Atomic Theory and Multiple Combining Proportions: The Search for Whole Number Ratios

John Dalton's atomic theory, with its postulate of compound formation through atom-to-atom combination, brought a new perspective to weight relationships in chemical reactions. A presumed one-to-one combination of atoms A and B to form a simple compound AB allowed Dalton to construct his first table of relative atomic weights from literature analyses of appropriate binary compounds. For such simple binary compounds, the atomic theory had little advantages over affinity theory as an explanation of fixed proportions by weight. For ternary compounds of the form AB2, however, atomic theory made quantitative predictions that were not deducible from affinity theory. Atomic theory required that the weight of B in the compound AB2 be exactly twice that in the compound AB. Dalton, Thomas Thomson and William Hyde Wollaston all published within a few years of each other experimental data that claimed to give the predicted results with the required accuracy. There are nonetheless several experimental barriers to obtaining the desired integral multiple proportions. In this paper I will discuss replication experiments which demonstrate that only Wollaston's results are experimentally reliable. It is likely that such replicability explains why Wollaston's experiments were so influential.