CHAPTER XXIII. MANUMISSION DURING THE EMPIRE {cont.). STATUTORY CHANGES. LI. IUNIA, AELIA SENTIA, FUFIA CANINIA.

Of these three statutes the first mentioned, perhaps the last in date, was essentially different in object from the others. It enlarged existing rights: they were restrictive. For this reason, and because some of the provisions of the lex Aelia Sentia seem to presuppose the lex lunia, it is well to deal with this law first.

Lex Iunia.

This statute defined the position of those who had been in libertate tuitions praetoris by the earlier law¹. It made them latins, giving them broadly the position of colonary latins, subject to certain dis­abilities of a very serious kind. Because of these restrictions they were called Latini luniani to mark them off from the others^{[1913] [1914]. The cases with which it dealt were, apparently, the slave freed by his bonitary owner[1915], the slave informally freed[1916] [1917], and the slave freed under 30[1918], though as to this case we shall see that there is doubt as to what is due to this lex and what to the lex Aelia Sentia6. Most of the points of difficulty under this lex will be more conveniently discussed later: here it is enough to mention a few points.}

Notwithstanding the language of Gaius^[1919] it is clear that a bonitary owner could give freedom by will^{[1920] [1921]. It is hardly so clear whether he could do it vindicta2. And it seems that manumission censu must have given civitas or nothing10. Apparently the entry of the man’s name must have been a nullity, of no more force than any other mistake of the Censor’s11. And it does not seem that it amounted in itself to a manumission inter amicos or per epistolam11.}

Only such a slave was protected and thus became a latin as was talis ut praetor libertatem tueatur^[1922].

Most, probably all, of the other cases of latinity we shall have to consider are of later origin. This type of status, having once been invented, had new groups added to it from time to time, by an economy of invention to which the Romans were prone. Just as the rules as to dediticii were made to apply to cases quite different from that for which they were invented, and Junian latins themselves are an extension of the idea of latinity, so there come to be latins under like rules who have nothing to do with the lex lunia.

There are cases of inferiority in manumission which it does not in any way affect. Thus a peregrine owner could not give the slave in any case a better status than that he had himself⁴. He could it seems use only informal methods. And it may be supposed that any latin owner might use the method per vindictam, and any colonary latin that by will. But we are without information⁸.

The only other topic to consider in connexion with this lex is its date. It is always called lex lunia by the classical writers⁸, and usually even in Justinian’s time⁷, but in one passage of the Institutes it is called lex lunia Norbana⁸. No direct evidence as to date exists, but as the Fasti give consuls bearing the names lunius and Norbanus for A.D. 19, this has been commonly accepted as the correct date. The matter has been the subject of much controversy’, of which some state­ment is necessary, though the point is not important enough to justify a long account. The same names are not found again in any one year, but in 82 B.C. one of the consuls is called Norbanus. This date is impossible: Cicero, writing later¹⁰, enumerates the modes of manu­mission, and could hardly have failed to mention so important a law had it existed.

¹ Post) p. 542. ² Date however doubtful, see Willems, Droit Pub. Rom. 113.

8 Aug. 40. ⁴ See the texts cited in nn. 6, 7, 8 on p. 534.

⁶ G. 3. 56. ⁶ So Vangerow, Lat. lun., ±sqq.

^I Ulp. 3. 3. ⁸ G. 1. 29, 31, 66.

9 Ulp. 7. 4. Ulp. 1.12, sometimes cited on this side, seems rather to support the other view. See post) p. 536, n. 3.

w Fr. Dos. 7; Ulp. 1.10. So G. 1.167 and Ulp. 11. 19, dealing with tutela of latins, seem to treat the lex as dealing only with slaves freed by bonitary owners.

^II e.g. G. 1. 22; 3. 56,eic. ¹² Ulp.l. 11. words quoted have little point unless they are an allusion to informal manumission. But this means that if the lex Aelia is earlier than the lex lunia, either Ulpian is wrong or a man freed informally would be a slave if he had done no wrong—free if he were a rascal¹. Moreover Gaius’ in dealing with the law as to the distribution of the goods of deditidi uses language which implies that latini (iuniani) existed at the date of the lex Aelia.

So far as the general question goes, opinion seems on the whole to favour the view that the lex lunia is the older⁹. But the contrary view has many supporters’. Karlowa⁸, following Brinz, argues strongly for it. He points out that though Gaius says the lex Aelia deals with latins under 30, he nowhere says that they got latinity by that law, which must have been the case if the lex lunia had already been passed. Indeed in one text he implies that they got it through the lex lunia’.

¹ Vangerow thinks (op. cit. 13) that a criminal slave freed informally was not protected by the Praetor. Er. Pos. 10. This would avoid the absurdity. He notes that the lex lunia gives latinity to all persons protected. G. 3. 56. He holds it to be only by a Sc.

² G. 3. 56, latini essent. ⁸ TJlp. 1. 12. So Vangerow, loc. cit.

⁴ Schneider, Z. S. 8. 5. 225 sg'«?., 6. 186 sqq., 7. 81 sqq.

⁰ Pu Caillaud, cit. Schneider, Z. S. S. 5. 241.

6 See, e.g., Girard, Manuel, 124; Mommsen, Staatsr. (8) 8. 1. 626; Pr. Pub. Rom. 6. 2. 248; Roby, Rom. Priv. Law, 1. 88, etc.

1 e.g. Cuq, Inst, Jurid. 2. 148; Kariowa, R. R. G. 1. 621 sqq.; Holder, Z. S. S. 6. 205 sqq., 7.4/isq. ⁸ loc. cit.

⁹ G. 3. 56. The fact that it is not called Norbana by early writers he thinks proves nothing: many consular laws are cited under one name. The point is, however, not that the absence of the name Norbana shews that the lex is not of a.d. 19 but that the fact that it is once so called 500 years after the assumed date proves little in favour of that date, especially as the mbb. differ and the Greek paraphrase is equivalent to Urbana. Schneider, Z. S. S. 5.225. Vangerow, op. cit. 9, points out, however, that Norbanus was also called lunius.

It is true that the lea; Aelia Sentia seems to speak of marriage^[1924] of those freed under 30, which implies latinity. Vangerow holds that the lea; Aelia Sentia spoke only of contubernium and that Gaius is antedating the expression uxorem ducere*.

On the whole, as Mommsen says’, while the priority of the lea; lunia is the solution which creates least difficulty, certainty is unattainable. But it is only certainty on this point that can give certainty as to the meaning of some of the obscure texts in which the classical jurists seem to be at odds on points connected with this legislation.

Lex Aelia Sentia, a.d. 4.

This is a comprehensive enactment dealing with the relations be­tween libertini and their patrons, and also imposing restrictions on manumission. It is only with these last provisions that we are con­cerned. There are four rules, which do not all start from the same point of view or protect the same interests, but have the common quality that between them they constitute the first inroad on the principle that a formal manumission by a quiritary owner makes the man a civis. The rules need separate consideration.

I. The manumitter must not be under 20, otherwise the manu­mission is void ipso гиге, the rule being prohibitory and nullifying⁴. It applies to all cases inter vivos or on death, and even soldiers’ wills are not exempt®. As the law does not divide days it is enough if he has completed the day before the 20th anniversary of his birthday. He cannot then be said to be less than 20 and the lea; does not require him to be more than 20®. The rule is in one respect very favourably construed. If the manumitter was 20 when he made a codicil in which he made a direct gift of liberty, it is immaterial that the will, confir­mation by which is needed, was made before he reached that age⁷. Usually the codicil is read into the will, the effect of which is in some cases to destroy the gift⁸. The text in the Code gives as the reason for laying down the more favourable rule, nec enim potestas iuris sed iudidi consideratur. This, which is not literally correct, since it is a question of potestas iuris, must mean that as the case is clearly not within the mischief attacked by the rule, and the rule itself is restrictive of a civil right, it is to be construed narrowly. The rule applies only to a manumission: thus a minor pledgee of a slave can give the assent without which the manumission is void⁸

As might be expected attempts were made to evade the lex. One at least of these was checked by a Senatusconsult which provided that a gift by a minor to a man of full age, in order that he might free, was void^[1925]. In the same way he could not, in his will, validly direct liberty to be given². Where a minor sold a slave ut manumittatur, the sale was void, even though the slave was delivered, and even though the intent of the minor vendor was that the manumission was not to take place till he was of age⁸. The point is that his judgment was not regarded as yet sound enough, and if the transaction was allowed to stand, he would be unable to change his mind. Where a common owner, a minor, abandoned his share to a common owner animo manumittendi, the receiver could not free—the transaction being null: nihil aget\ Where a minor released a debtor on his promising to free a slave, the stipulation was void, and there was thus no novation of the old debt⁶. It is evident that the Senatusconsult⁶ was somewhat general in its terms. Probably it prohibited what Proculus calls fraus legi⁷, and left a good deal of room for juristic interpretation⁸. The fact that there was a gradual development may perhaps account for the view attributed to the early Campanus, that if a minor requested his heres to free a slave of his (i.e. of the heres), this was valid and not affected by the lex^s. It is not easy to distinguish this from the last case: presumably the lex and the Senatusconsult were at first regarded as applying only to freedom given to the minor’s own slave¹⁰.

But where a filiusfamilias freed under the authorisation of his paterfamilias, this was valid whatever the age of the minor, for here the father was the true manumitter^[1926].

All this is subject to the very important exception that if causa was shewn to a body called the Consilium, the minor might with its approval manumit per vindictam, and as proof of the causa did away with the statutory bar, he might even free informally, with the effect of making the slave a latin¹². But, ordinarily, the manumission was done at once, on approval of the causa, by vindicta, before the magistrate whose consilium had approved: hence the manumission is sometimes said to be done apud consilium⁷³. This consilium was a council chosen by the magistrate who presided in it. It consisted, at Rome, of five Senators and five Equites, and it sat to enquire into causae on certain specified days. In the provinces it consisted of 20 Recuperatores, Roman citizens, and this class of business was attended to on the last day of the Conventus, the judicial Assize or Session¹. This particular business however hardly seems to have been looked on as judicial, since we learn that a person domiciled in one province could shew cause in this way, and manumit, in any other province in which he chanced to be^{[1927] [1928] [1929]. It was immaterial that the Praetor who presided was bis tutor*. The magistrate himself might be under 20: this would not prevent him from presiding, unless it were his own slave: in that case he could not do so in earlier classical law, as he would have to nominate the consilium*·.}

As to what was a sufficient causa, We have a considerable list, and we are told moreover that there was no hard and fast rule: the sufficiency of the causa would be determined in each case*. A cause duly approved, whatever it was, sufficed, and after the manumission it could not be called into question. Thus an enactment of Valerian lays it down that while a manumission by one under 20 without cause shewn was a mere nullity, one after cause shewn did not admit even of restitutio in integrum·. liberty is irrevocable*. This is only an application of a well-known principle. But a text of Marcian goes a little further. He tells us that Antoninus Pius laid it down, that when once the causa had been accepted, then, however defective it really was, the liberty must proceed : causas probatas revocari non oportere...nam causae probationi contradicen­dum, non etiam causa iam probata retractanda est^[1930]. This means presumably that there was no appeal: it would not prevent a magistrate from vetoing any further steps, where a fraud was proved*.

Apparently the only fixed requirement for a causa (and this was a creation of practice) was that it must be honesta causa, non ex luxuria sed ex affectu, non deliciis sed iustis affectionibus^>. Among the more obvious causae were blood-relationship of any kind or degree, the relation of nurse or paedagogus, foster parent or child, foster brother or sister^{[1931] [1932]. The causa might be notable services in the past, e.g. the protection of life or honour11.}

There is more complication as to those causae which contemplated the future. If the slave was over 18, desire to have him as a procurator was enough¹, provided that the manumitter had more than one slave^[1933]. It is laid down, though not without some doubts, that the desire to have the man as tutor was not enough: the reason assigned being that he who needs' a tutor is not fit to choose one^{[1934] [1935] [1936]. The reason seems hardly satisfactory. The enquiry into the sufficiency of the causa would include an enquiry into the fitness of the man. The argument of the text seems indeed to suppose that the cases in which manumission by a minor was allowed were those in which even an immature mind was able to decide, but it is obvious that this was not the principle at all. The truth is that for pupilli without testamentary or statutory tutores the law provided another well-known method of appointment.}

A common and much discussed causa was intention to marry. To make such a causa admissible it was required by a Senatusconsult (perhaps the one which dealt with fraus legi) that the minor should swear to marry the woman within six months. If he did not so marry, the manumission was null^{[1937] [1938] [1939], so that if she had a child in the meantime, its status was in suspense till the marriage or the expiration of the six months’. There were obvious limitations on this causa. Not more than one could be freed for· this purpose, and the manumitter must be of a class a member of which might reasonably marry a libertina[1940]. That the woman might marry a third person was no causa, and if no other was shewn, then, even though, e.g. on divorce by the third party, the minor married her within the six months, this did not save the manumission : it was simply void, and could not be saved by an ex post facto causa[1941].}

A woman freed matrimonii causa could not refuse⁸ or marry any other without the manumitter’s renunciation of his right⁹. It is said that she could not divorce, but this is contrary to the Roman conception of marriage, and the rule, as Julian says, really means that if she did divorce, she could not marry anyone else. No doubt the patron could divorce her¹⁹.

A woman could free on most of these causae, but not, it seems, matrimonii causa, unless she was a liberta, and a slave, e.g. a fellow­slave, had been left to her for this purpose^[1942].

There are other causae of a totally different nature which need separate treatment. If a minor was instituted heres on condition of freeing a certain slave, this was a sufficient causa: his iudidum was not in question®. If a slave was conveyed to a man ut manumittatur, whether gratuitously or for a price, it was provided by Marcus Aurelius, about A.D. 178, that the man should become free, though nothing was done, by the effect of the disposition³. It is clear therefore that if he was so delivered to a minor, there was no need for the minor to shew causa, since he could not help the freedom. Accordingly we are told in two texts by Papinian and Ulpian dealing with donatio ut manumittatur, that there was no reason to shew cause⁴. But another text of Ulpian says’ that where the slave was so given, either for nothing or for a price, the minor might prove by way of causa, either the lex donationis, or the intent of the transferor, otherwise shewn. If there was a price, there was obvious reason for shewing the causa, since it might involve a loss, but the text expressly covers also the case of donatio. The texts may perhaps be harmonised on the supposition that the expression causae probatio is here used untechnically, and the meaning is that where the manumitter is under 20, the Praetor presiding will require to be satisfied of the circumstances, and the matter can be referred over to the consilium, if need be, as in the case of sale for a price’. In the analogous case of a slave suis nummis emptus, there was a rule, a little earlier in origin, that if not freed he could apply to the Court and get an order directing the holder to free him⁷. We are told that this constitution applied even though the owner were a minor⁸. We are not told whether if the minor proceeded to free he must prove the causa, but from the argu­ment of Papinian in the case last discussed⁹ it is to be presumed that he must, since the liberty would not take effect of itself.

Another analogous case is that of fideicommissum. Here there are distinct cases. We are told that a minor could not free by direct gift by will, but that he could do so by fideicommissum, and that the gift would be valid, if the man was one as to whom the minor could have shewn cause, if he had freed inter vivos™. We are not told that the adult fidudarius must shew cause, and, indeed, the form of the texts is opposed to this. He had a perfect right to free, and the transfer to him could not be regarded as null as it could inter vivos¹. If, however, the slave attempted to put in operation the compulsory machinery, he would have to satisfy the court that a causa existed. Another case is that of a fideicommissum of liberty imposed on a minor. Here we are told by Papinian’, consistently with his view in the case of a slave donatus ut manumittatur, that the minor must prove the causa. In this case the man would not become free ipso facto without the intervention of a magisterial decree.

It has already been noted that the presence of causa nullified the statutory defect. Accordingly it justified some of those acts, by a minor, manumittendi causa, which without it were void. Thus a minor with causa could convey his part to the co-owner for manumission, though he could not without⁸. But the existence of causa did not do away with restrictions independent of the lex Aelia Sentia. Thus an infans could not free, whatever his causa was, for he could not be authorised and his tutor could not free. A pupil lus not infans could however free, tutors auctore, but not, says Paul, so that the peculium passed^{[1943] [1944] [1945] [1946] [1947] [1948] [1949].}

II. The slave must be over 30 or he does not become a civis⁹. There can be no doubt that the lex Aelia Sentia went as far as this: whether it went further and defined a slave freed under 30 as a latin is uncertain. The answer depends on the relative dates of the two leges. If the lex lunia was the later, the lex Aelia probably placed such persons in the same position as those informally freed’. The effect of causa is exactly as in the last case, i.e. if the man were freed vindicta, after cause approved, he became a civis’’. There is however some diffi­culty as to what happened if there were no causa. If he was manumitted by will directly, he became a latin: there could be no question of causa⁹. So too there is no sign of causa in relation to manumission censu. If the man was over 30 he became a civis: if he was not it was presum­ably void⁹. But the point is unimportant, for when these texts were written, the census was long obsolete in practice. If the manumission was informal, the man could not be a civis in any case, so that proof of causa would serve no purpose. We have seen that a man under 20

oh. xxm] Lex Aelia Sentia: Age of Slave 543 could not give liberty even by way of postponed fideicommissum¹. The same reason does not apply here, and we are told’ that a direct or fidei- commissary gift of liberty to take effect when the man reached 30 was valid. The distinction shews that the reason for refusing dvitas to slaves freed under 30 was not that till that age it was not possible to be sure of their fitness, but that till that age they were not fit to be entrusted with the responsibilities of citizenship.

The effect of manumission vindicta sine consilio is not clear: the only text on the matter is corrupt. As it stands it tells us that (Z&r?) sine consilio manumissum Gaesaris servum manere putat³. This is absurd. He cannot manere what he has not been. Nor is there any reason why he should become the property of Caesar: a derelictio would not have this effect, but would leave him res nulHus, and manumission, which leaves the manumitter patronus, is much less than that⁴. The text does not say who putat: it must presumably be the lex which is the subject of the preceding and the following sentences. To say that a lex putat in a text which is setting forth its provisions is perhaps unexampled⁴. Of the many suggestions for emendation⁸, the old one that the word was originally the name of some jurist is the most plausible. To make this sort of emendation rational it must be assumed that in early law, at least in the opinion of some jurists, manumission vindicta could not make a man a latin: it must be civitas or nothing. There are some circumstances which tend to make this possible. No specific case of manumission vindicta giving latinity can be found in classical texts, and though some are mentioned in Justinian’s constitution⁷ abolishing latinity, they all seem to be instances of that mass of legislation and practice, creative of latinity, which he tells us overlay the ancient law, and of which, as he also tells us, he was at pains to remove the traces⁸. Moreover manumission vindicta is an actus legitimus of extreme antiquity, and for this reason may have been regarded as a nullity if not completely operative⁹. However this may be, it is probable that practice early

* C. 7.4. 5.

’ G. 2. 276; 10. 2. 39. 2; 40. 4. 38.1; 40. 7.13. 6; 34. 6. 29; cited by Vangerow, op. cit. 38. s Ulp. 1.12.

⁴ It may make him servus sine domino (Ft. Dos. 11), but that is not quite the same thing.

⁶ The expression does occur, but not apparently as a reference to an explicit provision. See 40. 7. 25 ana Cicero, de Bep. 4, cited Vangerow, op. cit. 25.

⁶ Among them are: to omit Caesaris, to substitute Senates (suggesting regulation by Sc.), to substitute the name of a lex. These Vangerow cites and rejects on what seem adequate grounds (op. cit. 25 so.). He also rejects the suggestion of the name of a jurist (Cassius; Gaelius Sabinus) on less convincing grounds. He treats the whole clause as a gloss. He considers that the act, as it shews intention to free, is an informal manumission. This ignores the probable view that not every declaration makes the man free, but only one which comes within the conceptions inter amicos or per epistolam. Ante, p. 446. He also urges that the statutory bar to its giving civitas ought not to have prevented it from producing other effects. See also Krüger, ad h. I. Schneider thinks the text rational as it stands (Z. S. 8. 6.189; 7. 31 sqq.), but see Holder, Z. S. S. 6. 205 sag.·, 7. 44 sg. Justinian deals with this case apart from the other cases of latinity, 0. 7.15. 2. 7 0. 7. 6. 6, 7.

⁸ The hypothesis of later legislation might account for the obscure texts as to servus

pigneraticius Jr actuaries, post, pp. 574, 579.

⁹ Cp. 50. 17. 77; see App. iv.

544 Lex Aelia Sentia: Fraud: Degraded Slaves [pt. n developed disregarding these considerations, and from the generality of Gaius’ language in connexion with anniculi probatio¹, it seems possible that for him both formal and informal modes were on the same level, as to the present point.

III. Manumission in fraud of creditors or patron is void.

This will be dealt with fully in the law of Justinian’s time. Here it is enough to state a few general rules. The rule applied to peregrine manumitters though the other parts of the lex did not*. The manu­mission was absolutely void^[1950].

A manumission was fraudulent if the manumitter was, and knew himself to be, insolvent either before or as a result of the manumission, and it must be shewn that the creditors actually were injured^[1951]. Thus a manumission was not in fraud of creditors if the manumitter had a maritime venture under way, which at the time had become a total loss, though he did not know it^[1952].

Fraud on the patron would occur for instance if a libertus made it impossible for himself to render the due aids and services, or if a dying latin freed his slaves, or if a civis libertus did so when he had no children. We are, however, without any direct information as to this rule, and can only argue by analogy from the rules as to alienations in fraud of patron^{[1953] [1954]. The rule as to the patron does not recur under Justinian’s law and even traces of it are hardly discoverable’. There are very few references to it even in the classical law.}

IV. Certain slaves become on manumission deditidi.

These were slaves who had been punished by their master with chains or branding or imprisonment, or had been tortured for wrong­doing, and convicted or made to fight with wild beasts. On manu­mission they were in numero deditidorum, no matter how formal the manumission, or how complete the capacity of all parties in other respects^[1955]. This type was a mere addition to a pre-existing class, the deditidi, with whose origin we are not concerned.

The different possibilities as to form of manumission make some difficulty in this connexion. In the case of will the matter is plain, but it is clear that the manumission might be inter vivosi If it was formal and subject to no defect but the badness of the slave, he became a

ch. xxin] Persons in numero Dediticiorum

545

deditidus. But the manumission might be such as to have made the man a Jatin apart from his defect. If and when the view was adopted, that manumission vindicta need not make the man a ciws¹, there was no difficulty in the case of manumission vindicta of one under 30, sine causa. But informal manumission creates a dilemma. If the lex Aelia makes him a dediticius and the lex lunia is later, it follows that the rascal would be free, though a dediticius, while the honest man would be still a slave, though in libertate. The difficulty does not exist if the lex lunia is the earlier. On the other view, the solution of Vangerow^{[1956] [1957] [1958] may be stated, that the lex Aelia Sentia applied its rule as to deditidi only to formal manumissions, and that its extension to all forms was due to a later Senatusconsult. But this is rather heroic in view of the texts which say that for this purpose the lex Aelia did not distinguish between the forms8. It involves the further corollary that after the lex Aelia the informal manumission of such a degraded slave was a nullity, since it is clear that the lex lunia gave latinity to all[1959] [1960] who were protected by the Praetor. This is in itself not improbable, for, as Vangerow remarks, they would be just the persons to whom the Praetor might refuse his protection. But the texts give no hint of all this and much of it is, as we have seen, in contradiction with them. On the whole evidence the view that the lex lunia was the earliest seems to be the most probable.}

Their position was carefully defined. They were incapable of dvitas, and thus, for instance, if they satisfied all the rules of erroris causae probatio, though the other effects of the rule were produced, the deditician member of the union remained a deditidus⁸. They had no testamenti factio of any kind. They could neither make wills nor take under them⁶. Their property reverted to their patron on their death under rules which hardly concern us, but which seem to have been obscurely stated in the lex''. It appears to have provided that the bona were to go as if they had not been deditidi⁸. This might mean “as if they were still slaves.” But it was construed as meaning “ as if they had not suffered from the defect which made them deditidi." This interpretation, however, itself needed limitation. As it stands it would make the goods go as those of a latin in some cases and as those of a civis in the others. But this would be to give a right to the children, as well as a right of testation. Neither of these existed, and the rule of the classical jurists was that the goods were to go to the patron in any case, as those of a latin, if the man would have been a latin but for his offence,

otherwise as those of a dvis. In the first case this would be a reversion of peculium, in the other it would be a succession, the distinction being important in many ways¹.

If they stayed, or dwelt, within 100 miles of Rome, they were sold into perpetual slavery beyond that limit, and if then freed they became, so Gaius tells us, servi populi Romani. Though Gaius seems to say that these detailed provisions are in the lex Aelia Sentia, this is not the necessary meaning of his words, and there may have been senatusconsulta¹.

The disabilities resulting from this degradation are very grave, and Paul shews that very definite rules were laid down as to the cases in which it took effect. Thus torture without confession was no bar to complete liberty, nor was punishment by one under a fiddcommissum to free, since he could not make the position of the slave worse. Nor, for the same reason, was punishment by one of two owners, or by pledgee or pledgor of the slave, or by a master who was insane or a pupillus. But punishment by a subordinate whose act was authorised or ratified was enough. Here, however, if the master knew that the man was innocent, at any time before the punishment was actually inflicted, the facts would not be a bar to future complete liberty⁸.

None of these four restrictions applied, not even the last, in a case of manumission by will to provide a necessarius heres to an insolvent⁴, in order to avoid intestacy resulting from refusal of the heres to enter. We have already discussed the general principles of the law as to these heredes necessarii*, and a word or two here will suffice. The privilege was strictly and narrowly construed. If any other heres entered, the gift to the slave was not saved’, and it was only against the restrictions of the lex Aelia Sentia that the exception held good⁷.

Lex Fufia Caninia (b.c. 2)’.

Slaves were very numerous in the Augustan age—an individual dvis sometimes owned thousands—a state of things very different from that existing in earlier days, if tradition is to be believed’. It was a natural consequence that manumission became frequent. It appears indeed that the number of libertini became a public danger. Manu­mission by will was the most common, as it cost the owner nothing, and ensured the attendance of a number of grateful liberti at his funeral. The result was an undesirable increase in the number of libertini, and

¹ G. 3. 64—70. · G. 1.27; 1.160.

» P. 4.12. 3—8. That children of dediticii could become civet appears from the rules of errorie causae probatio, G. 1. 26,67,68. As not subject to the special disabilities of their mother they were presumably on the same level as ordinary peregrin of the region. See Ulp. 22. 2.

< G. 1. 21; Ulp. 1. 14; In. 1. 6. 1; C. 6. 27. 1; D. 40. 4. 27. « Ante, pp. 605 sgg.

e Ulp. 1.14. 1 28. 5. 84.pr. s gee Mitteis, Z. S. S. 27. 357.

⁸ See the cases mentioned in Apuleius, Apol. 17. See also Wallon, op. oil. 2. Ch. m. occasional ruin to heredes. The lex Fufia Caninia was passed to check the evil. It provided that a man with 2 slaves could free both by his will, with 2 to 10, one half, with 10 to 30, one third, with 30 to 100, one fourth, with 100 to 500, one fifth, and never more than a hundred¹. The maximum in each case is called the legitimus numerus³. The lex further provided that the power of manumission was never to be diminished by an increase in the number of slaves’. If more than the right number were freed, only the earlier, up to the legitimus numerus were free⁴. To prevent evasion the lex required that they should be freed nomina- tim”: the sc. Orphitianum provided that a clear description would do as well, if there were no ambiguity, for instance, “my cook,” if there were only one, or “ whoever shall be born of such and such an ancilla³.” If the gift broke this rule, e.g. a gift of freedom to “ all my slaves,” the whole gift was void⁷. So also if the names were written in a circle in such a way that it was impossible to say which came first, the whole was declared void under a provision of the lex annulling quae in fraudem eius facta sint. Other similar attempts to evade the lex were met by senatuscmsulta³, which have, however, left little trace. There may be, indeed, one case. The lex applied only to manumission by will or codicil’, and Gaius tells us that it left manumission vindicta, censu, and inter amicos quite free¹⁰. His epitomator makes a similar remark, substi­tuting in ecclesiis aut ante consulem aut per epistolam aut inter amicos, but he adds that if a man on the point of death freed a number of slaves inter vivos, in fraudem legis, the manumissions were valid only up to the legitimus numerus¹¹. Perhaps this is the effect of a Senatusconsult¹².

In calculating the number of slaves, fugitive slaves were taken into account, a rule for which Paul finds it necessary to give the reason that such slaves are still possessed by their owner¹Â·. We are nowhere told how common slaves were reckoned. As the common owners’ rights in the slave were in nearly every case pro parte¹*, it is probable that the slave counted only as a fraction.

None of these texts applies the rule in express terms to fideicom- missary gifts, and the enactment by which Justinian repeals the lex Fufia Caninia¹Â· is rather ambiguous. It is plain, however, that unless it did apply to them it must have been nearly nugatory, and Paul, dealing

I G. 1. 43; P. 4. 14. 4; Ulp. 1. 24. » P. 4. 1. 16.

⁸ e.g. a man with 30 to 43 could free 10. With 44 he could free 11. G. 1. 45; Ulp. 1.24.

< G. 1. 46; G. Ep. 1. 2. 2. Liberties in a codicil were treated as subsequent to those in a will, though the will were later, since they owed their validity to the will, P. 4.14. 2.

⁸ Ulp. 1. 25; P. 4. 14.1. ⁸ P. 4.14.1.

⁷ G. Ep. 1. 2. 2. ⁸ G. 1. 46.

⁸ G. 1. 44; P. 4. 14. 1. ¹⁸ G. 1. 44. ¹¹ G. Ep. 1. 2. 1.

¹² Probably later than Gaius. The text limits the rule to the case in which the is already ill (of the malady of which he dies): probably no other evidence was needed to prove fronts legi, G. Ep. 1. 2. 3, 4.

¹⁸ P. 4.14. 3. Huschke (ad h. 1.) points out that the lex says habet.

I» Ante, p. 379. ¹⁸ C. 7. 3.1.

with this lex, gives, as an illustrative case, qui ex ea ancilla nascetur¹, which could have been effective only as a fideicommissum, It has been suggested^{[1961] [1962] that as at the time of the enactment fideicommissa were novelties, it probably did not apply to them at first, but was made to do so by one of the senatusconsulta to which Gaius refers. The writer notes that the various senatusconsulta affecting fideicommissary gifts of liberty do not begin till the time of Hadrian. He thinks that Justinian’s enactment[1963], abolishing the rule, is clear for its application to fidei­commissa in later law, and he cites a text of Paul[1964] [1965] in the Digest which seems to shew that it applied in the time of Neratius. There seems little reason to suppose a Senatusconsult: such a case would be well within a possible juristic interpretation of the prohibition of fraud contained in the lex. And the language of Paul’s text is much in favour of this view3.}

Before passing to the law of Justinian, it may be well to discuss shortly the circumstances under which the status of latinity could arise^{[1966] [1967]. The following list has no claim to completeness.}

1. The slave informally freed by a competent dominus’. It has been shewn by Wlassak that the classical law knew of but two of these modes, per epistolam and inter amicos, and that manumission in convivio is of much later introduction^[1968]. He remarks also that there is nothing in the form of the rule in the lex lunia to prevent its application to methods of later introduction^[1969]. The form of manumission inter amicos is not very precise. In one, the record of which has come down to us^{[1970] [1971], the witnesses do not sign and are not named. The transaction was in Egypt and some of its provisions are coloured by Greek law11, but there is no reason to doubt that this was in conformity with Roman practice. Hence the idea would naturally appear that any public manifestation of intent sufficed. This accounts for the acceptance of manumission in convivio[1972], and the enactment of Justinian abolishing latinity gives other instances of the same thing, such as declaring apud acta that he is a son13, giving him or destroying the papers evidencing his slavery14, and perhaps also the direction in the will that he is to stand, pileatus, at the grave of the deceased1. It is observable that here it is indifferent whether the direction is by the deceased or the heres. Justinian provides that even if there was no intention to free but only to make a false shew of humanity, the men are to be ewes, but in this case they would not have been latins in earlier law. It may be added that the lex lunia required the manumission to be nominatim*, but all this means is that the slaves must be evidenter denotati*.}

2. A slave informally freed by a master under 20, with the approval of the Consilium⁴.

3. A slave manumitted under 30’.

4. A slave manumitted by his merely bonitary owner’. Neither Gaius nor Ulpian enumerates the relevant cases of bonitary ownership : the latter mentions, as an illustration, the typical case of a slave acquired by mere traditio. But the rule must have applied equally to other cases of Praetorian ownership. Such would be the case of one held by praetorian succession (bonorum possessio cum re), the case of a slave ductus under a noxal action, that of one received under a decree of missio in possessionem, or a bonorum venditio. The case of a slave handed over under a fideicommissum is no doubt on the same footing, unless he was formally conveyed. The case of an owner in integrum restitutus, in respect of a slave, might seem to be on the same level, since it is a praetorian remedy, contradictory of the civil law, and giving rise to actiones fictitiae and the like. But it is clear that the Praetor restored the old state of things so far as possible, so that in this case such a reconveyance would be compelled (either officio iudieis or by the Praetor himself in those cases in which he carried out the restitut/io) as would restore the quiritary ownership⁷.

5. By an edict of Claudius a slave cast out because of sickness became free and a latin, provided the master publice ejected him and, having the means, took no steps to have him looked after or sent to a hospital⁸.

6. If a slave had brought a causa liberalis against his master and lost, and the price of the slave was paid to his master by an outsider to secure his manumission’, the slave, on manumission, became only a latin, as a sort of punishment¹⁰. The date of this is not known: Justinian credits it to antiquitas. It must have been express enactment.

7. If an ancilla was married by her dominus to a freeman, with a dos, she became a latin. This may be no more than a case of informal

¹ C. 7. 6. 5. ² Ulp. 1.10; Wlaasak, loc. cit.

³ See ante, p. 460, and post, p. 556. ⁴ G. 1· 41.

⁰ G. 1. 17; Ulp. 1.12. ⁸ G. 1. 35,167; Ulp. 1.16; 22. 8.

7 4. 2. 9. 7, 10. 1; 4. 4. 24. 4. ³ C. 7. 6. 3; ante, p. 36. manumission. Justinian made it give citizenship^[1973], as such manumission did. But the rule may have also covered the case of fraud.

¹⁰ C. 7. 6. 8.

⁹ Post, p. 640.

8. Where an andlla was sold with a condition against prostitution, but was nevertheless prostituted by the buyer, or where there was a condition for re-seizure in the event of prostitution, and her old owner did so seize her, and himself prostituted her, she became free and a latin².

9. A liberties ingratus under the conditions already discussed’.

10. If a testator has given a slave liberty, conditionally, and while the condition is still pendent the extraneus heres frees him, he becomes only a latin⁴. The text refers only to extraneus heres: probably a suus heres, whose the slave was apart from the will, might ignore the restrictive effect of the condition®. The date of our rule is not known: Pomponius® quotes Octavenus as holding that if one freed a slave by will conditionally, and expressed the desire that the heres should not free the slave pending the condition, this direction was of no force. From this it has been inferred that our rule is as old as the first century of the Empire.

11. A liberta who cohabited with a servos alienus without her patron’s knowledge was enslaved, and became only a latin if freed by him⁷.

12. Slaves who detected rape were under certain circumstances made latins by Constantine. Justinian gave them civitas⁶.

13. A freewoman, sciens vel ignara, cohabiting with a slave of the Fisc, remained free under a provision of Constantine, but the children of the union were latins’.

There remain several cases of a doubtful kind.

14. Where a person was freed formally with an expression of intent that he should be only a latin, the effect seems to have been doubtful. Justinian enacts that such expressions are to have no effect¹⁰.

15. The sc. Silanianum may have contained a case, to be discussed later^[1974].

16. A pledged slave could not be freed¹². But, on a text, which is imperfect, most editors seem agreed that he became a latin if so freed, at least when the debt was paid¹â€™. But Justinian does not mention this case in his general enactment¹⁴.

17. If a woman freed a slave without her tutor’s auctoritas, this was not valid. But if auctoritas was given at the time an informal letter of manumission was written, it was held and finally decreed that this should suffice^[1975]. The text is obscure and may refer only to formal manumission, in which the tutor, though not present at the formal act of manumission, had been present and assenting when the mistress wrote a letter to the slave declaring her intention, but it is usually taken to mean that an informal manumission was good, and made the slave a latin, even though the tutor gave auctoritas only when the letter was written, and had altered his mind when it was received⁸. The latter view better fits the words of the text.

18. If a slave was under usufruct he could not be freed. A certain truncated text on the matter is commonly taken to mean that though the owner could not free the man vindicta, still, if he did go through the form, the man became a latin when the usufruct ended⁸.

19. It was a standing rule of manumissions that a manumissor could not give the slave he freed a better status than his own: it may be presumed therefore that a man freed by a Junian latin was himself a Junian latin⁴.

20. If a slave was freed conditionally by will, he did not become a statuliber till the heir entered. We are told, however, that if he was usucapted, in the meantime, the Praetor would protect his liberty⁵. In another text it is said that his spes libertatis is restored favore sui’. The language of the first text has led to the suggestion that the slave, on the satisfaction, became a latin’. This seems improbable: it is hardly consistent with the language of the other text. The help of the Praetor is referred to in other cases where the slave became a civis^B, and the difficulty resulting from the fact that when the heres entered the man was the property of another would suggest rather a fideicommissum than resulting latinity. But in fact the difficulty was disregarded favors libertatis.