sensu stricto (Title 11), and (3) legitimation (Title 10, § 13). Corresponding to these are described the modes in which patria potestas may be dissolved, and a person alieni may become sui iuris (Title 12), most of which may be grouped together under the three kinds of capitis deminutio, which is described in Title 16, rather out of its proper place. The guidance and control which children require in their tender years is naturally provided by the care of the father; but to meet the contingency of the latter's death while a child is still impubes, and especially to protect his proprietary interests, the Roman law established various forms of guardianship, to which the reader is introduced by a further distinction of persons sui iuris into those who have a tutor or curator, and those who act entirely for themselves (Title 13). The various forms of guardianship are treated in the succeeding titles, and the functions of a guardian, together with the limits of their necessary exercise, are touched on in Title 21. Title 22 relates to the modes in which guardianship is terminated, and Title 23 to curators, or persons appointed by the magistrate to manage the property of various classes of persons-minors over the age of puberty, lunatics, interdicted prodigals, and others who, from some mental or bodily infirmity, are incapable of adequately attending to their own affairs. Title 24 describes the circumstances under which guardians and curators have to give security for the due administration of their office, the nature of such security, and the liability of magistrates who neglect to exact it when they appoint these functionaries. Lastly, in Title 25, it is pointed out that the duties of guardians and curators are of a public nature, and cannot, as a general rule, be declined by any one called upon to undertake them; certain grounds of excuse however were recognised by law, and these are enumerated in detail.

The very exhaustive treatment which the status familiae, as contrasted with libertas and civitas, obtains in Gaius and the Institutes of Justinian has led Savigny to regard the 'ius quod ad personas pertinet' as merely family law, and to maintain that such was the meaning which those writers themselves attributed to it. He examines the view that the real subject of their first book is status, though not exactly in the sense in which we have understood the term; that is to say, he distinguishes status into natural and civil; by the latter he means the three great status to which alone we have given our attention; by the former he would express the differences which exist among persons as subjects of rights and duties on account of differences of age, sex, health, etc. This, of course, is a perversion

of the strict Roman idea of the term status; and his rejection of this view is no reason for also rejecting that which we have adopted as the true one. To Hugo's theory, that the real subject of the 'ius quod ad personas pertinet' is 'capacity of right,' or the three characters which correspond to the three forms of capitis deminutio (which coincides in the main with our own exposition) Savigny is more favourable, but he finds himself unable to accept it because it does not account for the facts, and therefore lacks the very essence of a tenable explanation: his main objections being that tutela has no relation to 'capacity of right' (Rechtsfähigkeit) but only to 'capacity for legal disposition' (Handlungsfähigkeit), and the omission of the important distinction between cives and peregrini, upon which something has been already said. If, says Savigny, we look more closely into the actual contents of the first book of the Institutes, we find that it corresponds very nearly to what he calls family law. It treats, in point of fact, of marriage, patria potestas, slavery, patronatus, guardianship (and, in Gaius, of manus and mancipium). On the other hand, the division into cives, Latini, and peregrini, important as it is in relation to capacity of right, is not to be found, because it belongs, in its real nature, to public law. The most serious objection, in Savigny's opinion, which can be urged against his own view, is the omission (in Gaius) of kinship, but this he does not consider fatal. It is remarkable that he did not see how slender was the coincidence between family law, as he conceived it, and the 'ius quod ad personas pertinet' of Gaius and Justinian. The former is in some ways of larger extent than the latter; thus it comprises the whole of the relations between husband and wife, and parent and child, though these belong to the latter only so far as they relate to patria potestas, and also the rules in respect of patronatus and colonatus, which find no place in the 'ius quod ad personas pertinet' whatsoever. In some points, on the other hand, its content is less. The classification of 'persons' in Gaius and Justinian is based upon two great divisions, one of which is iuris gentium-all men are either free or slaves-while the other— all free men are either ingenui or libertini-is iuris civilis; the first of these, prominent as it is in Gaius and Justinian, has nothing to do with family law, because the relation between master and slave is one which belongs properly to the law of things,' and which, as has been pointed out, is touched upon in the first book only by reason of the capacity of the slave, as distinct from other 'res,' of becoming a 'persona' by manumission.

Perhaps one reason why there has been so much dispute as to

the real signification of this division of law into personae, res, and actiones, is the supposition that it was a classification of great antiquity among the Romans themselves, a classification which they regarded as fundamental and not lightly to be departed from, and by understanding which therefore we shall obtain a deeper insight into their national legal habit, and comprehend more fully the leading distinctions of their system, and the interconnection of its parts. Such a supposition, however, is an error which it is important once for all to discard. Sir Henry Maine has pointed out' that there is no reason to suppose that the Roman lawyers set any extraordinary value on this classification. It was confined, he says, to their institutional treatises, or primers of law, the educational manuals placed in the hands of their beginners. We may add that Gaius' other wellknown work, the 'res quotidianae,' was composed on another plan, and that the order followed in the Institutes of Florentinus, who wrote after Gaius, was also quite different. The Twelve Tables again, and the Praetorian Edict as consolidated by Salvius Julianus, have no trace of the Gaian classification; the Gregorian and Hermogenian codes were arranged upon a different principle; so was the code of Theodosius II; so are the two larger works of Justinian himself. It may even be the case, as Savigny has suggested, that that classification was based on an individual, accidental opinion of Gaius himself; upon any view, however, we cannot suppose that by the ordinary Roman jurist the division was regarded as much more important than other current contrasts which Savigny instances, such as those of vi, clam, and precario, the three forms of domestic dependence, potestas, manus, and mancipium, the three capitis deminutiones, and the three classes of cives, Latini, and peregrini. If there was any traditional view among the Romans as to the true mode of classifying their legal rules, Sir H. Maine has shown, in the paper referred to, that it was probably the view which from the force of circumstances has prevailed in nearly every indigenous system of law under primitive conditions, and which assigns the first place in the code to judicature and rules of procedure, in the interstices of which the substantive law has in early times the appearance of being gradually secreted. The material law tends to become distributed into 'heads of dispute' in an order which seems to depend on their relative importance when it was finally determined, and in which, in more celebrated codes than one, a prominent place is given, in particular, to deposits and thefts.

1 'Ancient Ideas as to the Arrangement of Codes,' in the Fortnightly Review for May 1879.

Between the private law of Rome as stated by Gaius, and as stated by Justinian, there is far more difference in the matter of personae and actiones than in the intermediate department, the 'ius quod ad res pertinet,' except perhaps in such part of the latter as relates to inheritance or succession. The most important points of change in the subject now under our immediate consideration are briefly as follows: (1) The threefold division of libertini into cives, Latini, and dediticii, which was important when Gaius wrote, was practically obsolete in the time of Justinian, who formally abolished it; this necessitated a change in the law of manumission, which was greatly simplified, the statutes which had introduced these distinctions being to that extent repealed, as also was the lex Fufia Caninia, limiting the number of slaves who could be manumitted by testament; (2) of persons' alieno juri subiectae,' exclusive of slaves, there had in Gaius' time been three classes, viz. those in paternal power, those in manu, and those in mancipio. In the Institutes patria potestas is still a living reality, but manus and mancipium have become obsolete; hence a second simplification, the modes in which manus could arise, and some of the purposes for which it had been employed, being a somewhat complicated and difficult branch of law. (3) In Justinian legitimation of children not under patria potestas at birth is definitely treated as one of the modes in which this power may originate; in Gaius' time this occurred only by the fact of a Latinus acquiring the civitas, and by his children born before this event being subjected to his potestas; but in the later period there were no Latini; and accordingly the long discussion in Gaius (i. 65–96) 'de statu liberorum,' and on the modes in which a Latin could become a civis, has no counterpart in Justinian except the paragraph on legitimation in i. 10. 13; (4) with regard to adoption sensu stricto, Justinian completely altered its form, and also its effect unless the adopting person was a natural ascendant; with regard to adrogatio,' the form had been changed between the two writers, it being now effected 'principali rescripto,' not 'populi auctoritate,' as in the time of Gaius; but no other change of any great importance had been made in this branch of law. (5) Justinian entirely changed the mode of emancipation (i. 12. 6.) by substituting for the old fictitious sale a mere declaration by the parent before a magistrate. (6) In the department of guardianship the perpetua mulierum tutela,' with which, though still existing in his day, Gaius (i. 190) had expressed his dissatisfaction, had altogether disappeared long before the compilation of the Corpus Iuris. The expression 'fiduciaria tutela' has also changed its signification (see notes on

Title 19), and the mode of magisterial appointment, in default of a testamentary or agnatic guardian, is different from what it was in the time of the earlier writer. On the subject of curators Justinian is far fuller than the latter, and to his last two Titles (de excusationibus— de suspectis tutoribus et curatoribus) there is no corresponding matter in Gaius at all. This is probably to be explained by the fact that the conception of tutela had changed between the two writers, or at any rate had in the later period come to be regarded far more as a publicum munus,' from which exemption could be claimed only on very sufficient grounds, than in the age even of the Antonines.