``It was during the fall of 1943 that my Chairman at the technical University of Budapest called me into his office and told me that the subject of my dissertation was the problem of three bodies. When I asked him what, specifically, I should be doing he looked at me and said, `Go away, and solve it. Read Whittaker's book!'
The reader should be filled in with a few details at this point in order to comprehend the story. In those days in most European universities the Chairman of the department was the absolute ruler and king, in charge of the assistants, adjuncts, graduate students, etc. He decided who qualified as a PhD student without conducting a formal qualifying examination. (In fact this was the case also in the 1960s at Yale University where Dirk Brouwer, as the Chairman of the Department of Astronomy decided which graduate students qualified and which were sent away.) The next information the reader needs to realize is that the assistant in these days did not ever argue with the professor but did as he was told or just left the University.
So when I was given my assignment, I was first very happy that I `qualified' and ran to the library to read Whittaker's book. Apparently my king did not realize that my English was not adequate to read this book, but, of course, I could not disturb him with such minor details and started working on Whittaker's `Analytical Dynamics.' The chapters on the problem of three bodies were absolutely fascinating, and I could not put down the book until I read it several times with a pencil in my hand to figure out the missing details. I believe this was the point in my life when I fell in love with celestial mechanics. I started playing less soccer, cut back on movies, even reduced the time chasing girls an started thinking on what I learned from Whittaker.
The most important think I immediately realized was Poincar\'e's non--integrability principle. As the reader will see later on, this principle means that no general valid solution can be found for the problem of three bodies. So I went back to my ruler and after reminding him who I was and what my assignment was, told him that the problem cannot be solved and I even referred to Poincar\'e, partly showing off and partly hoping that he would not execute me on the spot. Well, he actually smiled, which I had never seen before, and told me to go and solve the problem of three bodies. I was standing in front of his desk and could not move. I realized that my love of dynamics was not reciprocated and that my infatuation led me to a catastrophy since I would never receive my PhD degree. My king must have noticed my silent desperation and said: `I told you to solve the problem of three bodies. I never told you to find the general solution or to solve the gravitational problem of three bodies. Read Radau's article in Comptes Rendus.'
So I ran to the library and immediately realized that this time I was supposed to read French, which I did with considerable difficulties. I also established that the variables invented by Radau allowed us to find several new particular solutions of the problem of three bodies when the forces acting between bodies were not inversly proportional to the squares of the distances (as in the case of gravity) but directly proportional to the cube of the distances. My dissertation was finished in March 1946 and a version was published in 1952 in the U.S.
My love affair was reinforced and I learned English and French. The most important thing I learned was that in science careful attention must be given to the formulation of problems and that generalizations sometimes offered new results. I am still looking for the physical application of this strange force law and I am convinced that not finding it is my limitation and not nature's.'