Attribution re no membership-respecting morphisms

I've attributed the third saying at http://boole.stanford.edu/dotsigs.html to Mike Barr. If there's an earlier attribution I should be using please let me know. Vaughan Pratt