> How about the positive integers with * as sum, with 1->3 as the only > nonidentity arrow? The unique cubical coalgebra is 1 but there is no > square coalgebra. Doesn't sound very functorial to me. Peter Johnstone