The accident was that here I wanted to stop adrenaline acting on the heart and here is the drug. So adrenaline to isoprenaline is just a methyl group being replaced by an isopropyl group. And so isoprenaline selectively stimulates the heart and dilates the bronchi, and that's all we knew. So, where do you start to go looking for an antagonist? So, all we could think of starting was with isoprenaline. Now, I had read a paper somewhere that you could make an antagonist out of a molecule, par doublement la molecule – doubling it up – so that's what we set out to do: to make doubled up versions of isoprenaline. So, basically, these are dibenzyl ethylamines. So, that's... we were working away with these, and getting nowhere in finding an antagonist. And then... so, let me tell you, isoprenaline, it is derived from an amino acid and it's of the phenyl group and the side chain and amine group. Two hydroxyl groups on the ring, so it's a dihydroxy. And it was known, by this time, that the action of adrenaline was terminated by one of these hydroxyl groups becoming methylated. And so a group out in Eli Lilly, in the States, were trying to make a long acting version of isoprenaline, and so they thought: it's these hydroxyl groups that are metabolically labile... let's replace them with something. So, they replaced them with two chlorine atoms, dichloro analogue of isoprenaline. Well, they tested it. It had ceased to be a bronchodilator, so they hadn't achieved their objective. But they did notice that once a preparation had been treated with DCI, as it's called – dichlorides or DCI – that it would no longer respond to isoprenaline. Now they didn't know about Ahlquist's work and so when they published the work they didn't talk about maybe this was blocking beta receptors, but it was published, and two months later... that was published at the end of... just about... end of about '58, beginning of '59, and two months later a paper appeared from a group down in Atlanta claiming that, in fact, the stuff was acting by blocking Ray Ahlquist's beta receptors. So, was this the lead we were looking for? So, John immediately made it, and we tested it, and in the preparation we were using, it was just as though we'd given isoprenaline. So at that time my knowledge of pharmacology was pretty well non-existent, so I didn't understand it; put it on the side. And then we went ahead. Now John, when he looked at this, pointed out to me that these chlorine atoms were quite large, and so big lipophilic... he said, 'You know, if... if I joined up two phenyl groups it would be the, sort of, same shape and size, so that's going from... to analine'. But, in the meantime, I had changed my method of assessing these compounds and found that, in this new preparation, that DCI wasn't... didn't have these isoprenaline-like agonist activity. And so suddenly we got all excited about DCI. John had this idea about analine so that was the first. I remember, from idea to making it, to it working, that's my recollection. I mean, and by Jove, we got something which was as good as DCI but without any agonist activities. So, and that was the first beta blocker and we went into man with it. It came from Alderley Park, so it was Alderlin. Do you remember? Pronethalol. Pronethalol, yes. But we learned a lot from that compound. Then, subsequently a better compound was made when we... By this time ICI, you see, had woken up to the possibility that maybe there was some money here, and put a whole bunch of chemists on with... Bert Krause, I remember, was the leader of this group and so they did a great job at ordering in the changes, and they produced propranolol, and that has been the backbone of beta blocking drugs ever since.