When you balance chemical equations in middle or high school, you rely on atomic weights to choose the right coefficients. At the end of the 19th century, published values often had errors of several percent, causing large deviations in calculations for complex compounds. Richards combined gravimetric methods—reducing silver chloride to metallic silver—with gas-density measurements and corrected carefully for temperature, humidity, and buoyancy. He reduced the uncertainty of many atomic weights to less than one part in ten thousand and revised the data for more than fifty elements. The numbers he produced remain almost unchanged in modern textbooks and underpin quantitative analysis and chemical industry standards. Moreover, he noticed that lead obtained from the radium decay series had a slightly different atomic weight, paving the way for the idea of “isotopes.” These achievements led to his selection for the 1914 Nobel Prize in Chemistry.

At the turn of the 20th century, determining atomic weights with high precision was one of the foremost experimental challenges in physical chemistry, providing the quantitative bedrock for theoretical chemistry and geochemistry. Richards founded a micro-analytical laboratory at Harvard, designing and refining his own Accession balances, adiabatic calorimeters, and vacuum drying systems. In the gravimetric reduction of silver chloride to metallic silver, he pushed the classical method of stoichiometric mass comparison to its limits, adding a 1 100 °C ignition step to expel residual moisture and adsorbed gases. Using the gas-density technique, he compared molar volumes pycnometrically and applied ideal-gas corrections, suppressing errors from vapor-pressure differentials to the 1 ppm level. By statistically combining these experiments, he demonstrated that while a chemical element is homogeneous in reactivity, slight mass differences might occur, thus providing empirical footing for Soddy’s later isotope concept. Richards’ values were also used to recalculate the Faraday constant and the Avogadro number, prompting revisions of the International Table of Atomic Weights. With reliable absolute atomic weights in place, the accuracy of thermodynamic data, reaction kinetics, and quantum-chemical calculations improved dramatically, cementing the quantitative turn of 20th-century chemistry.

In his gas-density work, Richards employed a closed-tube pycnometer, reading minute pressure differentials near 100 kPa with a mercury manometer to an accuracy of 1/20 000 atm, lowering the molecular-weight conversion error to 0.00005. In the gravimetric AgCl→Ag conversion he corrected for trace Ag2O formed as a side product by volumetric titration and eliminated halogen adsorption gains. He applied the Grubbs test to each measurement series, rejected outliers, and reported weighted means with standard deviations on the 10^-5 scale. Harvard Report 1911 No.21 gave an atomic weight for Pb of 206.085 ± 0.003 and noted a significant difference from Pb derived from the uranium series (~206.180), one of the earliest mass-based isotopic evidences predating Soddy’s 1913 isotope proposal. Richards also derived an Avogadro number of 6.062 × 10^23 mol^-1 from the decomposition heat of silver oxide, keeping the discrepancy with later X-ray crystal-density values below 3 ‰. His apparatus-calibration protocols and humidity-correction model are still cited in CODATA systematic-uncertainty evaluations. Such meticulous metrology, hailed by the Nobel Committee as “a new standard of precision,” underpinned his 1914 award.