As a journal editor who is somewhat anal about references, I agree with Steve Kates and others on this matter: first date of publication first, and any others in order. Barkley Rosser