Second replay (sent 1st February 2000)  
_____________________________________________________________________  

  Dear Editor, 

  I understand that, after the second negative report, you 
  consider the question of my paper definetively settled. 

  However, I would like to point out that the both reports are
  based on the same, more or less explicite assumption that "Bayesian 
  is equivalent to non-scientific". 

  Allow me to put the question in the following terms: 

  - The Editor agrees with the above assumption, and then I have
    little to say more.

  - The Editor accepts that, at least in principle, Bayesian 
    methods are suitable for physics data analysis, but then
    the paper should be reviewd by a referee who considers this 
    point of view valid, and judges the paper on the basis of
    weakness or inconsistences within this framework. 
    In this case I would be very glad, if requested, to 
    clarify specific points and to improve the paper. 

  Still arguing on the data analysis method, I would like to remember that
  the paper, with all the limitations it might have, addresses for the
  first time in HEP literature the though issue of `outliers', and tries
  to solve it in a scientific way, i.e. stating clearly framework and
  assumptions. And the referees were not able to show that this non-easy
  task has been treated and solved consistently elsewhere. 

  Finally, I would also like to point out that we should not forget the
  physics issue. Last week, during the SPSC committee, it was considered
  an important point that the picture of a positive and sizable value
  of epsilon'/epsilon survives this kind of sceptical analysis. 

  Your sincerely,

     Giulio D'Agostini.