Professor Philip Nolan:

The Chairman is right. We got involved in a discussion on the rest of the matters. That has been published, and I think it was on 21 May, so the work was done and then it was written up. The methodology for the estimation of time-dependent R and the model structure that gives us the other estimation-----