Gibbs and Einstein in 1902: Parallel Constructions in Statistical Mechanics
Résumé
This article reexamines the striking conceptual and formal parallels between Josiah Willard Gibbs’s Elementary Principles in Statistical Mechanics and Albert Einstein’s first paper on statistical mechanics. Although their derivations follow different paths, both arrive at the canonical distribution and construct closely related theoretical frameworks for thermal equilibrium. Each offers a mechanical grounding of the second law of thermodynamics and defines temperature as the quantity that equalises when systems exchange energy, thereby characterising equilibrium without recourse to kinetic theory. In both cases, the constructions rely on general mechanical principles—most notably energy conservation and Liouville’s theorem—combined with probabilistic reasoning, and deliberately set aside Boltzmann’s kinetic approach. Whereas Gibbs’s treatise quickly became a canonical reference, Einstein’s 1902 paper remained marginal in both physics and its historiography for more than half a century. Focusing primarily on that first article, while also considering the closely related papers of 1903 and 1904, the present study reviews the main historiographical interpretations of their reception. Particular attention is given to Paul Hertz’s two critical articles of 1910, which identified a close affinity between Gibbs’s and Einstein’s approaches and played a decisive role in Einstein’s subsequent disengagement from his early statistical programme, following a private meeting in Basel arranged to avoid public controversy. Drawing on new archival evidence concerning the rapid European dissemination of Gibbs’s book in the spring of 1902, the article reassesses the plausibility of Einstein’s later claim that he was unaware of Gibbs’s work at the time. It also highlights Max Planck’s extensive 1903 review, published in the influential Beiblätter zu den Annalen der Physik, which offered the first physically transparent discussion of statistical fluctuations within the Gibbsian framework.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |
