When one discussed development, at that stage you couldn't really tackle something like development. You could say I'm interested in development or I'm interested in the nervous system, but nobody actually knew how you could do it. The classical technology, if I can put it, of development, was essentially cut and splice. You take a piece out of an embryo at one point and… at one part and you move it around to another part. And I can remember that… and of course you either note whether it changes or it doesn't change. If it alters, that is, takes up the properties of its new environment, then we call it that it's been regulated, and it wasn't determined. And a whole language had been developed, which I can remember reading through in the ‘40s with… with people, with Joe Gilman, reading… Needham's book, the Biochemistry and Morphogenesis, and being totally involved in really very detailed discussions on the difference between evocation and induction. It was tremendously hard to discover whether induction and evocation were really two different things or was one part of the other. And whether an evocator was something different from an inducer, and so on. And so we had no other approach then to cut and splice. Of course what we knew is, that we now had genes to contend with. We knew what they are. We knew that genes made proteins, and so that just said, well, let's get started in… into this and look at how genetics or genes will now affect our knowledge of embryology. I have to say also that the biochemical approach to development had totally fizzled out. The search for the inducer had gone on, lots of people had tried, there were lots of things that… that had worked. There were, there was a huge area of biochemistry in development which simply consisted in assaying enzymes in eggs and in tadpoles, and the… the connections weren't there. There was a lot of what I would call theoretical work, in fact a few books written by Waddington at the time on organisers and genes — a very interesting book — and ideas that genes really control rate processes, but all of this had no body in it, it lacked… it lacked a material content and was simply verbalising our ignorance. And having… and the way that I think Francis and I worked is that I kept on thinking is, what… what experimental system would one attempt to do this? And I remember saying well, what we should do is decompose the problems of development into sub-problems, and just see whether we can tackle these problems in an experimental way independently, because the global problem is intractable at this time. And we then decided to ask what were the questions that you could decompose it into. So one of the questions was that… cell movement, it had become clear that I mean, cells move around and that cell movement may be an important thing in modelling… modelling the organs of… of animals. Polarity, which I think is still a very important problem, was in my mind the essential problem. You had to explain why one part of a cell became different from another. Why certain cells divided with certain planes of division, which of course one could see from all the classical descriptive embryology that polarity, I think, needed this explanation.