My first piece of real research was in number theory, directly inspired by Ramanujan, who was this Indian genius who lived in the early 20th century and died in 1921. And, so he had left open a lot of conjectures and I found some fresh conjectures very much in the Ramanujan style and I published those in the student magazine at Cambridge called Eureka. And so that was the earliest work I ever did, and the beautiful thing was that one of these conjectures called the crank conjecture was actually proved 45 years later and that's... so I was very delighted I lived long enough to see this conjecture proved; it was done by... by Frank Garvan who is a young Australian. And so that subject is still very alive, and in fact the whole subject of what I call 'Ramanujanology', which is the sort of number theory in the style of Ramanujan, is now rejuvenated, mostly the result of two people, George Andrews and Bearnt... Bruce Bearnt, two mathematicians who have taken up the systematic exploration of everything Ramanujan conjectured. They've gone through all the notebooks of Ramanujan, proving all the statements that he conjectured, of which there are thousands, so it's been a life work for Andrews and Bruce Bearnt, and so in... and they've attracted a very large number of graduate students over the years so there's now an army of young people involved in this kind of number theory, which is great fun, and of course Frank Garvan is one of them. So the ranks and cranks that I worked on as a student in Cambridge are still very much alive. So that's... that's a big joy for me; it's the beauty of mathematics, as opposed to physics, that it's forever. I published my selected papers recently in one volume, and I found out that when you publish your selected papers most of the physics is ephemeral, that you don't want to publish stuff that was written 10 or 20 years earlier, but the mathematics is permanent. So essentially everything I've ever published in mathematics is there, whereas only about a quarter of what I published on physics was worth preserving.