The obsidian was really a very interesting piece of work for me. Because I was doing my doctoral research in the Cyclades, I wanted to understand what was going on in the Cyclades and therefore the obsidian question became a very important one, perhaps more important than it would otherwise have done, and so there were these obsidian sources, two main obsidian sources on Melos, which I'd been able to visit and then it was suggested that the obsidian used in Malta might have come from Melos whereas it was known that there was a source nearer to Malta at Lipari just off the North coast of Sicily, and so that was a problem that it looked as if it would be very interesting to solve and also there's a lot of obsidian being found in Turkey and in the Near East and the question was where as all this obsidian coming from and how could you work that out? And so I had the good fortune to discuss this in St John's College with an old friend of mine, Joe Cann, who had been in the same class as I was. We were both in the same class in St Alban's School and he'd gone and he hadn't done national service. He'd got deferment and then in the end he hadn't done national service and he'd become a geologist and petrologist so he agreed to look at the question, and my job was to get good samples of obsidian from wherever I could so I got plenty from - from Melos and then I wrote to David Trump who was working in Malta and he gave me samples of what was being found in Malta. Then with Joe we went through the collections of the department of mineralogy and petrology in Cambridge, and there they had abundant obsidian samples from the sources, like Lipari and so on, and so we looked at it first of all from the standpoint of refractive index which obviously would be an optical technique and we couldn't really see much - much difference there. We looked at it systematically and there wasn't much to be done. On the terms of appearance there wasn't much systematic difference so Joe suggested that we should do trace element analysis by means of optical emission spectroscopy which was a fairly new technique then, a standard technique in mineralogy and petrology, and it involved getting quite small samples and grinding them up in a pestle and mortar and then he had access to a machine which was run by a technician when these were burnt in the machine, and then the spectrum produced was recorded on a photographic plate and then you could read the intensity of the spectral lines, the spectral lines being caused by the presence of trace elements of these various - so it was possible to get estimates, semi-quantitative estimates, for the quantities in parts per million of these various trace elements that were present in the obsidian. And so we looked systematically at the obsidian samples from Melos and those from Lipari and the other known sources that we could get our hands on, and there did, in the end, seem to be some systematic differences. If you plotted them on a graph of barium against zirconium in parts per million, you could separate the Melian sources from the Lipari sources and separate those from the Sardinian sources and those from the sources in - in Hungary, although from the - from the Pontine Islands, which are off the coast of Italy, they didn't differ very much from the Lipari ones until we looked, until Joe looked, at a new trace element we hadn't looked at before, which was caesium and he said - I was in - in Athens by then so he sent to me a message saying that caesium allowed one to distinguish between the two so I sent him a telegram saying, ‘Congratulations, caesium!’ and this really helped to sort of crack the problem and so it became possible having learned how to discriminate the materials from the sources, to characterise the verb we introduced, to characterise the materials from the sources, it became possible to look at the samples from Malta, from other archaeological sites, and do the trace elements analysis on them and see which they matched up with. And the Maltese specimens matched up with mainly with Lipari and the suggestion that there'd been contacts between the Cyclades and Malta as documented by the Melian obsidian, proved to be completely unwarranted. And then it got even more interesting really in the East Mediterranean where it was possible to show that obsidian at early sites down in the Levant like Jericho, pre-pottery Neolithic Jericho, the obsidian sources for these had to be in Central Anatolia and we knew some of the sources there and we took steps to try and work out the other sources and, in fact, my father went on holiday and was instructed to go and pick up obsidian where he could and did so, just by the roadside, and that identified the approximate position of one of the sources in Anatolia, so in this way it was possible using really very hard scientific methods to build up a picture of the trading networks that must have been operating, and so that was certainly the most important early paper, I think it was the first paper, it was published in the "Proceedings of the Prehistoric Society" for 1964, “The Characterisation of Obsidian and its Application to the Mediterranean Area". And, in some ways, it was just about one of the first really successful characterisation studies. There had been good work done earlier in England on stone axes using thin sections and the petrological microscope, and so the Neolithic axe trade had been investigated already in the 1950s by that means, but this was the first time that trace element analysis had been successful. The Germans had been doing it, Sangmeister, and his colleagues, on metal objects but they hadn't really been able to distinguish one source from another so it didn't really solve very many problems in any clear way, but in this case we were able to establish these things and then later on when it became clear that Melian obsidian was cropping up in Mesolithic deposits, really very early deposits, 8/9000 BC in the in the Peloponnese, at the time was the earliest hard evidence for seafaring, it Franchthi Cave had clearly been brought by sea, and so that meant that the Cycladic Islands or the Island of Melos was already being visited in the Mesolithic and then ultimately in the upper Paleolithic period. And in the Near East, it was very important because it made clear that all these different early farming villages were in some kind of contact and obsidian was being traded down the line through these various villages and so that was, I think it was an important paper, and, of course, it interested people quite widely. When I first went to America, I went and lectured on that because people like Robert Braidwood in Chicago who was excavating at Jarmo, he had by then sent us material for analysis and Jimmy Mellaart had sent material from Çatal Höyük and Kathleen Kenyon had sent material from Jericho, so it was possible to make very firm statements about these trading networks which had been suspected before but they'd never really been very well documented, so that was quite an exciting thing to be involved with. So Joe and I carried that forward and got more samples and did a further paper in the Proceedings in the Prehistoric Society, then when Jane - when Joe was busy with other things, we invited John Dixon who'd also been at St Alban's School but in a younger class, and had also been in St John's College, and he became a collaborator, so we did a whole series of papers which really, to a large extent, I think sort of sewed up the obsidian trade question. The work was gone into again by more sensitive analytic techniques, neutron activation analysis, and it still continues but I think the broad patterns that were established then remain valid. So that was an exciting thing to be involved in to the complete investigation of a trading system using specific scientific techniques that were capable of giving a very good answer.