where this was the case, the husband did not acquire any power over the wife, and the latter retained whatever property she previously possessed. It was therefore not marriage per se that created the power of the husband, for a woman might be legally married and not be under the marital power. But although this free marriage,' as we may call it (the term is not Roman, but invented by modern jurists), was legally possible, the custom, and in old days the almost invariable custom, of the people was to add to the marriage ceremony not essential to its validity as a marriage, but one which had important legal consequences. We may safely assume that there was originally no true marriage without the ceremony, but at the time of the Twelve Tables this was no longer the case. The ceremony created a relation which the Romans called Hand (manus), and brought the wife into her husband's power, putting her, so far as legal rights went, in the position of a daughter (filiae loco). It gave the husband all the property she had when she married. It entitled him to all she might acquire afterwards, whether by gift or by her own labour. It enabled him to command her labour, and even to sell her, though the sale neither extinguished the marriage nor made her a slave, but merely enabled the purchaser to make her work, while still requiring him to respect her personal rights. In compensation for these disadvantages the wife became entitled to be supported by her husband, and to receive a share of his property at his death, as one of the family heirs' (sui heredes), whom he could disinherit only in a formal way. She had by coming under his Hand passed out of her original family, and lost all right by the strict civil law to share in the inheritance of her father.

There were two forms of ceremony by which this power of the Hand could be created. One, probably

1 Some writers doubt whether this power of sale existed, and refer to a supposed 'law of Romulus' mentioned by Plutarch which devoted to the infernal gods whoever sold his wife. But the balance seems to incline in favour of the existence of the power.

the older, had a religious character. It took place in the presence of the chief pontiff, and its main feature was a sacrifice to Jupiter, with the eating by the bride and bridegroom of a cake of a particular kind of corn (far), whence it was called confarreatio. It was originally confined to members of the patrician houses. The other was a purely civil act, and consisted in the sale by the bride of herself, with the approval of her father or her guardian (as the case might be), to the bridegroom, apparently accompanied (though there is a controversy on this point) by a contemporaneous sale by the bridegroom of himself to the bride. The transaction was carried out with certain formal words and in the presence of five witnesses (being citizens) 1, besides the man who held the scales with which the money constituting the price was supposed to be weighed. The price was of course nominal, though it had in very early times been real.

These two forms have been frequently spoken of as if they were indispensable forms of marriage, so that marriage had always the Hand power as its consequence. But this, though it may probably have been the case in very early days, was not so in those historical times to which I must confine myself. And the proof of this may be found in the fact that if a woman was married without either of the above forms, she did not pass into the Hand of her husband unless or until she had lived with him for a year, and not even then if she had absented herself from his house for three continuous nights during that year 2. And where the Hand power had not been created, the property rights of the wife, whatever they were 3, remained unaffected by the marriage.

1 There has been much dispute as to this ceremony: I give what seems the most probable view. It may descend from a more ancient sale of the wife by her relatives to the husband, similar to that which we find in some primitive peoples.

2 This was in pursuance of the general rule that rights over a movable were acquired by a year's continuous holding: 'usus auctoritas fundi biennium, caeterarum rerum annuus esto.'

3 If she was in the power (potestas) of her father, she had no property of her If she was sui iuris, she was under guardianship.

own.

The period of three nights is fixed in the Twelve Tables, possibly as a precise definition of a custom previously more uncertain.

This was the old Roman system, and a very singular system it was, because it placed side by side the extreme of marital control as the normal state of things and the complete absence of that control as a possible state of things. Doubtless the marriages with Hand were in early days practically universal, resting upon a sentiment and a social usage so strong that women themselves did not desire the free marriage, which would put them in an exceptional position, outside the legal family of the husband. Nor can we doubt that the wide power which the law gave to the husband was in point of fact restrained within narrow limits, not only by affection, but also by the vigilant public opinion of a comparatively small community.

IV. CHANGE FROM THE EARLIER TO THE LATER

SYSTEM AT ROME.

Before the close of the republican period the rite of confarreatio practically died out, or was referred to as an old-world curiosity, much as a modern English lawyer might refer to the power of excommunication possessed by ecclesiastical authorities. The patrician houses had become comparatively few, and the daughters of those that remained evidently did not wish to come under the Hand power 1. The form of coemptio, which all citizens might use, lasted longer, and seems to have been not infrequently applied in Cicero's time. Two centuries later it also was vanishing, and Gaius tells us that the rule under which uninterrupted residence created the husband's power of Hand, and might be stopped by

1 Nevertheless it was retained in a few families for the purpose of providing persons who could hold four great priestly offices, since by ancient usage none save those born from a marriage with confarreation were able to serve these priesthoods. But its operation seems to have been restricted by a decree of the senate so as to apply only so far as religious rites were concerned (quoad sacra) (Gai Inst. i. 136).

the wife's three nights' absence, had completely disappeared (Gai Inst. i. 111). So we may say broadly that from the time of Julius Caesar onwards the marriage without Hand had become the rule, while from the time of Hadrian onwards the legal acts that had usually accompanied marriage, which placed the wife under the husband's control, were almost obsolete.

This was a remarkable change. The Roman wife in the time of the Punic Wars had, with rare exceptions, been absolutely subject to her husband. She passed out of her original family, losing her rights of inheritance in it. Her husband acquired all her property. He could control her actions. He sat as judge over her, if she was accused of any offence, although custom required that a sort of council of his and her relatives should be summoned to advise him and to see fair play. He could put her to death if found guilty. He could (apparently) sell her into a condition practically equivalent to slavery, and could surrender her to a plaintiff who sued him in respect of any civil wrong she had committed, thereby ridding himself of liability. One can hardly imagine a more absolute subjection to one person of another person who was nevertheless not only free but respected. and influential, as we know that the wife in old Rome was. It would be difficult to understand how such a system worked did we not know that manners and public opinion restrain the exercise of legal rights.

Such was the old practice. Under the new one, universal in the time of Domitian and Trajan, which is also the time of Tacitus, Juvenal and Martial, the Roman wife was absolutely independent of her husband, just as if she had remained unmarried. He had little or no legal power of constraint over her actions. Her property, that which came to her by gift or bequest as well as that which she earned, remained her own to all intents and for all purposes. She did not enter her husband's family, and acquired only a very limited right of intestate succession to his property.

This striking contrast may be explained by the fact that the disabilities which attached to the wife under the old system were not in legal strictness the consequence of marriage itself, but of legal acts which an almost universal sentiment and custom had attached to marriage, though in themselves acts distinct from it. A perfectly valid marriage could exist without these legal acts, and so far back as our authorities carry us, we find that a few, though probably originally only a very few, marriages did take place without them. Accordingly when sentiment changed, and custom no longer prescribed the use of confarreation or coemption, the power of Hand vanished of itself and vanished utterly. Had it been an essential part of the marriage ceremony, it would doubtless have been by degrees weakened in force and accommodated to the ideas of a new society. But no legislation was needed to emancipate the wife. The mere omission to apply one or other of the old concomitants gave the marriage relation all the freedom the parties could desire and perhaps more than was expedient for them.

We may now dismiss these ancient forms and address ourselves to the position of the wife under the normal marriage of later times-the so-called 'free marriage,' since this is the form in which the Roman institution descended to and has affected modern law 1.

V. LATER MARRIAGE LAW: PERSONAL RELATION
OF THE CONSORTS.

The following points deserve to be noted as characterizing the Roman view.

The act whereby marriage was contracted was a

1 I pass by the distinction between iustae nuptiae, which could be contracted only between Roman citizens, and the so-called 'natural' marriage, or matrimonium iuris gentium, which was created by the marriage of a full citizen to a half citizen or an alien (peregrinus), because the latter is of no consequence for our purpose, and practically disappeared when al! Roman subjects became citizens. It was a perfectly valid marriage, and the children were legitimate. As to their status, see Gai Inst. i. 78, 79.