The work on penicillin was really quite widely known about in Oxford, that Florey and Chain had the idea of working on antibiotic substances, and one of them was penicillin, and this looked very promising. I... I did go to one lecture, I think by Professor Gardner, in the very early stages of the attempt to isolate penicillin. This was given to the  [unclear] and that in turn lead to the isolation of a barium salt which wouldn't crystallise, but which had penicillin activity.

I was very lucky that I came into it really through meeting Ernst Chain in the... outside the path lab in South Parks Road in a very excited state just after the morning when they had done some most successful animal experiment in which they gave a streptococcal infection to eight mice, and four of these mice they gave penicillin, and four not, and all the ones who were given penicillin lived, and the others died. Norman Heatley stayed up that night with them, with the mice, looking after them so he may have told you this story from a closer observer.

Well, Ernst just said, 'We've found something really important that does really cure people' and he also said, 'And we'll... we'll soon be able to get it out, of course, and when we do, there'll... there'll be some crystals for you', because he knew that what I was very keen on having at the moment was the crystal to work on, about which I knew nothing, and so that I could see how far I could push the method, but not anything too complicated because that would take too long and the calculations took in any case rather longer than we at that moment hoped that... I joined in in their research, which was the ordinary organic chemist's kind of trying to break the molecule into pieces which were more easily recognisable, and the next research student that I had was Barbara Low.

One of my Somerville students helped with the practical work required taking X-ray photographs, and measuring the density and the unit cell sizes of the molecules present in the different breakdown products. We first examined penicillamine, which was the first of them, and then went on to various others, and one that was most helpful was penicillaminic acid. This one gave very nice beautiful crystals including a copper salt, and took measurements of it in the copper salt. And I tried, as usual, to make up a formula for it, and discovered very rapidly that I couldn't do it, because it appeared to have too many oxygen atoms to be attached to the crystals in a skeleton there. I went over to the DP, and I said to Norman Baker... Wilson Baker, as I was giving him the results, 'I suppose you know you can't write a structural formula for this compound?' He said, 'Oh no, can't you? No, I have been suspecting this for quite a time'. And the answer was that it contained sulphur instead of two of the oxygens, and this had been missed in the combustion analysis, which is an old phenomenon which happens from time to time - it happened to cystine, for example, in its early history. So then, penicillin, from that moment, had a sulphur in the formula, and became more and more understandable.