Now, while all of this was going on, we continued to do experiments in… in molecular genetics. Anand Sarabhai, one of my graduate students, who had what we call the luck of the Sarabhais, had in fact shown that amber mutants, these conditional lethals, cause chain termination, and that he could identify the chain terminated products directly without purification. And by doing this we were able to show, with a simple assumption, that a protein got started at one end and went to the other end and it was a topological argument that in fact the genetic map was co-linear with the protein map, not having to do sequences as Charles Yanofsky was doing at about the same time. Now, the amber mutants were very interesting, because what Anand had shown is that they chain terminated, and we then began to work on what were the amber triplets, and of course that is the only case where in fact we had completed the original program of molecular genetics without resorting to biochemistry. The idea was that we would analyse the structure of DNA by chemical mutagenesis. We'd analyse the structure of the protein and then we would be able to say, these triplets correspond to this protein, to this amino acid. And that was only completed in this one case of this… of the amber, the nonsense triplets. What… the experiment we did was a very… was a very straightforward one, and in fact it is still being used by people to teach present-day students the principles of analytical molecular genetics, because it's a very exceptional example of what you could say from the DNA almost by genetics alone. We had already shown that there were these chain-terminating triplets in the R2 gene. In the R2 gene we had one strain in which the mutant could not grow. We had a strain, of course, where everything grew. So what we did is by using a chemical reagent, hydroxylamine, we treated bacteriophage with this. The argument is that hydroxylamine changed a base cytosine to uracil or an equivalent of it, and of course it could change one in the strand that has been copied into messenger or it could change one in the other strand. Therefore, if you took a family of bacteriophages containing some mutants and you passed them through the non-restrictive strain, then everything would ultimately become expressed. But if you took them and passed them through the restricting strain, then only those which contained unaltered cytosines on the sense strand, only those would grow through. And all the ones in which the cytosine had been altered on the sense strand so it was now read incorrectly, that would make a mutant gene product and would not pass through. And that would mean that all the mutants that… that didn't pass through could be assigned a cytosine in the strand that was copied into messenger, and the ones that did, could be assigned the cytosine in the other strand. And that gave us a way then of decoding the triplet as to which way… whether it had a G or a C.