Roger then, I think, went off to Harvard and later to Stanford and we set about trying to crystallise the nucleosome, a single nucleosome without H1. H1 is the fifth histone, which we later showed with Thoma and Koller, it's absolutely essential for making a regular packed helix. That was a paper in '79, the paper got rejected by I don't know how many different referees but it turned out to be right in the end. How many different journals and referees... because they thought that a linear array would be enough, the answer... it really was at the basis of that. So the...

[Q] Len Lutter.

Oh, Len Lutter. So then we recruited Len Lutter and the idea was to try to find out what the nucleosome was, so he then... Lutter was an American Post Doc, a very skilful, very skilled at the bench and he began doing nucleosome preparations of different kinds. And he found, that if you added microchoccal nuclease then you could get a ladder, just as you went down in the one, two, three, four experiment, all the way down to single nucleosomes. If you went on digesting you found that the 200 base pair b, and turned into 160 base pair b, and so clearly something had happened. And during that time, I think I've forgotten which point we showed that, the H1, if there was H1 present, fell off. And there was H1 present because we did this from... originally from calf thymus chromatin, but I think that we got it from beef kidney in the end. Some people tell you, you can get better preparations from beef kidney, was it you who use to go to the abattoir to collect it?

[Q] No, no.

They used to have to...somebody... You had to tip the slaughterman to give you the kidneys.

[Q] Yes.

That was a good, useful source. And so we, so the... Ray Brown and Daniela Rhodes crystallised the nucleosome. It was quite an interesting crystallisation because it, if you actually started adding magnesium it precipitated. And they were trying to do things with manganese as well which had the same behaviour, and then Ray Brown and Daniela Rhodes, one day announced he's got crystals, but the crystals were at 40 millimolar magnesium; And I couldn't believe it. And then I realised what was happening, and I go back to insulin... this is not a digression this was actually part of the reasoning. You know, if you actually, insulin picks up zinc, this is the work of Dorothy Hodgkin, who worked out the structure, there's two zinc, there's four zinc and six zinc. And the different zinc they formed different... it's well known in organic chemistry they formed different... higher aggregates. And the... so it was clear that the... you could make six insulin because they were an aggregate of insulin molecules. And I realised that this must be an aggregate at 40 millimolar of the... so the... so what was happening before... so we managed to grow good crystals because if you... if you, so the... the low amounts of manganese they precipitate out and they don't give good crystals at all, no crystals at all. If you go on adding manganese, they salt in as the physical chemists say, so we actually determined the boundary at about 40 millimolar. It turned out to be dependent upon the exact length of the DNA that you chopped up. Normally it was about 100 and 160 turned out to be only a rough measure and the best crystals came with 147 bases of DNA, and these were crystallised.