quas licet ex ipsius lectione apertius cognoscere. Hoc am- 4 plius eae obligationes, quae consensu contrahuntur, contraria voluntate dissolvuntur. nam si Titius et Seius inter se consenserunt, ut fundum Tusculanum emptum Seius haberet centum aureorum, deinde re nondum secuta, id est neque

if there is not one to novate, the attempted novation is null; if there is one to novate, but the novating contract is void (e. g. 'si id quod tu mihi debeas, a peregrino, cum quo sponsus communio non est, spondes verbo stipulatus sim' Gaius iii. 179), the former is altogether unaffected.

The promise of a slave ordinarily created a natural obligation, and consequently, as Servius Sulpicius argued (Gaius, loc. cit.), it is hard to see why it should have no novative effect; the explanation given by Theophilus is ὅτι ποιεῖ νοβατίωνα οὐ μόνον τὸ τίκτεσθαι φυσικὴν ἐνοχήν, ἀλλὰ καὶ τὸ ὑπεῖναι πρόσωπον ἀπρόσωπος δὲ ὁ δοῦλος.

The rule that, if the parties to the new contract are the same, a novation ensues only if it contains 'novi quid,' holds only where the novated obligation was itself verbal. If it had been otherwise, novation would have been disabled from discharging one of its most valued functions, viz. the substitution of obligations pursued by actions stricti iuris for obligations enforceable only by actio ex fide bona. The Proculians were of opinion that the adiectio or detractio of a surety in the new contract was not sufficient to support a novation, Gaius iii. 178.

For condicio and dies see on Tit. 15. 2 and 4 supr.

Servius Sulpicius had held that a conditional stipulation novated an unconditional contract whether the condition was fulfilled or not. Gaius (iii. 179) thought that the old contract subsisted until the condition of the new one was fulfilled, but suggests that if the creditor sued upon it before such fulfilment he might be met by exceptio doli or pacti, and this was soon recognised as law, Dig. 23. 3. 50; ib. 83; 12. 1. 36; though Labeo (Dig. 23. 3. 80) had thought otherwise.

Among the praesumptiones or evidence upon which the jurists relied to prove that there was animus novandi was, according to the Sabinian school, the addition of a surety (cf. Dig. 2. 14. 30. 1); for the presumptions against such intention see Dig. 46. 2. 6. pr. and 1; 45. 1. 58. Justinian's own enactment is in Cod. 8. 42. 8; his statement that, unless the intention to novate was express ('nisi ipsi specialiter remiserint quidem priorem obligationem et hoc expresserint, quod secundam magis pro anterioribus elegerint' Cod. loc. cit.), the two obligations should subsist side by side is apparently subject to the qualification that when one was fulfilled the other was ipso facto extinguished; see Dig. 46. 2. 8. 5.

§ 4. When the res was no longer integra, an agreement between the parties to be off their bargain did not merely extinguish the obligation: it rather operated as a new contract which bound the one in whose favour performance had taken place to restore the other in statum quo, but which was unable to injuriously affect rights acquired under the

pretio soluto neque fundo tradito, placuerit inter eos, ut discederetur ab emptione et venditione, invicem liberantur. idem

previous contract by third persons: '[re secuta] non tam hoc agitur, ut a pristino negotio discedamus, quam ut novae obligationes constituantur' Dig. 2. 14. 58.

Among the modes quibus obligatio tollitur' described by Gaius is the commencement of an action, litis contestatio, which, if the action were a iudicium legitimum, and the formula was in ius concepta, produced a quasi-novative effect, termed by the commentators novatio necessaria; the very delivery of the formula in the action by the praetor to the iudex extinguished the defendant's debt, and substituted for it a new obligation, viz. the legal liability to be condemned if the plaintiff proved his case, Gaius iii. 180, 181: iv. 107. It differed, however, in its operation from novatio proper (novatio voluntaria), for it left the original obligation subsisting naturaliter, Dig. 12. 6. 60. pr., and, as a consequence, did not destroy accessory rights, such as guaranties, hypothecs, etc., Dig. 46. 2. 29. Of this process-consumption, as it is called, there are still traces in the Corpus iuris, but the general rule under Justinian is that litis contestatio no longer extinguishes the creditor's right, Bk. iv. 13. 10 inf., Cod. 3. 1. 13. 2 and 5; 3. 10. 1. pr. It must indeed have ceased to extinguish it ipso iure with the disappearance of iudicia legitima under Diocletian, A. D. 294; but now indeed it no longer enables it to be counteracted ope exceptionis; an obligatio is destroyed, not by the bringing of an action, but only by its adjudication, so that we cease to read of the exceptio rei in iudicium deductae, which is swallowed up in the exceptio rei iudicatae, Cod. 8. 41. 28: see on Bk. iv. 13. 5 inf. Novatio necessaria is also said by Gaius to be produced by judgment, sententia, res iudicata; for its effects, which do not belong here, see Poste's Gaius, p. 421.

Among modes of extinction operating ipso iure, which are not here noticed by Justinian, are physical impossibility of performance arising ex post facto without default of the debtor, Dig. 46. 3. 92; ib. 98. 8; ib. 107: cf. note on Tit. 19. 1 supr.; and in some cases death of one of the parties to the contract, as in societas (Tit. 25. 5 supr.) and mandatum, (Tit. 26. 10 supr.): cf. Gaius iii. 120, Dig. 4. 8. 32. 3, and Bk. iv. 12. I inf., which is important for a large class of obligations arising ex delicto. The operation of compensatio (set-off) is a matter of some little difficulty: see on Bk. iv. 6. 30 inf. Confusio (p. 290 supr.) operated in this way if the deceased's heir was the sole debtor or sole creditor to the obligation in question, because one of the prime requirements of an obligatio, two persons, is no longer satisfied, Dig. 46. 3. 95. 2; but this would be exemplified under Justinian only when the heres, being debtor or creditor of the deceased, did not make an inventory, see p. 292 supr. Where the heir was one of two or more correal or solidary debtors or creditors of the deceased it was otherwise: see Dig. 46. 1. 71. pr. cited in Excursus VII. inf.

est et in conductione et locatione et omnibus contractibus, qui ex consensu descendunt, sicut iam dictum est.

Of modes of invalidation whose effect is produced only ope exceptionis the most common are limitation (Bk. iv. 12. pr. and notes, inf.), capitis deminutio (note on Tit. 10. 3 supr.), beneficium competentiae arising upon a cessio bonorum (p. 395 supr.), and waiver or pactum de non petendo (Bk. iv. 13. 3 inf.) if absolute, i. e. not binding for a time only, or conferring rights upon the debtor only and not upon his heir. An informal acceptilatio was construed as a pactum de non petendo, if the creditor's intention was really to release the debtor, Dig. 2. 14. 27. 9. The obligations involved in the actiones furti and iniuriarum were dissolved ipso iure by agreement not to sue, Dig. 2. 14. 17. 1. Transactio or compromise, by which two parties who affirm that they each have claims against the other mutually surrender somewhat of their alleged rights in order to remove uncertainty and narrow the issues, implied a pactum de non petendo, Cod. 2. 4. 17; ib. 24. The same result ensued from compromissum, an agreement to refer a dispute to arbitration, Dig. 4. 8. 13. 1.

EXCURSUS IV.

ORIGIN AND DEVELOPMENT OF BONORUM POSSESSIO.

SERIOUS Speculation on this subject commenced about a century since with Hugo, who held that bonorum possessio was in origin the system applied in the succession to aliens by the praetor peregrinus, from whose edict it was gradually transferred to that of his urban colleague. But this, though at first sight an attractive hypothesis, seems untenable on account of the tenderness which the praetor almost ostentatiously showed for the civil law and the agnatic conception of kinship: a gentile system of inheritance would more probably have been based on cognation only. Niebuhr connected bonorum possessio with the possession of ager publicus (p. 343 supr.), with the succession to which alone he thought it was originally concerned; but his view is inconsistent with the very name bonorum possessio, and has practically nothing in its favour. A third theory, which originated in 1837 with Fabricius, finds the germ of the institution in the judicial regulation of Possession as preliminary to an hereditatis petitio, which was tried in the centumviral court by the procedure of sacramentum, and of which a prominent feature was the award of possession pending the proceedings (Gaius iv. 16, 17). Such an award would be a provisional determination of the right of inheritance, and it is suggested that it was so often accepted by the other litigant as substantial justice that the possession came to be more than interim possession, and tended more and more to be regarded as an independent and impregnable interest. Such bonorum possessio was iuris civilis adiuvandi gratia; but as the praetor's quasi-legislative activity extended itself the forms iuris civilis supplendi and corrigendi gratia were gradually added. Savigny and Huschke connect the praetorian scheme of succession in origin with the old usucapio pro herede (Gaius ii. 52-56), of which, according to them, the praetor laid hold as the starting-point of his innovations: he gave the possession to certain persons who (it was

admitted) had an equitable, though not a legal, right to the estate, and devised, in the interdict quorum bonorum, a remedy by which they could recover the deceased's property from other possessors, thus placing them in a position to become owners by usucapion; the practical assimilation of the bonorum possessor to the heir, and the introduction of the possessoria hereditatis petitio, being due to the identification of bona fide possession in most respects with bonitarian ownership. But this hypothesis fails to give any explanation of the bonorum possessio iuris civilis adjuvandi gratia, which is on all hands regarded as the earliest form of the praetor's intervention, and seems in other ways to harmonise ill with the general history of the Roman law of inheritance.

It would seem that more light is thrown on the solution of the problem by a consideration of the oldest evidence which we possess of the law regulating the devolution of a deceased man's property on intestacy. The Twelve Tables enacted, 'Si intestato moritur, cui suus heres nec escit, adgnatus proximus familiam habeto. Si adgnatus nec escit, gentiles familiam habento.' Here the term 'heres' is confined to the deceased man's agnatic descendant: he merely assumes on the ancestor's death the actual control of property of which his ownership was previously in a state of suspended animation (note, p. 355 supr.). The law does not give him the inheritance; it recognises it to be his already. On the other hand, the nearest agnate and the gentiles are not called 'heirs' at all: all that is said is, in the absence of sui heredes, 'familiam habento': they acquire nothing ipso jure, but must take possession in order to make their title good, whereas it is clear that the suus heres had possession as heir, without the necessity of any act on his part. It is a familiar rule that extranei heredes acquired the inheritance only by acceptance, and there is no doubt whatever that under the earlier law such acceptance was required in all cases to be formal and solemn (cretio); the cretiones with which readers of Gaius are acquainted seem to have been inserted in the will only for the purpose of compelling the instituted heir to accept within a reasonable interval 2. The formula of acceptance was 'adeo cernoque': it had to be uttered before witnesses in the domicile of the deceased, and

1 Hence when there was a suus heres in existence usucapio lucrativa was ipso iure excluded; the inheritance was already possessed.

2 Thus Ulpian (Reg. 22. 27) defines cretio as 'certorum dierum spatium quod datur instituto heredi ad deliberandum, utrum expediat ei adire hereditatem necne.' 3 Voigt, XII Tafeln, ii. p. 372, note 12.