“Marriage is ended by divorce, death, captivity, or the enslavement otherwise oc­curring of either party.”

So writes Paul (D. 24.2.1). Divorce and the death of the husband or wife were in the Roman world, as they are today, far and away the most common reasons for a marriage ending.

Divorce, as the Romans conceived it, is in some respects the reciprocal of mar­riage. The state did not intervene to regulate or even to register divorce, and there were no divorce courts as such because there was relatively little to settle. The spouses' property had remained separate during marriage, and so its division upon divorce was, in theory, easy. Further, custody of children was not an issue since as a rule they remained in their fathers control (Case 49), although he could arrange for them to be brought up by their mother.

Under these circumstances, the law was able to tolerate what may seem to us an extraordinary casualness about divorce. Just as entering marriage was chiefly a mat­ter of an objectively manifested agreement between husband and wife, so too, with­out much formality, divorce could be accomplished by either spouse at any time, through a simple process of terminating the “marital affection” upon which the con­tinuity of the marriage depended. Still, practicality dictated that divorce require more than a mere mental act of renunciation; and the jurists toy with a requirement that the divorcing spouse must notify, or at least try to notify, the other party.

Although divorce itself was a simple process, major complications were intro­duced by dowries, which husbands (or their heirs) frequently had to surrender when marriages ended. Literary evidence suggests that the burden of returning a dowry was often oppressive, mainly because husbands found it hard to gather up and transfer appreciable capital. Further, the return of the dowry was complicated by the Roman tendency to treat this as an appropriate occasion for a complete fi­nancial settling between husband and wife. The consequence was the creation of numerous legal bases whereby the husband could “retain” portions of the dowry. For the most part, there is no evidence that such retentions are of statutory origin, and the jurists seem generally to have crystallized them out of long-standing judi­cial practice; doubtless, lawsuits over retentions were not uncommon. The reten­tions based on the husbands prior expenditures on the dowry raise particularly im­portant questions about how Roman law preferred that husbands administer dowries.