categories: Re: complete atomic Boolean algebras

[Marta BUNGE <bunge@scylla.math.mcgill.ca>]

For an "upgraded" version of the correspondence 
see the newly appeared

M.Bunge, J. Funk, M. Jibladze, T. Streicher
"Distribution Algebras and Duality"
Advances in Mathematics 156, 133-155 (2000)

This is done for a bounded topos E--->S, replacing 
S = Coc_S(S,S) by Dist(E) = Coc_S(E,S) and caBA(S)
by "distribution algebras". Identifying E with S gives 
the correspondence in question. For just that one, it 
may be worthwhile to consult directly Mikkelsen's 
thesis combined with Pare's Theorem.

Marta Bunge