The last chapter presents von Neumann's model for measurements of canonical quantities. The model has many applications, and Sect. 7.4 applies it to resolve the velocity paradox of Sect. 7.1. However, we may not be comfortable with the resolution of the paradox. Velocity is not a canonical quantity, but it is gauge invariant: mv = p−eA/c. (See Prob. 4.3.) The model implies that a velocity measurement lasting a time T yields v with uncertainty ∆v ≥ /mT . (See Prob. 7.10.) Yet the same model implies that we can measure momentum instantaneously, with no minimum uncertainty, and without changing it. We can also measure p − eA/c instantaneously, with no minimum certainty, and without changing it; A is a canonical variable. But p − eA/c equals mv (except during the measurement) so why can't we measure velocity instantaneously? Likewise, in the laboratory we measure electric and magnetic fields, E and B, not the potentials A and V that appear in the Hamiltonian, Eq. (4.7). The magnetic field B = ∇ × A is a canonical quantity, but the electric field is not (although it is gauge invariant), because it depends on the derivative of A with respect to time:The question of measuring E troubled such physicists as Landau, Peierls, Bohr and Rosenfeld. Rosenfeld arrived at Bohr's institute in early 1931 just as the question had come to a head. He ran into Gamow and asked what was new. Gamow answered with a drawing he had just made. It showed Landau, tightly bound to a chair and gagged, while Bohr stood before him with upraised finger, saying "Please, please, Landau, can I just get a word in!" Landau and Peierls had come a few days before with a new paper to show Bohr, "but he does not seem to agree," said Gamow airily, "and this is the kind of discussion which has been going on all the time." Peierls had left the previous day, "in a state of complete exhaustion," added Gamow. Landau stayed on for a few weeks, and Rosenfeld discovered that "Gamow's representation of the situation was only exaggerated to the extent usually conceded to artistic fantasy" [1]. Landau and Peierls had considered measuring the electric field by sending a charged test particle through it [2]. The electric field deflects the test charge; the change in the momentum of the charge indicates the field strength. But an accelerated charge radiates and loses an unknown part of its momentum to the electromagnetic field. (See Prob. 8.1.) We can suppress the radiation by reducing the charge on the test particle; then the momentum of the particle changes more slowly, but the measurement lasts longer. An impulsive, accurate measurement of the electric field is impossible. Bohr felt uneasy, but could not refute the claim of Landau and Peierls.Ultimately, Bohr and Rosenfeld did refute the claim; it took them nearly three years [3], but they showed how to measure electric and magnetic fields instantaneously [4]. Bohr and Rosenfeld did not treat noncanonical observables in general. We will see that in the von Quantum Paradoxes: Quantum Theory for the Perplexed. Y. Aharonov...