Now, Anderson was a world authority in non-stoichiometric compounds. I don't have to tell you, Brian, but perhaps I'll say for some other people that one of the cardinal things about chemistry is that compounds have a defined composition, and the answer is: unless. That rule of Dalton's law of fixed proportions was true for organic compounds, so-called molecular compounds; molecules have their atoms combined in strict, regular proportions and you can work out the composition, and that is what analytical chemistry does, amongst other things.

But, supposing you have a solid... that rule does not hold. Instead of stoichiometric compounds, that's to say compounds with fixed composition, you can have non-stoichiometric compounds. And some of the most important properties of solids stem from their non-stoichiometry. Why are they catalysts? Why are many of them coloured? Not all of them, some are coloured for a particular reason. Why do they conduct electricity? What about a transistor? What about silicon chips? All of these properties depend on non-stoichiometry. That was just beginning to be understood. It was a heresy in most places that chemical compounds could vary in composition, but Anderson was one of the first and certainly one of the most important in trying to put this on a firm basis.

And it is, of course, obvious if you think about it that if a solid is in equilibrium with a gas, one of its components, oxygen for example, and you alter the pressure of oxygen you should, in principle, alter the composition; it should take up or lose some oxygen, perhaps only minutely, but that is the effect which influences the property of the solid. Now, one of the solids which is very non-stoichiometric is copper oxide, in old-fashioned nomenclature cuprous oxide or copper (I) oxide: two atoms of copper, one of oxygen, Cu₂O, but the proportion varies. The most common one, I suppose, is magnetite, magnetic oxygen... oxide of iron, which, as its name implies, is magnetic, it's a basis of compasses and so forth. There is no compound FeO, iron oxide, one atom of iron, one of oxygen; that is unstable under all conditions, although the crystal lattice has a sodium chloride-like structure, some of the metal places are vacant, they are vacant and so some of the ferrous iron (II) atoms have to be oxidised to iron (III). So the process of oxidation is altering the amount of oxygen in the lattice and therefore altering the oxidation of iron.

That happens in copper oxide as well, and my job was to study this and to study the composition. It wasn't a new problem, it was one which had been studied, there was a lot of controversy about the properties, but Anderson said, ‘The only thing to do, Norm, is to do it properly', to make the compound under defined conditions, hold it in a controlled atmosphere, alter the composition and follow the electrical conductivity and follow the thermoelectric effect, the thing that makes thermocouples work.

So my first job was to make an apparatus and before that I had to learn glassblowing, so I actually became quite a reasonable glassblower in the end, after a few bad mistakes. But I built a very complicated vacuum line with a large glass tube surrounded by a heating element, which I had to wind non-inductively, and then have leads on to a sample which would be current leads and voltage probes to measure the resistance.

So, that was, in fact, quite an elegant and difficult job to do because first of all I had to make the copper oxide, then had to make joints on to this, then assemble it, be able to heat one end of the rod whilst keeping the other end cold, measure the temperature at both ends, and take all of these electrical leads out through a pinch seal with 10 metal to glass seals on it. Now, that did defeat me, I couldn't do that, but JS Anderson, who is a master glassblower, was able to do it. And he had an interesting technique. He took electric light globes, which already have exactly that seal in it, and he used those, put them together, fused them in and that was the glass seal that I needed. So I measured those things and did a... it took two years, wrote a thesis which was accepted, and I got a good degree on that as well, and as a result of these things I was able to get a scholarship to come to England to study.