CHAPTER XXIV. MANUMISSION UNDER JUSTINIAN1.

Many of Justinian’s changes, not directly concerned with the law of manumission, had, indirectly, great effect upon it. It may be as well to enumerate the chief of these changes before stating the law systematically.

The rules in his time may be stated thus.

A. Form. Census is gone. Vindicta remains, having long since ceased to be, if it ever was in any reasonable sense, a judicial process. Manumission in ecclesiis still continues⁷. Manumission by. will of course still remains. The general effect of legislation of the later Empire having been to abolish the praetorian will, the question whether freedom can be given by it is obsolete. The place in the will is now immaterial⁸. Implied gifts are more freely recognised. Whatever may have been the earlier law it is now clear that appoint­ment of servus proprius, as tutor, implies a direct gift of liberty⁹. The rule is subject to some obvious restrictions. Thus, as it turns on an implication of intent, the rule does not apply where the facts negative this intent, e.g. where the testator thought the slave free¹⁰. And where the slave is appointed cum liber exit, the appointment is a mere nullity. Conversely the appointment of servus alienus without such words is a mere nullity¹¹, though there is one text which seems to say that the condition will be implied and even that such a gift amounts to a fidei- commissary gift of liberty¹².

1 Many rules being common to Justinian and the classical law, convenience has decided the question which should be discussed here and in earlier chapters. The same consideration accounts for some repetition.

2 C. 7. 25. ⁸ In. 3. 12.

⁴ C. 7. 5, 6; In. 1. 5. 3. He remarks that dediticii have disappeared already in practice.

⁶ C. 7. 15. 2. ⁸ C. 7. 3; In. 1. 7. 7 Q. 7. 15. 2.

« In. 2. 20. 34.

⁸ Ante, p. 463. The chief text (26. 2. 32. 2) gives liberty at once but delays the tutela till the libertus is 25. This is due to Justinian, but how far the alteration goes is uncertain.

10 26. 2. 22. ¹¹ In. 1. 14. 1. ¹² 26. 2. 10. 4; ante, p. 463.

In the same way a gift of the hereditas to servus proprius implies, in Justinian’s time, a gift of liberty¹. Thus where a slave is instituted with a gift of liberty, and, in the same will, the gift of liberty is adeemed, this cannot be construed as taking away the institution, for it is a rule of law that a hereditas cannot be adeemed. The institution stands good, and this implies a gift of liberty, so that the slave takes the inheritance, with liberty’. Justinian bases his rules on a presumption of intention, but he is hardly logical, for although a legacy to a man cannot take effect unless he is free, he does not allow a gift of liberty to be inferred from a legacy®. The fact is that the inference from another gift is not accepted unless in addition to the benefit to the slave, and favor libertatis, there is also some other public interest to be protected. In the case of institution, there is intestacy to be avoided: in the case of tutela there is the interest of the ward.

An inference from a form of words not amounting to a formal express gift is on another footing. Here the intent must be absolutely clear. Thus the words, in libertate esse iussi, do not suffice*. And mere intent to free is not manumission. Thus where it was clear that certain slaves were destined to look after a temple about to be built, if they were not actually freed they were the property of the heres^{[1976] [1977] [1978] [1979] [1980].}

The refusal of classical law to make implications left room for many doubts, nor were these removed as a matter of course by the mere admission of such implications. Accordingly Justinian finds it necessary to settle a number of doubts by express provision. It is not to be expected that any clear principle shall be discovered in relation to these numerous specific decisions on points of detail and construction, but they must be set forth as illustrative of the manner and tendency of his changes. Where the institution was in a will and the liberty in a codicil, the ancients had doubted, since the institution would not have been good without the gift of liberty, and as this was in a codicil there was in effect a gift of the hereditas by codicil. Both are now to be good®. Where A made his child heres and freed a slave, and then made a pupillary substitution in favour of the slave without any gift of liberty, the ancients had doubted, as the institution and the liberty were in different grades. Justinian declares that the slave is heres necessarius’’. A made two heredes: one was his slave, but had no gift of liberty. He then left the slave to a third party. The ancients doubted as to the result, whatever the order of the gifts. Justinian

554 Informal Manumission under Justinian [pt. n directs that the institution is to take effect notwithstanding the legacy^[MCMLXXXI]. If a slave was left as a legacy, and, later in the will, there was a pupillary substitution in his favour without a gift of liberty, the effect was doubtful. Justinian decides that the legacy is in suspense, till it is clear whether the substitution takes effect or not, which is the view that had formerly been held where the substitution had been accompanied by a gift of liberty’. A servus proprius is instituted pure and given liberty under a condition. If it is in his power and he fails to satisfy it, he loses both: if it is not in his power and is not satisfied, he is to be free nevertheless, but not to have the hereditas unless the estate is insolvent, in which case he is heres necessarius³.

Fideicommissary and direct gifts can now be made to unborn persons, so that they shall be bom free, and, if there are twins or more, all will be free⁶. The latter part of the rule does not look very rational: we should have expected the first born to be free“.

As to the informal modes, Justinian legislates elaborately. He enacts that some shall be valid with witnesses, others without, and the rest shall be void⁷. Those valid are to have the same effect as manu­mission vindicta and are legitimi modes. Those allowed with witnesses are:

1. Per epistolam, five witnesses writing their names on the letter, quasi ex imitatione codicillorum. If the slave is absent the letter makes him free only when he receives it⁸.

2. Inter amicos, also with five witnesses in imitation of a codicil. The act must be formally recorded by the master, and the slave must get the testimony signed by the five witnesses, and also by publica persona, i.e. a tabellio³.

3. Formally recording the slave as a son, apud acta, involving of course an official witness¹⁰. We have seen that in this and the other case now to be stated, latinity had resulted, though probably not in classical law“.

4. Giving to the slave, or destroying, in the presence of five witnesses, the papers evidencing his slavery¹².

сн. xxiv] Rules as to Age and Consent 555

Other informal modes allowed were¹;

5. If by order of the deceased or of the heres they stand around the funeral couch, or walk in the funeral procession, pileati, i.e., wearing the cap of liberty’.

6. If a slave woman is given in marriage to a freeman with a dos’.

In the other cases, apart from informal manumission, in which latinity had been conferred⁴, civitas was now to result.

B. Effect and Requirements.

I. All valid manumission makes the slave a civis, and a declaration that a slave freed is to be a latin is to have no effect⁵.

II. The master must be 20 years of age. To this rule there are, however, some exceptions.

(а) A slave may be made necessarius heres if the master is 14·.

(б) A minor may free vindicta before the consilium for a cause approved by that body. An impubes needs auctoritas tutoris⁷.

(c) Justinian allows it to be done by will at 17, as a man could* make a will for all other purposes at 14. This via media seems to be adopted rather hastily. It is not mentioned in the Digest or in the Code, and it is not long preserved, since in 549 it is provided that a man may free by will at 14⁸.

(d) The rule does not apply if the slave has been received inter vivos, from a competent person, on a condition to manumit⁹.

It will be observed that there is in Justinian’s law no limit of age in the case of the slave: he may be an infant, just as a mad slave may be freed, though a mad dominus cannot free¹⁰. ^[MCMLXXXII]

[pt. n

“natural.” It seems more in the Roman way and more in keeping with Justinian’s language to say that so mean a creature was not to be allowed to spurn Roman citizenship. The rule is not prominent in the texts and in the one in which it is clearly laid down, Justinian proceeds to state an important exception: in an addictio bonorum libertatis causa a slave may sometimes refuse the liberty¹.

IV. The manumission must be nominatim¹. The rule remains, though the chief point of it is destroyed by the repeal of the lex Fufia Caninia. The survival of the rule is the more remarkable, in that its existence is expressly set down by the classical writers to that lex*. It is somewhat confused with the rule that liberty cannot be given to an incerta persona, which is itself based on the lex Fufia⁴.

I C. 7. 2.15. 2a; post, p. 625. ² 40. 4. 24.

⁶ G. 2. 239; Ulp. 1. 25; P. 4. 14. 1. ¹ G. 2. 239; In. 2. 20. 25.

⁹ In. 2. 20- 25; C. 6. 48.1. According to the Sc. Orphitianum, it is enough that he is so described that his identity is clear, 40. 4. 24, ante, p. 460.

6 40. 4. 21, 54.j)r. 7 34. 5. 28—30; 40. 4. 37.

8 Fr. Vat. 227. The manumission may be in the will and the description in a codicil or vice versa, 40. 4. 37.

9 C. 6. 38. 4.

w Bernstein, Z. S. S. 4. 177sqq. He is discussing, generally, various types of alternative gift and obligation.

II C. 7. 4. 16; post, p. 610. ¹³ 40. 5. 22. 1, 46. 5.

choice, and entitle a particular man to his freedom¹. An analogous case is that of a fideicommissum, charged on a beneficiary, to free a certain number of slaves. What is the effect where the gift is less than the value of all the slaves? Ulpian^{[1983] [1984] appears to decide that as many of them must be freed as the amount will cover and then asks the question —which ? The answer is rather in the manner of Tribonian. There is nothing to indicate any right of choice in anyone. But the text says that the order in the will should determine. If this will not serve, the matter is to be determined by lot, or by an arbitrium, on their merits. It is not to be left to the Praetor, lest suspicion rest on him, of ambitio vel gratia. It is difficult to believe that this is Ulpian. Indeed it would seem that in classical law the principles which we have already discussed would require that he should be compelled to free them all, even though they are worth more than the gift[1985].}

Such a gift may be made in favour of an unborn person^[1986]. Here too Justinian changed the law. The earlier law is not absolutely clear, but on the whole, Paul’s text is very strongly in favour of the possibility of such gifts, by way of fideicommissum^[1987], and Justinian elsewhere^[1988] states a rule in terms which assume, as a matter of course, that such gifts were possible in early law. The doubt which Justinian suggests in his enactment^[1989] dealing with the matter is perhaps not as to whether such a gift could be made, but whether, if it were made, the child was born free. He enacts that both direct and fideicommissary gifts may be made in favour of unborn persons, at least if conceived, and this whether the mother be given freedom or not. The effect will be, he says, that they will be born free. It is difficult to apply this to fideicommissary gifts, since these require an act of manumission, but Justinian is not so logical that we can be quite sure that he troubled about this: it may be that he meant them to be free ipso facto in such a case, but to be liberti of the heres. Whether “ bom free ” is to be taken as meaning ingenui is not clear: it seems hardly probable. What he says is that cum libertate solem respiciat. The logical difficulties are obvious : they are discussed elaborately but not to much purpose by the early com­mentators^[1990]. It is in no way inevitable that Justinian should have considered these words as meaning ingenuitas, though they cannot in strictness mean anything else.

V. Manumission must be by the owner. We have already¹ con­sidered this rule. No great change of principle seems to have occurred under Justinian, but the rules require some further discussion and illustration.

We have seen that an ownership liable to determine still entitles its holder to free. Thus a heres under a trust to hand over the hereditas can free before doing so, being liable for the value of the slave whether he knew of the trust or not^{[1991] [1992] [1993]. The rule is no doubt classical, but the text has been mutilated by the compilers, who speak of the validity of the liberty as a case of favor libertatis. There could be no meaning in this for classical law: the case is on all fours with that of one who has agreed to sell the hereditas and who frees a slave before handing it over. He is owner and the gift is good, but nothing is said about favor libertatis*. But Justinian’s fusion of legacy and fdeicommissum, and his provision that a real action was always available, led to a good deal of confusion[1994] [1995] [1996].}

There must, however, be real ownership. To free another man’s slave is a nullity: in some cases it is penalised®. We are told, however, that where A manumit B’s slave, B can have, if he likes, the value of the slave instead of the man, and this is said to have been saepe rescriptum*. So far as this represents classical law it presumably means no more than that the old owner could, if he preferred, treat the matter as a sale: it does not imply that the manumission was good or would be validated by the quasi-purchase. But it is not unlikely, judging by other texts, that this is what it means for Justinian. Thus while we are told that manumission of a servus alienus was not confirmed by subsequent inheritance of the slave’, we are told elsewhere that if X is directed to free a servus hereditarius and so be heres, and he does manumit, nihil agit, so far as freeing is concerned, yet he has manumitted and thus satisfied the condition. And the text adds: post aditionem manumissio...convalescit^[1997]. This last clause must be an addition by Tribonian.

If a man freed a slave of his own by vindicta thinking he was alienus, or if the slave or both were under the mistake, the manu­mission was good’. The text gives as one of the reasons that, after all, he is free voluntate domini, which is hardly the case if the master thought his act was a nullity.

In one text we are told that a coheres rogatus manumittere, can, by a rescript of Pius, free the slave even before partition, (when he cannot be sole owner,) where the other heres is an impubes non rogatus. But this is part of the law as to the enforcement of fideicommissa not carried out^[1998].

Other exceptional cases may be noted. If a man is free at the time when a will is made, a fideicommissary gift of liberty to him is good, if he is a slave at the time of the death or of the satisfaction of any condition². But this is not a real exception to any rule laid down, for he will be the property of the person who actually frees him. It is only noticeable in that it is a case in which it is permissible, favore libertatis, as it seems, to contemplate supervening slavery³.

Where A gave a slave to his wife mortis causa, (which was valid,) and instituted the same slave with liberty, if the institution came last and was intended as a revocation of the gift, the slave was a necessarius heres. If it was not so intended, or the gift came last, the gift prevailed, and the wife got the hereditas through the slave⁴.

The exceptional cases in which Justinian allows effective manu­mission by a person not dominus do not represent any change of principle. They are no more than attempts to do equity, in a particular case, without any thought of the relation of the decision to ordinary legal rule. It still remains true that manumission cannot be by agent. But, whatever may have been the law before, it is now clear that manumission can be carried out by a filiusfamilias on behalf of his paterfamilias by any method inter vivos⁶. An extension of this prin­ciple is due to Justinian. In 530 he provides³ that ascendants of either sex might authorise their descendant of either sex, whether in potestas or not, to free on their behalf—a rule which follows, as he says, the general breakdown of the ol 1 narrow conception of the family. It is noticeable that, notwithstanding its date, there is no trace of this extension in the Digest, published three years later.

As a part of his rearrangement of the rights of the father in acquisi­tions of the son he provides⁷ that if a slave is given ab extraneo to a filiusfamilias to free, he may do so, the father’s usufruct established by Justinian being disregarded, as unreal, in such a case, where the whole ownership is merely formal, since there is a duty to free at once⁸.

VI. Manumission in fraud of creditors is void. Nothing is said by Justinian in this connexion about fraud on the patron². The reason

560 Manumission in Fraud of Creditors [pt. n for the disappearance of this from the rule is not clear. No doubt Junian latins were abolished, and it was in their case that the rule was most important, since all their property went to their patron when they died. But this does not account for the omission : it would require all the rules as to fraud of patron to disappear, which they do not^[MCMXCIX]. It has been suggested that the omission is linked with the general rearrange­ment of patronal rights². But if Justinian had intended a definite change in the law he would probably have said something about it. It has also been suggested that the matter is sufficiently provided for by the rules as to revocation of acts done in fraudem patroni*: we are told that omne in fraudem patroni gestum revocatur*. And elsewhere Justinian tells us that when alienation is inhibited by a lex or other agency the words cover manumission⁵. But the title in the Digest³, though of some length, never mentions manumission, and the application of the above text to it conflicts with the important rule, shortly to be considered, that liberty was irrevocable. The lex Aelia Sentia makes the manumission void ab initio, and the distinction is clearly recognised in matter of alienation⁶. Some modern commentators appear to ignore the reference to’the patron in connexion with this provision of the lex. But Gaius and Ulpian are quite explicit’.

The general principle is that the manumission, to be void, must have been intentionally fraudulent⁸. Thus where an insolvent gave liberties “ if my debts are paid ” there was a general agreement among jurists, though Julian doubted, that this could not be fraudulent*. In another text in which the same rule is laid down, Julian seems to have no doubt¹⁰, but probably his conformity is due to Tribonian. It has been suggested^[MM] that Julian was inclined to hold the gift void on grounds independent of the lex Aelia Sentia, as not having been seriously meant. But the gift obviously was seriously meant, and the whole structure of the text brings Julian’s view into connexion with the lex. It is true that the latter part of the text expressly negatives fraud, but this, again, does not look like a part of the original text¹³.

According to the Institutes the gift is fraudulent if the owner is insolvent and knows it, or knows that he will become so by the manu­mission, though the text hints at an abandoned view that the fact was ch. xxiv] Manumission in Fraud of Creditors 561 enough without knowledge of it¹. The source of the text is a passage from Gaius, which is cited in the Digest, where, however, nothing is said about knowledge². It has been suggested that Gaius did not require this, but it is at least possible that he is citing an older view only to reject it, the doctrine of the Institutes being his own* as it certainly was the Sabinian⁴. However this may be, it is clear for later law that consilium fraudis was needed. Thus we are told that if a son frees volente patre, the gift is void if either knows of the insolvency®. The rule applies only to an ordinary voluntary manumission. Thus it has no application where the manumission is of a slave received ut manu- mittaiur⁶: such a person would be free without manumission if the direction were not carried out’. So also where the manumission is under a fideicommissum*, or is in return for money⁹.

If it is given fraudandi animo in a codicil, it is bad though at the date of a previous confirmatory will the testator was solvent¹⁹, but if he was solvent when he made the codicil, the fact that he had been insolvent at the date of the w'ill was immaterial.

Besides intent, there must be actual damnum to the creditor— eventus as well as consilium¹¹. Thus insolvency at aditio might destroy a gift designed in fraud, but solvency at aditio would always save it¹². Two cases raise some difficulty here. It was possible for a heres who doubted the solvency of an estate, and yet wished to save the fame of the deceased, to agree with the creditors, before entering, that they should accept a composition, and it was provided, apparently by Marcus Aurelius and Antoninus Pius, that if a majority of the creditors agreed, the composition could be confirmed by magisterial decree, and thus forced on the other creditors¹⁹. Scaevola in two texts¹⁴ discusses the question whether under such circumstances manumissions in the will are valid. It is clear that legacies are not unless the estate shews a profit to the heres. But he lays it down that liberties are valid unless they were given in fraudem creditorum. It is not clear that there was any eventus damni, since the creditors when they made their agreement knew of these liberties. The point is, however, that the heres could offer more if these slaves were assets.

Another noteworthy case is that of solvency of the heres. Some jurists held that this would save the liberties, but the view which pre-

¹ In. 1. 6. 23. If more than one were freed the gift might be void only to the extent of the excess, taken in order unless differences of value made it more favourable to liberty to alter the order, 40. 9. 25.

² 40. 9. 10. ⁹ A. Faber, Jurisp. Pap. Sci. 199.

⁴ 40. 4. 57. » 40. 9. 16. 5.

⁶ 40.1. 10; 49.14. 45. 3; Fr. de i. Fisci, 19.

1 Post, p. 628. ⁸ 28. 5. 56; Fr. de i. Fisci, 19.

⁹ C. 7. 11. 5. As to fideicommissary gifts,post, p. 565. The texts say nothing of collusion. 1« 40. 9. 7. pr. n C. 7. 11. 1.

“ 40. 9.18.pr. i· 2.14. 7.17—10. The texts give further details.

“ 40. 4. 54. 1; 42. 8. 23.

vailed was that it was immaterial¹. In a text which seems to say the contrary it is clear from the context that a non has dropped out². Another text declares to be governed by the same principle the case in which the liberty is conditional on the payment of money, and a third party is willing to pay it, so that the estate suffers no loss’. It has been suggested that the reason for this rule is to induce the heres to enter⁴. This is open to the objection that as the estate is rendered solvent by the entry, the creditors have no interest in getting the gift declared void, and the heres has, as we shall see shortly, no power to do so. Moreover the principle would not apply to the second case, in which the heres would not lose, as he gets an equivalent, and yet the principle to be applied, whatever it is, is common to both cases. It seems more probable that it merely represents a close adherence to the idea that the state of the actual hereditas and the intent of the testator are the only material things. But this makes the rule applied in the case of a gift, “ if my debts are paid,” all the more remarkable. It can only be justified on the ground that the use of this formula negatives fraudulent intent. But as we have just seen it might have been used as rather an ingenious way of injuring the creditors, and Julian’s doubt seems to be fully justified.

The rule applies to soldiers’ wills’ and, unlike other provisions of the lex, it applies to manumissions by peregrines⁶. Most, of the texts apply to direct gifts by will, obviously the commonest case, but the rule applied equally to gifts inter vivos⁷, and to those by way of fidei- commissum³.

A creditor for the purpose of these rules is anyone who has an action, or an inchoate right to sue, even though the debt be ex die or conditional, so that there is no present liability⁹. There is, however, one distinction to be noted : a claim on account of legacy or fideicommissum is a sufficient debt¹⁶. But if the debt is merely a conditional legacy or fideicommissum due from the manumitter, this is not enough¹¹, probably because, as there has been no negotium or other legal act between them, the legatee or fideicommissarius is not a creditor till the condition arises¹². If the debt which makes the man insolvent is conditional, we are told that the slave is a quasi statuliber, pending the arrival or failure of the condition¹³. Elsewhere we are told that in the case of an absolute debt the slave is a statuliber till it is certain whether the creditor will

¹ 40. 4. 57; C. 7. 2. 5. ² 40. 9. 5. pr. _

⁸ 40. 9.18.1. Sed si heres locuples non projicit ad libertates nec qui dat pecuniam prodesse potest. The case in C. 7.11. 5 is different: there is there no damnum and no animus.

< Accarias, Precis, § 71. ⁶ 40. 9. 8. 1.

⁶ G. 1. 47. A sc. under Hadrian. ⁷ 40. 9. 5. 2.

⁸ C. 7.11. 7. Roby considers on the authority of 28. 5. 84.1 that this rests on a sc. under Hadrian.

⁹ 40. 9. 8. pr., 16. 2, 27. pr. ¹⁰ 40. 9. 27. pr.', C. 7. 11. 1; i.e. due from the testator,

n 40. 9. 27. pr. 12 Cp. 44. 7. 5. 2. 40. 9. 16. 4.

ch. xxiv] Manumission in Fraud of Creditors 563 use his right¹. This suggests that the more accurate way in which to state the law is that the gift is bad if there is animus, if there is damage to the creditor, and if the latter takes steps’.

The creditor may be an individual civis, a corporation, or the Fisc, animus fraudandi being as necessary here as elsewhere’. The Fisc does not seem to have had any privilege in the matter in classical law⁴, and as civitates and the Fisc are not creditors in the strict sense of the lex, it seems that special enactments were necessary to bring them within the lex^s.

If the master was insolvent at the time of manumission, and after­wards pays off all his creditors, new creditors cannot attack the gift, since there was no intention to defraud them. Julian is quite clear that the animus and the eventus must apply to the same creditor’. The following text adds as a note of Paul on Papinian that this does not apply if there is proof that the money to pay off the old creditors was derived from the new⁷. In another text however, from Papinian, and apparently from the same book⁸, a general rule is laid down that new creditors can attack the gift. It is possible that this was Papinian’s view, corrected elsewhere by Paul. In any case it is clear that the rule of later law is otherwise : proof must be forthcoming that money of the second creditor has been used to pay the first. It has been suggested* that the texts may be harmonised by supposing that Papinian is dealing with a case in which there is intent to defraud future creditors as well, while in Julian’s case the intent is to defraud the present creditor. The texts shew no trace of any such distinction. And when we remember that Julian, in speaking of fraudulent intent¹⁰, speaks of it merely as knowledge of insolvency, it is difficult to resist the impression that the determination of the animus to the one creditor is considered by him to result from the fact that he is the only creditor, and not from any mental act of the manumitter. It is difficult to see indeed how his intent could be made out. On the whole it seems more probable that Papinian’s text is a little too widely expressed¹¹.

We have seen that the lex makes the manumission absolutely null¹’. If it is set aside the man never was free and thus he is fairly called a statuliber in the intervening period¹⁸. There are, however, some texts > 40. 7.1. 1...

⁸ As this will occur if at all after aditio, the gift is in effect conditional. See A. Faber, op. cit. 218.

⁸ 40.9.11; 49.14.45.3 ; 0.7.11.5. < As to 40. 9.16. 3; see past, p. 564.

⁶ 40. 9.11. The nullity was dependent on certain steps being taken: the Fisc itself could not take these steps.

⁶ 42. 8.15. ⁷ h. L 16; cp. h. t.10. 2. ⁸ 40. 9. 25.

⁹ A. Faber, op. cit. 202. Bodemeyer, op. cit. 23, 24., ¹⁰ 42. 8. 15.

¹¹ A having two slaves, S and P, and no other property, promises “ S aut P.” Julian says the lex prevents his freeing either. If he frees one the other may die. Scaevola confines this to the case of his having no other property, as otherwise anyone who had promised “a slave of mine ” could free none at all, 40. 9. 5.2, 6.

>^a See the references on p. 544, ante. ¹⁸ 40. 7. 1.1.

which look as if the gift were only revocable. Thus we are told that the Fisc can revocare in servitutem¹. But this is correct, for in the meantime the man has been or may have been in libertate. Other texts in which the creditor is the Fisc say retrahi placuit*. Others having nothing to do with the Fisc use similar language⁸. These can hardly be more than mere loosenesses of language. The view that the case of the Fisc is on a special footing in this matter⁴ is negatived by the fact that many of these texts do not refer to the Fisc, and on the other hand there are texts dealing with the Fisc which declare the gift absolutely void'. The view that the case of the Fisc and of civitates was regulated on this point by special enactments rests on little evidence: there is no reason to suppose that the constitutions’ and senatusconsulta did more than declare the lex to apply. It is highly improbable that the Fisc would be placed in an inferior position, or that a revocable liberty would be casually introduced in this way. On the whole we must assume that in all these cases the manumission is either void or valid.

Apart from some special provision it would seem that the nullity of such a gift ought to be capable of being pointed out at any time. There were such provisions. Liberty begun in good faith was protected after a lapse of time which varied from time to time and will be considered later’, while if it began in bad faith it was not protected at all⁸. Similarly the status of one apparently free was not to be disputed five years after his death, though it might be up to that time⁹. These rules seem to account sufficiently for the text which tells us that the slaves we are dealing with were statidiberi dum incertum est an creditor iure suo utatur^w. However other explanations have been suggested. According to one view it must be within one year from the sale of the goods, this being the time within which fraudulent alienations must be revoked¹¹. But there seems no ground for assimilating the void to the voidable in this way, and there would be difficulty in applying the rule to the case, which might occur, of nullity on this ground where the goods were never sold at all¹⁸. Others take the view that he had ten years, arguing from a text of Paul which says that a slave freed in fraudem fisci is not to be recalled into slavery si diu in libertate fuisset, id est non minus decennio¹³. But neither Paul nor Aristo whom

^I 40. 9.16. 3; 49. 14. SO. ⁸ 49. 14. 45. 3; Fr. de i. Fisci, 19.

⁸ “ Revocabitur quemadmodum si in fraudem manumisisset” (42. 8. 6. 5; 5. 2. 8. 17); “per legem Aeliam rescinditur ” (42. 8. 15; 40. 9. 5. 2); “ libertates in fraudem creditorum revocari ” (C. 7.11.1); “si fraudem se feeisse areditoribus ut revocet libertates" (C. 7. 8. 5). Rescinds is less significant than the others. See A. Faber, op. eit. 204.

�? Ante, p. 563. ⁸ 40. 9. 11. 1; C. 7. 11. 5.

⁸ 40. 9. 11. 7 Post, p. 648. ⁸ C. 7. 22. 1.

⁹ C. 7. 21. pass.; post, p. 651. The protection did not apply to those in fuga at latitantes, C. 7. 21. 8.

w 40. 7. 1. 1.

^II 42. 8. 6.14; see Accarias, Precis, § 71.

H 40. 4. 57. ⁸⁸ 40. 9. 16. 3.

ch. xxiv] Manumission in Fraud of Creditors 565 he is quoting could have laid down this positive rule of time: it is clear that it is the work of the compilers, or at any rate of some later hand. If that is so it seems to establish a privilege of the Fisc, for there is strong reason to think that a private person could not attack the status of a person apparently free, after the lapse of five years from a traceable prima facie valid manumission by his dominus¹.

There remains one point: the manumission being void, who is entitled to have the nullity declared ? Clearly the creditors can, and the language of one text seems to shew that they alone can². There is no question of a popularis actio. As the slave was in possession of liberty, the proceeding would take the form of some sort of claim to him. If no heres enters, so that the bona are sold, the creditors can of course make a claim. If the heres enters he will clearly have a sufficient interest. But though as we have seen the question may be raised although there is no sale’, it is clearly laid down in three texts that the heres of the manumitter is barred from bringing proceedings⁴. Thus it is difficult to say what happens if a solvent heres enters on an insolvent estate. The heres, the only person interested, has no locus standi. The creditors, secure of payment, will hardly move. It has been suggested’ that the heres himself can proceed directly in this case but the contrary texts seem too strong. If the creditors have a claim it is possible that they may transfer their right to the heres, e.g., by authorising him to pro­ceed as procurator in rem suam, but even this seems barred by the wide language of the text. Probably the proceeding is a special one organised under the lex, and not an ordinary vindicatio in servitutem, so that the creditor’s right does not depend on any claim to the slave but on the mere fact that he is a creditor. If that is so, he need not wait for a decree of missio in possessionem, and he would not be barred by the mere fact of entry of a heres, so that the heres might enter only on the undertaking of the creditor to proceed. The texts seem to mean that if he did not take this precaution he would have no remedy.

It may be added that there is no reason to suppose all creditors need join, and that the manumitter could not himself impeach his own manumission inter vivos⁶.

There is some difficulty as to the application of the rule about fraudulent manumission in the case of fideicommissary gifts. The texts make it clear that, after some dispute, the rule was settled that in the case of such gifts the animus was not considered—the eventus alone determined whether the gift was valid or not⁷. What is the reason of this ? It must be remembered that it was only by adoption of the Sabinian view⁸ that the rule was reached which made animus

¹ Post, p. 650.

» 40. 7. 1. 1.

⁵ Accariaa, Precis, § 71.

⁸ 40. 4. 57.

» 40. 4. 57. « C. 7. 8. 5.

⁴ 40.12. 31; C. 7. 8. 5; 7.16. 7.

’ 40. 5. 4.19; C. 7.11. 1, 7.

566 Manumission must be in Perpetuity [pt. n material in direct gifts. The texts shew that the non-application of this to fideicommissary gifts was not a mere oversight, but was a positive decision. It is perhaps idle to speculate as to the unrecorded reasons which led to this view. It may perhaps be due to the notion that the testator in abstaining from completing the gift may be regarded as having tacitly subjected his direction to the condition of solvency of the estate. If that is the explanation the rule does not depend on the lex Aelia at all, and any such dependence cannot be made out on the texts.

VII. Manumission must be in perpetuity. Any limit of time which it was sought to fix was simply struck out^[MMI]. This idea of irrevocability, already mentioned, can be illustrated by many texts. Civitas once obtained cannot be added to or subtracted from by any subsequent manumission®. If a man frees a slave under a fideicommissum contained in a codicil which is afterwards shewn to be a falsum, or by way of fulfilling a condition which is afterwards shewn to be, from any cause, not binding, the manumission is still valid’. A slave is given to a legatee under a codicil which is declared false. There has been actual conveyance so that the legatee is certainly owner. If he has freed the slave the gift is good⁴. Even where the gift is such as to work a fraud on a third party, if validly given it is irrevocable. Thus where a son was under a fideicommissum to free a slave praelegated to him, when accounts had been rendered to the heredes, and he freed him before this was done the manumission was valid’. So when a heres under a fidei­commissum to hand over the hereditas frees a slave, the gift is valid’.

The case of restitutio in integrum is discussed in many texts. The general principle is that there is no help to a minor adversus libertatem’’. The following text says, “except ex magna causa on appeal to the Emperor.” What would be a sufficient magna causa does not appear, and it is likely that the words, which purport to be Paul’s, are misapplied by Tribonian⁸. Thus liberties which have taken effect, by the aditio of the minor, are not undone by his obtaining restitutio in integrum*. If the minor is under a fideicommissum to free and does so, and afterwards gets restitutio, the liberty is unaffected¹⁰. A slightly more complex case is that in which a slave is substituted to a minor. Here if the minor repudiates, the slave is necessarius heres and free. If now the minor is restitutus, the liberty, having taken effect, is not affected. But if the ch. xxiv] Manumission must be in Perpetuity 567 minor has accepted in the first instance and then obtained a decree of restitutio, Papinian thinks the slave will be neither heres nor free. But Ulpian disagrees, remarking that a rescript of Pius and one of Caracalla (?) have decided that on such facts the substitute is free and necessarius¹. The reason for Papinian’s difficulty is no doubt the rule semel heres semper heres: praetorian relief cannot alter that. No doubt the rescripts were necessary. It will be seen that at the end of the text Ulpian uses this principle to shew that in the first case the slave is still technically heres.

The principles are the same if the matter is wholly inter vivos. If a minor is led by the fraud of his slave, or of anyone else, to free him vindicta, with cause shewn, the manumission is good². If a minor sells a slave and the buyer frees him, and the minor is restitutus, the liberty holds’, even though the slave managed the affair in fraud⁴. If a minor over 20 sells ut manumittatur and the man has been freed, it cannot be undone, and if a minor acquires under this condition, he cannot be restitutus after he has freed’. We are further told that if in proceedings between the minor and a slave the latter has been declared free, there can be no restitutio in integrum, but only an appeal⁶. As between the parties the man must be presumed free.

The Querela inofficiosi testamenti raises similar points. In estimating the estate, for the purpose of the pars legitima, the value of freed slaves is first deducted⁷. If a will is void direct liberties are null⁸. As to fideicommissary gifts, if the slave belonged to the fiduciary, the gift, when once carried out, is good, as we have seen. If the slaves were in the hereditas direct gifts are good in the same way, if the testamentary heres enters for his share, though the will is upset by bonorum possessio contra tabulas, but not otherwise’. A will upset by the querela is not void, but voidable, a distinction which might have been thought material, but, as in other respects, the point is not logically treated, and the result is much as if the will were void. In a case of simple successful querela in which the will fails, the direct liberties fail¹⁰. But we are told that fideicommissary gifts must be carried out¹¹. The statement is directly quoted by Modestinus from Paul, and is thus probably genuine, but it is odd to find fideicommissary gifts treated as more binding than direct. The rule looks as if the will were pro tanto treated as a codicil binding on the successful litigant, since fideicommissary gifts might be in a

¹ 4. 4. 7. 10.

²If no cause were shewn it was simply void, C. 2. 30. 2, 3.

» 4.4.48. ⁴ 4. 3. 7.pr.

» 4. 4. 11.1. This has been retouched though it may represent classical law. Gradenwitz, Z. S. S. 23. 346.

⁶ C. 2. 30. 4. ¹ 5. 2. 8. 9. As for the Quarta Falcidia.

» 40. 4. 25. But see post, p. 609. » 37. 5. 8. 2.

¹⁰ 5. 2. 8. 16, 9, 28; unless the ins ti tut us also takes on intestacy, C. 6. 4. 26 b. n 5. 2. 9.

codicil while direct could not. But it is not easy to see why in that view other fideicommissa were not binding, and perhaps the text really means only that, if they have been carried out, they are good. This is confirmed by the further rule it contains, shortly to be considered, as to compensation, but it is not what the text says. If, as may, ex magna causa, be the case, the querela is allowed after five years, the manu­missions are good¹. This is in accordance with a principle of which we have seen and shall see other traces, i.e. that apparent liberty cannot ordinarily be disputed after five years from a traceable prima facie valid manumission by the dominus*. A constitution of Severus and Caracalla says that where a will has given fideicommissary liberty to slaves of the hereditas, and, there having been delay, a decree of the Praetor has ordered them to be carried out, and this has been done, the liberties are to remain good even though the will is upset by the querela brought by a son⁸. This is in accord with what has already been said, but there is a point in which it takes the rule a little further. It enacts that the validity of the fideicommissary gifts carried out shall not depend on their having been given by an undisputed owner: it is equally so if they were slaves of the hereditas. It might have been thought that these would be treated like direct gifts⁴.

If, as may happen, the will stands partly good, after the querela, all the liberties both direct and fideicommissary stand good’.

In many of these cases in which the principle of irrevocability causes liberties to be good, under circumstances which create injustice, there is an obligation to give compensation, but any general rule is not easily made out. In the case of liberty given under a false codicil Hadrian provided that the libertus must pay 20 solidi to the owner who lost him by the manumission⁶. In one text it is said, apparently in error, that the slave’s value is to be paid⁷. How far the rule could be extended to analogous cases was, it seems, disputed. Papinian says, in his Quaestiones, that constitutum est that the same rule must be applied where one frees under a condition in an institution, and the >5.2.8.17. ³ Post, p. 650. »0.3.28.4.

⁴ An imperfect admission of the principle that the ownership of the heres was good for the time being. Cp. 4. 4. 31; 37. 5. 8. 2. It may be that in C. 3. 28. 4 the institutes was also a filius.

⁵ 31. 76. pr.; 37. 7. 6; 44. 2. 29con­ditione’’. Similarly unreal is the case of a manumission void because the manumitter was vi coactus“. So, too, new enslavement for ingrati­tude is not an exception’: freedom does not involve incapacity ever to become a slave.

But though manumission could not be in diem, it might, as we have seen, be conditional or ex die¹. The authority on conditions is confined almost entirely to conditions on direct manumissions. In strictness it appears that only in such a case did the status of statuliber arise^{[2011] [2012] [2013], but, from the very few texts that mention the matter, it may be inferred that similar principles applied to fideicommissary gifts of liberty subject to conditions’. We must remember that even if there is no condition, a slave in whose favour such a gift has been made is in loco statuliberi*, and this is not altered by the presence of a condition: in such a case, cum sua causa alienetur1. Thus where there is a conditional fideicommissum of liberty, anyone to whom the man is conveyed must give security to restore him for the purpose of the manumission when it becomes due, nam in omnibus fere causis fideicommissas libertates pro directo datis habendas*. Thus there is nothing to add as to them in the matter of conditions. With regard to manumissions inter vivos there is more difficulty. We have already considered whether and if so how far conditions could be imposed on manumission per vindictam'’. As to informal manumissions there is nothing in their nature to exclude either tacit or express conditions, and the later part of a text already considered8 seems to say that they might be made mortis causa, and revocably, in the sense that they were not to take effect unless the expected death occurred. In such a case no doubt alienation would revoke the gift as it did a donatio mortis causa, and the slave could in all probability be usucapted. It is not easy to say what the law was as to an ordinary dies or condition. One text vaguely suggests that a slave freed ex die was at any rate to a certain extent in the position of a statuliber*. But it is not dealing exactly with such a case, but with one transferred, ut post tempus manu­mittatur, which is a different thing, and it expressly adds that these persons are not, in all respects, like statuliberi. In one text it is said that if the father of a woman accused of adultery manumits, by will, a slave in her service, before the 60 days have expired, the man is a statuliber, i.e. the gift will fail if the testator dies before the days are up10. In a neighbouring text we are told that if the woman manumits inter vivos, within the time, the gift is void—which seems to imply that there is no power of suspension11. But this was probably written of manumission vindicta. It is true that the most authoritative definition of statuliberi is in terms which cover a manumissus sub conditione}

572 Manumission: Conditions [pt.n,ch.xxiv

inter vivos^[MMXIV], but, on the other hand, the language of Festus⁸ and the whole drift of the title on statuliberi seem to ignore this case. The absence of texts makes it impossible to say what the law really was. If such things occurred, no doubt the slave was still a slave, but he was probably not a statuliber, and would not carry his status with him. It is likely that an alienation would be regarded as annulling the intent to free, which had not yet operated, and probably also the slave could be usucapted. But all this is obscure, and perhaps the right inference from the silence of the texts is that such things did not occur. There could be little use in them. On the other hand there were obvious advantages about manumissions mortis causa, and conditions on such gifts were reasonable enough.