Abstract

This paper revisits the debate on the reality of atoms. At the turn of the 20th century, many physicists treated the atomic hypothesis with scepticism, claiming that atoms were fictional entities. While many, such as Ostwald and Poincare, changed their minds after the publication of Thompson’s and Perrin’s experiments, some, such as Mach and Duhem, continued to oppose the reality of atoms despite the experimental support. I argue that at the heart of this debate are methodological arguments that influenced physicists’ stances both before and after experimental evidence in favour of the reality of atoms. Ostwald and Poincare were able to accept the reality of atoms since the atomic hypothesis became scientific on their terms in light of being experimentally testable, with the multiple ways of calculating the number of atoms in a volume being particularly convincing. Conversely, Duhem and Mach continued to reject the reality of atoms since they held that science should offer explanations that do not go beyond the observable. I evaluate the arguments on both sides and reflect on how philosophical stances impacted on what scientists were willing to accept as genuine scientific evidence.