Dalton’s idea was right, but his details were wrong in some cases. For instance, on closer examination it turned out that the water molecule was composed of 1 oxygen atom and 2 hydrogen atoms. For this reason, the water molecule may be written H₂O, where H is the chemical symbol for a hydrogen atom, and O for an oxygen atom.
It is still a fact that a quantity of hydrogen combines with eight times its mass of oxygen, so the single oxygen atom must be eight times as massive as the 2 hydrogen atoms taken together. The oxygen atom must therefore be sixteen times as massive as a single hydrogen atom. If the atomic weight of hydrogen is 1, then the atomic weight of oxygen is 16.
At first it seemed that the atomic weights of the various elements were whole numbers and that hydrogen was the lightest one. It made particular sense, then, to consider the atomic weight of hydrogen as 1, because that made all the other atomic weights as small as possible and therefore easy to handle.
The Swedish chemist Jöns Jakob Berzelius (1779-1848) continued Dalton’s work and found that elements did not combine in quite such simple ratios. A given quantity of hydrogen actually combined with a little bit less than eight times its mass of oxygen. Therefore if the atomic weight of hydrogen were considered to be 1, the atomic weight of oxygen would have to be not 16, but 15.87.
Jöns Jakob Berzelius
As it happens, oxygen combines with more elements (and more easily) than hydrogen does. The ratio of its atomic weight to that of other elements is also more often a whole number. In working out the atomic weight of elements it was therefore more convenient to set the atomic weight of oxygen at a whole number than that of hydrogen. Berzelius did this, for instance, in the table of atomic weights he published in 1828. At first he called the atomic weight of oxygen 100. Then he decided to make the atomic weights as small as possible, without allowing any atomic weight to be less than 1. For that reason, he set the atomic weight of oxygen at exactly 16 and in that case, the atomic weight of hydrogen had to be placed just a trifle higher than 1. The atomic weight of hydrogen became 1.008. This system was retained for nearly a century and a half.
Throughout the 19th century, chemists kept on working out atomic weights more and more carefully. By the start of the 20th century, most elements had their atomic weights worked out to two decimal places, sometimes three.
A number of elements had atomic weights that were nearly whole numbers on the “oxygen = 16” standard. The atomic weight of aluminum was just about 27, that of calcium almost 40, that of carbon almost 12, that of gold almost 197, and so on.
On the other hand, some elements had atomic weights that were far removed from whole numbers. The atomic weight of chlorine was close to 35.5, that of copper to 63.5, that of iron to 55.8, that of silver to 107.9, and so on.