Owen had previously pointed out that marsupials differ from ordinary placental mammals in having four (in place of three) pairs of cheek-teeth at the hinder part of the series which have no milk, or deciduous, predecessors, and are therefore, according to the usual rule, to be regarded as true molars, in contradiction to premolars, in which such deciduous predecessors are generally developed. He considered, however, that all the premolars in the kangaroo (and therefore presumably in other marsupials) as well as the incisors or cutting teeth, and the canines or tusks, were preceded by milk-teeth. Flower, on the other hand (who it is only just to add had a much fuller series of specimens of young marsupials on which to work than was available to Owen), was enabled to show that in the Marsupialia only one pair of teeth in each jaw, at most, is preceded by a milk-tooth. The tooth, in question, is the fifth from the posterior end of the series, and whereas in the adult animal it differs in character from those behind it, its deciduous predecessor resembles the latter. The replacing tooth was further considered to correspond with the fourth or last premolar of placental mammals, while the replaced tooth was regarded as the only one in the entire series corresponding to the milk-teeth of placental mammals. This view rendered it necessary, of course, to regard all the four pairs of cheek-teeth behind this abnormal one as corresponding to the true molars of placentals, as had been done by Owen, thus making, as already mentioned, marsupials to differ from ordinary placentals by possessing four instead of three pairs of these teeth.