AbstractIn 1815, the British physician William Prout had advanced the theory that the molecular masses of elements were multiples of the mass of hydrogen. This “whole number rule” (and especially deviations from it) played an important role in the discussion whether elements could be mixtures of isotopes. F. Soddy's discovery (1910) that lead obtained by decay of uranium and of thorium differed in mass was considered a peculiarity of radioactive materials.The question of the existence of isotopes came up when the instruments developed by J.J. Thomson and by W. Wien to study cathode and canal rays by deflection in electric and magnetic fields were steadily improved. In 1913, Thomson mentioned a weak line at mass 22 accompanying the expected one at mass 20 when he analyzed the mass spectrum of neon. Subsequently Aston obtained the mass spectrum of chlorine with masses at 35 and 37. Still in 1921, Thomson objected heavily to the idea of isotopes.The isotope problem was finally settled, but more accurate mass measurements showed that even isotopic weights differed to some extent from the whole numbers. Based on earlier ideas of P. Langevin and J.‐L. Costa, F.W. Aston and A.J. Dempster developed the idea of packing fractions and mass defects due to the transformation of a portion of the matter comprising the atomic nucleus into energy.While the determination of the exact isotopic masses had improved over the years, the accurate determination of isotopic abundances remained a problem as long as photographic recording was used. Here especially A.O. Nier pioneered using dual collectors and compensation measurements. This was the prerequisite for the discovery that isotopic ratios varied somewhat in nature. M. Dole discovered the fractionation of oxygen isotopes by photosynthesis and respiration. Today 13C/12C‐ratios are employed to detect adulterations of food and in doping analysis, and 14C/13C‐ratios obtained by accelerator mass spectrometry are used for dating historical objects, just to give some examples. © 2005 Wiley Periodicals, Inc.