I decided in 1982 that... by which time we had worked out how DNAs packaged in chromosomes, at least at the first couple of levels. The nucleosome and then the higher order structure, the 300 angstrom fibre and that of course applied to bulk chromatin. But already in the last few years, before that, people had been exploring chromatin by enzymes seeing if there were patches which were... part of them which were accessible to enzyme actions and so on. And the concept grew out of what was called active chromatin. That was chromatin in which the DNA was either being transcribed, turning into RNA or about to be transcribed, poised to transcribe. And the way this was explored was by... as I say, by enzymes, particularly by DNAs one and so on. And there were the beginnings of understanding, a question was that also people found that there were HMG proteins, another class of protein which were found in 300 angstrom fibre a particularly in embryonic forms and so on. But basically I thought that the bulk understanding, at least for the first two levels, and there were higher levels in which the 300 angstrom fibre could be folded but I couldn't see how that was accessible. So I became intrigued by active chromatin and I looked for a system, this was deliberate, I looked for a system, I read the literature, on transcription obviously and also on transcribing chromatin and I looked for a system which we could access biochemically, but to produce material on a large scale for structural studies. And I read about the work of Roeder, that's Bob Roeder who was then at the... St Louis in the USA and Don Brown who was then at the Carnegie Institute of Washington, which happens to be Baltimore, oddly enough, on the system of transcription which is used to make 5S RNA and Xenopus. Xenopus the frog, as a biologist insist on calling it a toad and I'll call it a frog. It's a South African toad. It is a very important animal in biological studies, there's a whole groups work on Xenopus, it's one of the classic animal models. The 5S RNA genes are switched on by a transcription factor called TF3A, TF3A is transcription factor A of polymerase three, there are three different polymerases in your carriers. Bacteria only have one, but there are three different ones all closely related but they are used for different classes. Class two is what makes ordinary; switches on ordinary genes which make... are messenger RNA... polymerase three is what makes RNA which is the final product, such as RNAs going into the ribosomes and so on, and the... so although it wasn't in the main class of... but this is one which is accessible because they showed that there were large quantities of TF3A in immature Xenopus oocytes. And it happened that in our lab was Hugh Pelham newly arrived, he is now the deputy director and he'd worked with Don Brown and he's shown that TF3A was the same as a storage protein, identified by some French work as 40 kilodaltons in size which was thought to be a storage protein for 5S RNA, so this was a real conundrum. The TF3A was both a transcription factor but it also bound to the product of transcription 5S RNA itself. So it bound both the 5Ss RNA gene and the 5S RNA product. Well, we found out it's only in 2003, two years ago, we found how it bound to RNA, actually that's yet another story, it's a quite important postscript to it all, but I'll come to that. But this intrigued me, but the thing that intrigued me most, or rather that attracted me most, was there seemed to be large quantities and... we had Xenopus in the lab because John Gurdon was in the lab so these were taken from oocytes and the way that you do this is that you can extract the material... you stimulate, you put in factors which will promote ovulation in the female frogs and then what the experts do would be to actually cut him open and take out the oocytes, extract them and you can sew them up again, apparently that's what Gurdon did, but cruder people simply kill the frogs and take out the oocytes, so you can get large quantities.