It is called, ``Numbers can be just what they have to''. It appears to be a preprint and done, unfortunately, on a typewriter. I have written to Charles asking him to ask Colin if he wants to post it, but Colin, in a Philosophy Dept, isn't into computers and almost certainly doesn't have in computer readable form. You can of course write him at Dept of Philosophy, Case Western, Cleveland, OH USA 44106. There is something I would like to add. I got one query from someone who didn't believe that it was possible to use categories or toposes as a foundation because a category has a set of objects and a set of morphisms (or a class, but that's not the issue here) and you cannot talk about the domain or codomain functions and the rest of the structure unless you have some underlying set. Right? No! You have no more need (or exactly the same need) to have an underlying set as you do when axiomatizing sets. If you need to have a set to talk about a function, then you need a set to talk about the membership relation. In both cases, it would be nice to have a set; it puts us on familiar territory, but when founding mathematics you have to start somewhere. It is as legitimate to start with a pre-existing notion of morphism, domain, codomain and so on as with a pre-existing notion of set and membership. Are your intuitions well-founded? How do I know? I think mine are, but I cannot prove it to my satisfaction, let alone prove that yours are. That's the way it is with foundations. Michael ++++++++++++++++++++++++++++++++