Why I wanted to breed them is very simple. I expected that the serology which I was using at the time would be very difficult on wild mice. On the inbred mice it was facilitated by the fact that you had many individuals of the same kind so you could sort out the complexities, but in the wild... mice, they are all heterozygotes. That means they have four different genes instead of two in the H2 complex, so it was much more complex... expected to be and turned out to be that way. But eventually all these problems were managed and I was typing on wild mice.

The first thing I discovered is that the antigens could be sorted out into two categories. Some were very restricted in their distribution among the individuals while others were wildly shared between individuals, and I called the former the private antigens and the latter the public antigens. That turned out to be very helpful then in the formulation of the two-locus model because on the basis of the two-locus model it turned out that the private antigens are specific for an allele and that each inbred strain had only two private antigens, which was another evidence that there were only two loci. And it helped in the interpretation of the H2 and the polymorphism in the wild mice, so we discovered that polymorphism was indeed tremendous and were able to find out many things about the nature of the polymorphism and also about the wild mice themselves... how they lived, what social organisation they had and so on and so on. I'm not going to go into that... I would like to just mention to what these discoveries led, because at least for me then it became quite important.