CHAPTER XIII. SPECIAL CASES (cont.}. SERVUS PIGNERATICIUS, FIDUCIAE DATUS, STATULIBER, CAPTIVUS.

XVI. Serves Pigneraticius.

The law concerning a pledged slave derives some peculiarities from the fact that, while on the one hand the rights acquired by the pledge creditor are slight (being essentially no more than the right to hold the slave without deriving profit from him), on the other hand the institu­tion is only a praetorian modification of the old fiduciary mancipation, under which the creditor became owner.

A pledged slave is still in bonis debitoris^[MLXXXVI], and thus a legacy of my slaves includes those I have pledged, but not those pledged to me². The debtor retains the actio servi corrupts. The pledged slave is treated for the purpose of the Sc. Silanianum in all respects as if he had not been pledged⁴. But there are many respects in which the creditor’s interest comes effectively into play.

IfÂź the pledge creditor kills the slave, the debtor has the actio Aquilia against him, or, if he prefers, he may bring the action on the contract⁶. If on the other hand the debtor kills the slave, the creditor has not the acii’o Aquilia, even utilis, but is given an actio in factum⁷. If the slave is killed by a third party, the pledger has the actio Aquilia, and the creditor is allowed an actio utilis, because in view of possible insolvency of the debtor he has an interesse. A text⁸ credited to Paul, hints, in a rambling manner, that the creditor’s action is given only in the case in -which the debtor is insolvent, or the creditor’s remedy, apart from the pledge, is time-barred, and says that, in that case, the debtor has it only if the debt is less than the value of the slave.

The case of theft of a pledged slave presents difficulties: they have nothing to do especially with slaves but cannot well be left undiscussed, as one or two of the most difficult texts deal with slaves. Many texts shew clearly that the pledge creditor has actio furti if the thing is stolen², even by the debtor³, and that he will himself be liable to the action if he exceeds his right⁴. Beyond this it is difficult to get a clear doctrine: there are divergences on all material points. In seeking the basis of the pledge creditor’s interesse, it is natural to think of his obligation custodian praestare, either absolute, except for vis maior*, or only diligentian eaactan praestare*. And two texts of Ulpian⁷, in one of which there is an appeal to older authority, seem to start from this point of view, but the first shews that there is an interesse independent of obligatio*, and the words from hoc ita to competit are probably from Tribonian. The responsibility is limited to the case of culpa, and there is some reason to think that at least in the case of pignus, this limita­tion dates from the later Empire⁹. What Pomponius approves is the earlier part of the text. In the other¹⁰, the words iten locati pignorisve accepti may be interpolated, but, indeed, the whole text looks corrupt¹¹.

It is clear on other grounds that the obligation whatever its extent cannot have been the sole basis of the interesse. Had it been so, there would have been, at least, doubts as to the right of action where the debtor stole the thing, since in that case the obligation did not exist¹².

¹ Interpolationen, 89 sqq. Do they represent even later law ? G. points out that where, as here, two can sue, though substantially only one sum is due, the normal course is for the first plaintiff to give security for defence of the person liable against the other.

²13. 7. 22.j>r.; 47. 2. 14. 16, 15. pr.; In. 4. 1. 14; 4. 2. 2; etc.

⁸ 13. 7. 3; 41. 4. 5; 47. 2. 12. 2, 67.pr.\ C. 7. 26. 6; G. 3. 200, 203; P. 2. 31. 19; etc.

< 13. 7. 4,5; 47. 2. 52. 7, 55.pr., 56, 74; C. 9. 33. 3; etc.

⁵13. 7. 13. 1,14; C. 8. 13. 19. « C. 4. 24. 5, 8; In. 3. 14. 4.

1 47. 2.14. 6, 16. » 47. 2. 14. 6.

⁹ As to the nature of custodia the most acceptable opinion seems to be that which may still be called the orthodox view. (See for this, with variations, Pernice, Labeo, 2. 1. 345; Lehmann, Z. Sav. Stift. 9. 110; Biermann, same review, 12. 33; Girard, Manuel, 655. Contra, Ferrini, Archivio Giuridico, 53. 260.) According to this doctrine, custodia meant originally the obligation to keep the thing as against thieves, but not as against robbers. This was gradually modified, till except in a few cases, it was no longer to be distinguished from the obligation diligentiam maximam praestare. This development explains such texts as speak of custodiam diligentem (e.a. 13. 6. 5. 5) and the like. But there is room for difference of opinion as to the date of this change. Pernice attributes it to the influence of Julian. Biermann is, in general, of the same opinion, but considers that in regard to pignus, the change is of the later Empire, since Diocletian still distinguishes in this connexion between culpa and custodia. Paul and Ulpian make the same distinction in other connexions (P. 2.4.3; D. 19.1. 36; 47. 2. 14. 6, etc.}, but this Biermann regards as mere historische Reminiscent. But that kind of reminiscence is more characteristic of Tribonian than of Ulpian or Paul, and as they state the distinction clearly it is likely that the law of their time admitted it. It seems highly probable that the assimilation in most cases of the liability custodiam praestare to that diligentiam praestare was a development of the later Empire, except in the case of commodatum, which had special rules. It is to be observed that nearly all the texts which state the newer doctrine are confused in form (e.g.

¹⁰ 47. 2. 14. 6.

¹¹ Doubts in the case of commodatum are plentiful: 13. 6. 5. 6; 47. 2. 14. 16, and many in C. 6. 2. 22.

« 13.fi. 21. pr.

But many texts give the action in that case and none denies or doubts its existence^[MLXXXVII]. Again, in all cases except where the thief is the debtor, the creditor must set off what he recovers against the debt’, while in the other cases of interesse based on obligation, the plaintiff keeps what he recovers⁸. Again, if the right were based on obligation, the creditor’s action would exclude that of the debtor⁴, but, here, both parties have the action⁸. Ulpian tells us indeed", that there had been doubts as to the creditor’s action if the debtor were solvent, but himself holds, with Julian, Pomponius and Papinian, that he has an interest in all cases, a rule for which the Institutes give the reason, quia expedit ei pignori potius incumbere quam in personam agere'¹. The text says nothing about obligation. Again, if the right were based on obligation it would not exist, as it does, where there is a mere hypothec, since here the obligation does not exist". It seems clear, then, that, at least on the dominant view, the creditor’s right does not depend on his liability. There is, however, one difficulty. Two texts of Paul’ allow the creditor to recover the value of the res. This is consistent with the view that his right rests on obligation, and not easy to reconcile with any other view. But one of them definitely gives the action against the debtor, which is inconsistent with that view¹⁰, and though Paul here notes that in that case the recovery is limited to the amount of the claim, this itself shews that there is another basis of interesse. There are, however, a number of texts, mainly from Ulpian^[MLXXXVIII], which limit the creditor’s right in all cases to the amount of his claim.

What is this other basis of interest ? The mere fact of possession would not suffice, if indeed the pledge creditor can be said strictly to have possession¹⁸ But his right, narrow as it is, exceeds mere possession. It has an economic content. He has a right to keep the thing until his claim is satisfied, and thus he will win, not only in possessory proceed­ings, but also even if the owner vindicates the thing. It is this ius retentions which bases his actio furti¹³. This is why his interest is limited to the amount of the debt, and why, if there are several things pledged together for one debt, the action, if any one of them is stolen, is limited only by the total amount of the debt¹⁴. Paul’s texts giving a right to recover the whole value¹⁸ do not seem to have been retouched : they may perhaps be based on a recognition of the right resting on

⁴ G. 3. 203, 6 ; D. 47. 2. 12. pr. etc. « 47. 2. 12. 2.

⁸ 47. 2. 19. 6, 62. 8.

¹⁸ k. t. 88.

¹⁸ h. t. 15. pr., 88.

284 Pledged Slave: Delicts: Acquisitions [pt. i liability, but their explanation is more likely to be found in the known genealogy of pignuts. The rule of fiduda, by which the creditor, being owner, recovered the whole value, simply passes over to pignus. Perhaps as early as Pomponios¹^ the more logical view appears which limits his action to the amount of his claim. Later on, the pledgee is so far assimilated to other holders who custodiam praestant, that his right based on liability is recognised, without excluding the other. But this is of late law, and, even in the Digest, the dominant doctrine ignores it. It may be noted that in several other cases, in some of which there would not be a possessory right, the ins retinendi gives an actio furti², on the analogy of pledge⁸.

In one text already mentioned⁴, Ulpian seems to give the debtor (the owner) an action only if the thing is worth more than the debt.

The rules as to noxal liability for the slave have already been discussed’. It may be added that the actio ad exhibendum lies against the creditor and not against the debtor: he has no power of producing the man⁷.

As to acquisitions, the rule is simple. The slave cannot acquire for the pledge creditor in any way, by traditio, stipulatio, or otherwise, not even possession, though the creditor possesses him for interdictal purposes’. All fruits and acquisitions ex operis are the debtor’s, and go to reduce the debt, any balance over the amount of the debt being recoverable by the debtor⁸. All other acquisitions, e.g. hereditas or donatio, go of course to the debtor. And the rules as to the effect of acquiring or purporting to acquire expressly for any person other than the owner are in no way exceptional in this case. The debtor cannot acquire possession through the slave since he is possessed by another, except so far as bars usucapio by the creditor, and allows the debtor to usucapt things of which possession had already begun¹⁸.

The creditor must not misuse the thing pledged and thus if he prostitute a pledged andlla, the pledge is destroyed¹¹. Unusual expenses may be added to the charge, and thus, if a pledged slave is

I 47. 2. 14. 6, 7. » h. t. 14. 1, 15. 2, 54. 4, 60; In. 4. 1. 14.

8 Modestinus (Coll. 10. 2. 6) denies actio furti to depositee even with ius retentionis. The reason is not clear and the rule is not in the Digest. The rule probably began with pignut.

* 47. 2. 46. 4. ⁶ h. t. 12. 2. « Ante, pp. 116,125.

7 41. 3. 16. 8 41. 1. 37. jm·.; 41. 2. 1. 15. » C. 4. 24.1, 2.

¹⁰ 41. 2.1.15. Or the slave himself. The rule does not of course apply to hypothec, n 13. 7. 24. 3.

captured and redeemed, the creditor’s right revives when he has paid off the lien of the redeemer, and he may add the amount paid to his charge¹. In the same way he may add to the charge any reasonable expenses incurred in training the slave in either necessary arts, or those in which the debtor had already begun to train him, or those to his training in which the debtor has assented^{[1089] [1090]. Expenses incurred in paying damages for a delict by him can also presumably be added, for we know that the creditor is compellable to pay them or abandon his pledge[1091].}

XVII. Slave held in Fiducia.

Upon this case the texts give us little information: the institution was obsolete in Justinian’s time, and, as we have learnt from Lenel, many of the texts which really dealt with fiducia cum creditore have been applied by the compilers to pignus, with or without alteration. But of these there are very few which have any special importance in regard to slaves as opposed to other chattels. Such slaves were technically the property of the fiduciarius. Thus a legacy of “ my slaves,” by the debtor, did not cover those he had so conveyed^[1092]. And what such slaves acquired in any way was the property of the fiduciarius. But this ownership was little more than nominal, for he must account for all such receipts, setting them off against the debt, and being liable for any balance’, having however the same right of charging expenses as the creditor in a pignus*. Moreover, a thing given in fiducia could be left per precep- tionem, at least according to the Sabinians, the heirs being bound to free it, though in general such legacies were confined to the property of the testator^{[1093] [1094]. The Sabinians did not consider this form to be confined, as legacy per vindicationem was, to the quiritary property of the testator, but applied it to anything in his bona*. But this is, as Gaius observes, a still further extension, for technically such a slave is not in bonis debitoris. As we have seen, however, the rules of account make this formal rather than real; Gaius, in fact, treats the transaction as essentially a pledge, the heirs being under a duty to reduce the thing into possession[1095].}

As the creditor was owner he could have no noxal action for any­thing done by the slave, but it may be inferred from some texts discussed in the chapter on noxal liability that he had a right to an indemnity¹, the amount of which could be added to the debt. The debtor could, however, abandon the slave instead of paying, leaving the original debt intact, but only if he was unaware of the character of the slave he was mancipating². Conversely it seems to follow that the creditor was noxally liable for anything the slave did, but of course this surrender while it ended the security did not destroy the debt, and did not impose on the creditor any liability for the value of the slave’. If the slave stole from the debtor there ought to be a noxal action against the creditor, but such an action would be of little use. For if the creditor surrendered the slave the debt remained, and he was usually in a position to require fresh security. If he paid the debt, because it was less than the value of the slave, it would seem that he could add the money so paid to the debt. But direct authority is lacking.

XVIII. Statuliberi.

The most important points in relation to these will arise for dis­cussion under the law of manumission, but something must be said here as to their position while still slaves. It is not necessary to define them with any exactness. Broadly, they are persons to whom liberty has been given by will under a condition, or from a day, which has not yet arrived⁴.

The main principle as to their position is that till the gift of liberty takes effect in some way’, they are still slaves of the heir⁶, for all purposes. Thus they may be examined under the See. Silanianum and Claudianum, if the heir is killed⁷. Children of an ancilla statulibera are slaves of the heir⁸. Statuliberi are subject to the ordinary incidents of slavery, with the restriction that no act of the heir can deprive them of their prospect of liberty, on the occurrence of a certain event, and some other restrictions shortly to be stated. Thus they may be sold, legated, delivered by traditio, adjudicated, and even usucapted, but always carrying with them their conditional right to liberty’. They may be pledged, but arrival of the condition destroys the creditor’s lien¹⁰. A usufruct may be created in them¹¹. They may be noxally

1 Ante, p. 125. ² Arg. 41. 2. 62. 3. ’ Arp. 9. 4. 17. 1. Ante, p. 116.

⁴ 40. 7. 1. One as to whom it is doubtful if his freedom is in fraud of creditors is a statuliber. Ibid. See Festus, s.v. Statuliber. And see^ost, Ch. xxi.

8 As to cases in which the gift took effect without satisfaction of the condition, post, Ch. xxi.

⁶ U.2.2; D. 40. 7.16, 29.pr. ’ 29.5.1. 4. 8 40.7.16; C. 7. 4. 3.

9 30. 81. 9; 40. 7. 6. 3, 9. 1; U. 2. 3; C. 7. 2. 13; etc. The condition may be such that

alienation destroys the spes, ag. “ to be free if my heres do not sell him,” 40. 7. 30. A sale may be with or without 40. 7. 3. 7, 6. 6, 27, 35.

10 20. 1. 13.1. ^u 33. 2. 20.

ch. xni] Statuliberi. Privileged Position 287 surrendered, and this will discharge the dominus, without affecting their hope of liberty, and if the condition be satisfied during the litigation, the dominus is entitled to absolution¹. As they belong to the heir, they cannot receive a legacy under his will, unless the condition is satisfied at his death, or the legacy is under the same condition as that under which they are to be free². So, too, a statuliber can acquire for the hereditas³.

But there are respects in which their position differs from that of an ordinary slave. Though they can be sold, the sale may not be under harsh conditions⁴: the heres may do nothing to make their position worse⁵. He may be validly directed to maintain them till the condition arrives, cibaria dare, a special enactment of Severus and Caracalla forming an exception to the rule that a legacy cannot be made to your own slave sine libertate³. It would seem to be aimed at preventing the heres from abandoning a slave from whom, as he is about to be free, no great profit can be expected⁷. A legacy of optio servi or legatum generis does not give the legatee a right to choose a statuliber, so long as the condition is possible⁸. Such a choice is hardly likely since he takes his spes with him. But in any case, the rule is merely one of construction: a testator who has made both gifts cannot be supposed to have meant the choice to cover the man freed, and the rule that if the condition fails, he may be chosen, rests on the view, of Q. M. Scaevola, that a gift of which the condition has failed is to be regarded as completely non-existent.

Slaves can ordinarily be tortured as witnesses but statuliberi may not, at least from the time of Antoninus Pius, in ordinary pecuniary cases, though they may in a case of adultery -without prejudice to their ultimate right to liberty⁹.

Where his value is material a statuliber is reckoned at his value as such, e.g., in actio furti and condictio furtiva¹³, and for the purpose of the lex Falcidia. If the slave so freed dies, he is reckoned as part of the hereditas at his value as a statuliber, if, as the event turns out, the condition fails, but if the condition arrives after he is dead, he is not reckoned at all“. A text of Julian deals with a similar question in relation to condictio fu/rtiva. If a thief, or his heir, is sued by condictio furtiva, and the thing stolen ceases to exist, e.g. a stolen slave dies, the plaintiff is still entitled to judgment¹². But Julian says¹⁸ that if a slave,

1 40. 7. 9. jpr·) 2; 47. 2. 62. 9. In this case liberty was attained by paying 10 to heres. He must give this to plaintiff unless it comes out of peculium. This is because the heres might in that case have forbidden the payment without barring the liberty. Post, Ch. xxi.

2 31. 11. pr. 8 40. 7. 28. 1..

4h. t. 25. Such as, ne intra loca serviant, ne unguam manumittantur.

8 h. t. 33; C. 7. 2. 13. ⁶ 30. 113. 1.

⁷ No mode of enforcement is stated. Probably failure would ground appeal to the Emperor under the rule then newly laid down by Antoninus Pius, ante, p. 37. See post, Ch. xx.

a 33. 5. 9. 1. ⁹ 48. 18. 8.1, 9. 3.

io 13. 1. 14. pr.; 47. 2. 52. 29, 81. 1. ¹¹ 35. 2. 11. 1.

¹² 13. 1. 8.1, 20. The thief is always in mora. ¹⁸ 13.1. 14. pr.

left conditionally, is stolen, the heres has condictio furtiva so long as the condition is unfulfilled, but if, pending the action, the condition arrives, there must be absolution. He adds that the same is true if the gift is one of liberty, since the plaintiff has now no interesse and the thing has ceased to belong to the thief, without dolus of his. The second reason is none, for it is indifferent in this action whether the thief is still owner or not¹. The first reason is hardly satisfactory. If a stolen slave is dead, we are told that the action must still proceed, because there may be other interests than the personal value of the slave, e.g. the loss of an inheritance to which he was instituted, on which his master has been prevented from making him enter’. The same reasoning might have applied here. In another text Ulpian says the same thing⁸ about actio furti for a statuliber. But he confines the rule to the case in which the condition is satisfied before aditio, so that the slave never was the heir’s. In the case of condictio, Julian allows release if the condition is satisfied at any time before judgment. This requires altogether different reasoning. It is no doubt, as Windscheid says^{[1096] [1097] [1098] [1099] [1100], an application of the rule soli domino condictio competit. There is nothing remarkable in allowing the cesser of ownership after litis contestatio to affect the matter, in view of the tendencies of classical law, but it seems somewhat unfair in condictio furtiva. It is in fact an application to this action of the principle applied to real actions (though ill evidenced in the texts[1101]), that the plaintiff cannot recover unless his interest continues to the time of judgment.}

As to criminal liability there is some difficulty. In one passage we are told, by Pomponius, that statuliberi are liable to the same criminal penalties as other slaves⁶. But Modestinus and Ulpian say^[1102] that, on account of their prospect of liberty, they are to be punished as freemen would be, in the like case. The contradiction is absolute. Ulpian attributes the rule to a rescript of Antoninus Pius⁸, and as Pomponius is much the earliest of these jurists and wrote some work under Hadrian, it may be that this text states the older law, before the rescript, and that its insertion in the Digest is an oversight of the compilers.

We have seen that statuliberi may be bought and sold, but it is clear that one who wishes to buy a slave will not be satisfied with a statuliber. We are told in one text that if Titius owes Stichus ex stipulatu, and hands him over, the promise is satisfied though he is a statuliber. This view is credited by Ulpian to Octavenus⁹. It deals of course only with the case of a promise of a particular man, who must be taken tails quails. Thus Africanus tells us¹ that if a man hominem promisit, and delivers a statuliber, this is not performance. The receiver can sue on the stipulation, without waiting till the condition is satisfied. Africanus adds that if in the meantime the condition fails, there is no loss and therefore no right of action. It is in relation to sale that this matter is most fully discussed. There are several possible cases.

(a) The slave is sold with no mention of the fact that he is entitled to liberty on a condition. Ulpian says there is weighty authority for regarding this as stellionatus, but whether this be so or not, there is an actio ex empto². But subject to the ordinary rules as to notice, there is also the ordinary remedy evictionis nomine, i.e. the express or implied stipulatio duplae³. This of course depends on the buyer’s ignorance, and is subject to one deduction which may be important. The condition may be si decern dederit, or the like. In that case if the money has been paid to the buyer, as it would ordinarily be, he must allow for it, unless it has been paid out of the buyer’s property, for instance out of the slave’s peculium*.

(b) The vendor says that the man is a statuliber, but does not say what the condition is. Here if he knew what the condition was, but the buyer did not, he is liable, not evictionis nomine, but ex empto³. The text repeats itself. The first statement looks like Scaevoia’s own, which the compilers proceed to amplify, and justify, which Scaevola very rarely does. It is not obvious why the buyer should have any action at all, and this is perhaps what struck the compilers. The rule seems to be that if the vendor gives the buyer to understand that he is buying a statuliber, he saves himself from liability on the warranty but if he does so loosely, perfusorie, laying no stress on it, knowing all the time that it is a likely contingency, he may very well have deceived the buyer, who may bring the actio ex empto. As this is a bonae fidei indicium, the index will ascertain, without any exceptio, whether there was dolus and the buyer was deceived.

(c) The vendor states a condition, but states one entirely different from the actual one. Here the liability is evictionis nomine³. Thus where a slave was to be free on accounting, and there was in fact nothing due on his accounts, and the vendor said the liberty was con­ditional on payment of so much money, he was liable as for eviction, the man sold not being a statuliber at all, but already entitled to his liberty⁷.

1 46. 3. 38. 3. ² 40. 7. 9. 1. ’ 21. 2. 39. 4, 46. 2, 51. 1.

⁴ 44. 4. 2. 7. If the man has already paid the vendor the money or part, this can be recovered by the vendee, but only if he releases the liability ex evictione, 44. 4. 3.

⁴ 21. 2. 69. 5. 6 21. 2. 69. 2.

⁷ h. I. 4. Same rule where one whose liberty was unconditional was sold as a statuliber, h. I. 1.

(d) The vendor states the condition but states it inaccurately so that the buyer is prejudiced. There were evidently different opinions here, the dispute being not exactly as to what the remedy was in this case, but as to what cases ought to come under this head, rather than the last. The case discussed is that in which the vendor says there is a condition of payment, but overstates the amount. Here, on the authority of Servius, the view prevailed that this was not a case for the eviction penalty, but only for actio ex empto^[1103]. In a text of Paul the same rule is laid down, but the remark is added that, if there has been an express stipulatio duplae, the action on this arises. This seems contradictory and the grammatical form of the sentence suggests that it is compilers’ work². A similar case arises where the slave is to be free on accounting, and there is money due on his accounts, and he is sold as decern dare iussus. Here if what is due is less than ten, an actio ex empto arises. If it is the same or more, there is no prejudice, and thus no action’. On the other hand, if the sum is stated correctly, but it is payable to a third person, so that it does not pass to the buyer, this is essentially a different condition, and the facts come under case (c), giving rise to the eviction penalty⁴.

(e) The vendor excepts generally the case of his freedom : here if he is already free or is now entitled to liberty, there is no liability at all. It may be presumed that if he has led the buyer to believe that the liberty is not yet due, there will be an actio ex empto⁵.

All these liabilities depend on prejudice to the vendee. We have already seen this, in the case of the liability ex empto⁵: the same rule is laid down for the liability evictionis nomine. Thus where the condition was si Titius consul foetus fuerit, but on the sale the condition declared was si navis ex Asia venerit, and this latter event occurred first, there was no liability at all⁷. Africanus seems to add that the same rule applies where, though the slave was entitled to liberty on the easier condition, he actually does satisfy the condition stated in the sale. He takes the case of a gift of freedom in one year, the condition stated in the sale having been of freedom in two years. Here, if the slave does not claim the liberty for two years, there is no liability. So also if there was a condition to pay 10 and the vendor says 5, and he pays the 10⁸. Neither of these cases is clear. The second is, as stated, no illustration of the proposition, for the condition named is less onerous than that which actually exists. The case is rendered a little confusing by the fact that if the money is not payable out of the peculium, the larger payment, while more onerous to the slave is also more beneficial to the buyer, who will receive the money. But that does not affect the matter. In the first illustration, the decision is unfair, since the slave is free from the year, and this makes a great difference as to the destination of his acquisitions, during the second year, even though he be in that year a bona fide serviens. There is some authority for reading the text differently¹, and making the real condition two years, and that stated, one. This brings the two illustrations into line, but makes them illustrate only the obvious proposition that if there is no possibility of prejudice to the buyer there is no liability.

The various rules as to immutability of status apply only if the man actually is a statuliber. But he is not a statuliber till an heir has entered under the will. Thus the rule is laid down by Ulpian that, if, before entry, he is delivered by traditio, or usucapted or manumitted, his hope of liberty is lost². But elsewhere he tells us that on such facts, if a heres does ultimately enter under the will, the man’s position as a statuliber is restored favore sui³. He says this only in relation to usucapio, but it is presumably general⁴. Marcellus is quoted by Marcian⁶ as laying down the same rule for usucapio¹¹. Indeed this, and manumission by the usucaptor, are the only cases which are likely to happen, for it is not easy to see how a slave in such a position can be effectively transferred by anyone before aditio⁷.

XIX. Capt/vp.

The circumstances under which a man became a captivus should properly be discussed later in connexion with modes of enslavement, and those under which he regained his liberty with the modes of release from slavery. But as the matter stands somewhat apart from the general law of slavery, it seems best to take it all together. The sources deal almost exclusively with the case of a Roman subject captured by the enemy.

The principle governing the matter is that persons captured become slaves⁸. In general the capture will be in a war, and those captured will be part of a force. But they may be persons taken in the hostile country when the war breaks out⁹, and it is not always

¹ Mommsen, adh. I..

²40. 7. 2. pr. He loses by manumission the position of libertus orcinus.

⁸40. 7. 9. 3. ⁴ Post, Ch. xxi.

5 40. 5. 55.1. He says it might be regarded as due to their culpa that they were so dealt with; except in the case of children: he does not seem to mean that such culpa would bar.

6 The favourable rule is clearly late and is not in the Edict. It seems that the point was decided by cognitio of the Praetor, 40. 5. 55.1. As to whether persons so reEeved were cives or Latins, post. Ch. nm.

? Probably the words sive tradetur are incautiously adopted from the earlier part of the text where they are rightly used.

Âź In. 1. 3. 4. They must be actually removed to the foreign territory, 49.15. 5.1.

» 49. 15. 12. pr.

the case that there is a war: persons who are found in a State with which Rome has no agreed friendly relations are liable to be made captives, though there is no declared war¹. But if it is a war it must be one with a foreign people². Those taken by pirates or robbers, or in civil war, remain free³.

If captured by the forces they become, it is clear, the property of the State⁴. Whether under any circumstances they belong to an individual captor is not clear³. Where they become the property of the State, they do not necessarily become servi publici populi Romani. In many cases they are given freedom⁶. Often they are sold, su6 hasta, or sub corona⁷. Some are made servi populi, with or without a view to their manumission if they properly carry out the duties entrusted to them. Some remain the property of the State but without the status of servi publici, being set to meaner labours, and often, no doubt, intended to be sold in course of time⁸.

The person captured may have been before his capture a slave or a freeman: if he return he is restored to his old position by postliminium⁹, subject to some important restrictions, and in some cases to a redeemer’s lien, both of which will require detailed discussion. While in captivity he is a slave: if he die captive he is regarded as having died at the moment of capture, though there were doubts as to this in classical law¹⁶.

So far as the doings of the captivus during his captivity are con­cerned there is nothing to be said: he is a slave and the ordinary rules of slavery apply to him¹¹: the possibility of postliminium does not affect the matter, any more than the possibility of manumission does in other cases¹². But the case is different, with his property and family left behind. Here the provisional nature of his status is freely expressed in the rules, which can hardly be stated so as to present a logical appearance. It is necessary to consider the state of things during his life in captivity, the effect of his death in captivity, and the conditions and effect of postliminium.

The general principle governing the rules as to transactions and events during the captivity is that the status of the captive is in suspense and the destination of the acquisition, etc., will be determined by the event of death or return. But it is clear that this rule is

1 h. t. 5. 2. See Sueton. Tiberius, 37. So captures might be made by unauthorised raids on such territories.

8 49. 15. 24. s h. t. 19. 2, 21.1, 24; C. 7. 14. 4.

< 48.13.15; Livy, 26.47.

8 Girard, Manuel, 281. The rule of the foreign capturing State would not necessarily be the same.

6 Livy, 6.13; 26. 47; 32. 26; Halkin, Esclaves publics, 17.

7 Marquardt, Vie Priv4e, 1.196.

8 See Mommsen, Droit public, 1. 275; Sta&tsrecht (3) 1. 241, as to the rights of the general in command. See also Blair, Slavery among the Bomans, 17.

» C. Th. 5. 7.1. io G. 1.129.

n See the emphatic language of In. 1. 3. 4.

18 An exception in case of wills to be considered shortly.

ch. xm] Position of former property o/Captivus 293 a gradual development, which in some parts of the law is far from complete: in some, indeed, there is no trace of the rule of suspense; the captivity ends the right.

The son or slave of a captivus can acquire and the effect of an acquisition is in suspense^[MCIV]. There is no conflict of opinion, and the rules are assimilated to those applied in the case of transactions by a servus hereditarius⁹. No doubt the rule applies to all cases of direct acquisition³. This condition of suspense raises difficulties both theo­retical and practical as to the interim ownership. There is here no such conception as the hereditas in which acquisitions can vest, and thus it is not easy to say to whom any acquisition is made in the meanwhile. Paul and Pomponius are clear that the property is not in the captivus*. On the other hand Javolenus says that in retinendo iura... singulars ius est⁹. And Diocletian⁶ and Justinian⁷, speaking of pro­tection of the property, use language which attributes interim ownership to the captive. This is in fact a mere question of language: the real difficulty is that, whether he is owner or not, he is not there to protect his own interests. Some protection must be devised. Certain forms of pillage can no doubt be dealt with criminally, and the fact that no seizer can make a title is pro tanto a protection. But these cannot suffice. The earliest protection of which we know anything is provided by a lex Hostilia, certainly early but not mentioned in any extant text earlier than Justinian, a fact which has led some writers to doubt its authenticity⁸. It authorises action on behalf of those apud hostes, perhaps a popularis actio⁹, for the case of spoliation. But it probably plays but a small part, and in later times we find a more effective remedy in the power, of those who would succeed to the property, to apply to have a curator bonorum appointed in their interests, who gives security to a public slave¹⁰. This curatio does not seem to have been an ancient institution. It is mentioned only by Diocletian and Ulpian^[MCV], the latter speaking of it as a well recognised institution¹². There is no trace of it in the Edict. Moreover, while it is clear that such a curator can sue, and be sued, as defensor¹⁹, Papinian says that if a heres has given security for a legacy and is then captured, his sureties cannot be sued as there is no person primarily liable under the

stipulation^[MCVI]. Ulpian sees no such difficulty in an analogous case². The reasonable inference is that Papinian did not know of this application of curatio. It seems to have been a development from the better known case of curatio bonis in the interest of creditors. There could be no bonorum venditio in the case of a captive debtorÂź, but the creditors could apply, under the general edict⁴, for the appointment of a curator bonis'“. Such a curator has no functions except to protect the property: his powers and duties are in the field of procedure, and it does not appear that he could, by contract or the like, create rights or duties for the estateÂź.

As a captious is himself possessed he cannot possess. It follows that capture definitely ends possession by the captive himself, and thus interrupts usucapio by. him. Nor is there any question of suspense: if he return he does not reacquire possession except by retaking, and his retaking has no retro-active effect, even though no one has possessed in the meantime⁷. Thus his possession is a new one, a fact which may be material, if for instance he has learnt in the meantime that he is not entitled: this will prevent him from usucapting, and bar the actio Publiciana. If the possession was not by himself, but by a son or slave, the rule is the same⁸, unless the matter was one of the peculium. But there is a difference of opinion if the res is peculiaris, held at the time of the capture, or subsequently received. According to Labeo’s view⁹, the rule is the same in this case. This may be the logical view, at least in the case of a slave, since his capacity is purely derivative, and the captive, himself now a slave, has none. But the view which prevails is that of Julian, justified by obvious considerations of convenience, that in this case the possession continues, if the slave still holds, and ripens to ownership by usucapion at its proper time¹⁰. Julian, it may be remarked, says that the usucapio is in suspense so that it will be effective, if the captive returns, but he doubts for the case of his death^[MCVII]. This turns on difficulties as to the fictio legis Corneliae, to be considered later. Marcellus thinks¹² on the other hand that, if he dies, since his death is then supposed to have occurred at the moment of capture, it ought to make no difference whether he or the slave possessed, since from the moment of capture there was a hereditas, and the hereditas iacens vias by his time capable of possession¹⁸. But this view is not

ch. xin] Captivus: Possession: Family Relations 295

accepted. Marcellus goes indeed so far as to hold^[MCVIII] that whether he dies or not there ought to be usucapion of what he has possessed, presumably on the ground that the fiction of postliminium was that he had never been away. But this ignores the real point, namely, that the fiction of presence and the fiction of possession are not the same thing: we shall see that there are many things that the postliminium does not ipso iure undo. From the fact that a captivus has restitutio in integrum, within an annus utilis of his return, or even before his return, if a curator has been appointed to his property², we know that usucapio runs against him ipso iure, that actions by or against him may be barred by time, and so forth. If, as we are told, he does not lose his right as suus or legitimus heres or his right to bonorum possessio contra tabulas, this is precisely because these claims are not subject to any statutory limit³. The case in which a slave of a captivus stipulates and takes a surety, is on the same footing as that of such a contract by servus hereditarius*. This branch of the matter may be left with the remark that if a man has given security for his appearance in court (iudicio sisti), supervening captivity is an excuse and his sureties are not liable⁵.

The law of family relations is governed by the same principle. Paul indeed says⁸ that a captive ceases to have his children in potestas. But this means only that their status is in suspense, just as is that of a captus filius familias⁷. Thus no tutor can be given to one whose pater is captive, and (though there were doubts) Ulpian holds that such an appointment is not merely suspended in operation, but absolutely null⁸. On the other hand the tutela is ended by captivity of the tutor or of the ward, so completely that, though it is possible for the tutor to regain his position, his sureties may be sued⁹. Consideration of the purposes for which a tutor is appointed will shew that any rule of suspense would cause intolerable inconvenience. But if a person otherwise entitled to legitima tutela is a captive the person next entitled is not let in; a praetorian tutor is appointed¹⁹, and thus though we learn that the old tutela is recoverable by postliminium, this cannot be retrospective^[MCIX].

As we have seen, a captive does not lose his rights of succession¹², and there are many texts laying down the rule, for various cases, that if, upon a death, there exists a heres who is a captive, though he cannot himself make a present claim, he excludes from claiming those who would be entitled if he did not exist¹³. In many other ways his existence 296 Position of Captivus: Marriage [pt. i is recognised. Thus his mother must apply for a curator bonorum to be appointed to him, if she wishes to preserve her rights under the Sc. Tertullianum, just as she would have had to get a tutor appointed to an impubes, not captive¹. A captive, or his slave, may be validly instituted², though there can be no question of entry. Similarly, since he is not dead, there can be no entry on his hereditas*. But a child born in captivity to a captivus is no relative to him apart from postliminium*.

It is clear that capture of either party dissolves a marriage, and that it is not restored by postliminium, but only by renewed consentÂź. Paul is reported as saying that, if the wife refuses this renewal of consent without just cause, she is liable to the penalties resulting from causeless divorce’. It may be doubted whether this remark is really from Paul. The evidence for the existence of definite money penalties for causeless divorce in classical law is very doubtful⁷. The cesser of the marriage might seem to be explained by the fact that there can be no connubium with a slave⁸, but this would not of itself account for the refusal to treat the matter as in suspense. And its insufficiency is also shewn by the fact that if the wife is a libertina, freed for the purpose of marriage, the marriage still subsists though the patron be captured, according to a rule laid down by Julian and reported by Ulpian’. The real reason is one of convenience, and the rule brings into strong relief the de facto nature of marriage as conceived by the Roman Law.

But though the marriage has ceased it does not follow that there is complete liberty to marry again, and the texts create some difficulty¹⁰. Justinian in a Novel¹¹ observes that captivity is such a dissolution of marriage as had involved no penalties. But he adds that he takes a humaner view, and lays it down that the marriage is to subsist so long as it is certain that the captive, male or female, is alive; and the other party cannot contract another marriage without incurring the penalties for causeless divorce. But if it is uncertain whether the captive is alive or dead, there must be five years’ delay, after which, if the uncertainty still exists, or the captive is dead, the party at home may remarry, without fear of any penalties, as if there had been a perfectly valid repudium. If this stood alone there would be no difficulty: it is a typical piece of Byzantine legislation. But though it looks like new legislation, and probably is so, as to the continuance of the marriage, it certainly is not absolutely new as to the bar to remarriage. For it is later than the Digest, which contains two texts which speak of a similar rule. Paul is reported as saying¹² that the wife is free to marry post

^I 38. 17. 2. 23, 30. ² 28. 5. 32. 1. ⁰ C. 8. 50. 4.

⁴ 38. 17. 1. 3; 49. 15. 25; C. 8. 50. 1. See post, p. 308.

« 24. 2. 1; 49. 15. 8, 12. 4, 14. 1. « 49. 15. 8.

7 See especially the tenor of C. 8. 38. 2. ⁸ Nov. 22. 7; D. 23. 2. 45. 6.

⁰ 23. 2. 45. 6. 1« Karlowa, R. R. G., 2. 120.

^II Nov. 22. 7. ¹² 49. 15. 8. ch. xm] Position of Captivus: Marriage 297 canslitutum tempus, and Julian as laying down a rule^[MCX] which is sub­stantially that of the Novel except in two respects. He does not say that the marriage continues, but only that it may seem to continue from the fact that remarriage is barred. And he is more explicit as to the date from which the five years are to run: it is the com­mencement of the captivity. Karlowa is of opinion that the five year rule was contained in the lea: Papia, or in connected legislation, since the text of Paul in which he mentions the rule’, is in his commentary on that law, and thus, that the text of Julian, notwithstanding its florid style, is genuine so far as that rule is concerned³. But it is generally thought that the allusion in the text of Paul is interpolated, and Lenel⁴ treats it, and nearly the whole text of Julian, as compilers’ work. And this seems the more probable view. The Novel⁶ refers to earlier legislation dealing with the matter, and we have some of this in the Code. Theodosius provides that on divorce by a wife without cause, she may not remarry at all within five years’. Julian’s opening remark is probably no more than an allusion to the fact that those whose husbands were captured were forbidden, though not absolutely prevented, from marrying for a certain time for reasons sufficiently obvious; the rule applied primarily to widows and divorced women⁷. The compilers build on it an extension of the limit of five years, if it is uncertain whether the husband is alive or dead, to this case of captivity, the rule being a prohibition, but not a bar. The Novel declares the marriage still on foot and applies the five year rule to both sexes equally.

A captive father cannot assent to his son’s marriage, and public convenience makes it necessary to dispense with his approval though in the event this may cause children to be added to his familia without his consent⁸. If he does not assent, says Tryphoninus, at least he does not dissent. This text imposes no delay. But Ulpian appears⁹ as holding that a son can marry only after three years. Paul lays down a similar rule¹" for son or daughter where the whereabouts of the father is not known, and Julian observes^[MCXI] that if a son or daughter of a captive, or absens, marries before the three years are over, the marriage is good if the spouse is one to whom the father could not have objected. These texts do not tell a consistent story, and it is perhaps impossible to extract the development of the law from them. There is no trace of any earlier legislation on the matter, and the jurists could hardly have established the positive term of three years. The texts contain many errors of grammar and peculiarities of diction, which suggest Tribonian, and

perhaps the right solution is that classical law allowed marriage without consent, and Justinian allowed it only after three years of captivity, unless the person was one to whom no exception could be taken by the father¹.

If the person captured was a slave, old rights in him cease to exist, subject, of course, to postliminium². Thus if he has been pledged, the pledge does not exist for the time³. Ownership is gone: he is for the time a servus hostium*. Usufruct in him is gone, though it may be restored by postliminium^. If in the meantime the period of nonuser has passed, it would seem that some form of restitutio in integrum may be necessary since usufruct needs positive enjoyment to its retention, not, like dominium, mere absence of adverse possession. But the text seems to negative this requirement, and the Edict does not mention this case. The slave loses for the time his old characteristics. He becomes incapable of possessing⁶. He ceases to be a servus poenae, if he was one before⁷, and if he has been from any cause incapable of manumission, the defect does not apply to him till postliminium¹. But several texts bring out the suspensive and provisional character of these rules. The slave still exists. Thus an actio de pecidio on his account does not become annalis, so long as he can possibly return with post­liminium. A legacy of him, made before or after he was captured is good, and the heir must give security, not for his value, but for his delivery on his return⁸.

The situation is of course completely changed if the captive dies apud hostes. The matters in suspense are now decided, and the nature of the settlement calls for a good deal of discussion. Matters are in general adjusted as if there had been no captivity⁹. The captive’s children become sui iuris, and though for Gaius it is doubtful from what date their independence is regarded as beginning¹⁰, he stands alone in this doubt. Ulpian indeed says merely that they become sui iuris¹¹. But the Digest is quite explicit, and the texts do not seem to be interpolated. Tryphoninus says that on the death of a captivus his children are sui iuris as from the day of capture¹⁹. Julian lays it down

¹ See Accarias, Precis, § 84. Karlowa thinks (R. R. G. 2. 121, following Bruns) that in the text of Julian a “ non ” has dropped out. The three years rule is classical for captives (hence Paul's doubt as to absence); the compilers extended it to absentee and Julian’s allusion to these is an interpolation. He remarks however that Bechmann reverses this. It may be noticed that if a non is inserted in Julian’s text, the resulting rule is so severe as to give little relief, and that all Karlowa’s argument is equally in favour of the more complete interpolation.

² 35. 2. 43; C. 8. 50. 10. 12, etc. ⁸ 49. 15. 12. 12.

⁴ 40. 7. 6. 1; 49. 15. 5. 2. ⁶ 7. 4. 26.

⁸ 41. 3. 11. 7 49. 15. 12. 16.

⁸ 30. 47. 2. A legacy of A or B where B is a captivus is treated like one in which he is a fugitivus, ante, p. 271.

³ 24.1. 32. 14 in fin.\ 24. 3. 10. pr.; 38. 16. 1. 4, 2. pr.; 38. 17. 2. 7; C. 8. 50. 4.

¹⁸ G. 1. 129. « U. 10. 4.

¹³ 49.15. 12. 1, and thus can have a hereditas, 38. 16. 15.

ch. xni] Death apud Hostes: Succession 299

that what a son of a captive stipulates for, or otherwise acquires, is his own if his father dies still a captive¹. And it is repeatedly laid down generally that in all parts of the law the effect of the death is the same as if it had occurred at the moment of capture^{[1112] [1113] [1114]. So, on his death, since (as we shall shortly see) his will operates, he is restored to his place in his father’s succession, and thus the other representatives, even his own children, may be excluded[1115]. Papinian discusses the case of a son or slave, of a captive, who stipulates in the name of the father or master, and considers the effect if the captive die in captivity[1116] [1117] [1118]. He observes that though, in a simple stipulation, the effect would be different in each case, since the slave would benefit the heir of the dominus, and the son would benefit himself alone, yet here they are on the same level. Both are bad. In the case of the son this is because he stipulates alii, non sibi: it is in fact for a non-existent person. In the case of the slave it is a stipulation for his dead master, which, as we have seen, is void8. One Case is peculiar. We are told by implication, by Paul8, that where a slave is captus, the actio de peculio on his account becomes annalis on his death, but it is evident that the year is not counted from the capture.}

It is in connexion with the succession to the dead captive that the most difficult questions arise. The man dies a slave and cannot in strictness have any will or indeed any inheritance^[1119]. We are told ex­plicitly by Ulpian that his will becomes irritant on his capture⁸. As he is a slave he cannot make a will in captivity^[1120], and though as we shall see there is relief against the destruction of a previously made will, codicils made during captivity are not confirmed by it: they are not valid even for the purpose of creating fideicommissa, since at the time of making them he had not testamenti factio*. Whether, in very early law, succession to such a person did not exist at all cannot be said. Perhaps the relief developed pari passu with the notion that a captive suffered capitis deminutio maxima¹â€œ. However that may be, it was provided, directly or indirectly, by a certain lex Cornelia¹¹, that the succession to such a person should be regulated as if he had died in civitaie¹³. The exact nature and scope of this provision cannot be clearly made out from the texts. The rule is sometimes stated as a direct

300 Succession to Captivus: lex Cornelia [pt. i

provision of the lew^[1121]. Sometimes it is called the beneficium legis Comeliae⁶, sometimes the fictio legis Cornelias³. While many texts speak of the rule as creating a hereditas⁴, others do not use this ex­pression, and Ulpian, in two texts, declares that it is not, strictly speaking, a hereditas, giving, as his reason, that a man who dies a slave cannot have a hereditas, and that one who could not make a will cannot properly be said to have died intestate. He adds that as succession is given to those to whom it would have been given, if he had died in the State, it is treated as if it were a hereditas’¹. It is widely held that the rule is not a direct provision of the lew⁶, but a juristic interpretation of something therein, and conjecture has gone so far as to assume that this lew Cornelia is identical with the lew Cornelia de falsis, and that it contained a rule punishing the forgery of the will of one who was apud hostes, from which the jurists inferred that the will of such a person would be valid⁷. For this opinion there seems to be no evidence what­soever⁸, and it is difficult to hold it in face of the texts which say that the lew did contain a direct and express provision on the matter⁹.

It is not easy to say what this provision was. It is fairly certain that it was in a form which left room for doubts, since it is observable that a large proportion of the texts dealing with the matter are from collections of Quaestiones and Disputationes’⁰. It is also most probable that it was in the form of a fiction^[1122], and, from Ulpian’s language¹², that it did not speak of heres or hereditas. Thus Paul tells us not that the lew but that the fictio legis Corneliae et heredem et hereditatem facit¹³, though elsewhere both he and Papinian use more unguarded language¹⁴. One or two of the texts give us what purport to be explicit statements as to the content of the lew. Ulpian says that the lew confirms the will as if he had died in dvitate¹⁶. Julian says that by the lew Cornelia the state of things is to be that...quae futura esset si hi de quorum heredi­tatibus et tutelis constituebatur in hostium potestatem non pervenissent¹⁶. Paul says it confirms wills, legitimae tutelae and legitimae hereditates¹¹. On all this evidence it seems clear that the lew itself, without speaking of heres or hereditas, confirmed the succession to him as if he had not been captured, a provision which exactly accounts for Ulpian’s purist ch. xm] Succession to Captivus: lex Cornelia 301 objections, which do not in the least require that the whole rule be one of juristic inference¹. The lex also dealt with tutela, as to which provision there is some difficulty. It had nothing to do with provision for tutelae in his will: it is expressly distinguished from the provision affecting wills². It cannot have referred to a tutela held by or over the captive: he was dead, and in any case such tutelae were ended by the capture³. The tutela* is legitima. Karlowa⁶ suggests that the provision referred to tutela. of his children left behind. But it is difficult to see any need for this: it could not be retrospective, and these children and their agnates had suffered no capitis deminutio. More probably it was tutela over liberti. The lex enabled the children to claim tutela as liberi patroni, the captus not having really been patron at the time of his death, apart from the fiction.

The lex left open the question at what date the will was supposed to operate, and the lawyers developed the principle, which as we have seen, came ultimately to be applied generally⁶, that the case was to be treated as if he had died at the moment of capture⁷. The texts do not state this refinement as part of the lex, which also left open the question of the application of the rule to pupillary substitutions— secundae tabulae—a matter which gave rise to much discussion. Apart from these substitutions the working of the rule is fairly simple, and can be illustrated very briefly.

Its general effect is that the succession is to go to those who would have had it if the deceased had never been captured⁸, there being a right to enter as soon as it is known that he is dead⁹, and if there are no heredes under the lex, the property goes to the State”. It is only wills made before capture which are thus validated¹¹. The will can produce no more effect than it would if he were in the State. Thus where a man was captured while his wife was pregnant, and a child was born and died, the will was null, as it would have been had he not been a captive¹². The validation of the will does not amount to post­liminium, and thus if a child is born apud hostes, and returns, but the father dies there, the child can have no claim under the lex Cornelia: he is a spurius, the father being regarded as having died at the moment of capture¹³. If a filius familias miles dies apud hostes, the rule applies,

^I See also Julian in 28. 1. 12. ² P. 3. 4a. 8; 49. 15. 22. pr.

³ Karlowa, loc. cit. ⁴ P. 3. 4a. 8.

⁵ loc. cit. ⁶ Ante, p. 298.

’ 29. 1. 39; 38. 16.15; 49.15.10. pr.; h. t. 11. pr.; an enactment of Severus and Caracalla almost states this refinement as a part of the lex, C. 8. 50. 1.1.

³ 49. 15. 2'l.pr. 9 C. 8. 50. 4. ’0 49. 15. 22. 1.

^II In. 2. 12. 5, etc. Codicils made during captivity could not be confirmed by anticipation in a previously made will, and as having been made by one without testamenti f actio, they were invalid to create fideicommissa, 49. 15. 12. 5.

¹² 49. 15. 22. 4.

¹³ 38. 17. 1. 3; 49. 15. 9, 25; C. 8. 50. 1. 1. A slave made heres will be free and heres whether he wishes to be or not, and a son will be heres at civil law, though as Julian observes, the expressions necessarius and suus are not strictly applicable, 28.1.12.

302 Succession to Captivus. Pupillary Substitutions [pt. i and if before his death his father has died, leaving a grandson (by the miles), and the grandfather has omitted the nepos, his will fails, but that of the soldier does not, if he has omitted his son, both rules being due to the fact that he is supposed to have died at capture while still a filius familias¹. If a child, captured with his parents, returns, but they die in captivity, he has a right to succession by the lex Cornelia’.

The effect of the rule, on pupillary substitutions, is discussed in several texts some of which give rise to difficulty³. If the father is captus and dies, and then the impubes dies, the pupillary substitution takes effect, as the lex covers all inheritances passing by the will of the captive⁴. To the objection that in fact the child was sui iuris before his father’s death, Papinian answers that the accepted principle is that on his death the captive is regarded as having died at the moment of capture³. It is likely that the rule was not so settled till the classical law. In the following text Papinian seems to adopt the view that he has here rejected⁶, but it is probable that, as Cujas suggested, the words nihil est quod tractari possit do not mean that there can be no question, i.e. of pupillary substitution, but that the case can give rise to no difficulty. Probably there has been abridgement.

If the son alone is captured after the father’s death and dies impubes the lex applies and a substitution will be good, but this is really an extension of the lex, since the person who dies is not the actual testator, and the lex says only that the will of the dead captive is to be confirmed. Accordingly Papinian says the Praetor must give utiles actiones⁷. If the father die after the son’s capture the lex has no application, for this would be to give the provision more effect than it would have had if he had died at the moment of capture, when he would have had no bona*: he has, in law, predeceased his father, and that destroys pupillary substitutions’.

There is difficulty in the case where both are captured. There are two texts, both obscure. Papinian says¹⁰:

Sed si ambo apud hastes et prior pater decedat, sufficiat lex Cornelia substitute non alias quam si apud hostes poire defuncto, postea filius in civitate decessisset.

The only rule stated in this text is that if both die in captivity the substitution fails while if the son returns and dies impubes it takes

¹ 29.1. 39. The traditional explanation of an obscure text. See Pothier ad h. I.

² C. 2. 53. 5. The text adds (it is an application of a rule already stated), that he has restitutio in integrum, like any other captive, against usucapio, etc., within an annus utilis. In such a case it is an extension of the restitutio, since it is not he against whom tima was running and he is not an ordinary successor. The rule is not confined to mUites, 4. 6.15. pr.

³ Muhlenbruch-Gliick, 40,449 sqq.\ Buhl, Salvius lulianus, 1. 267 sqq.

* 28. 6. 28. ³ 49. 15. 10. or. « X. t. VI. pr.

1 28. 6. 28; 49. 15. 10.1.. 8 28. 6. 28.

⁹ So, if he had been a pubes, captured while his father was alive, the lex has no application, though the father ultimately die first.

w 49.15.11.1.

ch. xm] Succession to Captivi. Pupillary Substitutions 303 effect, a decision which agrees, as Miihlenbruch observes, with the presumption, where an impubes and a pubes have died at unknown dates, that the impubes died first. But its opening words seem to foreshadow treatment of a case in which both die apud hostes, and as it stands the word prior serves no purpose. Among the proposed explanations^[1123] is that of taking non alias to mean similiter, and taking the text to mean that though both die in captivity, the substitution can take effect, if it be shewn that the father died first. But this is arbitrarily to change the meaning of words, though the rule itself would not be unreasonable. The earlier commentators, cited by Miihlenbruch, are not quite satisfied with it, and suppose that the father was captured first and the words et prior pater decedat are a gloss. This is very conjectural. Perhaps the suggestion of the insertion of the word sint after the first occurrence of hostes is best², though that does not account for the word prior. Better still is it to admit that we cannot reconstruct the original text: it is impossible to be satisfied with it as it stands.

The other text is from Scaevola³:

Si pater captus sit ab hostibus, mox filius, et ibi ambo decedant, quamvis prior pater decedat, lex Cornelia ad pupilli substitutionem non pertinebit nisi reversus in civitate impubes decedat, quoniam et si ambo in civitate decessissent veniret substitutus.

Here one would expect the substitution to be effective, since if each had died at the moment of capture, the father would have died first. Miihlenbruch⁴ considers however that the rule is correctly stated in the text. He remarks that the hereditas is not delata till the actual death, that the impubes could have had no property, and that thus no one could inherit from him. The point as to delatio is hardly material, but this seems the right explanation of these texts. The child is a captive who had nothing at capture, and, not having returned, he has no post­liminium. He can thus have acquired nothing by hereditas or otherwise. Thus there is nothing on which the substitution can operate⁶. The allusions to the date of the father’s death suggest doubts as to the reference back to the date of capture. But as it is plain that the text now under discussion is not as it was originally written⁷, it is also possible that these allusions here and in the other text are hasty interpolations. Many emendations have been suggested, starting from different points of view, but none of them is satisfactory⁸. The only thing certain is that Scaevola did not write it as it stands⁹.

The next topic for discussion is postliminium. The prisoner of war who, under certain conditions, returns to Roman territory, is restored to his old legal position, with some limitations, this right of postliminium being suspended, if the captive was redeemed for money, till the redeemer’s lien is paid off. There are thus three topics: the conditions of postliminium, its effects, and the law as to a redeemer’s rights.

As we are concerned only with captivi, we shall not consider a voluntary change of State in time of peace. Most of the requirements for postliminium can be shortly stated. Not every captive who escapes is said postliminio redire'. He must actually have returned to the territory or to that of a friendly State^{[1124] [1125], though it matters not how he effects his escape, whether by evasion, vi aut fallacia, by dismissal, exchange or recapture[1126]. He must have returned as soon as it was possible for him to do so[1127]. He must come to stay, not having any intention of returning to the enemy[1128]. It may be that this requirement is of classical law though it is expressed by the traditions of republican Rome, which however are not contemporary. Regulus was declared not to have postliminium, not because he had sworn to the Carthaginians that he would return, but because he meant to keep his oath, non habuerat animum Romae remanendi[1129]. The captives who were sent by Hannibal to Rome, on the same mission, and who stayed there, and shewed that they had never meant to go back, notwithstanding their oath, had postliminium, though they were declared ignominiosi and intestabiles for not keeping their promise[1130].}

Discreditable circumstances might bar postliminium. Thus, one who had surrendered when armed had no postliminium*. One who had been deditus by the pater patratus had no such right, though apparently this had been doubted^[1131]. There was a tradition of a difficulty in the case of Mancinus, who had been so surrendered to the Numantines, but they had refused to receive him. It was held, on his return to Rome, says Cicero¹⁰, that he had no postliminium. Else­where the same authority throws doubt on the decision, or, rather, holds that the man had never ceased to be a civis¹¹, since there could be no such thing as deditio, any more than there could be donatio, without an acceptance. There is no postliminium for a transfuga, i.e. one who treacherously goes over to the enemy, or to a people with whom we have only an armistice, or one with whom we have no friendly relations¹. As we shall see shortly, this rule is not applied in the case of slaves. But this exception is strictly construed: the rule applies in all its severity to the case of a filiusfamilias, notwithstanding his father’s rights^{[1132] [1133] [1134]. In strictness of course a person who never was a Roman civis cannot have postliminium. Even in the case of children born to captivi, this rule is so far applied that if the child reaches Rome, but neither parent does, he has no civitas: if he and his mother come he is a civis and her spurius: if he and both parents come, he is a filiusfamilias*.}

A further suggested requirement of postliminium has given rise to some discussion. It is laid down by some writers that to obtain postliminium the captive must return eodem bello, i.e. before the con­clusion of peace^{[1135] [1136]. This view seems to rest mainly on three texts. Pomponius tells us that a captive si eodem bello reversus fuerit postli­minium habet*. Paul says that if a man returns during indutiarum tempus, i.e. during an armistice, he has no postliminium*. Finally he tells us[1137] that if a captive returns after peace has been declared, and on a fresh outbreak of war is recaptured, he reverts to his old owner, by postliminium. The first text of Paul is of no weight: an agreement for an armistice involves, as he says, the maintenance of the status quo so long as it exists: indutiae and peace are not the same thing. The other text of Paul looks more weighty, since, as Accarias says, if he is still a slave of the old master, he has not been freed. But he was not freed by his old master, and the more reasonable inference is that postliminium gives him restitution only against his own State, apart from some special agreement. It is the allusion to the peace that is regarded as making the text important in the present connexion. But it may be noted that if the war is still continuing, my slave escaping to the enemy is a transfuga, and by virtue of the special rule in such cases will revert to me if recaptured[1138]. Paul purposely takes a case to which this special rule would not apply. Neither of these texts is conclusive. That of Pomponius remains. Against it may be set a text of Try- phoninus9, which says that there is postliminium on return after peace is made, if there is nothing in the treaty of peace to exclude it, and goes on to make the same remark about those who were caught in the foreign country at the outbreak of war. There are other considerations. Returns after peace must have been not uncommon, and one would have expected discussion of the case of those who had not postliminium. Yet the texts, though they tell us that not every escaped prisoner has postliminium, say nothing about the state of things where it does not arise. And that postliminium is the common case appears from the fact that many texts speak of return of captives as giving postliminium, without more1. It may be added that while we have reference to agreements that there shall be no postliminium, after the peace*, we have none the other way: what we have are agreements that prisoners shall be allowed to return’, which is a different matter. As Karlowa says[1139] [1140] [1141] [1142], a war ends either by surrender of the enemy—deditio—or by a treaty of amicitia, and he points out that we are told that between states in friendship there is no postliminium, since the subjects of each state retain their rights in the other[1143]. But the reason shews that the writer is dealing with cases arising after the foedus, not with persons who were captives at the time of the treaty.}

It may also be observed that the principal text ’ is not conclusive: it lays down a right of postliminium if the captive return eodem bello, and it is only in that case that a general proposition is justified: the fact that the treaty might exclude the right compels the limitation. Never­theless the text is regarded as so conclusive as to require an emendation of the above cited text of Tryphoninus which is even more conclusive the other way⁷. For his it is proposed to read non iis. This drastic measure is defended as being necessary to account for the words: quod ideo placuisse Servius scribit, quia spent revertendi dvibus in virtute bellica magis quam in pace Romani esse voluerunt, and for the contrast estab­lished later on in the text. But the remark of Servius is quite plain as the text stands. It is the practice of exclusion by treaty that he is justifying. And the contrast is between those who left Roman territory in time of peace and those who did so during the war, and no more⁸. If the contrast had been that supposed by Karlowa, the jurist would have said tarn in pace quam in bello and not tarn in bello quam in pace. The other contrast is an unreasonable one: a person who is in the other country at the outbreak of war is to be allowed an unlimited time for return, but one captured in the war is not. And the absurdity is emphasised by the jurist, who notes that these persons are there, suo facto”. There is no justification for so altering texts as to create this

ŃĐœ. хш] Postliminium: Conditions and Effects 307 distinction, especially as those captured without war by a State with which there are no friendly relations are clearly under no such restriction. Altogether the requirement that, to obtain postliminium the captive must have returned before the peace, is not made out. It is impossible to be certain on the point, but the most probable view seems to be that there is no such rule, but that treaties of peace sometimes exclude postliminium for later returns.

What is the position of one who returns without postliminium ? There is but little authority. Transfug am, we are told, iure belli recipimus^[MCXLIV]. This seems to mean that a slave transfuga reverts to his master. If he is not a slave he is at best a prisoner of war; but he is ordinarily capitally punishable’. Apart from this case we know really nothing. Most probably their previous Roman condition is simply ignored, and they take the position, whatever it is, which they would have taken had they belonged originally (either as slaves or freemen) to the State from which they came.

We have now to consider the effect of postliminium. The general effect is to put the civis in the same position as if he had never been captured³. The captives, according to the traditional formula, recipiunt omnia pristina iura*, and, it may be added, obligations too⁵. Moreover the effect is retrospective⁵: fingitur, says Justinian, semper in civitate fuisse⁷, and this expression clearly represents the classical law. It is, says Paul, ac si nunquam ab hostibus captus sit, and the same conception is involved in the notion of pendency of rights, already considered⁸. It is not necessary to repeat, by way of illustration of the general rule, what has already been said. It may be observed in addition that he has actio furti for what has been stolen in the meantime’, and probably the same is true of actio Aquilia. It may also be presumed, though the texts are silent, that he is noxally liable for what has happened in the meantime. An actio de peculio and the like will probably lie against him, though the texts are silent, and those dealing with restitution of actions, to and against him, where time has run, are not specific as to the date of the transaction¹⁰. But there can really be no doubt. A person who manages any affair for the captivus has an action on negotia gesta on his return^[MCXLV]. The captive can sue, on his return, on contracts

made, and securities taken, by his son or slave¹, and, in general, he is in the position of one succeeding to a hereditas iacens¹.

The questions as to the effect of postliminium, on the marriage of the captivus or of his child, and on tutela, have already been discussed’. There remain however some topics to consider. The rules as to the effect of postliminium on patria potestas are in the main fairly simple. The father, returning, regains potestas over his children, and acquires it over those of whom his wife was pregnant at the time of the capture, and even over the issue of any lawful marriage contracted by a son so born^{[1146] [1147] [1148] [1149] [1150]. In the case of a child born to a captive there was evidently some difficulty. There could be no postliminium for one who was never captus. The law seems to have been settled by a comprehensive rescript of Severus and Caracalla, part of which is preserved in the Code. The practical result is that postliminium is given to those bom in captivity, in right of their parents, but not otherwise, so that if both parents return the child is in potestas, with its parent’s status, and with rights of succession[1151] [1152]. But if the mother alone returns, with the child, the father dying captive, the child has no succession to the father, since the lex Cornelia, as it assumes the father to have died at capture, cannot give rights to those born in the captivity6. It is thus connected only with the mother, to whom it stands in the relation of any other child sine poire''. It is noticeable that no text mentions the case of the father returning with a child bom in captivity, the mother having died apud hostes: the chief texts especially require the return of both parents[1153]. Indeed since the marriage ended at capture, and the mother is supposed to have died then, it is difficult to see how the child can be regarded as having been born in wedlock.}

His will made before he was captured is valid^[1154]. Postliminium does not validate what he has done as a captive, and thus a will made during captivity is void. The same is true of codicils even though an earlier will has confirmed them, since when they were made he had not testamemii factio. But Justinian alters a text of Tryphoninus, in a dubious way, so as to make him say that on grounds of humanity, they are to be enforced, if he returns with postliminiumℱ. On the other hand, he or his slave may be instituted, and if he return he may make or order the necessary entry¹¹.

ŃĐœ. хш] Postliminium: Effects: Possession 309

As to ownership, a returned captive acquires at once all that is in eodem statu; and as to those rights which have been lost by usucapio or by non-use, he has, as we have seen, an annus utilis within which to recover them by utiles actiones. This is provided for in the edict as to restitutio in integrum for those over 25¹. If a son has died in the meantime, his acquisitions vest in the returned captive’.

He has ceased to possess and thus, subject to what has been said as to possession by sons and slaves peculiari causa*, he has no possession till he has actually retaken it, even though no other person has possessed the thing in the meantime*. Postliminium has no application here⁶. This may have the result of barring his usucapion altogether. For it is a new possession, and if, since the purchase, he has learnt that there was no title, there can be, on well known principles, no usucapion’. It is immaterial whether he was already usucapting or a slave has received the things not ex peculiari causa during the captivity. If however he is still in good faith he will usucapt. The question whether he can add his former possession or must begin afresh is not resolved by the texts. Most writers hold⁷ that the earlier time is quite lost. But the texts are far from conclusive⁸. They make it clear that possession is interrupted, and that it is not restored by usucapion. But they do not clearly deny that in such a case the two possessions can be added together, if both were begun in good faith, as in the case of a buyer, whose possession is equally a new one. They speak of nova possession secunda possession, which it no doubt is, but they also speak of reciperata possessio¹¹. They say that there can be no usucapio, if the second possession was begun in bad faith“, and that, in any case, postliminium ei non prodest. But the words are added ut irideatur usucepisse'*, which are limitative in form and are not inconsistent with allowance of accessio temporum. But the accepted view is that of most modem systems of law in analogous cases, and the silence of the texts as to accessio is in favour of it¹*.

In the case of slaves, postliminium does not operate quite in the same way as in the case of freemen. It is a case of postliminium rerum, the principle of which is that ownership reverts only in those things which are useful in war, except arms and clothing, which, it is said, cannot have been lost without disgrace¹⁶, a slave, of any sex or age, being regarded as ^{[1155] [1156] [1157] [1158] [1159] 310 Postliminium, where Captive had been a Slave [pt. i a thing useful in war, since he can be employed in carrying messages or advising, and in many ways other than actual fighting1. No reason is assigned for this limitation’. In any case the rule has the odd result of discouraging the private recapture of such things, since the actual captor gets nothing out of it. A slave, though a thing, has a will of his own, and hence it is easy to see possible conflict between the traditional principles of postliminium and regard for the interests of the master. In three points the latter prevailed.}

(Đ°) A transfuga has no right to postliminium³, but a slave transfuga does revert to his master. This is only in the latter’s interest. If he has been a statuliber, and the condition has arrived during his absence, he loses the benefit of it⁴, but the text is not very clear as to what does happen. Presumably he is punished as an ordinary transfuga: at any rate he does not revert, as his owner would not have had him after the arrival of the condition in any case.

(б) A freeman does not recover his rights \yy postliminium, unless his return is with intention to remain: a slave reverts by postliminium in any case’.

(c) A freeman has returned as soon as he is in Roman territory, but as it is not on his own account that a slave has postliminium, he does not revert till he is in his master’s possession, or in that of someone, as a slave·.

Apart from those points the case is simple. If other conditions are unchanged, not only does his owner’s right revive⁷, but so do minor rights, such as usufruct⁸. If he was a serous poenae before, he is one still⁹. A serous furtivus, captured, recaptured and sold, remains a res furtiva, incapable of usucapio by the buyer¹⁰. If he was a statuliber, he is a statuliber still, but if the condition has arrived during his captivity he gets the benefit of it¹¹. A legacy of him or to him or an institution of him before or during the captivity is effective on his return ^u. More­over the captivity has operated no capitis deminutio: he is eadem res, and thus an eaceptio rei iudicatae, which applied to him before the capture, applies to him still¹â€™. If he has been pledged before capture the creditor’s right revives¹⁴.

¹ 49. 15. 19. 10.

⁸ Karlo wa, op. cit. 2.125, associates it with the alleged rale that there was^stZtmtmwn only in the same war: he treats it ap an inference from that rale. There is no obvious connexion: it is not here confined to the same war.

‱ 49. 15. 19. 4. ⁴ h. I. 5, 6. ⁶ h. t. 12. 9. « h. t. 30.

^I C. 8. 50.10,12; Festus, s.v. Postliminium; 49.15. 28, etc.

⁸ 7.4. 26. If he returns after his owner’s death he is in the hereditas for Falcidian purposes, 35. 2. 43.

» 49. 15. 6. io h. t. 27.

^IIh. 1.12.10. And thus the child of a statulibera Captiva whenever conceived is ingenua if the mother returns with postliminium, 40. 7. 6. 1, 2. The application of the rale to one conceived in captivity is a humane extension. Cp. C. 8. 50.16.

¹² 30. 98. 1« 44. 2. 11. 4.

¹⁴ 49. 15. 12.12. If before the captivity he had been sold on the terms that he was not to be freed, or was subject to another impediment to manumission, the prohibition still applied if the conditions were unchanged in other ways, 49.15.12.16. Ante, p. 72. Post, Ch. xxv.

ch. xm] Postliminium. Redemption 311

It remains to consider how far these rights are suspended in the case of redemption for a price, by a third party. In the Christian Empire the provision of a fund for the redemption of captives seems to have been a usual form of charity. There is no trace of anything like organisation of such funds till the time of Justinian, who provided that if a man left all his property for this purpose, by making captivos, generally, his heirs, the institution was to be valid, the estate being kept in perpetuity for this purpose, the annual profits of all kinds, from rents and sales of produce, being applied without any reduction for the cost of administration. This was to be managed by the Bishop and the Oeconomus of the testator’s place of residence, who were to have all the rights and liabilities of the heres\ Somewhat later he authorised the local churches to alienate any land, which had been given to them without any prohibition of alienation, for the purpose of redemption of captives^{[1160] [1161] [1162]. Again, he provided that if any provision was made by will for the redemption of captives, and the will did not say who was to carry out the redemption, the Bishop and Oeconomus were to see to it. If someone were appointed, and neglected, after two warnings from the Bishop, to carry it out, his benefit was forfeited to the Bishop, who was to administer the whole for charitable purposes8. A provision inspired by the same spirit is that which makes neglect to redeem any ascendant a just cause of exheredatio, while if the ascendant dies in captivity, the neglectful descendant’s share is forfeited to the Bishop and goes to redeem captives. And, generally, where any person is entitled on intestacy (whether a will has also been made in his favour or not) to succeed to any extraneus who is a captive, or, not being a relative, knows that he has been instituted heir, neglect to ransom is punished by similar forfeiture of benefits to the purpose of redemptions, the rest of the will standing good[1163] [1164]. Justinian had also provided that donations for the redemption of captives were to be exempt from the rules as to registration8. It may be presumed that all these provisions as to the duties of relatives, etc., apply equally to the payment of money due to an actual redemptor, under the rules now to be considered.}

A complex situation arises where a man has redeemed a captive, by paying a ransom. The general rule applied here is that the ransomer has a lien on the redeemed captive, and postliminium, with its various results, is postponed till the lien is ended’. There is no lien except for actual redemption money. Thus there is no lien if he is simply 312 Condition of Captive redeemed for a Price [pt. i captured from the enemy^[1165], though at a certain cost’. Nor is there any lien if the redemption is of relatives, pietatis causa, even though redemption is for a price, and the payer afterwards seeks to re­cover it⁸.

The state of things so long as the money remains unpaid cannot easily be expressed in terms of any other institutions. It is apparently rather illogical. We are told that it is a state of pledge, not slavery⁴, and yet we know that the captive has not yet postliminium from his captivity in which he was a slave. The practical meaning seems to be that the lien in itself has no enslaving effect, no reference being intended to the provisional slavery involved in capture. The significance of the proposition is shewn in the rule that when the lien is ended the old status is restored: -the man is not a libertus and owes no obsequium to the redemptor⁶. The disabilities under which he suffers are not the result of the lien, but of the fact that he has not yet postliminium. He can serve no militia*. Apparently he cannot validly marry⁷. He cannot in strictness enter on a hereditas, but, favore ingenuitatis, he is allowed to do so, or to receive a legacy, so that the money may be applied to the release from the lien⁸. A child redeemed with the mother is under the lien⁹. An enactment of Diocletian lays it down as undisputed law that a child bom to a woman under such a lien is not itself subject thereto¹⁰. But Ulpian seems to imply that the pledge covers such issue^[1166], and Tryphoninus must have been of the same opinion, at least where the person redeemed was originally a slave¹⁸. But this is only a case of the dispute already considered as to whether partus is covered by pledge¹⁸. Pledge is not the only close analogy. The transaction is, from the captors’ point of view, a sale: the redemptor is constantly spoken of as buying the captive, and it is part of the argu­ment of Tryphoninus in the text just mentioned that the partus is to be treated as sold with her. Here too the same point has already been considered, with special reference to eviction and usucapion¹⁴. The better view is that in later law the pledge covers the partus, though the classics are not agreed¹⁸.

The texts are not clear as to what advantage the redemptor can claim from a redeemed ingenuus. He can assign the lien, but not so as to increase the amount payable by the redemptus, the person who takes ch. xmj Redemptio for a Price: Rights of Redemptor 313 it over having an action against the redemptor for any excess^[1167]. We must assume that the redemptor is bound to maintain the man, and we have seen that his lien covers only what he has actually paid to the captors⁹. Unless we assume the redemptor to be a philanthropist of a most unselfish kind, and therefore the case of redemption kept by the law within very narrow limits, we must suppose that he may employ the man. We shall see that in later law, the captive may pay off the lien by labour’, but this of itself does not prove that he can be made to work. As to acquisitions through an ingenuus so held we have no information. A pledgee does not take acquisitions, and the language of pledge is constantly used in this connexion⁴, with the implied warning that it cannot be a true pledge, since the man is not a thing’. Thus he is so far in the holder’s potestas that he cannot witness his will, and this, not slavery, is given as the reason⁶; and the reason why the interdict quern liberum does not apply to him is only that the holder non dolo facit⁷. The discharge of the lien is called luitio, the primary meaning of which is discharge of a pledge⁸. On the other hand, his purchase is regarded as an emptio⁷, and where the redemptus was originally a slave the redeemer becomes, as we shall see later, his owner, in a limited sense¹⁰. Tryphoninus discusses the case of a statuliber under a condition of payment, and decides that the money can be paid by the redemptus out of any part of his peculium except what is acquired ex operis suis or ea: re redemptoris^[1168]. This does not however imply that the holder acquires, like a bonae fidei possessor¹⁷, only ex re sua and ex operis servi, at any rate where the person actually was a slave: a special constitution created logical difficulties in that case¹â€™. We shall see that the redemptus may free himself by labour, but we hear nothing of his freeing himself by means of his acquisitions. This may mean no more than that such a dealing would come within the general rule¹¹ as to repayment of the money. The texts clearly contemplate his paying the money himself¹⁴, and it may be that his acquisitions ex re redemptoris and ex operis suis go to his holder, while other things go to himself.

The lien may be ended in various ways. Actual payment of course suffices¹⁶. Children redeemed with a man or woman may be freed from the lien by independent payments; either the sum specifically paid for them, or if there was no allotment, then a proper proportion¹⁶.

314 Redemptio for a Price: Lien, how ended [pt. i

It may be ended also by tender and refusal of the redemption money^[1169], or by any remission of the debt, which, however informal, cannot be revoked’. It is no doubt on this principle of remission that it is ended by the redeemer’s marrying the captiva³, indeed we are told that remission results from cohabitation with her⁴. The same result follows from his instituting the captive as his heir’. The pledge is ended, and the right to the money forfeited, if the redemptor pro­stitutes, knowingly, the woman redeemed⁸. It seems further that, at least in later law by a constitution of A.D. 408⁷, five years’ service ends the lien, at least in the case of a rivis captured. It is clear from some of the rules just laid down that the lien is not affected by the death of the holder. In earlier law the death of the captive, though it of necessity destroyed any practical lien, left the debt standing and pre­vented the heirs from succeeding till they had cleared it off⁸, the result being that they were worse off than if he had died still a captive. But Ulpian here mentions, and elsewhere accepts without comment⁹, so that it is clearly the later law, a doctrine more favourable to the successors. The death ends the pledge: the redemptus gets postliminium and is restored to his old status so that the whole obligation is blotted out.

The effect of luitio is to bring into operation the ordinary post­liminium¹³. Heavy penalties are imposed by Honorius, by the enactment of 408, on those who detain captives on whom there is no lien, or the lien on whom is from any cause ended. If the undue detention is caused by an agent, the principal being away, he is to be liable to deportatio, or even to condemnatio in metallum: if it is by the principal himself, he is liable to deportatio and forfeiture. To assist in enforcing this law the local clergy are to watch over such cases, and the curiales of the neighbouring localities are liable to penalties of 10 aurei, they and their apparitores, if they fail to see to the carrying out of the law^[1170].

Where the redeemed captive is a slave, there are special rules of some difficulty. Here too, though the slave is property, there is no lien except for money paid for redemption¹²: recapturers must give him up at once and have no right in him¹⁸. On repayment of the money he goes back to the dominus¹³, and any old rights in him, e.g. pledge, revive. ch. xiii] Effect of Redemptio of one who had been a Slave 315

In fact a pledgee can pay off the lien, and add the sum to his charge, just as a creditor with a subsequent charge can confirm it by paying off a prior incumbrancer¹. Apparently the lien cannot be paid off pro parte by one of common owners, but if all pay their share, or one pay the whole in the name of all, the lien is at an end; in the latter case the payment will come into account in the actio communi dividvmdo. If he is acting for himself or for some of the others, then, as to their shares, the lien is at an end, and the common ownership restored : as to the others it is an assignment of the lien, and the payers are in the place of the redemptor*.

For convenience the right of the redemptor has been called a lien. In fact it is a great deal more. Tryphoninus tells us that a certain constitution protinus redimentis servum captum facit^{[1171] [1172] [1173]. We have no information as to what this constitution was. Karlowa[1174] [1175], in view of the form of the allusions to it, thinks it to have been a general provision, and he considers it identical with the constitutio Rutiliana, which Julian applies to the alienation of a woman’s property without tutoris auctoritas1; the form of the allusion, here too, being such as to suggest that it bad no special application to that case[1176] [1177]. But it may be noted that the Rutilian constitution is cited as making usucapio possible. The present constitution is cited as causing dominium to pass and so making usucapio impossible. Moreover in that case the thing is the property of the alienor, the mode of conveyance being defective: here the defect is of a different kind, consisting in the overriding right of the old owner. All that they have in common is that the effect of the transaction can be set aside on the repayment of certain money. One of the allusions (not cited by Karlowa) looks as if the enactment dealt specially with this case: at is de quo quaeritur lege nostra quam constitutio fecit civem Romanum dominum habuif.}

Whatever its nature, the effect of the constitution is to set up an exceptional state of things. There is an ownership in the redemptor, and another ownership in the old dominus, liable to come into operation at any moment. Cases of ownership which are to come to an end on failure of some condition, etc., are not uncommon, especially in relation to sale and donatio. But, in such cases, there is, in classical law, a need of reconveyance. In our case postliminium operates with no such need⁸.

316 Rights of Redemptor of Captured Slave [pt. i

The redemptor is set in the position which the hostis held : if he does acts in relation to the man, which if done by the hostis would not be effective, as against the rights of the old owner, by the Roman law, how are these to be looked at, seeing that he is not actually a hostis, but a Roman aims ? The question is considered in relation to several states of fact.

A statuliber, to be free on paying a certain sum, can pay it out of peculium to his master for the time being¹. Such a master ordinarily derives title directly or indirectly from the donor of the freedom, and, at least, if he gave value, has a remedy, if he was not informed of the prospect of liberty². How if he is a redemptor, as to whom none of this is true ? Tryphoninus says* the man is free on payment to the redemptor. But in ordinary cases he cannot pay it out of any peculium but that which passed with him¹. There is no such peculium here. Whether the redemptor bought him cum peculio so that his peculium represents that apud hostes, or did not, at any rate it does not represent that which belonged to the donor of liberty. Nevertheless, says Tryphoninus, he is allowed, favore libertatis, to pay it out of any part of his peculium, except what is acquired ex operis or ex re redemptoris. This is a sort of rough justice: it must not be understood to imply that the redemptor (and owner) acquired only what a bonae fidei possessor would have acquired.

The constitution applies to a purchase in the ordinary way of business: it does not require that the buyer shall know that he is redeeming a captive. If the purchase was made without that knowledge, can the buyer, since he is a bonae fidei emptor, usucapt the slave to the exclusion of the old dominusl The difficulty is that by the constitu­tion he is owner and a man cannot usucapt his own. Tryphoninus, arguing from the view that the constitution is not designed, by making him owner, to make his position worse, concludes that on such facts, though the Conception of usucapio is not applicable, the old owner’s right to pay off the charge will be barred by the period of usucapio*.

This topic leads the jurist to another. If the redeemer can usucapt, can he manumit ? Tryphoninus remarks that, of course, manumission by the hostis, whose place he has taken, would not bar the old owner, and asks whether a manumission by redemptor will free, or will merely release his right, and cause the man to revert to his original dominus. Clearly the redemptor, in the case in which time has barred the old owner’s claim, can free, and Tryphoninus observes’ that even under the old law, if the redeemer had bought him knowing that he was (a captive and) alienus, and had sold him to a bona fide buyer, the

I Post, Ch. XXI. « Ante, p. 289. » 49. 15. 12. 11.

â€? 49.15. 12. 8. » 49. 15. 12. 9. oh. xm] Rights q/Redemptor of Captured Slave 317 buyer could usucapt and manumit, and thus the right of the original owner would be destroyed. Therefore he holds that the redeemer himself can manumit. He does not rest his view on the technical ownership created by the constitution, but rather on the fact that if the old owner never pays the redemptor the slave will be in the position of being incapable of manumission through no fault of his own. The argument is not convincing: the same thing is true of ordinary pledged slaves^[1178] and of many others. No doubt the ownership created by the constitution is really the deciding factor. The result, for which the strange provision of the constitution is to blame, is at any rate in appearance unfair, and Tryphoninus tries, with little success, to put a better moral face on it. Indeed his view as expressed would lead logically to a requirement of notice to the old owner.

Ulpian discusses the same case more shortly and without much reasoning⁸. He holds, somewhat doubtfully, favore libertatis, that the manumission frees him and that postliminium operates, not so as to restore him to his old master (hoc enim satis impium est), but to cause the libertus to be indebted to his old owner to the amount of his own value as a slave. Ulpian makes no reference to the constitution, and indeed, while Tryphoninus seems to be struggling with a logical necessity, leading to a power of manumission which he thinks in­equitable, Ulpian thinks the logic the other way, and frees the man only favore libertatis. The language of Tryphoninus throughout the discussion does not suggest that the constitution he is discussing is an ancient one, a republican senatusconsult, as Karlowa thinks’, but rather a new one the working of which is not yet clear. Nothing that is known of the dates of the two jurists makes it impossible for the text of Tryphoninus to have been written after that of Ulpian. It seems not impossible that Ulpian is writing under, or with reference to, a regime under which the redemptor is in an anomalous position, since he has acquired what Roman law recognises as the subject of ownership, from one whom Roman law recognises as its owner (quod ex nostro ad eos pervenit illorum fit*), and yet is not himself recognised as owner⁶.