Maxwell's color statistics: From reduction of visible errors to reduction to invisible molecules

• The standard account of Maxwell's kinetic theory of gases is challenged.
• A new cross-disciplinary or cross-domain account of Maxwell's statistics of gas molecules is proposed.
• Maxwell applied in color research statistical techniques borrowed from his seniors' astronomical and experimental work.
• Maxwell sought an objective mathematical representation of color as a perceptual and not a physical phenomenon.
• Quetelet's social statistical ideas entered Maxwell's molecular research after the formalism had been applied to color relations.

This paper presents a cross-disciplinary and multi-disciplinary account of Maxwell's introduction of statistical models of molecules for the composition of gases. The account focuses on Maxwell's deployment of statistical models of data in his contemporaneous color researches as established in Cambridge mathematical physics, especially by Maxwell's seniors and mentors. The paper also argues that the cross-disciplinary, or cross-domain, transfer of resources from the natural and social sciences took place in both directions and relied on the complex intra-disciplinary, or intra-domain, dynamics of Maxwell's researches in natural sciences, in color theory, physical astronomy, electromagnetism and dynamical theory of gases, as well as involving a variety of types of communicating and mediating media, from material objects to concepts, techniques and institutions.