It is notorious that the classical law on the acquisition of possession by persons in potestas is far from clear. Justinian's law appears to be that where possession is acquired in respect of the peculium the paterfamilias acquires possession immediately the person in potestas takes the thing. The paterfamilias need not have authorised the taking and he need not know of it. Where possession is acquired otherwise than for the peculium, however, the paterfamilias only acquires where he has authorised the taking or when he learns of it. It is probably the dominant opinion that this scheme also represents the classical law, but it has long been maintained by some that in classical law the paterfamilias automatically acquired possession of whatever a person in his potestas took and that the existence of a peculium was either completely irrelevant, as has recently been maintained by Nicosia or at least was relevant only in certain circumstances, such as in determining the beginning of a usucapio, the view more recently defended by Watson. A question arising in the course of this dispute is whether the rules, whatever they were, were the same for acquisition of possession and acquisition of ownership by traditio and I propose to consider this question and then deal briefly with the rules of acquisition of possession.