“It is not always the magnitude of the differences observed between species that must determine specific distinctions, but the constant preservation of those differences in reproduction.”