Pierre Weiss

Pierre Weiss was born in Mulhouse in 1865, when Alsace was still French. He was eight when the Franco-Prussian War transferred his home city to the German Empire; he spent his adolescence in a German-speaking gymnasium and his undergraduate years at the ETH Zürich, taking his engineering degree in 1887. He picked up a doctorate in physics at the Sorbonne in 1896 and spent the rest of his career shuttling between Zurich, Rennes, and Strasbourg — every move following the shifting political borders of his native region. Few physicists of his generation worked across three national scientific traditions; Weiss worked in all three and was fluent in all three.

His 1907 paper "L'hypothèse du champ moléculaire et la propriété ferromagnétique" is the founding document of modern ferromagnetism. Pierre Curie had already shown that ferromagnets lose their magnetisation above a critical temperature, and that paramagnets obey χ = C/T. But Curie's law couldn't explain why a piece of iron retains magnetisation when the applied field is removed, or why ferromagnets exhibit hysteresis. Weiss proposed a single sharp idea: inside a ferromagnet, each atomic magnetic moment feels not just the external field but an enormous internal "molecular field" produced by the average alignment of all its neighbours. This mean-field self-consistency turned Curie's law into the Curie–Weiss law, χ = C/(T − T_c), and predicted — correctly — that the transition to ferromagnetism would be a genuine phase transition at T_c. The molecular field is fictitious as a microscopic entity (exchange interactions, not classical dipole sums, actually do the work, as Heisenberg would show in 1928), but as a mean-field device it gave quantitative predictions that matched iron, nickel, and cobalt.

Weiss needed a second idea to explain everyday ferromagnetism: if every magnet were uniformly magnetised, every piece of iron would lift nails on contact. He proposed that bulk ferromagnets are divided into tiny regions — he called them *domains* — each uniformly magnetised but pointing in different directions, so the macroscopic average cancels out. An external field doesn't rotate atomic moments; it grows the domains aligned with the field at the expense of misaligned ones, via boundary motion. Domain boundaries were first seen experimentally by Francis Bitter in 1931 using magnetic-powder patterns, twenty-four years after Weiss had predicted them. Weiss himself also pioneered high-field electromagnet design — the Bitter magnets at his Strasbourg laboratory reached the highest sustained fields in Europe through the 1920s — and served as director of the Strasbourg physics institute until the 1940 German invasion forced him south to Lyon, where he died later that year. The mean-field approach he invented became the template for every subsequent order-parameter theory, from superconductivity to superfluid helium to the Ising model: average everything, solve self-consistently, get the phase transition.