HISTORY OF THE RECOMMENDED ATOMIC-WEIGHT VALUES FROM 1882 TO 1997 - ACOMPARISION OF DIFFERENCES FROM CURRENT VALUES TO THE ESTIMATED UNCERTAINTIES OF EARLIER VALUES

International commissions and national committees for atomic weights ( mean relative atomic masses) have recommended regularly updated, best values for these atomic weights as applicable to terrestrial sources o f the chemical elements. Presented here is a historically complete lis ting starting with the values in F. W. Clarke's 1882 recalculation, fo llowed by the recommended values in the annual reports of the American Chemical Society's Atomic Weights Commission. From 1903, an Internati onal Commission published such reports and its values (scaled to an at omic weight of 16 for oxygen) are here used in preference to those of national committees of Britain, Germany, Spain, Switzerland, and the U .S.A. We have, however, made scaling adjustments from A(r)(O-16) to A( r)(C-12) where, not negligible. From 1920, this International Commissi on constituted itself under the International Union of Pure and Applie d Chemistry (IUPAC). Since then, IUPAC has published reports (mostly b iennially) listing the recommended atomic weights, which are reproduce d here. Since 1979, these values have been called the ''standard atomi c weights'' and, since 1969, all values have been published, with thei r estimated uncertainties. Few of the earlier values were published wi th uncertainties. Nevertheless, we assessed such uncertainties on the basis of our understanding of the likely contemporary judgement of the values' reliability. While neglecting remaining uncertainties of 1997 values, we derive ''differences'' and a retrospective index of reliab ility of atomic weight values in relation to assessments of uncertaint ies at the time of their publication. A striking improvement in reliab ility appears to have been achieved since the commissions have imposed upon themselves the rule of recording estimated uncertainties from al l recognized sources of error.