CHAPTER VII COMMERCE

Trade and business are the theme of this chapter, and contract, agency and security are the legal modalities under which they will be discussed. As to the theme, it is only fair to warn the reader that trade and business, though the chapter about them will be long, were insignificant in importance in ancient times compared with the land.¹ As to the modalities, correspondingly, the Roman law exhibits gaps, compared with modern systems, which would be very surprising if this warning were not borne in mind.

1 Modem commerce is regulated largely by big block statutes (such as our Sale of Goods Act, 1893). For Roman commerce there was singularly little statute; apart from the aedilician edict on the sale of certain goods the law of commerce was built up, like all the other branches of the law, out of the old ius civile by a com­bination of ius honorarium and juristic interpretation.

2 Company law remained at a primitive level of development. Partnership grew out of, and retained characteristics of, ancient family common ownership; limited liability never got beyond an embryo stage; and legal personality was not granted to business corporations.³

3 The concept of agency, by which â€?if an agent enters into a contract on behalf of his principal with a third party, he creates rights and duties directly between principal and third party, and himself incurs neither’,⁴ was (it is not very clear why) an awkward hurdle for Roman jurisprudence, and so never systematically developed.

4 There was no law of patent or copyright, no protection for property in ideas?

5 Accounting remained primitive. Accounts were of course kept carefully enough, especially by the tax-farming companies and by bankers; but they were never more than lists of receipts and expenditure (the Romans did not invent double-entry), and so were of little use for economic planning—even agri­cultural.⁶

6 Banks performed important functions in the community; they held money on deposit, made payments on behalf of customers on the basis of written instructions, and made transfers of funds as between their own customers or persons known to them. But they never got as far as the negotiable paper instrument or the overseas book-transaction between unknown persons.⁷

7. Above all, ancient society as a whole never hit on the notion of putting capital to work to increase productivity. Modem capital, it has been said,⁸ is based on production and spent on more production; ancient capital came either from rents or from money-lending at interest, and was spent on unproductive ends. The difference can very clearly be seen in the field of mortgage. One of the major features of modern business development has been the use of the mortgage to acquire capital for industry, whereas in Rome mortgages were not used for such a purpose, but only as one (and not the most common) of the ways of giving security for a debt.⁹

★

Probably the oldest, and therefore the longest-lived, kind of Roman contract was the "stipulation’, an oral request and prom­ise.¹⁰ One party â€?stipulates’, that is, says â€?Do you promise that such-and-such shall be given (or done)?’, and the other party duly promises. Any form of words would do (though there was one solemn ancient form, using the word spondeo, which was confined to Roman citizens), so long as the promise was a relevant and consistent answer to the request.

*

stipulation; â€?real’ contracts 209 Another kind of contract was the â€?real’ contract, which came into being not verbis but re, by the handing over of a res', deposit for safe-keeping, deposition, was one main case, and loan the other. These arrangements were very old, like stipulation, and they retained a feature characteristic of the early Roman society in which they were bom: they had to be gratuitous, for no fee or reward. If you charged me for keeping something safe or for lending me a slave or a spade, that was a perfectly good contract, but it was â€?letting and hiring’, not deposit or loan. This is the old noblesse oblige:¹¹

â€?loan is an affair of good-will and officium.⁹

The labour force, the tools, the precious objects for display, passed round the community as required, and the reward was simply the reciprocal obligation on the other man to do likewise when called on.

With deposit, Rome remained in this climate of thought. Not that deposit was of trivial or declining importance; given the slow rates of travel in the ancient world, the shortage of police, the need to go off to vote or trade or litigate, men had a frequent need to call on the goodwill of a neighbour to look after their property. One of the nastiest and most dishonourable things a man could do was to deny a deposit, to brazen it out that he had never received the thing. Juvenal’s thirteenth satire is cynical consola­tion to someone who had suffered in this way:¹²

�What, at your age, with sixty years behind you, you get mad because a friend will not hand you back a solemn (sacrum) deposit?’

As for cash, that you deposited at the bank, and it is significant of the ancient economic attitude that during the whole of our period that meant actual deposit for safe-keeping, sealed up in bags or chests.¹³ Only late in Roman law is there talk of â€?irregular deposition, whereby the bank took and used the money and was ready on demand to return not the actual coins but the equivalent sum.

was being held in deposit for the purpose of purchasing land; it could not be doing both.¹⁴

One of the two kinds of loan was in many ways akin to deposit. (For there are economically two kinds of loan, for which the Romans, unlike ourselves, conveniently had different names: if I lend you a slave or a spade I remain its owner, and want the self­same slave or spade back—and that was commo datum; but if I lend you ten pounds or a bottle of wine it is for you to consume, which can only be done by my making you owner of it, and what I want back is the equivalent—that was mutuum.) Com­modatum, loan for use, without payment, belonged to the â€?mutual help’ sphere of social ideas. You could sue for the return of your deposited or loaned object by the actio depositi or the actio com- modati, and the only important difference between them was in the degree of liability for the object while it was in the hands of the â€?bailee’, the man to whom it had been entrusted. The commo- datary, who benefited from the arrangement, was liable not only if he failed to restore the object unharmed through his fault or negligence, but even if it was stolen or damaged through no fault of his (barring certain extreme cases like violent robbery or earth­quakes). The depositee, on the contrary, who was doing you a favour and not benefiting—for he must not use the object, which would be theft—was liable only for fault or negligence.¹⁵ Con­viction in an action on deposit resulted in legal infamia, whereas in an action on commodatum it apparently did not; the reason for the distinction is not known,¹⁶ but it may have been a quid pro quo for the stiffer formal liability that faced the commodatary any­way. â€?No fault’ situations might involve liability, but could not reasonably involve â€?infamy’.

Commodatum was not commercially important; neither was the other sort of loan, mutuum, as such.

commodatvm; mvtvvm 211

could not charge for the loan), if you wanted to contract for interest it had to be done by stipulation. As people do not, in the way of business, lend money for nothing, the important contract for money and debt was therefore the ordinary verbal stipulation. (Hence there was no such thing as an actio mutui; the creditor sued with a standard formula as on a stipulation, such as the formula certae creditae pecuniae which was quoted in Chapter III.)¹⁷ Consequently we need not bother with the question whether the gratuitousness of mutuum was due to another ancient attitude of Roman society, the hatred of usurers, because usury and mutuum had nothing to do with one another, and mutuum was gratuitous simply because it was â€?real’.

As for the hatred of usurers, it is certainly possible to produce a long chain of celebrated passages showing that people regarded with distaste the professional moneylender, the man whose whole livelihood was derived from loans at interest, such as Plautus’ Curculio, where the moneylender is equated with the brothel­keeper as a plague on society, or Horace’s Epode on Alfius.¹⁸ But the much more interesting fact is that everybody did lend money at interest quite without embarrassment, even the highest nobility (who had most to lend). â€?Here we are with a civil war on’, says Cicero in 49,¹⁹ â€?with Pompey under siege by a Roman army; and yet there is the City the same as usual, the courts in session, the games in preparation, and, as usual, the great and good are clocking up their interest’. Pliny, the ultra-respectable, tells a friend that though his property is mostly in land, he has some funds out at interest.²⁰ And it might even be a legal duty, as for guardians, who were required to put the funds of their wards out at interest. Rates of interest were not normally much greater in Roman times than we are accustomed to, though they were a bit greater, and fluctuated over rather wider limits; there is a curious tendency amongst scholars to exaggerate them.^2oa In 62 bc Cicero wrote:²¹

�Actually there’s plenty of money to be had at six percent, and one thing about my achievements to date is that I’m regarded as a good risk.’

On the other hand, from Atticus’ uncle Caecilius, a hard man:²²

�not even his relations can squeeze a shilling at less than twelve per cent.’

On 15 July, 54 bc such a lot of cash was going into bribes for the consular elections that:²³

�money’s gone up from four per cent to eight per cent. I can hear you saying “I can’t say I mind that”. What a noble fellow! What a public-spirited citizen!’

According to Tacitus, who gives a confused account of the matter,²⁴ the Twelve Tables had laid down a statutory maximum rate of interest, unciarium faenus, which is now usually believed to have meant one hundred per cent per annum,²⁵ not in­conceivable or without parallel in an early agricultural com­munity; and there were penalties for â€?usurers’ (presumably people who exceeded the maximum). At some later date interest had been forbidden altogether, which naturally became a dead letter. In Tacitus’ own day the situation was supposed to be governed by a law of Julius Caesar, but he does not tell us its provisions, merely saying that it too was not observed. We hear in the Digest about the non-actionability of interest above a legal maximum,²⁶ and this was centesimae usurae, which had nothing to do with the old unciarium faenus but was one per cent per month, i.e. twelve per cent per annum. Within this limit it was the business of governors of provinces in Cicero’s day to settle the maximum for their province; what he did (probably the normal thing) was to make it the same, twelve per cent.²⁷ According to the Gnomon of the Idiologos that was the rate in Egypt under the Principate.²⁸ According to Ulpian interest over one hundred per cent and compound interest were void;²⁹ perhaps what he meant was the two things taken together, i.e. compound interest that raised the total owing to more than twice the original debt, for compound interest was not in itself prohibited—Cicero allowed addition of interest to principal annually in his province.^{3 0} The Digest also tells us that a judge who had to assess interest in a bona fide action had to take into account the â€?custom of the region’?¹ In addition there were rules about interest on debts not settled by due date, the details of which we cannot embark on?² There was no National Debt; the state did not, except in rare moments of emergency, borrow from private individuals. On debts of individuals to the treasury the normal rate in Paulus’ time was six per cent?³

The activities of the hated professional moneylender are illustrated by a group of the Transylvanian Tablets, which record certain transactions by a small-time practitioner of this necessary but despised trade in the Dacian mining villages, one Julius Alexander?* In ad 167 we find him in a moneylending partner­ship with another man;³⁵ their entire capital was a bit under three thousand sesterces and the partnership was for four months only—just to make a little quick profit out of local poverty, no doubt. In two documents of a few years earlier we get his actual contracts of loan, at the standard maximum of twelve per cent interest; in one case he is lending and in the other he is himself borrowing. Here is the lending contract?⁶

�For sixty denarii to be given on the day requested (for repay­ment) in good coin: promise called for in good faith by Julius Alexander, duly given in good faith by Alexander, son of Cariccius, who declares that he has received the sixty denarii in cash as a loan and owes them. And for the interest upon them from this day per thirty days one per cent to be given to Julius Alexander or to whomsoever it may concern, promise called for in good faith by Julius Alexander, duly given in good faith by Alexander, son of Cariccius. Surety for payment of aforesaid principal and interest, properly and in good coin: Titius Primitivus. Transacted at Albumus Maior, 20 October, consulship of Rusticus for the second time and Aquilinus.’

*

To us, to whom entering into a contract means signing on a dotted line, so that the layman would say that, on the one hand, if there is no writing there is no contract and, on the other, a piece of writing if produced cannot be gainsaid, the Roman idea of writing as being inessential, mere evidence, is at first perplexing. The reader might be pardoned for turning with relief to the next kind of contract we must come to, the contract â€?by writing’, litteris (in which, as Gaius explains,³⁷ the writing did constitute the contract), and for wondering why it was not brought in first, and perhaps supposing that it must have been the commonest form of Roman legal bargain. He would, however, be quite wrong.³⁸ It was a very specialized arrangement, and though it existed in Cicero’s time the absence of later evidence for it except Gaius’ rather inadequate description³⁹ suggests that it actually faded out of use. It was not made â€?by writing’ in general, but by entries in account books; A entered a sum of money as owed him by B, and that constituted the contract (but, though Gaius does not say this, there must have been required some evidence of reciprocal agreement or consent by B). Its purpose was auxiliary: to turn contracts from one form into another (eg. a bona fide consensual contract of sale into a stricti iuris written contract) or from one person to another (e.g. to transfer a debt). Cicero’s defence of Roscius the actor ought to be our best guide to the contract litteris, because it turns a good deal on entries in account-books; but we cannot tell whether he is talking of entries as contract or entries as evidence (he is making a case and perhaps deliberately muddying the waters, and we have not got the plaintiff’s side).⁴⁰ It is possible that this form of contract proved unsatisfactory because the Romans did not have double­entry bookkeeping; it does not seem to have been particularly used by the banking houses, who kept the most systematic accounts.

Gaius notes that in the Greek-speaking world (which had long been used to treating contract as a writing) deeds of hand, acknowledging debt and promising payment, were accepted as contractual in force.⁴¹

★

We come to the final group, known as the �consensual’ contracts, which include three sorts of bargain of particular importance for commerce: sale, hire, and partnership. Every contract is con­sensual in the sense that it involves agreement between people, but these particular ones were consensual in the special sense that nothing but consensus was needed to constitute the contract—no

writing, no particular words, no presence of parties, no handing over of anything; none of these were of the essence, and the contract was â€?on’ from the moment of agreement. It is a great question of Republican legal history just when these fully con­sensual contracts achieved acceptance,⁴² but we can talk about them freely because they were there at least by Cicero’s time. They contained nothing that could not be done by stipulations, but for many purposes they had great advantages. They were bilateral, creating both rights and duties in both or all the parties by a single act of agreement (as is implied by the double name of two of them; â€?buying-selling’, emptio venditio, and â€?letting-hiring’, locatio conductio). More important, they gave rise to bona fide, instead of stricti iuris, actions, so that any balancing of sums or interests and any other agreements relevant to that in dispute could be taken into account by the judge, and the formula needed no exceptiones.

Some documents of sale have already been quoted in connection with slavery; here are just two more, again from widely separated parts of the Roman world. The first is from Dacia, its date ad 159:⁴³

�Andueia, daughter (? son) of Bato has bought and received by mancipation half a house, the right half going in, w^Thich is at Albumus Maior, in the hamlet of the Pirustae, between the neighbours Plator Acceptianus and Ingenuus son of Callistus, for three hundred denarii from Veturius Valens. This half­house, object of contract, with its fences, surrounding walls, bounds, entrances, doors and windows, locked with keys and in best condition, Andueia daughter of Bato is to possess lawfully. And if anyone makes eviction of said house or any share in it, so as to prevent Andueia daughter of Bato or anyone whom it may concern from holding, possessing or usucapting it lawfully, then for proper payment of such sum as represents the value of what has been prevented, promise has been called in good faith by Andueia daughter of Bato and duly given in good faith by Veturius Valens. And as price of the said half house Veturius Valens declares that he has received and holds from Andueia

daughter of Bato three hundred denarii. And it has been agreed between them that Veturius Valens shall pay taxation on the aforesaid house until the next census. Transacted at Albumus Maior, 6 May, consulship of Quintillus and Priscus.’

The second is from somewhere in Egypt, and its date is ad 77:⁴⁴

�C. Valerius Longus, trooper of the Aprian Squadron, has bought one horse, Cappadocian, black, for two thousand seven hundred Augustan drachmas from C. Julius Rufus, centurion of legion XXII. That said horse eats and drinks as veterinary animals customarily do, apart from [...] described openly visible on its body, and if anyone makes eviction of him, that double [? single] the value shall be properly paid in good coin as is customary, promise was called for by C. Valerius, duly given by C. Julius Rufus. And C. Julius Rufus, centurion, declares that he has received and holds from C. Valerius Longus, purchaser, said two thousand seven hundred Augustan drach­mas, and [has transferred said horse to him]. Transacted at [...], 7 June [or 9July], consulship of the Emperor Vespasian, eighth time, and Domitian, son of the Emperor, fifth time.’

We observe that it was normal to add to the contract of sale stricti iuris stipulations for undisturbed enjoyment and quality of the goods, even outside the cases required by the aedilician edict, and to record receipt of the price. It may be that the bona fide ele­ment in sale left purchasers a bit unhappy, so that they preferred the security of strictum ius for their essential guarantees (though the standard clauses may just reflect the hidebound practice of local notaries). A number of anecdotes in Cicero illustrate the problems in the field of house-buying. We have already had the story about the house burdened with an undisclosed demolition order; there is another about one sold without express mention of a servitude (it was sold to a man who had owned it once before and so was alleged to know anyway).⁴⁵ In these cases the bona fide nature of the action gave all needed scope to the judge, but the case of the unfortunate C. Canius was different.⁴⁶ He bought a

217 �little property’ at Syracuse in the belief, induced by fraud, that it was a marvellous place for fish; but he allowed the price to go down as a book debt against him, a contract litteris—stricti iuris. So when he discovered there was no regular fishing for miles he had no remedy, for, says Cicero:

�my friend and colleague C. Aquilius had not yet invented the formulas de dolo malo*

A famous rule of sale in Roman law was that, unless there was express agreement to the contrary, the risk of accidental destruc­tion of, or damage to, the object passed to the buyer as soon as the contract was made. It is true that the seller was under a duty to transfer it to the buyer, and was liable (on what seems the best view) for custodia of it until he did so, but this still leaves damage or destruction by violence or natural forces—if the house was destroyed by an earthquake the buyer nevertheless had to pay the price. Risk has to lie somewhere, of course, but normally it passes when ownership passes,⁴⁷ whereas in Roman law transfer of ownership awaited the necessary traditio or mancipation. And about transfer of ownership there is one additional and truly vexing problem which at first sight looks very fundamental but surely cannot have mattered as much in practice as it seems: Justinian in his Institutes gives a rule (not copied from Gaius) that even when the object of sale has been transferred (by traditio, he says; but then he had abolished mancipatio), ownership of it only passes when the price has been paid or security given for the price.⁴⁸ Justinian quotes the Twelve Tables for this rule, but he ends his statement by adding lamely that â€?if the seller accepts the good faith of the buyer, then ownership does pass straight away’. A Digest text from Pomponius gives the same rule—with the same feeble pendant that reduces the rule to a nullity (but of course the pendant might be an interpolation) ;⁴⁹ and in the Digest even Gaius seems to imply the same.⁵⁰ But in his Institutes, Gaius says firmly that res nec mancipi handed over by one who owns them become the property of the buyer at once.⁵¹ Virtually every possible view has been expressed by scholars about this set of passages.⁵² It is really a question about sale on credit, and it

may be best to hold that the rule existed but did not long have practical importance, because (a) every seller in his senses would require some security if the price was not paid at once, otherwise he just would not convey, and (b) the praetor protected possession ad usucapionetn, so ownership for that short period did not matter.

Another thorny set of problems (not only for Roman law) concerns mistake in consensual contracts like sale. Many kinds of mistake can be analysed; a typical one is if you think you are buying this bed and the other man thinks he is selling you that bed. The subject is deep and complex and we cannot embark on it.⁵⁵

Most things could be objects of a valid sale, including inheri­tances, servitudes, and the right to collect debts. You could have sale of something from a stock of things, or sale of a future thing, like the next harvest,⁵⁴ or a real gamble (though we do not hear much about lotteries, only of the small â€?raffle’ sort occasionally at imperial festivals in Rome). You could have sale subject to all sorts of conditions; two need mention. In diem addictio was in effect â€?sale subject to a better offer’:⁵⁵

�If he gets a better offer the vendor must notify the first bidder, so that if someone has added more he himself may add yet more.’

And the pactum displicentiae was in effect â€?sale on approval’ :⁵⁶

�Labeo has a question: suppose I have handed you horses to try for sale on condition that you return them in three days if not satisfied, and you run a vaulting-race on them and win it and then refuse to buy, is there an action of sale against you?’

A common field of application of this was the sale of wine with a right to taste it.⁵⁷ As to one kind of transaction which the texts never discuss under (and which, it is assumed, was not amongst the possibilities of) emptio venditio, scholars sometimes make a song and dance. This is what is nowadays called emptio generis, sale not of a specific thing or part of a specific stock but just â€?a horse’—as opposed to â€?one of those horses’—or â€?twenty dozen Arretine cups of standard size’. Now this is just the kind of sale that modern large-scale commerce mostly deals with, and its absence from the texts on emptio venditio is sometimes said to correlate with the lack of mass-production and mass-consumption in the Roman world. But it must be remembered that the bargain could always be achieved by stipulations, and there is some evidence that it was:⁵⁸

�If someone calls for promise of “one hundred modii of good African wheat” or “one hundred amphorae of good Cam­panian wine”, this, it seems, is a stipulation for a “thing un­certain”, because you can find something better than good, so that “good” does not signify any absolute standard. But if you call for promise of “best quality” you are taken to mean something the goodness of which is the highest grade of goodness, and consequently that adjective signifies something 44 ^ · >> >

certain.

It is interesting to be told that â€?English experience in connection with bulk sales, which are usually governed, not by the Sale of Goods Act, but by standard contracts, seems to show that... parties prefer to set out all the terms afresh. The stipulatio was peculiarly fitted for this purpose.’⁵⁹

For a contract to be an emptio venditio there had to be a money price—otherwise it was barter; it had to be a real, not a nominal, price—otherwise it was gift; and it had to be definite price. We cannot go into the detailed application of these rules.⁶⁰

A regular feature of Roman life was the auction sale. One aspect of it was the selling up by public auction on behalf of the treasury of the property of condemned persons (including Emperors) ;⁶¹ but the private auction was common and important, especially for land or the contents of inheritances—guardians, for example, were under a duty to get the best price for property of their wards. During our period the custom was for these sales to be done through an intermediary, argentarius coactor, the banker-auctioneer.⁶² A recent study has elucidated the triangular legal relationship between buyer, seller and middleman.⁶⁵ Caecilius lucundus, the banker and moneylender of Pompeii, some of whose files survived its catastrophe, was engaged in these transactions; they could be for quite substantial sums, given the middle-class ambience of that small city. Here is a specimen, of ad 54

�One thousand nine hundred and eighty-five sesterces, which sum was object of stipulatory contract by L. Caecilius lucundus on account of auction of box-plantation of C. lulius Onesimus, payable by next 15 July: this sum, less auction fee, C. lulius Onesimus has declared himself to have received from M. Fabius Agathinus on account of L. Caecilius lucundus. Tran­sacted at Pompeii, 10 May, consulship of M’. Acilius Aviola and M. Asinius Marcellus.’

The standard purchase tax was calculated on the price as paid (or promised) by the purchaser to the middleman.⁶⁵

The reader will naturally enquire how far the Romans used deposit in our common modem sense—the advance on the price, which the buyer will forfeit if he does not go through with the bargain. It can serve many purposes in different systems, from simply giving a man a â€?first refusal’ on something to being the actual clinching moment of a contract. Rome used for it the Greek word arrha⁶⁶ (or its Semitic form, also used in Greek, arrhabo); in the Greek law of sale arrha was clinching, and this role persisted sturdily, at least in Egypt.⁶⁷ Greek law stuck to sale as a handing of something over for cash; mere agreement to buy and sell imposed no duties and grounded no actions. Con­sequently arrha was a form of security that the agreement would be turned into actuality. The Romans, once they had developed consensual sale, in which the consensus was the contract and grounded the actions, did not need arrha in a clinching role. It was used, indeed, and Plautus is full of it (perhaps because his plays were written before consensual sale had been fully recog­nized),⁶⁸ but in our period it was treated as merely evidentiary:⁶⁹

�Sale is contracted when the price has been agreed on, even if it has not been paid and even if no arrha has been given. For

auctions; deposits; locatio condvctio 221 what is given under the name of arrha is evidence of a contract of sale?

(Gaius sounds here as if he is warning against an erroneous opinion into which people—in the East, for example—might be prone to fall.) Roman Egypt had a â€?law of arrhabo’ which was penal; the vendor was liable for double the deposit if he failed to fulfil (the buyer simply forfeited the deposit).⁷⁰ Naturally, if you did give a deposit it counted as an instalment of the price;⁷¹ and in general, payment by instalments was perfectly proper if the parties so contracted,⁷² as was also agreement that if the whole price was not in by a given date the sale was rescinded.⁷³

The contract of locatio conductio covered such a variety of transac­tions that we have met some of its rules in two different chapters already—tenancy in Chapter V and hire of labour in Chapter VI. We are left with what one might call the â€?business’ aspects of it— the bargains for carrying out some specific task for a reward, such as transport of goods or people, cleaning, repairing or storing goods, building houses, and manufacturing things out of customers’ materials. This branch of locatio conductio is not sharp- edged; it fades off into sale,⁷⁴ and may be difficult to distinguish in certain cases from hire of labour and hire of premises.⁷⁵ Like sale, it was a consensual, bona fide contract and there had to be a money reward; but to a much greater degree than with sale there entered into this contract difficult problems about custody and risk, as to the Roman law’s solutions for which modem scholars are still much at odds;⁷⁶ there is no doubt that those solutions underwent modification with time, but great doubt as to which direction the modifications moved in. It must be kept firmly in mind that people could make what arrangements they liked about these points if they chose specifically so to contract; what we have to ask is, what was the law in our period about custody and risk if no specific provisions had been agreed? Every­one who held something under this contract would be Hable to make good his deHberate fault or carelessness (on the fault of his employees there is doubt owing to suspicions of interpolation in a

text of Ulpian).⁷⁷ Gaius tells us that fullo and sarcinator, the Roman cleaner and repairer of clothes, respectively, were liable for custodial* which is much more severe, because it includes situations in which a man would not normally be said to be at fault—notably, theft. Now many scholars believe that Gaius’ fullo and sarcinator were intended by him as typical cases, and that therefore all â€?bailees’ under the contract of locatio conductio were liable for custodial but not everyone agrees. Unfortunately, whatever Justinian’s professors found about this in the texts they used for the Digest, they altered it all to fit in with a quite different scheme of relative liabilities, and this makes it impossible to be sure what the classical texts said. The dividing line, at any rate, is theft. On one side of it (certain liability) we hear a lot about lack of care⁷⁹—the careless laundry that lets your clothes be nibbled by mice, the careless navigator, and so on—and it was firmly stated that incompetent workmanship counted on this side of the line.⁸⁰ On the other side (certain non-liability), are the events called casus and vis maior, that is, accidents of nature or of human violence such as robbers stealing goats, passing armies, fires, blight and landslides. About ordinary wear and tear we are not informed; one text makes it sound as if this was normally specified in contracts, where appropriate.⁸¹

From this point on it will be better to look at one or two com­mercial situations more specifically, beginning with building. First, we are told (it is not self-evident) that building contracts were locatio, not sale.⁸² You might begin with a mandate to the builder to quote you a price;⁸³ then you would agree the contract, including of course the price and sometimes a completion date.⁸⁴ The contract might be for a lump sum, per aversionem, in which case all risk was on the builder until the moment of final approval by the client; or, in the modem way, the payment might be by stages based on quantity survey of the amounts completed, per pedes tnensurasve, and the builder then took the risk only of the uncompleted parts (and the client must not deliberately delay the necessary surveys).⁸⁵ The builder was not in any event liable for casus unless the contract stated so.⁸⁶ There is talk also of payment by the day, but it would not be likely for house-building;

building; transport 223

perhaps it was for small things like garden walls or sheds. There might or might not be approval by the day, but this was not locatio operarum: approval was necessary if the workman was to qualify for his price.⁸⁷ â€?Approval by client’, w^re are told, meant â€?satisfaction as a good man would judge’;⁸⁸ and there is one passage, on the consequences of changes due to client’s im­promptu orders on the site, which will remind the modem architect vividly of his own problems.⁸⁹

If we look next at the transport system of the Roman world we shall be involved in one or two sets of legal rules besides those of locatio conductio. Maritime arrangements were pre-eminently a field in which the Romans were preceded, and always outclassed and outnumbered, by the Levantines and Greeks, and it used to be held by everybody and is still widely held that the Romans just borrowed the set of maritime rules which they found existing in the Mediterranean world, especially two typical branches, the law of bottomry loans, nauticum faenus, and the law of jettison or â€?average’, iactus. That there was a group of rules going by the name of the â€?Rhodian Sea Law’, known in our period to the Romans⁹⁰ and surviving into the Middle Ages,⁹¹ cannot be denied, but there are reasons for being sceptical whether the Romans borrowed these rules (notwithstanding the existence in the Digest of a title (14. 2) â€?On the Rhodian Law of Jettison’) and did not rather reach similar results on a basis of their own legal principles of mutuum and locatio*²

Nauticum faenus, or traiecticia pecunia, was the loan of money to a man to enable him to buy a cargo to ship and sell abroad, risk being on the lender so that he could only sue for his principal and interest on the safe return of the ship, and the cargo (and often the ship too) being security for the loan.⁹³ It was a form of maritime insurance. Great profits could be made, but the risks were great, and so the interest that could be charged was not subject to any statutory maximum, provided that the creditor took the whole risk (i.e. that the contract really was nauticum faenus and not just an ordinary loan).⁹* There is in the Digest an argument based on a set of facts which include what looks like a standard bottomry contract:⁹⁵

�Callimachus received a bottomry loan from Stichus, slave of Seius, in the province of Syria, city of Beirut, destination Brindisi; sum in credit: two hundred sesterces for each day of the voyage; pledged as security: the goods bought at Beirut for carriage to Brindisi and any goods to be bought at Brindisi for return carriage to Beirut. And it was agreed between them that on arrival at Brindisi Callimachus should before 13 September following buy other goods, load them, and sail back to Syria, or that if by due date he had not bought other goods and left that city he should repay the whole sum as if the voyage were now terminated, and provide all necessary expenses for those persons carrying the money to enable them to take it to Rome. Promise for proper performance of all these things was called for by Stichus, slave of L. Titius, duly given by Callimachus.’

lactus, jettison’, is the rule whereby, if goods have hastily to be cast overboard to save a ship (or handed over to pirates to ransom it), the man whose property was sacrificed is entitled to a con­tribution from those whose property was thereby saved. In the Byzantine â€?Rhodian Sea Law’ this principle of â€?all in the same boat’ was carried very far on a kind of partnership notion, but the Roman law of our period seems to have kept it within narrow bounds (insisting, for example, that the rules only applied if the ship was saved), and to have thought of it as arising out of the contract: the loser had an action against the captain who had ordered his things to be thrown over, and the captain was entitled to recover from those whose goods he had saved, the whole being judged on a bona fide basis:⁹⁶

�A number of merchants had on board the same ship a variety of cargoes; in addition there were numerous passengers, slave and free. A great storm rose and they had to jettison. The following questions arose: must all parties contribute to make up the loss, even those whose cargoes added no weight to the ship, such as gems or pearls? And what is the basis of apportionment? And must there be contribution on behalf of free passengers [x.e. without cargo]? And by what action is all this to be achieved?

The answers were: all who had a [pecuniary] interest in the jettison taking place must contribute, because things thereby saved owe contribution; therefore the owner of the vessel is himself Hable for a share. The sum of loss is to be apportioned pro rata to the values of the respective cargoes. No financial estimate can be made of free persons. The owners of the goods sacrificed will have an action on their contract of hire with the captain, i.e. the ship-master/

This passage illuminates the ordinary sea transport arrangements (as well as perils)⁹⁶⁴ in the Roman world. One finds very little reference to exclusively passenger vessels—perhaps only the â€?packet-boat’ on the busy Brindisi-Dyrrhachium crossing;⁹⁷ and merchants normally voyaged with their cargoes. There is much else of interest in the Digest on freight transport. One text is about lack of care by a lighterman transferring cargo; the importance of these navicularii on the rivers in general and es­pecially at river mouths, like Ostia, is well known.⁹⁸ Another concerns the â€?irregular locatio⁹ of grain ships (compare the â€?irregular depositum⁹ in banks) whereby several shippers poured their grain into a common hold and were entitled to dehvery by quantity.⁹⁹ And a third, a rare and valuable passage, quotes the cargo quantities in a ship carrying oil and grain from Cyrene to Aquileia.¹⁰⁰

Road transport makes its appearance a little more frequently than the usual disdainful references to it might lead one to expect. It was slow, inefficient, subject to brigands and to frequent tolls; apart from the cursus publicus, the government post and supply service, it was probably not used much for long-distance travel, which rehed on rivers and the sea. But it existed; you had to get goods to the rivers and the sea. The ox-waggon was ubiquitous, and so was the mule-waggon: â€?all vehicles on the roads are drawn by pairs of them’, says Varro, who also mentions trains of pack­donkeys bringing produce to the sea.¹⁰¹ Thus in the Digest we get pack-mules hired with agreement as to maximum load,¹⁰² the mule-waggon paying toll at a toll-bridge,¹⁰³ and road trans­port of columns, tree-trunks and wine-vats,¹⁰⁴ not to mention

P

the case of the careless cabby (whose fare, tipped out and injured, was someone’s slave).¹⁰⁵

Travellers were offered certain special protections. The praetor’s edict propounded two actions against â€?shipmen, innkeepers and stablekeepers’.¹⁰⁶ One was a special action of theft ;¹⁰⁷ the other said:¹⁰⁸

�Whatever property of any person shipmen, innkeepers and stablekeepers have received for safe custody, unless they return it I will give an action against them?

These were all people at whose mercy the traveller was. He could not stop to choose—and usually there would be no choice.¹⁰⁹ He had to carry his money on him, and the goods he carried might be valuable and might be someone else’s. Innkeepers were notor­iously Hable to be in league with local thieves,¹¹⁰ and shipmen were a no less shady lot;¹¹¹ and all of them were likely to be institores or exercitores, agents for an absent owner. The special HabiHty under theft is straightforward: the ordinary actio furti lay only against the thief (or accompHce), but this praetorian action lay against the innkeeper, etc., for theft committed by any of their servants or the permanent inhabitants of their premises (not by other travellers, if Ulpian’s statement is free from interpola­tion).¹¹² But the special liabiHty for safe custody is a subject of extreme controversy, directed mainly to (a) whether it applied only to goods specifically accepted for custody, and (i) why it was needed at all, ifâ€?bailees’ under locatio conductio were Hable for custodia in any case, as many think they were.¹¹³ The texts are contradictory, as can be seen by comparing Digest 4. 9. 5 pr. with 47. 5. 1. 4 on question (a) and with 4. 9. 3. 1 on question (6). What foUows must therefore be understood to be very dogmatic and very far from an â€?agreed opinion. First, this HabiHty was fundamentally concerned with the baggage of traveUers, not with carriage of freight in general;¹¹⁴ it is a mistake to think that it was originaHy or ever part of shipping law as such. Secondly, whether or not â€?bailees’ were in any case Hable for custodia, innkeepers, etc., were singled out because of the special position of travellers. The traveUer’s contract was for lodging or conveying

himself; it was not obvious that he was entitled on that to an actio locati for the baggage. What is more, locatio conductio being a bona fide contract, allowing considerations of set-off, an actio locati for the baggage might become entangled with argu­ment about the bill. Or the traveller might be carrying someone else’s valuables for which only their owner could have an actio fiurti. Thirdly, there was probably no special pact of custody.¹¹⁵ If it seems unfair on the innkeeper, if his guest was travelling with priceless undeclared ropes of pearls, it should be recalled that the English common innkeeper’ was under just such an unrestricted liability until the 1860’s. Perhaps the Hotel Proprietors Act, 1956, may afford a guess as to what actually happened; the notice, which every hotel must display, makes the hotel liable â€?to make good any loss or damage to a guest’s property even though it was not due to any fault of the proprietor or staff’, but the liability is limited to £100 per guest, except for items specifically deposited. Now Labeo discusses a case in which:¹¹⁶

�the manager of a repository put up a notice that he did not receive gold, silver or pearls at his risk.’

It may be conjectured that innkeepers, etc., protected themselves in the same way, t.e. by saying â€?of the following things I do not accept custodia⁹. It is a guess, but we do hear of shipmasters excluding liability for damage in this way.¹¹⁷

The question arises in a particularly acute form in the case of this institution of the law: how real in practice was the protection it afforded to travellers, who ex hypothesi needed to get on about their business? There were no grand hotels;¹¹⁸ the rich travelled via their own chains of private villas or those of their friends,¹¹⁹ so these provisions would benefit (if at all) middle-class people, the business community. We have no record of any actual case of prosecution under this heading; and yet it is not plausible that this kind of rule, a special invention, not just an assertion of some old principle, was for nothing and to no effect. The traveller had one thing on his side: the innkeeper would always be there to proceed against; and one must not underestimate the willingness of the â€?man in the street’, in a reasonably orderly society, to pursue his rights.

Repositories were provided by private enterprise for the storage of foodstuffs and other goods in a very general way; see the placard, from somewhere in Rome:¹*⁰

�In these private repositories, owner Q. Tineius Sacerdos Clemens,... are for hire grain space, lock-up space, close storage, safes, column-safes and space for safes, from today and from i July.’

They were naturally commonest in the big commercial centres like Ostia,¹²¹ and Rome itself. A couple of further passages in the Digest give hints as to the law about them.¹²² It seems that, as usual, the owner let out the running of the premises to a con­tractor, who did the detailed letting of space. This middleman, the horrearius, was liable to his customers for custodia, naturally enough (though we have seen that he might post notice of exceptions); he could not push his liability back on to the owner unless their contract specified this.¹²³ A fragment from a work on the duties of the prefect of the watch adds that if the tenants of space in a repository did not appear or pay the charges for a long time permission could be obtained to open their safes and take an inventory.¹²⁴ The government itself provided similar premises, and there survives a famous â€?notice of terms’ of an imperial repository:¹²⁵

�In these repositories of the emperor [...] Caesar Augustus are for hire grain space, lock-up space, safes and space for safes with service of custodians from today and i [January]. Rules of the repository: Anyone wishing to retain for a further year the safe or whatever else he rents must have rent paid up and give notice before 13 December.... No liability is undertaken for gold and silver. All property stored in these repositories will be subject to a Hen to the contractor against due payment of rent.... If anyone renting space in these repositories leaves his property there without making it over under seal to the custodian, the contractor will not be Hable.’

repositories; company law 229

As there was a contractor in here there is no reason to think that the rules laid down were different from what would have applied in a privately owned repository.¹³⁶

A discussion of the Roman law of partnership should contain what can be said about Roman company law. It is not very much. Except in one category, that of the publicani, we do not hear of big commercial enterprises; perhaps the best evidence for the sort of joint businesses that did exist is provided by the potsherds from the â€?Monte Testaccio’, that extraordinary hill near the Protestant cemetery in Rome composed entirely of the remnants of the jars in which foodstuffs, mostly from Spain and Africa, were brought to Rome and offloaded at the Tiber wharves.¹²⁷ Both the potters’ stamps and the painted abbreviations of names of the producing firms survive in some numbers, and when they are not just a single name they are combinations like â€?the two Aurelii Heraclae, father and son’, â€?the Fadii’, â€?Cutius Celsianus and Fabius Galaticus’, â€?the Caecilii and freedmen’, â€?the two Junii, Melissus and Melissa’,¹²⁸ â€?the partners Hyacinthus, Isidore and Pollio*, â€?L. Marius Phoebus and the Vibii, Viator and Restitutus’. It does not follow, of course, that â€?Snooks and Son’ may not be a huge firm, but the Monte Testaccio pattern suggests small â€?workshop’ businesses into which men brought their sons and their skilled workmen (freedmen, in fact) to share the profit and the loss—a sort of â€?partiary’ arrangement in lieu of a wage, perhaps. As a matter of legal history it is generally agreed that societas derived ultimately from the ancient automatic common ownership of family property by undivided heirs. Voluntary consortium of brothers goes on being found all through our period;¹²⁹ and very similar to it is the most characteristic form of partnership, the societas omnium bonorum, partnership of entire property, which was not commercial at all but based on the family notion. Even when business partnerships became common, their legal rules retained some of the familial atmosphere.

Societas was a consensual, bona fide contract. It could be between any number of people, for many purposes (not necessarily financial), though not every joint activity was in law a societas.¹³⁰

Apart from omnium bonorum, we hear in Gaius of slave-trading as a typical partnership activity;¹³¹ moneylending we have met in the Transylvanian Tablets and shall meet again; the facts in Cicero’s speech pro Roscio comoedo were about a societas between a man who had a gifted slave and Roscius the actor—the latter to teach his art to the slave and the partners to share the proceeds of their protege’s stage career. In the Digest we have men running a school in partnership, and an architect teaming up with a land agent.¹³² And Cicero’s pro Quinctio, it is important to recall, was about an agricultural partnership. The agreement must not be for a criminal purpose: robbers could not have an action pro socio for division of spoils.¹³³ Otherwise any terms whatever could, as usual, be agreed between the partners, especially about shares of profit and loss, except that it was not lawful so to arrange that one partner shared losses but took no profit at all (and anyone who wonders why such a rule was needed should remember the patron-freedman partnerships of the Monte Testaccio and the position of the necessarius heres). In default of express agreement shares in profit and loss were assumed to be in the same propor­tion.¹³⁴ Every member must contribute something, but it need not be money or goods, but could perfectly well be skill, know­ledge or standing. Agreement was between a fixed number of specific people, each to each; there was no possibility ofâ€?sleeping’ partners with financial shares just coming in and out, nor did a partner’s heir automatically succeed him in the firm.¹³³ Indeed, the death, capitis deminutio or selling up of any partner brought the whole business technically to an end, and so did the retirement of any one partner. This did not matter much, for tacit agreement of the rest to continue was enough; what did matter (and brings out the climate of the whole concept) was that any litigation on the partnership brought it to an end—the actio pro socio, in other words, was an action to liquidate the firm, and you could not wrangle in the courts about subsidiary questions without des­troying the business.

Nor was that all. Two important legal features of modem companies are agency and limited liability. In Roman law a partner was not an agent for all;¹³⁶ his acquisitions and contracts accrued to and bound only himself. He was naturally liable pro socio to bring acquisitions into the agreed joint management and exploitation, and his colleagues were liable pro socio to reimburse him, according to their shares, for his contracts, but that is quite a different matter from these things automatically accruing to or binding the firm as such. Secondly, in Roman law no rule was ever made confining the liability of a partner for the debts of the firm to the extent of his financial contribution. Partners were liable for the firm’s whole debts in proportion to their shares in the partnership, or, to put it another way, here as always a man’s debts were owed in full, but he could have an action to recover from his colleagues pro socio according to the proportions agreed in the partnership contract. Naturally, you could not legitimately retire from the firm in such a way as to escape your liabilities, but the structure made it all too possible for partners to let their colleagues down:’³⁷

"It’s his ruddy freedmen; they’ve walked off with the whole show. You know what I mean; the firm’s pot never gets to the boil, and the second things begin to run down hill your pals get out from under.’

It does not look as if the partners in the Transylvanian money- lending business regarded each other with much bona Jules; we find them almost at the expiry of the partnership taking stipulations:¹³⁸

�A partnership of moneylending was entered into between Cassius Frontinus and Julius Alexander, from 23 December last, consulship of Pudens and Pollio [ad 166], to continue to 12 April next. Terms: that any profit and loss from said partner­ship be borne in equal proportions. Into which partnership Julius Alexander contributed, in cash or out at interest, five hundred denarii, and Secundus, serous actor of Cassius Palumbus, contributed two hundred and sixty-seven denarii on behalf of Frontinus.... In which partnership if anyone is found to have committed fraud by malice he shall pay one denarius per as, twenty denarii per denarius. At end of term, with deduction of debts, each is to recover the sum above stated, or, if there is a profit, to divide it. For these things so to be done promise called for by Cassius Frontinus, duly given by Julius Alexander. Concerning which two pairs of tablets have been signed. Also owed to Cossa: fifty denarii, which are due to him from the above partners. Transacted at Deusaris, 28 March, consulship of Verus, third time, and Quadrates’ (ad 167).

A bit more is to be said about financial business, though it is in a way parenthetical, because finance was not necessarily carried on in partnership. It was essentially money-changing, money-holding and money-lending, and its operators appear under various names, probably with rather different functions which cannot be very clearly distinguished because they over­lapped in practice. The wealthy families all did moneylending with their own spare cash, and this sometimes took the form of a small private bank, kalendariutn, in charge, perhaps, of a slave— or even several, on their widely scattered properties—for the convenience (or exploitation) of tenants.¹³⁹ Then there is the daneista or faenerator, the professional moneylender. Higher up we come to nunwnularii and mensularii,¹*⁰ who were bankers, doing both deposit and lending business, but (as the names suggest) in a small way; and at the top are the argentarii, the big-scale bankers, who might well be publicani doing private business. A picture can be got from Cicero of their international functions. They made payments to public officials in the provinces on presentation of government bills of exchange,¹⁴¹ and, for known customers, similar payments on private drafts; the story of how young Quintus was paid his allowance at Athens University through Atticus’ bankers there can be traced in the letters.¹⁴² Another picture, of later date, is furnished by the files of Caecilius lucundus, who not only carried on at Pompeii the auction business which we have seen, but also ran a laundry under license from the municipality¹⁴³ and paid public dues on behalf of other similar contractors.¹⁴⁴

The argentarii were sufficiently important to generate some special rules of law. Women were excluded from the pro­fession.¹⁴⁵ A partner here was agent for the firm, in the sense that you could sue any of his colleagues on a contract made by him.¹⁴⁶ The praetor gave a special action to enforce a kind of agreement which was no doubt commonly undertaken, called receptum argentarii, namely that the banker would pay a customer’s debt over to his creditor (not, apparently, by the book transaction /itteris);¹⁴⁷ the agreement effectively transferred the debt, so that the creditor could now sue the bank if it was not settled.¹⁴⁸ Another set of special rules appears in the Digest title 2. 13 â€?On Disclosure’. The books of argentarii were regarded as unimpeach­able legal evidence, and, on grounds of public policy (we are told), were the subject of an edict¹⁴⁹ in which the praetor required bankers to disclose their entries as evidence on behalf of anyone to whose case they were relevant, provided he swore that his application was not vexatious; if disclosure was wrongfully withheld there was an action.¹⁵⁰ (Nummularii, says Pomponius, are not covered by the edict, but they ought to disclose like argentarii, for their business is essentially the same.)¹⁵¹ Yet another rule is given by Gaius, discussing bona fide actions:¹⁵² business between a bank and its customers was on a basis of the bona fide contract of mandate, not stipulations, but in claims against customers bankers must do their own calculation of debits and credits and sue only for the balance. If they overclaimed they lost all—a fierce rule, which shows why their books had to be so accurate. The Digest invents a typical banker’s letter to cus­tomer:¹⁵⁵

�From your bank account with me, at present date you have a balance on various transactions, in hand at the bank, of three hundred and eighty-six and appropriate interest. This sum which you now have in credit with me, uncontested, I shall refund. Any document under your hand for any sum on any transaction, still remaining with me, will be treated as null and cancelled.’

The only substantial exceptions to the general legal rules about partnership were designed for the publicani*⁵* These companies, to whom was farmed out in the Republican age the collection of all indirect taxation and the direct tributum of some of the wealthiest provinces (and companies they had to be, for individuals did not dispose of the great sums required), were in­dispensable in the absence of an extensive civil service. Their general history is well known; in their heyday, the time of Cicero, they wielded political power; the early emperors took much of the taxation out of their hands and deprived them of political influence, but they still farmed all the indirect taxes for a long time, and had to be subjected to new regulations by Nero. In the second century ad they gradually dropped out of the taxation structure—though not entirely, in our period, for Septimius Severus is still found struggling with their abuses. Cicero’s speeches, especially the Verrines, are the main evidence for their elaborate international organization. We must summarize the modifications in partnership law which facilitated their business and the rules made to curb their excesses. In the Republican period the taxes were farmed by the censors on the basis of their lustrum, that is, for five years at a time, and this continued under the Principate;¹⁵⁵ so the societates publicanorum were, as such, quite short-lived, since they had to be reconstituted anew each time, and the censors had power to exclude anyone from participation. It is probable that the same groups tended to reassemble, but presumably they could not carry any accounts over.¹ * *^a Within the period, however, they had two advantages: death of a partner did not dissolve the company (unless he happened to be the managing director), neither did litigation.¹*⁶ It is customary to add a third point, that people could come in and out as share­holders, participes. The evidence is in truth very thin. It depends first on a well-known remark of Polybius (in the second century bc), that every citizen in Italy had a stake in the state contracts,¹*⁷ which must at the very least be a fantastic exaggeration, and secondly on a quotation by Cicero of one of Verres’ edicts:¹*⁸

\. let him not admit as partner nor give a share..

with a late, but ancient, gloss saying that it meant â€?having a fixed share and no action for division like a partner’.¹**» This latter is evidence that if there were shareholders they did not come in and out, buying and selling from day to day, but were in for the whole term of the societas. The idea of the little man having his flutter on equities can be dismissed, but it is quite likely that the very wealthy nobiles participated on a â€?sleeping’ basis, helping the companies to meet the huge security that was demanded for these public contracts. A fourth point might be added: the senatus­consultum Macedonianum did not apply, so there was no bar to filii familias being partners.¹⁶⁰ There is question also as to a fifth point: were the companies of publicani at some stage recognized as corporations? We shall not tackle the difficult jurisprudential problems about the nature of â€?legal personality’ and how far the Romans had such a concept;¹⁶¹ but the municipalities and certain collegia¹⁶² were corporations in the sense that they could have a common chest, own property, manumit slaves, receive legacies and so on. Whether the publicani gained the same privilege really rests on Digest 3.4.1 pr., where they are equated by what purports to be Gaius with â€?certain collegia⁹ as being permitted to â€?corpus habere’.¹⁶³ It is a much suspected text, and not a strong enough peg to hang an assertion on. They certainly had a common chest and common slaves; in their heyday, the age of Cicero, the juris­prudential question had hardly been formulated.

The only title in the Digest about the publicani, 39. 4, concerns certain abuses:¹⁶⁴

�The audacity and temerity of the factions of the publicani are known to all.’

The praetor promised a special action for theft or damage com­mitted by the servants of the publicani, whether slave or free, and in the case of slaves the employers did not have the choice of noxal surrender—they must pay up.¹⁶* Nero’s new rules for curbing the â€?avarice of the publicani⁹, the most important of which was to bring suits against them into the courts extra ordinem, met with the scorn of Tacitus, who declares that they were soon evaded.¹⁶⁶ It is possible, however, that the reduction in the profitability of the occupation that resulted from the rules helped to condition their gradual abandonment of it. There was still trouble, though, at least in Egypt; one of the responsa of Septimius Severus runs as follows:¹⁶⁷

�To Isidorus, son of Deius: The audaciousbehaviourof Comon will be examined by his excellency Fulvius Plautianus, the praetorian prefect, my [?] kinsman. As for Apion the publicanus, if he is not implicated in the charges against Comon you will have the prefect of the province as judge.’

The publicans were sinners to the last.

We have one more kind of consensual contract to examine: mandate; but it can best be appreciated in the light of another typical Roman institution which was not mandate nor even technically a contract at all, but had effects of the same kind, namely negotiorum gestio, acting in a man’s affairs unasked and for nothing. The notion may seem odd, for nowadays we should be unlikely to meddle in our neighbour’s affairs unasked, at least in anything of such magnitude that it might lead to litigation after­wards. But in Rome the same considerations as made safe-keeping important applied to this also; the slowness of communications might put your friends and neighbours, when absent, in peril of assaults upon their property and families of which they were unaware until too late, so it was a part of officium for you to take steps, without authorization:¹⁶⁸

�lest through lack of defence they suffer possession or selling up of their property.’

You must announce yourself as their agent, defend actions on their behalf, and so on; propping up a dangerous building and undertaking the medical expenses of a slave are cases used as typical in the Digest—things, in fact, which might involve you in expenses as well as litigation. People prepared to do this were entitled to reimbursement if their friends later refused to â€?stump up’, and their friends were entitled to reimbursement if they mis­handled affairs—or made a profit. So the praetor offered the actio negotiorum gestorum, both ways, to and against those who had undertaken without authorization the affairs of an absent or dying man.¹⁶⁹ A recent writer has argued for the view that in our period negotiorum gestio was not always gratuitous and unauthor­ized,¹⁷⁰ on the ground that the action between a man and his general agent or procurator was negotiorum gestorum (not, or not necessarily, mandate or locatio conductio). The procurators of Cicero’s day seem to have been mostly procuratores absentium, appointed to look after a man’s affairs at home while he was abroad or vice versa; this general appointment may, however, have been a mandate, but not regarded as constituting specific authorization in emergencies (that is to say, procurator a manda­tary, but sometimes having to perform emergency acts of negotiorum gestio). Alfenus in Cicero’s pro Quinctio, of whom more will be said, was such a procurator absentis, and he took emergency steps which it was not certain his principal would ratify (but Alfenus was not a paid employee). The texts which couple procurators with negotiorum gestorum rather than man­date are all susceptible of interpretation as emergency situa­tions; and Gaius is firm that negotia ger ere with authorization is mandate.¹⁷¹ (Also banks, which look much the same as procura­tors, were under mandate, not negotiorum gestio.) The lesson seems to be that the borderlines in practice were fluid; mandates could be more or less specific, and procurators were of different kinds.

Mandate, then, was the contract by which one man undertook affairs of another on instructions. It was consensual, bona fide, and (being an exercise of officium)¹⁷² in theory gratuitous. It constituted the mandatary your agent, though as usual his transactions bound himself and not you, and so there had to be mechanisms whereby each party could secure proper per­formance, transfer over of acquisitions, reimbursement and so on; hence the actio mandati, both ways. The lawyers erected an elaborate structure of categories of mandate, but it is of no modem interest. Liability in mandate was probably only for dolus, deliberate breach of faith,¹⁷³ and conviction in the actio mandati resulted in legal infamia.

The figures who appear most as mandataries in the Digest title, 17. 1, are the surety and the procurator.¹⁷* The procurator, as general agent, if employed at a salary, was presumably a status inferior, and your contract with him should have been locatio conduction if he did it free he was either your freedman acting on the basis of operae owed to you, or else a status equal accepting your instructions, and taking perhaps an honorarium ex gratia; and the commission might be designed to benefit him as well as you. But the borderlines were fluid; it is as important not to envisage all procurators as humble instruments of the nobility as it is not to cast them all in the mould of the equestrian procura- tores Augusti of the Principate. Amongst the many letters in Cicero’s correspondence commending the procurators of his friends to the good offices of provincial governors there are to be found not only obviously humble agents, probably freedmen and probably salaried, such as Caerellia’s procurators in Asia,¹⁷* but also obviously high-status ones, like L. Oppius the banker at Philomelium, â€?my close friend’, says Cicero:¹⁷⁶

\. whom I specially commend to you, the more so first because I am so devoted to him and second because he looks after the affairs (negotia procurat) of L. Egnatius Rufus, my best friend of all amongst the equites*

And the facts in Cicero’s defence of Quinctius reveal another far from humble procurator. Quinctius and Naevius quarrelled over a partnership; Naevius got an order for possession against Quinc­tius in the latter’s absence in Gaul, and Quinctius’ procurator, Sextus Alfenus, at once tore down the possession notices and announced himself ready to defend actions against his principal. Alfenus was a friend of both parties, a relation of Naevius, a rich eques,¹⁷⁷ and had often before acted for Naevius in his absences.

Not every mandate was a general agency; it probably origin­ated in the social custom of asking one’s friends to do particular commissions. There are plenty of vague mandates in Cicero which he obviously did not intend to be contractual at all, such as to Atticus to see to his affairs in his absence, or to Tiro to make payments and collect moneys due.¹⁷⁸ There are, on the other hand, precise ones. Thus, a certain Vettienus was commissioned to act as purchaser of a country lodge for Cicero:¹⁷⁹

�He’s just written: bought for thirty thousand, to whom do I want conveyance made, payment date 13 November.’

And Fadius Gallus got into hot water for buying the wrong sort of statues for his exacting friend.¹⁸⁰ We also find Cicero himself under a mandate to buy something for M. Marius from the auction of an estate to which (as it had piquantly turned out) Cicero was part heir:¹⁸¹

�1’11 take good care of your mandate. You’re a clever chap; you’ve given the commission to the one man who has an interest in it fetching as high a price as possible. Very knowing of you to fix an upper limit,... for now I know your ceiling I shall make very sure you don’t get it for less.’

Pliny’s letters contain mandates of much the same sort, includ­ing one (interesting because of its slightly â€?distant’ tone) to an architect about rebuilding a small temple on one of his estates.¹⁸²

We are led in the end to pose a rather subversive question about the social reality of the principle of gratuitous services in our period. One need not deny the likelihood that in early Roman rustic-aristocratic society many things that later became subject to contract were done on the noblesse oblige principle, nor the likelihood that this coloured the contractual rules when they arose. And though it seems odd to us, for whom contract and â€?consideration’ go hand in hand, that such neighbourly acts as safe-keeping or making a purchase for a friend could become contractual and subject to litigation without sacrificing the aura of neighbourliness and ceasing to be gratuitous, not all that is odd is unhistorical. Nevertheless there are reasons for suspecting that in our period, which begins with the already complex and sophis­ticated Ciceronian age, gratuitousness and noblesse oblige in contract were an old tradition less and less honoured in the observance, as services became more and more specialized and what had once been amateur became professional. Public service, for instance, from Augustus on, was not in the least gratuitous: equestrian prefects and procurators, and even the grand senatorial generals and governors, received thumping salaries.¹⁸³ And when we look at the supposedly gratuitous contracts we find the need to make many qualifications and exceptions (excluding only the innocent commodatum). Mutuum, for example, was strictly loan without consideration; but in practice money was not lent for nothing, and little understanding of the financial pattern of Roman society would be achieved by anyone who confined himself to the gratuitous contract of mutuum. Or consider depositum: repositories did not work under it at all but under locatio conductio, for a rent; and as for banks, you could get interest on money in the bank, even if this was not common—we may say that it must have been by special pact because â€?irregular depositum⁹ is post- classical, but the social fact is that it at least sometimes happened.¹⁸⁴ Or we can go back to mandate. Caecilius lucundus was not in the auction business for nothing: he paid up â€?less fee’, mercede minus. Merces is supposed to be the sign-manual of the humble locatio conductio, but it is scarcely to be imagined that lucundus was thought of as a mandatary in banking matters but a paid employee under his other hat as auctioneer; he was in fact carrying out mandates for a reward. Barristers we have discussed; they often worked for a fee, but no one dared to call them employees. And how—once more—about procurators? It depended on status, and changes in status—on individual dignitas. Some procurators were wage-earners, some were not; there are texts about salary of persons who appear to be acting under mandates which require much agility to argue away, and we have faced Ulpian’s curious list of â€?professional’ people suing for fees extra ordinem. The diffi­culties with which scholars have struggled in trying to sort out negotiorum gestio and mandate and their relation to different kinds of procurator perhaps reflect a social fact: that the distinction between the gratuitous services of status-equals and the paid services of status inferiors had partly ceased to be real even in Cicero’s day and grew steadily more unreal. The jurisprudents continued to assert flatly that such-and-such a bargain must be gratuitous to constitute such-and-such a contract, because it was their conceptual framework and otherwise they would have been obliged to re-draw the boundaries of the whole system; but make­shifts were found, and the cognitio extraordinaria, about which they did not have to make the rules, came to the rescue.

*

With mandate and procuratorship we are already within the field of agency, a topic which deserves a general coup d⁹ceil because in modem commerce the principle of agency, that a man can be a legal conduit-pipe between two other men, is extremely important. If we examine Roman law from two aspects, (a) acquisition and alienation (can A’s transaction with C as agent for B result in B acquiring ownership of something directly from C or vice versa?), and (6) binding and benefiting in contract (can A’s transaction produce a contract directly between B and C?), we shall see that it started from a fundamental notion that had both a negative and a positive side: a man could not be a conduit­pipe, unless he were a slave or a filius familias, in which case he could not be anything but a conduit-pipe. It is not possible for us to acquire through an extraneous person’, and â€?A stipulation is void if we stipulate for conveyance to anyone except to one whose power we are in’: these are the maxims quoted by Gaius,¹⁸⁵ and with this basic concept the Romans found it hard to part, though they were resourceful in expedients to facilitate agency in practice in spite of it. Thus, in contract we have already seen the rules by which (a) a dominus or paterfamilias was made liable on the "binding’ transactions of his slave or filius, either fully or to the extent of peculium, and (b) a principal was made Hable on the "binding’ transactions of free persons sui iuris, as well as slaves and sons, if he put them in charge of businesses or ships. Such free persons sui iuris were agents, but one way only; the principal was bound by, but could not benefit by, their transactions; they acquired contractual rights only for themselves, and he must invoke his contract with them to make them hand over. We hear of some further advances: if a firm of partners put in a manager, people who had contracted with him could sue any partner in the firm,¹⁸⁶ and there is some (unfortunately rather dubious) textual evidence to suggest that by the end of our period partners were treated as agents for each other,¹⁸? which nearly reaches the point of making the â€?firm’ a legal entity. For

Q

the municipalities, which had corporate status, the rule was that receipts and discharges could be given by their actor publicus, but had to be signed or sealed by the magistrates?⁸⁸ So much for â€?binding’; as for â€?benefiting’, one’s slave or Jilius familias could certainly make contracts â€?benefiting’ oneself—he could, for example, stipulate?⁸⁹ But there agency stopped: no extranea persona, no one not in your potestas, could stipulate so as to benefit you directly.

In property (acquiring and alienating ownership) we come back to the procurator or general agent (noting that he was not neces­sarily covered by the actio institoria). Could a procurator directly acquire for or alienate from his principal? We are hindered from knowing the truth about this by the malignant fate which has damaged the text of Gaius’ Institutes at the two crucial points?⁹⁰ At II, 64, speaking about alienation by non-owners, he says some­thing (which is lost) about a procurator, and the general opinion is that at any rate by his time, and at least in some circumstances, a procurator was competent to pass ownership in his principal’s property; but we know nothing about the essential point, how far knowledge or authorization by the principal would have been needed. As for acquisition, Gaius says at II, 95, â€?we cannot acquire through an extraneous person; but there is just some question as to the possibility of acquiring possession through a —’. The word is lost, but is probably â€?procurator’, for what Gaius still found so doubtful had become settled by legislation by the time of Ulpian at the extreme end of our period?⁹¹ And if you could get possession through a procurator you could reach ownership by usucapio; you would have to know and authorize.

Akin to agency, and equally important in modem commerce, is delegation and assignment of debt. A owes money to B and is ready to pay, but B, who owes money to C, would like to settle by getting A to pay straight to C—and so on. Much business depends on having fluent and flexible means of achieving this kind of arrangement; Roman law seems again to have been rather stiff-jointed about it. Obligations were personal, not easily to be transferred; also, it has been suggested, there was reluctance, when the law of debt was harsh, to make a debtor accept a new creditor (who might be stickier than the old) without his consent.¹⁹² The â€?novating’ stipulation could produce a delegation or an assignment in practice: â€?That which you owe to me do you promise that it shall be paid to X?’, or â€?That which X owes to me do you promise that it shall be paid to me?’; though technically this resulted in a new debt altogether. Then there was an ingenious â€?penal’ stipulation to get round the rule that you could not stipulate to benefit a third party: â€?Do you promise that such-and- such a sum shall be paid to me if you do not give such-and-such to X?’ There was constitutum debiti alieni, the promise to pay some­one’s debt for him, of which one form was the banker’s receptum which we have seen. But the main device was a special kind of mandate, in which A, to whom B was debtor, gave C a com­mission to undertake his suit for the debt, with an agreement that C need not hand over what he got; C was procurator in rem suam. This did not require the agreement of B. It was â€?cession of action’, and it was useful in a number of other ways also.

*

A reading of the Roman law sources gives one the impression that the giving and taking of security was a universal feature of Roman life; it constituted a network of interrelationships in which everyone was perpetually enmeshed. The special characteristic of Roman as compared with modem society in this regard is usually said to have been that the security they gave and took was comparatively rarely â€?real’ security (money or land or other objects of value) and comparatively frequently â€?personal’ security, that is, the personal standing and credit of their friends and patrons brought in as a guarantee of their own transactions. The cement of their daily financial relationships was people, not things; we need not labour again the relevance of the concepts of ojficium and dignitas to this pattern,¹⁹³ but one must only utter the warning that there is no sound basis for estimating the relative frequency of these two kinds of security; sometimes both were used (and required).

Security was a regular feature of numberless transactions in both private and public law. In the private law, loans of money were secured, usufructuaries had to give guarantees for proper use of what they took over, marriage arrangements commonly involved security for the due return of dowry, sale on credit involved security for the price, and so on and on. In public law, all contractors with the state had to give guarantees for their contracts in a double form, both â€?personal’ and â€?real’—praedes and praedia.¹⁹* A curious surviving contract for building a wall at Puteoli shows that contractors with the municipalities were under the same obligation:¹⁹⁵

�Contract for building wall in front of the Temple of Serapis, far side of road: the successful bidder for the contract is to give sureties and put properties under seal according to the judg­ment of the magistrates.... Payment date: half the price will be paid as soon as adequate properties have been put under seal; the remaining half will be paid on completion and approval of the work.’

The earliest of the municipal charters of Italian towns after the Social War, the lex municipii Tarentini, requires praedes and praedia from candidates for municipal office, as guarantee for their proper handling of public funds,¹⁹⁶ and this must have been general, for the charter of Malaca, of Flavian date, has the same provision, besides one for contractors (the rules of which are said to be the same as those applied by â€?those who are in charge of the treasury at Rome’).¹⁹⁷

Looking at the mechanisms of security, we shall consider first â€?real’ security or the pledge or mortgage, and begin by repeating what was said earlier, that Roman mortgage was simply and solely a kind of security for debt, and was not used as a means of obtaining capital and plant for, or a long-term investment of funds in, industrial enterprise. Perhaps the oldest Roman form of real security vasjiducia.¹⁹* This was the actual making over of full ownership of a res mancipi to your creditor, by mancipation or cession in court, with a pact offides et fiducia, faith and trust, for its reconveyance if the debt was paid, and usually another pact (lex commissorid) that if the debt was not paid by the agreed date the creditor might sell up the mortgaged thing to meet it. Fiduda, though antique, was clearly found useful, for it lasted all through our period and has left a curiously large body of documen­tary evidence. There is, for example, the â€?formula Baetica’, the standard contract that hung in some notary’s office in Spain (which we can only summarize);¹"

�A has received the estate X in good condition by mancipation, and the slave Y by mancipation, from B, on transaction of Jiducia. Pact between A and B: above pledges shall stand as jiducia until all sums credited, lent, promised, etc., by A to B have been duly paid; if not paid by due date, above pledges are to be sold for cash by A or his heir....’

The structure is essentially the same in an actual mancipation of two slaves as security for a loan of one thousand four hundred and fifty sesterces from Pompeii in ad 61 ;²⁰⁰ it contains also an oath that the slaves are not already pledged for debt or owned jointly with anyone else, and after the clause permitting sale it enacts that if the sale fetches a sum inadequate to cover the debt the remainder will still be owing, but if it fetches more than enough the surplus will be refunded to the debtor. To these testimonies we can now add several from Herculaneum, which have moved their editor to suggest that in first-century Campania Jiducia must have been actually the standard form of â€?real’ security. Here are two passages:²⁰¹

�M. Nonius Fuscus swore by Jupiter and the spirits of the gods and the spirit of Nero Claudius Caesar Augustus that the said slave woman Nais is his and owned by him and he has possession of her and does not own her jointly with anyone else and that she is not pledged publicly or privately. Whereupon M. Cominius Primus accepted said slave woman Nais by mancipa­tion for one sesterce on transaction of jiducia for a debt of six hundred sesterces. Libripens (balance-holder) T. Blaesius Saturninus.’

\. re-mancipation of three slaves by M. Nonius Crassus to M. Nonius Fuscus.. ?

(Very strikingly exemplified in all this is the slave as a commodity, being passed from hand to hand like a pound note.) Fiducia, with its transfer of full ownership, gave massive protection to the creditor and was hard on the debtor, who wholly lost, while the debt was owed, a chattel of perhaps much greater value than the debt (for if he was badly in need of money he might be squeezed into giving something valuable xsfiducid). It is not surprising that conviction in an actio fiduciae for recovery of such a pledge was infaming.

The other kinds ofâ€?real’ security had much in common; they were p/gnws and hypotheca.²⁰² The whole of Book 20 of the Digest concerns them, and treats them essentially as a unity. Pignus was pawn—the handing over of something (not confined to res tnancipi or even to corporeal things) to the creditor, who thereby had possession, not ownership, to be returned when the debt was discharged—usually, no doubt, with a â€?commissary’ pact to allow the creditor to sell if the debt was unpaid (or the debtor could sell it if the creditor agreed). Hypotheca was a lien.²⁰³ The debtor in this case did not hand anything over; he continued to occupy his house or use his slave and so on, but they were forfeit if the debt was not paid. As with modem mortgages, property could be saddled with more than one charge of this sort, the creditors having rights according to the chronological order of establishment of the successive hens. (In later law there arose an elaborate and vexatious system of â€?privileged’ creditors, particu­larly the treasury.) Based on this principle of hypotheca, and probably its original and always its main use, was the landlord’s general Hen on his tenant’s goods for the rent, which was auto­matic;²⁰⁴ but the same kind of general hen could equally well be estabhshed by express contract.²⁰⁵ The creditor who held a pignus, having legal possession, was entitled to the praetor’s interdicts that protected it—that is, he had aâ€?real’right to it, and, as will be remembered, it was theft for the debtor to take it away even though he was technically still dominus of it. (In the Eastern parts of the Roman dominions, including Egypt, the standard pledge was a bit different, a form we call â€?antichresis’, which means that the creditor had the right to use the pledge—live in the house, take the fruits of the land and so on—but must set the proceeds off against the debt.²⁰⁶ Putting oneself into paranione

pignvs; hypotheca; â€?personal’ security 247 for a loan was really a kind of â€?antichresis’.)²⁰⁷ In the case of hypotheca, as far as Roman rules were concerned, the landlord­creditor, not yet having possession,²⁰⁸ had to be given a means of asserting his right to get the pledge from the debtor if he defaulted. Already in Republican times he too was given an interdict by the praetor, the interdictum Salvianum, and later (but at least before the codification of the edict under Hadrian) he got an action by the formula Serviana, a â€?real’ right, to recover his pledge from any­one.²⁰⁹

About the details of â€?personal’ security, important though it was, not very much needs to be said. In general it is worth noting that sureties are ubiquitous in our surviving documents, many of which record the transactions of people far below the grand walks of life; which shows that the concept of officium was not confined to the â€?upper crust’. The mechanism of suretyship was the ac­cessory promise. â€?Do you promise x?’, said the creditor to the debtor. â€?I promise’. â€?Do you promise the same x?’, he said to the surety. â€?I promise’. The effect was to make debtor and surety (or sureties) equally liable for the debt; the relationship of debtor and surety between one another was one of mandate. The oldest kinds of surety, the sponsor and fidepromissor (who were the only kinds in Cicero’s day) could only be brought in to support the stipula­tion, the â€?verbal’ contract; but from the time of Labeo (say, the beginning of the Principate) there came in a third, more flexible kind, the fideiussor, who could make a promise in support of a bargain of any kind—a â€?real’ or â€?consensual’ contract, sale and so on. There was a lot of uninteresting legislation about sureties and co-sureties of which Gaius gives quite a long account, but one very important question concerning all three kinds of surety seems unsettlable on existing evidence. Payment of the debt by any party, debtor or surety, naturally released the rest, and so did discharge (acceptilatio) by the creditor. More crucially, a standard opinion of scholars is that suit against the principal debtor, if it got to the vital moment of litis contestation released the sureties,²¹⁰ so that the creditor must ask himself carefully whether to take the risk of suing the debtor and finding him insolvent, for he could not then turn to the sureties. (He could, of course, sue the debtor

for part, and turn to the sureties for the remainder.)²¹¹ But what of suit against the surety first? It was perfectly possible; it was a blow to the debtor’s dignitas since it implied his insolvency, and might be actionable as defamation, but then the debtor might have sacrificed his dignitas and proclaimed his insolvency. The question is, did it automatically release the debtor? There is no good evidence that suit against afideiussor barred a subsequent suit against the principal debtor; the obligation of the fideiussor was technically an indepen­dent one, being based on a quite separate stipulation of his own?¹² But it is usually held that suit against a sponsor or jidepromissor did release the principal debtor because their obligation was not an independent one: they promised â€?the same x\ on the same stipulation just repeated. However, the soundness of this argu­ment has been doubted, and the texts do not settle the matter.²¹³ Dignitas probably supplies the answer in practice; a surety, even if his principal had â€?let the side down’ by insolvency, would pay up without suit to protect his own standing.

In order to round off the picture of Roman security we must briefly describe one other field of its application: it was required very frequently by the legal authorities themselves from parties to litigation or when litigation was likely—from the usufructuary for proper use, from the guardian for proper management, from the heir for due payment of legacies, in damnum infectum and operis novi nuntiatio and so on. Gaius gives an account of the â€?praetor’s stipulations’, and the Digest contains some titles about them.²¹⁴ The guarantee demanded by the courts was in some cases just an unsecured promise on a stipulation, but in others a promise with security—which had to be â€?personal’ security:²¹⁵

�Praetorian guarantees require persons to enter in support of them; neither by pledges nor by the payment into court of money or gold or silver is correct security given.’

We have seen how litigants took contractual bail from one another for initial appearance before the praetor; it was required obligatorily by the praetor for reappearance whenever actions had to be adjourned, and in proceedings for â€?aggravated iniuria' the amount of bail demanded by the praetor for reappearance in effect determined the damages in the suit.²¹⁶ Agents also had to give bail for the attendance of their principals.²¹⁷ Agents again, and guardians and caretakers, might have to promise iudicatum solvi, i.e. not only that they would make a proper defence but that if they were condemned their principal would pay up. This was the point at which Quinctius’s procurator, Alfenus, jibbed :²¹⁸ he refused to satisdare iudicatum solvi, which has raised the same suspicion in the minds of modem readers as no doubt filled the breasts of Quinctius’s adversaries, that the case of Cicero’s client was perhaps not as cast-iron as the great barrister feigned to believe. If the agent was plaintiff in a suit he had to give security â€?for ratification’, namely that his principal would accept the action and the judgment as properly concluded on his behalf and not bring any further suit in the same matter on his own account (for, as will be recalled, the agent’s suit was technically his own). We meet this security amplius non peti several times in Cicero;²¹⁹ perhaps the best example comes in a letter to Atticus (though the background is not clear):²²⁰

�1 have spoken to Acutilius. He says he has had no written communication from his procurator and is surprised that there has been this quarrel because the procurator refused to give security “that no further suit will be entered”.*