It was very difficult because of this complexity of the region. But in the meantime I was getting suspicious about the... what the real complexity of the system was because I remembered that when I did the somatics studies, the tumor variant studies, that I always got two regions. I didn't get eight regions or six or whatever. The antigens were grouping into two groups. I didn't test all of these antigens, but those that I tested seemed to be on two molecules. Nothing more. So that would mean two genes would be the simplest interpretation. So was this complexity that the recombinants seemed to be showing the many regions, was this real? Or could it be some artefact? Well then we realized, when we continued, we realized that some of the antigens seemed to map into different places depending on which recombinants you tested. That sometimes it was at one end and the same antigen in another recombinant would be mapping to the other end. Now, you can think of some explanations how this could be... could happen even if you keep those eight regions. But I thought that the simplest, the most parsimonious explanation would be the one that the tumor variants suggested there were only two regions, not counting the Ss. This was always considered to be very clearly separate, and the IR was a question mark. But the serologically detectable regions... six serologically detectable regions could be just an artefact. Just an illusion.

We discussed it with Don Shreffler with whom I was working at that time. Many times... Don was very conservative. I was anything but conservative, I guess. And... but primarily because of my experience with the somatic cells I was pushing for simplifying the interpretation and explaining the apparent mapping into different regions by saying it's a cross reactivity. The antigens are shared between the two different molecules at the different end on the H2 complex. So that in different... in some recombinants, in some H2 types you get a reaction with one at this end and another at this end, and this is the reason why the ambiguity arose in the test. So in the end he was invited to write the review, which he was very slow to always to write anything. And I was eager to get that review and he... well he told me we will write it together. So I wrote actually the paper and I brought it to him. And I remember to this day he just looked at it, and had... take a look... took a look at it and said 'Oh, I see. You put the two-locus model into it'. Well, it was the case, yes. He went along and it was published, and this is what I presented at the meeting in Asilomar, the so called two-locus model. I said, the regions are artefact except the K and D. There are two M... H2 loci and the rest is non-existent... it's an artefact of cross-reactivity. Now, that the audience understood and they realized that... and I think simultaneously there was already a rumour or it was already coming out, that McDevitt with the help of all these recombinants of which two pairs were crucial and they were produced by Stimpfling, Shreffler and myself. And together when you put those pairs together they gave you... you could test... you got a combination where the ends were identical and the middle part was different. So you were testing for the middle part. And McDevitt was getting an idea that the IR gene might be in the middle part and using these recombinants he could map it there.

So here was a case. Two loci, which are the H2, K and D, and in the middle is the IR gene that controls an immune response in some basic way. So suddenly everything was simple. Everything was clear and people knew that something... they could smell that something important was hidden there. Because as soon as McDevitt published his data one laboratory after the other was reporting response to this antigen, to this antigen, to that antigen, all of these responses are controlled by an IR gene which is in the H2 complex. So suddenly there was a tremendous demand for all the recombinant strains and everybody wanted to map his favourite antigen and everybody was calling the three of us... send me, and when are you going to send me? And we were sending anything. We had often, I must say, often I wanted to do some experiments, but somebody was begging so I didn't have enough. I sent him the ones that I set aside for my own experiments, and people were mapping.