And that is what I want to tell you. In the interior of the metal it is right; that at each point that had to be two opposite currents. On the surface, near the surface, it still seems to be right. A circle near the surface cannot get completed. It turns into the wall. And instead of going round that, that way, it will go round that way, all the way around the interior wall, jumping that way. And the old argument still holds without any modification. But here comes the correction. According to quantum mechanics, according to Heisenberg, this circle and this series of circles, have different energies. Energies and frequencies are connected in quantum mechanics and it is easy to see that this has a simple frequency and this has some complicated frequency, which leads to a change and calculable energy. Therefore, this current and this current will belong to electrons of different energy, which at a certain temperature will be represented with different probabilities, the electron being more probably here than here. The cancellation of currents near the surface is incomplete and that incompletion then leads quantitatively to the result of Landau. One of my pleasant memories is the occasion when I reported on all that to Heisenberg who was wrong. He was very much interested. I certainly could not detect in him the least trace of displeasure, he was just plainly interested and agreed. He kept me for dinner and I remember from that dinner to other events. Heisenberg who was a great pianist played before dinner a piece from Bach's Well-tempered Clavichord. I tried to imitate it and I succeeded, after a lot of attempts, only partly. That was the agreeable part of the evening. There was an embarrassing part. We had dinner and after dinner there came something that I thought was cheese and I tried to put it on bread. And the wicked Heisenberg left me and allowed me to notice as I bit into it, that it was not cheese, but something sweet. That was his only revenge for my having proven him wrong.