Bragg was then working in the Schuster Laboratory and the laboratory was packed with his staff, nearly all of whom were engaged in crystallography. Well, of course, this was wonderful work, which Bragg was one of the great experts, but I simply could not interest myself in crystallography. There was so much of it, and I realised later on that this was typical of me. I never... I was not then and never did really become interested in beginning research on subjects which had always been well explored. I always seemed to have this initiative and inkling to start new things. So I tried very hard to interest myself in crystallography. There's so much of it. Massive books and so many experts, that I thought it was not for me.

Two things then happened. I made the great mistake of trying to continue my work on thin films, but it was hopeless. I was given a room in the new physics building. Oh goodness, what that room was. In the basement, looking out on a brick wall and I thought of the lovely room I had worked in in Bristol, and I asked Burrough if he would make me another equipment so that I could continue my work. It was a very bad mistake. There were no facilities in Manchester, no glass blowing, no liquid oxygen or liquid nitrogen, and Burrough did send me the apparatus, but it got smashed in transport. So on that Christmas vacation, I did return to Bristol and complete another item of work on thin films, but meanwhile... and I can't remember how this happened.

Douglas Hartree, the professor of applied mathematics had become friends of the family of the girl with whom I'd fallen in love and whom I was to marry. Joyce Chesterman, well-known family in Bath. And I remember they would pick me up on their way back from the south country and it was Hartree who had built this amazing machine, the first differential analyser. The first integrated machine. A superb mechanical device and he asked me if I'd like to spend some time working with him on that machine, and I was delighted to do so. It was in the basement of the Schuster building, and though Hartree was the professor of applied mathematics, he did spend most of his time in the physics department, and there I met Dr. Arthur Porter, who became a life long friend and he was the person who built the original Meccano model of this differential analyser. I worked on that and helped Porter and Hartree solve a number of equations using this machine. I don't think I ever published any of the results for any of my work on that, but in later years as the modern computers took over, I'm very happy to have had association with that elementary machine.