I tried to find representations but what I did then was to look at infinitely high momenta for the particles, which is very much like looking at things on the light cone. And I looked for those currents that survived, that didn't go like: one over p sub z, or one over p sub z squared, but things that went like: one, as p sub z went to infinity, the momentum in the z direction. And I looked at the algebra of all those things. So what I got was the set of Fourier components oftime… set of Fourier components of the time components of current densities–so in other words charge densities. And those form of course a… an algebra, and that's still correct, there's nothing wrong with that. I looked for representations of it, but I looked for representations of it without pairs, and those are trivial. Those are just non-interacting, single objects. They have no… we… we showed after a long agonizing search that you don't get anything out of that; you have to put in a certain number of pairs. And it's also useful to look at the longitudinal components. And both of those things were done then subsequently in the work of Bjorken, and Feynman's work which sort of popularize Bjorken's ideas, and so on. Now we could have used the transverse Fourier components of the time components of the currents and found representations that were non-trivial by allowing higher values of the quantum numbers. What I did was to restrict the quantum… the total quantum numbers to the ones for the… for the particle itself. In that case we might have gotten some very interesting stuff but we… but we didn't do that.