Scientific modeling and the nature of science: The case of Thomson's atom

Thomson's atomic model is commonly presented in textbooks as an oversimplified and largely qualitative “plum pudding” picture, obscuring its original conceptual and mathematical structure. In this paper, we present a historically grounded reconstruction of Thomson's atom based on excerpts from original papers, showing how it was a dynamically constrained model grounded in classical mechanics and electrodynamics, explicitly addressing atomic structure, stability, and chemical properties. Using this historical example, we discuss key features of scientific modeling, including explanatory scope, empirical constraints, the epistemic role of mathematics, and conceptual continuity across subsequent models, showing that Thomson's atom provides an effective case study for promoting critical reflection on modeling and constitutes a step toward a gradual understanding of the nature of science in the teaching of physics.