Mr. Coffa's admirable and lucid exposition of the history of mathematical randomness, leaves, to my mind, nothing to be desired. I am not sure whether, or how, its end result is of use to the theoretically minded statistician, but it is certainly interesting enough in its own right to need no excuse.

I tend also to agree wholeheartedly with Mr. Coffa's treatment of physical randomness, or at least with his conclusion. I can easily imagine that he will not have succeeded in convincing those who are sure that epistemology is built into physics - say in quantum mechanics - but he has certainly given even these people some food for thought. I can even swallow the peroration of his last paragraph whole, and agree that one might hope and expect that physics can make its claims without a reference to knowledge or to randomness.