effect, and that the atomic masses of pure isotopes, or nuclides, are multiples of the hydrogen mass, to within about 1%.
In the 1860s Stanislao Cannizzaro refined atomic weights by applying Avogadro's law (notably at the Karlsruhe Congress of 1860). He formulated a law to determine atomic weights of elements: the different quantities of the same element contained in different molecules are all whole multiples of the atomic weight and determined atomic weights and molecular weights by comparing the vapor density of a collection of gases with molecules containing one or more of the chemical element in question.
In the early twentieth century, up until the 1960s chemists and physicists used two different atomic mass scales. The chemists used a scale such that the natural mixture of oxygen isotopes had an atomic mass 16, while the physicists assigned the same number 16 to the atomic mass of the most common oxygen isotope (containing eight protons and eight neutrons). However, because oxygen-17 and oxygen-18 are also present in natural oxygen this led to 2 different tables of atomic mass. The unified scale based on carbon-12, 12C, met the physicists' need to base the scale on a pure isotope, while being numerically close to the chemists' scale.