Tim Richmond came in 1978 to try to work on the high-resolution structure, to try to get better crystals. And we had got better crystals by this time, thanks to Len Lutter's work. And they diffracted to quite high resolutions, about six or seven angstroms at the time. And with Tim Richmond, we solved the structure to seven... seven angstrom resolution, wasn't it, yes. In which you could see the DNA very clearly, the two turns of the DNA, and you could identify the different domains. You could see alpha helices just on the edge, you could see the alpha helices... of H3 straddling the dymer position on the DNA and so on. It was really rather... And so that was the nucleosome but Tim Richmond went off to Zurich... there was a hiatus in all this because a man called [Evangelos] Moudrianakis working at Johns Hopkins questioned and said it was all wrong, it was all rubbish. Because he had solved the structure of a histone octamer, he thought, and found that it was much bigger than the histone octamer we put into our model. And this, apart from anything else almost, it was published in Science, no less, it was done by a single heavy atom, it was also done by, oh, what's... the solution, what do you call solution averaging around, there's a word for it... beg your pardon, in X-ray crystallography.

[Q] Yes.

You take the shape, you carve out the shape. There's a man who worked it out, what was his name? Anyway, it was a standard technique and they claim to have done it.

[Q] Solvent flattening.

Solvent flattening, now, this is meant to be X-ray crystallography. One of the referees of that paper was Alex Rich and so it was published in Science, so our paper was regarded as wrong and the model regarded as wrong. And Alex Rich enjoyed a certain bit of schadenfreude because he'd been beaten on the tRNA, he enjoyed that, he told me he refereed it. He said, 'Well, you have to settle this.' And of course, the whole thing was rubbish because they'd done everything wrongly, the heavy atom was on a two-fold axis of symmetry and they hadn't realised it and solvent flattening. And so... and it almost cost Tim Richmond his job, he was... by this time he was going to move to Zurich where... where the Head of the ETH Zurich, you know, the Federal Institute of Technology, was a friend of Moudrianakis at the time. And said to this chap, was unreliable. Luckily, he... I interceded and said, 'Moudrianakis is wrong, totally wrong and we can demonstrate that.' And it turned out in the end and it was all solved by a chap at Pittsburgh, whose name I've forgotten; but he introduced... he was one of those who introduced solvent flattening. And so our nucleosome structure was right at that level but was quite one of those episodes that shows that you actually need, well, we had a perfectly good high-resolution structure but it was the... we didn't have heavy atoms at the time, did we?

[Q] No.

So that was the nuclear... nucleosome structure and also the higher order structure which was a string of nucleosomes. Tim went on to... Tim went on, yeah, so Tim... now, this was done, this was done by random... DNA extracted from chromatin. So 100... 147 base pairs was a... it was 147 plus or minus by two or so you couldn't control the exact length of the DNA. And we could see, within the crystal structure seven angstroms. The ends of the DNA of two different nucleosomes came together like this, so clearly it was dependent upon the length exact crystal, and we did shrink the crystals and so on to try and get them packed tightly. So what Tim did when... after he'd gone... after he went to Zurich was to use defined length DNA, of defined length... absolutely fixed length. And he used 146 and 147. And he crystallised it using exactly our crystallisation conditions, and he solved the structure. He could do it by ice water replacement because we had the seven angstrom structure, and he did that. I'm not sure that he used heavy atoms at all. And by that time also, by... by that time the people had introduced fast freezing, so you could preserve structures to high resolution and... Absolutely... well, it... it was important, he made a dimeric. And he made a dimeric, there are two ways you can make DNA, a DNA sequence dymeric with two-fold symmetry, that was important. You see, if you have a two-fold axis perpendicular to the DNA, there are two-fold axis in the plane of the bases and there's another two-fold axis between the bases. So therefore you could make... so that's why he made 146 and 147, which would mean that the two-fold symmetry would be in two different positions. So one would be totally symmetrical, the other case could be twisted to one side, he solved them both. And so eventually, this led to a high-resolution structure of the nucleosome. And that was ultimately published in about the early '90s, I think. I took quite a few years... and so it was quite an achievement. The... it's all based upon the basic idea on Kornberg's biochemistry and our crystallography were the foundations for that.