Preliminary observations on the use of the rat as a model for the pig in the determination of apparent digestibility of dietary protein