I was there from '47 to '49. The arrival at Caltech was a crushing disappointment. Not that the place was not nice, it was just lovely, but it was a crushing disappointment first of all because Karman had gone, in fact he had gone to live in Paris. It's a nice story in itself why Karman came to live in Paris, but the main problem was that this union of mathematics and "hands-on" was no longer there. I had a course from one person who described how to build big rockets and how to fly faster, and he was sure, if he paid somebody millions, he's going to fly a rocket airplane and maybe die, but enough people are prepared to do it. So the sound barrier did not exist obviously. On the other hand there were mathematicians who were totally terrified of the sound barrier: something terrible will happen, some singularities will occur which are very, very difficult. As speed becomes faster, the equations change from being elliptic to being parabolic to being hyperbolic, and these equations behave very, very differently, and the core of my teaching of partial differential equations was to emphasise that the transition of one to the other didn't occur mildly. I had learned that from Paul Levy in analysis, and it was what people were teaching at Caltech. Therefore, in a way, I'd gone to a place, into a field where I didn't belong either, because again, I couldn't find this combination... Although at Caltech there is a single department for mathematics, physics and astronomy. Well, I was not in one of those departments. In fact, I was in the school of aeronautics, because at that time applied mathematics didn't exist and was part of aeronautics. But Caltech was well-known and recommended to me knowingly by Braue in many ways, and what you said is very important. Millikan, who single-handedly founded Caltech as a great school in the '20s, had been very burnt in Chicago by his colleagues in mathematics who never talked to him, never listened to him, were in a different world. So he decided in this new school that he was creating this would no longer happen, and so he specifically had the single division of mathematics, physics and astronomy, which was always run by a physicist, and mathematicians were in a certain sense a first class service department. They were not there to do what they liked, but to do things that the physicists liked. In that respect Caltech was certainly closer to what I felt was a proper relation between these fields, even though I prefer them to be more equal, than say Chicago, or Harvard or Princeton would have been in that period.