We can use viruses as models for evolution. In other words we can say well, in early prebiotic phases there has been a sufficiently rich environment in which complexes like the present viruses could reproduce and start evolution. That may be and that has to be shown if it is possible.  But the present viruses we have most certainly are somehow linked to their host, at least they cannot exist without the host. Now, to take the example of the AIDS virus, the HIV. You know that the HIV... the human immune deficiency virus so I look at the human and I compare it with the simian, that means monkey virus... specific virus. Now, in human virus we know at least two sub-types. People try to even to divide it further into sub-types, but there are two essential and very fatal types of viruses, that's the HIV-1 and the HIV-2, and they probably have a different origin. One looks more like having originated in central Africa, the other more in western Africa... the HIV-2... they have many relations, similarities, to the simian viruses.

Now the first question is: when did this virus occur? What is the evolution of the virus? Well, there have been many speculations about it because the virus in the Western world only appeared in the '80s. One thought: well, perhaps it's a new composition or something new. That question meanwhile is entirely clarified. Actually we did ourselves work and we used the method, you were involved in that, which we call statistical geometry. In other words, if you study those sequences which are evolved very far, which means which vary with the high mutation rate, you cannot use the classical methods of constructing trees, you have to... well, we have talked about the sequence space, you have to project the sequences into the sequence space and then look at their correlations among them.  And we have worked out a very potent method by which we can say we can study events which are very far back and distinguish it from events which appeared recently. So it's a more sophisticated technique which we also use for determining the age of the genetic code.