Part of the confusion was directly affecting the interpretation of the H2 system because on the basis of irreproducible data, the H2 was becoming again complex. As I explained at the beginning, there was a serological complexity which was real but the interpretation was incorrect and it had to be reduced by the two locus model to its right proportions. But after the reduction the H2 began to expand again that people were mapping IR genes and were placing different IR genes into different regions and they were discovering other phenomena, which were controlled by the H2, mapping them and again generating new regions. So it appeared that in addition to the K D I S regions, there were suddenly other regions like one region was called B, another was called J and so on.

And my function seemed to be to puncture all these bubbles that were appearing and to remove these regions. I sometimes felt like Mephistopheles in Goethe's Faust where he said, when Faust asks him... when he first appears, he asks him, 'Who are you actually?  And he says, 'Ich bin der Geist, der stets verneint!' ('I'm the spirit which always negates'). So that was the function I seem to have assumed, regions appeared and we did experiments to show that it's not so. The B region could be explained as an artefact of interaction. The J region... there was supposed to be an antibody against J and it was supposed to be involved in the suppressor cells... suppressor cell function.  Well we never succeeded in producing... reproducing the antibody and neither did any other people. The whole actually... the whole area of the suppressor cells turned out to be irreproducible. So in those times one of our contributions was to keep the H2 system... the MHC of the mouse in a believable state and that was the state which eventually was confirmed by the cloning of the genes and the biochemical works. Of course it turned out to be that there were other genes in the MHC and that brought me... but for these I always had the opinion that they were not related to the MHC function and I might come back to that question later.