has enough energy to ionize a mercury atom it gives up most of its kinetic energy and does not have sufficient energy to overcome the retarding potential between N and G. Further variations in voltage and electron behavior confirmed this interpretation. Bohr quickly pointed out that the energies involved were too low to ionize mercury atoms, but corresponded to the orbital transitions predicted by his new theory of the atom. This was soon accepted as Franck and Hertz's discovery.
In spite of initial disagreements on interpretation, both Franck and Hertz and Bohr assumed as established fact that: there are electrons; electrons are constituents of atoms; heating a wire causes electron emission; there is a linear relation between applied voltage and electron acceleration; that electrons travel in trajectories; that they collide with atoms; that there are both elastic and inelastic collisions, that the energy lost by an individual electron is absorbed by an individual mercury atom; and that this absorption induces structural changes in the atom. All of these 'facts' were subject to earlier, and sometimes later, controversies. Here they function as presuppositions. To compare this functioning with our earlier analysis, consider the propositions:
An electron, emitted at D, that has an inelastic collision with a mercury (F1)
atom looses kinetic energy.
The electron travels in a trajectory from D to the atom and from the atom to G. (F2)
However, in the light of wave-particle duality and the Schrödinger equation, we know that (F2) is not really true. So it might seem that we should substitute
Electrons seem to travel in trajectories. (F3)
In the context of the Franck-Hertz experiment (F3) does not work. This is not an instance of an historical limitation, which has since been overcome. As Cartwright has argued, similar presuppositions are operative in the design and execution of the Stanford Linear