thermodynamics nor electrodynamics had been grounded in a mechanistic foundation21. There was a looser linguistic, rather than ontological, unification in the primacy core mechanical concepts had in other branches of physics. Alternatives to mechanistic foundations were developed. There were physical alternatives: electrodynamic foundations based on the electromagnetic theory of mass; and thermodynamic foundations based on energetics. There were also proposals for epistemological foundations such as positivism, conventionalism, and hypothetical deductive reasoning. Boltzmann led the opposition to these proposed innovations and expressed his reasons by using Hertz's terminology of pictures.
I merely wish to work against the thoughtless attitude that declares the old world picture of mechanics an outworn point of view, before another such picture is available from its first foundations up to the applications to the most important phenomena which the old picture has for so long now represented so exhaustively, especially when the innovators have not the least understanding of how difficult it is to construct such a picture. (McGuinnes, 1974, p. 258)
The development of the Bohr-Sommerfeld (B-S) atomic theory (or program) was very much in the Boltzmann tradition, using mechanics as a tool, but not as a foundation. On the presumption that the physics is familiar, I will briefly summarize a neglected aspect of this development, Bohr’s conceptual crisis. Bohr believed that a physical account must precede and ground a mathematical formulation and that physical accounts related to differing mathematical formulations must have a basic coherence if the theories are to be used together. There was a significant difference between the way the Bohr-Sommerfeld program was treated by Sommerfeld and Bohr. Sommerfeld and his students treated it as a closed theory and focused on the problems that could be solved within the theory. Bohr and his associates treated it as a stepping-stone towards a more adequate theory.22 The close interaction between experimenters and theoreticians that has increasingly characterized atomic and particle physics began here. This personal orientation was reinforced by Bohr's professional role. As the presiding figure in the atomic physics community he encouraged a close collaboration between experimenters and theoreticians and repeatedly modified his theoretical formulations to accommodate experimental results. I will try to bring out the unifying role of the fundamental concepts that experimenters and theoreticians shared.
The experimental sources that contributed to the development of atomic physics can be roughly divided into two types: those that obtain information from the light emitted, absorbed, or modified by atomic processes; and those that obtain information by hitting a target with a projectile and examining the debris. In neither case did formal
21 Klein, 1972, summarizes the difficulties turn of the century physicists encountered in attempts to make mechanics foundational.
22 For discussions of Bohr's early work that stress is concern with conceptual problems and the interaction between experiment and theory see Hendry (1984), chaps 3-5; Röseberg (1984), chap. 3; Folse (1985), chaps. 2-3; MacKinnon, (1982, chap. 5), (1985) and (1994); Chevalley (1991), Introduction; Faye (1991); Darrigol (1992), Part B; and Petruccioli (1993).