Dalton noted that the chromium oxide molecules that he was studying had masses of oxygen that were not the same: he found that there were constant ratios to these mass differences, however, so there had to be some sort of "units" (atoms) making up the ratios of these particles.