2. THE INFORMAL CONTRACTS

 (a) The Real Contracts

The individual contracts. In Justinian’s classification there are four real contracts: mutuum, commodatum, depositum, pignus.

Mutuum, the pattern of the primitive ‘real’ debt,^[74] was a loan for consumption, not simply for use, i.e. a loan of things, such as money, food, and drink, which can ordinarily be used only by being consumed.^[75] It accordingly involved a transfer of ownership, and obliged the borrower to return not the thing itself but its equivalent in quantity and quality. It was, as we have seen, a unilateral contract, and was actionable by the condictio, a stricti iuris action. The obligation being simply to return, no claim for interest could arise from the mutuum itself.

Any agreement for interest had to be clothed as a separate stipulatio.

Commodatum was a loan for use only. Depositum was the hand­ing over of a thing for safe-keeping and not for use. The two contracts were otherwise closely similar. The borrower or depositee received neither ownership nor possession but simply detention.¹ And both contracts were gratuitous. If the borrower was to pay for the loan or the depositor for the deposit, the transaction was either hire or what was eventually called an innominate contract.

Pignus was, as we have seen,² the giving of real security by the transfer of possession.

Obligations re in Gaius.

The classification which we find in Gaius is therefore a transi­tional one. The emergence of agreement as the common factor of contractual obligations causes him to distinguish the con­tractual from the non-contractual obligation re, but he still sees the foundation of both in the old idea of the ‘real’ debt; and the old unity is for him still preserved in the single action which sanctions such debts, the condictios Commodatum, depositum, and

¹ See above, p. 112.

² Above, pp. 151 f.

³              See further below, pp. 223 if. pignus, on the other hand, have no connexion with the old idea.

Relative unimportance of the real contracts. These contracts are less important than their prominence in the Institutional classifica­tion suggests. Commodatum and depositum, being gratuitous, could be of no commercial significance, and a loan or deposit between friends will only rarely give rise to litigation. The existence or not of such a contract would, however, be indirectly relevant if a third person took possession of the thing. For it would deter­mine which party could sue and for what.^{[76] [77]} Pignus, on the other hand, is obviously a business transaction, but its importance lay less in its contractual aspect than in its effect on third parties.^[78] Again, a loan of any substantial sum of money will usually be a business transaction, but, if it is, the lender will expect to receive interest; and since for the exaction of interest on a mutuum a stipulatio was necessary, it would obviously be con­venient to embody both the undertaking to repay the capital and the undertaking to repay the interest in the same stipulatio, and thereby to enable both to be claimed in the same action.

Standards of care. In mutuum the borrower was bound to return the equivalent of what he received, no matter what became of the thing itself. He was the owner, and the risk of loss or damage was on him. In the other three contracts, however, it was neces­sary to determine what amount of care the recipient must show in looking after the things or, in other words, what degree of fault would make him liable for its loss or damage.

It is probable that the classical lawyers distinguished three degrees of liability: for dolus (fraud or bad faith), for culpa (fault or negligence), and for failure to exercise custodia. Dolus and culpa are abstract and generalized. They represent a failure to con­form to the objective (though not of course precisely definable) standards of, respectively, good faith and the care shown by the reasonable man {bonuspaterfamilias'). Custodia, on the other hand, which was probably the oldest of the three, is defined in a much more concrete and casuistic way. A man was liable for loss caused in certain typical ways (e.g. ordinary theft) whether or not he had taken reasonable care to prevent it, but he was not liable for loss caused in certain other typical ways (e.g. theft with violence). It can, however, be approximately defined as a strict liability (i.e. liability irrespective of fault) for all loss not caused by vis maior (which included both superior force and what we call acts of God).

It is not certain to which contracts the requirement of cu­stodia originally applied, since in the law of Justinian it had been confined to a few special cases and replaced by liability for culpa. At the same time, within the concept of culpa, a distinction had been made between culpa levis (slight fault) and culpa lata (gross fault) and, within culpa levis, a further distinction between what modern commentators have called culpa levis in abstracto and culpa levis in concrete.

Culpa lata is hardly to be distinguished from dolus. It is, says a text, ‘not to understand what everyone understands’; it is careless­ness so gross as to suggest bad faith.

Culpa levis in abstracto (judged by an abstract or objective standard) is the original, undifferentiated culpa—a failure to show the diligentiaoi a bonus paterfamilias (reasonable man).

Whatever the standard applied, mora (delay) would displace it. If a party through his own fault failed to fulfil his duty at the proper time, he was strictly liable for any loss or damage which occurred thereafter, even if occasioned by vis maior.

The texts are often in conflict as to the standard required in particular types of contract, and it is difficult to detect any con­sistent principle. For Ulpian the incidence of culpa levis or culpa lata depended on whether the party in question took a benefit under the contract or not. Thus most texts make the depositee liable only for culpa lata, whereas the borrower in commodatum and the pledgee in pignus were liable for culpa levis (in abstracto), and perhaps originally for custodia. Though there are exceptions, this principle is useful as a rough guide. Again, we are told that the reason why the contract of societas (partnership) required diligentia quam suis rebus was that a man has only himself to blame if he chooses a careless partner (the argument being presumably that societas is a peculiarly personal relationship). But this is an elusive principle: it could apply equally to commodatum.

(b) The Consensual Contracts

The consensual contracts were four: emptio venditio (sale), locatio conductio (hire), societas (partnership), mandatum (man­date).

(1) Sale {emptio venditio)

Sale claims a more detailed treatment than the other con­tracts, not only because it is the fundamental commercial contract, but also because the Roman law of sale has had great influence on modern Civil law, and even on the Common law.

Formation of the contract. As we have seen, the essential elements of the contract were that the parties should be agreed on a thing and a price.

(i)   The thing. The typical object of a sale was a res corpor alls, movable or immovable, but it might also be an incorporeal thing, such as a praedial servitude, or even a right of action against a third party. In short, there could be a sale of any right which the seller was capable of transferring to the buyer. But if what was being transferred was not a right but simply the benefit of the ‘seller’s’ services or of the use of a thing, the contract was not sale but hire (locatio conductw). For while the incidents appropriate to a sale of a corporeal thing could be extended without much difficulty or incongruity to incorporeal things, the same could not be said of bargains for services or for the use of a thing. But the drawing of a line between the two types of transaction did inevitably involve some fine dis­tinctions. If I engage a goldsmith to make me a ring, is this a sale of the ring or a hire of his services? Opinions differed, but the answer was eventually held to depend on which of us sup­plied the material—if it were I, then the contract was hire, if the goldsmith, then it was sale. The underlying principle is that the typical sale is an agreement to transfer the ownership of a thing, and it is only when the maker supplies the material that there can be such a transfer. It must be for a similar reason that it was further held that if I agree with a builder that he will build a house on my land, the contract is hire even though the builder supplies the materials. In such a case there is indeed a transfer of ownership, but it occurs, quite independently of any agreement, by the operation of the principle of accessio.

The thing must exist at the time of the agreement If it did not, either because it had never existed or because it had by then been destroyed, there was no sale. This was an application of the much wider principle, expressed in the maxim impossibilium nulla obligation that there could be no contract to do the impos­sible.

The most restrictive requirement was that the thing must be identified, in the sense that it must be either specific (‘my

03 slave, Stichus’, ‘that cask of wine’), or what is nowadays some­times called semi-specific, i.e. part of a specified mass (‘ten gallons of the wine in that cask’) or one of a number of specific things (‘one of my slaves’). There could be no sale of ‘generic goods’, i.e. of things defined only by reference to their genus (‘a cask of wine’, or ‘a slave’). Such a transaction could, of course, be made effective by means of two stipulations, but it could not, it seems, constitute emptio venditio. This rule looks like a survival from the primitive cash sale, but it is surprising that so thriving a commercial society as imperial Rome should have retained it. The explanation lies perhaps in the ease with which a stipulation could be made.

Even within these limits, however, a sale of some ‘future things’ was possible, e.g. next year’s crop from a specified field. But such a transaction presents a problem. If the crop fails, the principle that there can be no sale of a non-existent thing should make the contract void, and the seller should not be able to claim the price. This will be a proper result if the contract was intended to be conditional on the crop’s coming into existence, i.e. if the parties intended that the seller should take the risk of there being no crop. But they may well have intended that it should be the buyer who took this risk, the price being fixed accordingly. For the interests of the merchant, who buys, and the farmer (particularly the peasant farmer), who sells, are different. The farmer seeks protection from the vagaries of the weather and the market, whereas the merchant makes his fortune by risking his capital. The farmer may be willing to accept a low price in return for a secure income, and the mer­chant may be willing to risk a big loss in return for the chance of a big profit. The difficulty was met by making a distinction between emptio rei speratae (the sale of an expected thing) and emptio spei (the sale of an expectation). If the parties must have in­tended that the seller should take the risk of total loss, there was an emptio rei speratae, and the ordinary principle applied: if there was no thing there was no contract. But if the intention must have been that the buyer should take the risk, there was an emptio spei. The transaction was construed as the sale not of a res but of a spes—not of a physical thing but of an expectation. The difference between the two will commonly be reflected in the way the price is formulated. In an emptio rei speratae the

price will be proportionate to the actual yield (e.g. so much a bushel), whereas in an emptio spei the price will be fixed at so much per acre or so much for the whole crop, the same price being payable whatever the yield.

(ii)  The price. There must be a money price. This requirement meant that permutatio (exchange or barter) was not sale. The Sabinians had held otherwise, but the Proculians prevailed. Their argument was that if the price were not in money it would be impossible to distinguish buyer from seller, the im­plication being that since their duties, and the actions by which they were enforced, were different, the law would then be un­workable. It seems at first sight surprising that so old a transac­tion as barter should have been excluded from the Roman list of typical contracts, but it must be remembered that if part of the price were in money the Proculian objection would be met and the contract would be one of sale. And an exchange in which there is no balance payable in money is not likely to be either common or commercially important.¹

The price must be fixed (certum). There could be no sale for what English law calls a ‘reasonable price’, i.e. a price to be fixed by the parties in their subsequent dealings or, in default of this, by the court. Moreover, the price was sufficiently fixed only if it were either known or immediately ascertainable at the time of the agreement (e.g. ‘today’s market price’). To this rule there was one exception, disputed in the classical law but admitted by Justinian, that the price was sufficiently certum if it were left to be fixed by a named third party. But if, in the event, the third party did not fix it, the contract was void.

The requirement that the price be fixed is more easily defended than the requirement that the thing be specifically identified. Many of the incidents of a contract must, as we have seen, be supplied by the law, but in its essentials it should be the work of the parties. The law should not make their bargain for them. And one of the essentials of a sale is clearly the price.

For the same reason, the parties were free to fix their own price, subject only to the rule that it must be seriously intended and must not merely mask a gift. Inadequacy of price was

¹ Permutatio fell under the eventual heading of innominate contracts. See below, pp. 189 ff.

therefore ordinarily irrelevant, but might invalidate the trans­action if the parties were husband and wife.¹

In the late law, however, there appeared the doctrine of what is called laesio enormis (literally ‘abnormal injury’), the intended scope of which is uncertain, since it appears only in two con­stitutions concerned with individual cases. Both are sales of land and in both the seller is allowed to rescind the sale on the ground that the price agreed was less than half the real value of the land. Later civilians have usually confined the doctrine within these limits, refusing to extend it to sales of movables or to cases in which it is the buyer who complains that he has paid too much. In this form, though with modifications in detail, it survives in some modern Civil law systems, especially those derived from the French. Even so, its application presents con­siderable difficulties, and we may think that the classical law (like the Common law) chose the better course in refusing to consider the adequacy of the price unless bad faith or incapacity was shown. In the sphere of contracts, more than in other parts of the law, it is less important that the law should in every case be just than that it should be certain. The law of contract pro­vides the framework for commercial life, and in commercial life the taking of risks is inevitable. It is probably better that some sellers should be saddled with unfair, though not fraudulent, contracts than that a man who has taken a risk and lost should be able to escape from his bargain.

(iii)  Consent. This element was found, in the eventual analysis, in all contracts, but the Roman law, unlike the modern Civil law, did not isolate and generalize such concepts. It preferred to deal with them casuistically in connexion with individual contracts. It is thus that some problems of defective consent are discussed particularly in connexion with sale.

Consent involves the meeting of two minds, the concurrence of two intents. The first need is therefore to determine what these intents are. In doing this one meets the possibility of a divergence between a man’s real intent and the manifestation of that intent—between what modern lawyers sometimes call subjective and objective intent. For example, there may be an apparent agreement to buy and sell a horse but the buyer may have had in mind horse A and the seller horse B, neither party ¹ Since gifts between husband and wife were void. See above, p. 90. being aware of the disagreement. There is here, subjectively, no consent. But it may be that the natural interpretation of what passed between the parties, the interpretation of the reasonable bystander, is that they were agreed on horse A. Objectively there is consent. Modern systems differ. The older view, resting on the philosophical doctrine of the autonomy of the will (i.e. that the binding force of a contract derives from the human will, which is its own law) requires subjective consent. The more recent view asserts that the validity of a contract comes not from the individual will but from the law, and that the law is concerned with a balancing of interests. It emphasizes the difficulties of proof and the importance of stability and certainty in commercial transactions: each party should be able to assume that the other will be held to the objective interpretation of the transaction.

The Roman lawyers, with their habitual disregard of ques­tions of evidence, give little attention to matters such as this, but seem tacitly to assume a subjective interpretation, qualified only by such principles as that a man may not profit from an ignorance which comes from his own gross carelessness.

Granted that the intents of the parties are known, the second need is to determine to what extent those intents must concur, or, in other words, what defects of consent will vitiate a contract. Such defects are of two kinds. Either there is no consent, as to the whole or some part of the transaction, or even though there is consent, it has been obtained in such a way that the law will not enforce it. Defects of the first kind arise from error (mistake), those of the second from dolus (fraud or bad faith) or metus (duress).

In a bonae fidei contract such as sale, both dolus and metus came within the wide heading of bad faith so as to enable the iudex to hold that ex fide bona the innocent party should not be held to his contract. And it was bad faith not only if one party actively deceived the other on some material point, but even if he did no more than passively to acquiesce in the other’s self-deception.

Error gives rise to greater difficulties. It may be classified either with reference to the states of mind of the parties or with reference to the content of their mistake.¹ The states of

¹ Such classifications, although the materials for them, and to some extent the

mind of the parties may be defective in two principal ways. Either both parties make the same mistake (e.g. both believe that the cup they are dealing with is gold when in fact it is brass) or, more commonly, one party has one intent and the other another (e.g. the case of the horses, above). Within this latter category there is indeed a further distinction: either the mistake of one party is known to the other, or else neither is aware of the divergence of intent. But to acquiesce in another’s mistake is, as has just been said, bad faith, and can be more simply and readily dealt with as such.

Obviously not every mistake can be allowed to prevent the formation of a contract. To use the modern term, not every mis­take can be allowed to be ‘operative’. To produce this drastic result the mistake must be as to some fundamental matter. But what meaning are we to give to ‘fundamental’? This is the problem of the content of mistake. The traditional Civil law treatment, which has had a considerable influence on the Common law also, simply classifies the types of operative mistake which are to be found in the Roman texts.

Error in negotio occurs when one party thinks he is entering into one type of transaction and the other thinks he is entering into another (e.g. sale as opposed to hire). There can then be no contract. Error in pretio (where the parties intend different prices) and error in quantitate (where they intend different quantities) are, however, only partially operative. Neither party can enforce the contract at his own figure, but each can, if he wishes, enforce it at the other’s. Thus, if the seller intended a price of io and the buyer a price of 5, the seller can enforce the contract if he is prepared to accept 5, and the buyer if he is prepared to pay 10. Error in persona occurs where one party is mistaken as to the identity of the other. Surprisingly, however, it is hardly mentioned in the texts. Error in corpore is mistake as to the identity of the thing sold, hired, &c. (e.g. the case of the horses, above). It must be distinguished from error in nomine where the parties are in fact agreed on the particular object, but give it different names or descriptions.

In all these cases the mistake must obviously be of the second category mentioned above: it must result from divergent intent, terminology, can be found in the Roman sources, inevitably distort the casuistic approach of the Roman jurist.

In error in substantia, however, it may fall into either category. This, the most controversial and elusive type of mistake, occurs when the parties are agreed as to the physical identity of the thing sold (i.e. there is no error in corpore} but are mistaken as to some essential characteristic. The jurists seem sometimes to be thinking in terms of the philosophical distinction between sub­stance and accident, but the illustrations offered of operative mistake (where bronze is mistaken for gold, but not where gold­alloy is mistaken for gold; vinegar for wine, but not sour wine for good wine; a silver-plated table for a solid one, but not— according to another text—a gilt dish for a gold one) make it difficult to spell out a coherent principle in terms of the philo­sophical doctrine. Again, the term substantia is sometimes re­placed by materia, but a simply material test cannot account for the opinion that the purchase of a female slave in mistake for a male is void. The texts seem to have been altered, and the original doctrine or doctrines are probably irrecoverable. Its modem descendant, however, is of considerable importance, especially in French law. ‘Substance’ has there thrown off any material connotation and is that essential quality of the thing which determined the buyer to buy. The purchase of an imita­tion for a genuine Rembrandt, of a modern reproduction for an antique table, or of an old horse for a young one, will therefore all be void. Since the mistake may be, and usually will be, that of the buyer alone, there is an obvious danger that the doctrine may be used, as it were as an after-thought, in an attempt simply to escape from a bad bargain. It must therefore be shown that the seller knew at the time of the contract that the quality in question was in the buyer’s eyes essential. English law is even more restrictive. If it admits the doctrine at all—and this is very doubtful—it requires that the mistake be the mistake of both parties.

Effects of the contract, (i) The passing of title. This was, strictly, an effect not of the contract but of the consequent conveyance. For we have seen^[79] that Roman law insisted on the distinction between contract and conveyance, with the result that owner­ship did not pass to the buyer when the contract was made, but only when the thing was actually conveyed. Moreover, at least in the law of Justinian, it was necessary not only that the thing be delivered, but also that the price be paid or security given for its payment. The effect of this rule should have been to protect the seller against the buyer’s insolvency, since until the price was paid he could assert title to the thing, no matter into whose hands it had come, but we are also told that even the giving of credit was sufficient, and this seems to deprive the rule of all effect. For it is only when the seller has given credit (i.e. has delivered the thing without requiring payment of the price) that he needs the protection provided by the rule.

(ii) The passing of risk. Outside the contract of sale the rule is that the owner takes the risk of accidental loss or damage (res perit domino}. For example, we have seen that the borrower in a contract of commodatum was, at least in the law of Justinian, liable for loss or damage only if it were caused by his negli­gence. Otherwise the risk was on the lender. In sale, however, the rule was different. The risk (pericidum) passed to the buyer as soon as the contract was complete even though he did not be­come owner until the thing was conveyed. In other words, pro­vided the seller looked after the thing with due care in the period between contract and conveyance, he could claim the price from the buyer no matter what happened to the thing. And conversely the buyer had no claim against the seller if, without his fault, the thing was destroyed or damaged before convey­ance.

The need to define the moment at which the contract was complete so as to pass the risk gave rise to a number of detailed rules, but the broad principle was that nothing should remain to be done except the delivery of the thing and the payment of the price. For example, a sale of twenty bottles from a larger stock was not complete until the particular twenty were identi­fied; and a sale of ‘all my wine’ at so much a bottle was not complete until the bottles had been counted and the price thereby determined.

The separation of title and risk is often cited as an illogicality and a defect in the Roman law, and most modern systems have in consequence rejected it. But the Roman rule corresponds to the underlying economic facts. Until the contract is complete the seller retains an economic interest: in the example given above of a sale of twenty bottles from a larger stock, until the particular twenty are identified the seller has a power of choice which may be economically valuable. And similarly, once the contract has been completed, the seller, though he still has a legal interest in the thing, has no further economic interest in it. If there is a rise in the market price it is the buyer and not the seller who is entitled to take advantage of it by selling the thing.¹ And conversely, of course, the buyer takes the risk of a fall in the market price. It is for the same reason that he takes the risk of loss or deterioration.^{[80] [81]}

With the risk, on the same principle, go any accruing fruits or other benefits (commoduni). For example, if between the sale and conveyance of a mare she foals, the buyer has a right to the foal. This right is, of course, in personam', the seller still owns the mare and therefore owns the foal, but he is under a duty to convey both to the buyer.

Duties of the seller, (i) Care and delivery. The seller, we have seen, was bound to deliver the thing and until delivery to take care of it. He may originally have been liable for custodia, but at least in Justinian’s law he was required to show only the care of a bonus paterfamilias.

(ii) Warranty against eviction. The typical sale is an agreement for the transfer of the ownership of a thing, but the fact that the seller is not owner does not in itself vitiate the sale. He is required only to abstain (in this as in all aspects of sale) from bad faith, and to maintain the buyer in undisturbed possession until, if ever, he becomes owner by usucapio. If therefore the seller knows that he is not owner and the buyer does not, the buyer has a remedy, but on the ground of the seller’s bad faith, not of his lack of title. Whereas if the seller is in good faith the buyer has no remedy unless and until he is evicted by the owner. In most cases, of course, this will be a distinction without a difference, since the buyer will usually become aware of the seller’s lack of title only when the owner asserts his title, but exceptionally the rule can cause hardship. For example, if it is certain that the seller was not the owner but uncertain who is, or if the owner cannot be found, the buyer can neither claim against the seller nor, since he is aware of his lack of title, resell. The origin of the rule is much debated, but its survival into the classical law is surprising. It has disappeared from modern Civil law.

(iii) Warranty against latent defects. The original principle of the civil law was caveat emptor.¹ The seller was not liable for any defects in the thing unless he had by stipulatio expressly under­taken such liability. The development of the idea of good faith imposed a considerable qualification on the rule by making him liable for any defects of which he knew but which he had not revealed. Moreover, since the requirement of good faith was inherent in the contract itself, the seller could not contract out of this liability. Such a high standard of honest dealing is not imposed on the English seller.

The further development of the seller’s liability came from the Aedilician Edict. In exercise of their jurisdiction over the market the Aediles issued edicts regulating market sales of slaves and beasts of burden. Sellers were required to display on a board a statement of any physical defects and, in the case of a slave, certain other defects also (if he was a vagabond or a runaway or burdened with noxal liability²). If a defect appeared which had not been so declared the buyer, if he sued within six months, could claim rescission of the sale by the actio redhibi­toria, and, if within twelve months, could claim the difference between the price paid and the actual value of the defective slave or animal by the actio quanti minoris. In both actions the knowledge or ignorance of the seller was irrelevant: liability was strict. The seller could, however, exclude all such liability

¹      The maxim is not, however, Roman.

²                                                                                                        See below, p. 223.

if he made clear his intention to do so, though he would still be liable for bad faith.

By a process which remains obscure, this Aedilician liability was extended to sales outside the market, and eventually, but only by Justinian, to sales of every kind of thing, including land. This all-embracing strict liability, with its two actions, survives in the Corpus luris alongside the older liability under the civil law for what had been expressly warranted and for bad faith. The forms of action were no longer significant, but it was presumably intended that where more than one remedy was available, the buyer should choose whichever was most advantageous to him.

Duties of the buyer. The buyer’s principal duties were to pay the price (a duty which was concurrent with the seller’s duty to deliver, in the sense that if one party sought to enforce the other’s duty he must either have performed or be ready to perform his own), and to compensate the seller for any expenses he incurred in looking after the thing between contract and delivery.

(2)   Hire {locatio conduct™)

Scope and character. Into the framework of this single contract the Roman law fitted a wide range of transactions. There may at first sight seem to be little in common between a lease of land and a contract of employment, but in the Roman analysis both were locatio conduct™. Within the single Roman category later civilians distinguished three types: l.c. rei, l.c. operarum, l.c. operis. In l.c. rei one party {locator) places a thing, whether movable or immovable, at the disposal of another {conductor) for his use or enjoyment; in l.c. operarum the locator places his services {operae) at the disposal of the conductor', and in l.c. operis the locator places out a piece of work {opus) to be done by the con­ductor, the work having always, it seems, a physical object—a slave to be taught, a house to be repaired, and so forth—so that the texts speak of the thing being placed out to be worked upon.

The key to an otherwise confusing terminology lies in the idea of locatio as a placing out. It is for this reason that, although both l.c. operarum and l.c. operis are contracts for the performance of work, in the former the work is done by the locator and in the

183 latter by the conductor. In the former the locator places out his services, in the latter he places out the job to be done.¹ In a somewhat similar way we sometimes speak in English of placing or letting out the contract for, for example, the construction of a building. The difference of substance between the two contracts can be defined as a difference of control and responsi­bility. Whether a contract for the repair of a house is l.c. operarum or l.c. operis will depend on the degree of control which the house-owner is to exercise over the carrying out of the work. If he merely specifies the character and quality of the work to be done and leaves its detailed execution to the other party, the contract will be l.c. operis, but if he is to have supervision of the work from moment to moment it will be l.c. operarum. In a different context English law makes a distinction on these lines between an independent contractor and a servant.

Such is the civilian classification, but neither in terminology nor in substance are its distinctions clearly made in the Roman texts. This is not to say that the jurists were unaware of the differences between hiring a thing and hiring a man’s services, or between hiring a man’s services and commissioning the performance of a piece of work, but rather that they never arrived at a systematic analysis of them. This may seem sur­prising. The virtue, we have said, of the Roman system of specific contracts was that it made possible the elaboration of a scheme of incidents appropriate to each type of transaction, and the three types of locatio conduct™, although broadly similar, were sufficiently different to call for differing incidents. This is obvious where the contrast is between the lease of land and the hire of services, but it is also true where the contrast is between l.c. operarum and l.c. operis. If, for example, a house under con­struction is accidentally burned down and work has to begin again, or if the work is held up by bad weather, the incidence of the risk of these misfortunes must depend on the character of the contract between the parties. The jurists, of course, knew this, but they did not, as they did elsewhere, arrive at a set of clear-cut distinctions.² Moreover, the threefold classification

¹              This has a consequence in terms of actions. In the former the person doing the work will have the actio locati and in the latter the actio conducti.

²              For this reason the outcome of the texts on such matters as risk and the duties of the parties cannot be simply stated. These matters are therefore omitted in the discussion of the effects of the contract, below.

towards which they were working does not represent the limit of useful analysis. Within the category of l.c. rei the incidents appropriate to the hire of, for example, furniture will differ from those appropriate to a lease of agricultural land. For this reason modern Civil law systems commonly distinguish either between contracts in which the lessee is entitled to the use of the thing and those in which he is entitled both to its use and to the enjoyment of its fruits, or, more simply, between those relating to agricultural land and those relating to other things.

These shortcomings are only an aspect of what seems to us a wider lack of thoroughness in the Roman treatment of locatio conductio. Buyer and seller are commonly economic equals, but landlord and tenant or employer and employee may well not be. In our eyes, therefore, there is a greater need to regulate their relationship in order to prevent the economically more powerful party from abusing his position. In the ancient world this inequality was more marked: the agricultural tenant was commonly a small-holder who eventually, in the later Empire, degenerated into a serf (colonus glebae adscriptus), and the dis­parity of bargaining power between employer and employee was accentuated by the much greater extremes of wealth and poverty, and by the institution of slavery. But the jurists were not social reformers. They were conservative members of the wealthy class, and saw no need to make things other than they were. To blame them for this is to blame them for having lived when they did, but the consequence was that they devoted less interest to locatio conductio than its social and economic impor­tance seems to us to deserve. Their treatment of it is much less thorough than their treatment of, for example, legacies.¹

Formation of the contract. The rules as to the thing, the price, and consent were much the same as those of sale, in so far as they could be applied. The insistence on a money price did, however, yield sufficiently to admit the share-cropping tenancy (colonia partiaria). And the scope of the contract for services was limited by the exclusion of what we should call the professions and the liberal arts. For it offended Roman ideas of propriety that persons of good social standing should be paid a reward. The

¹ It is true that much of the modern English law on the subject derives not from ‘lawyers’ law’ but from statute; but imperial legislation is almost silent.

contract for their services was, if anything, mandate. The social distinction between professions and trades and between arts and crafts was, however, differently drawn in the ancient world. Painting and medicine, for example, were commonly the occu­pations of slaves.

Effects of the contract.¹ In l.c. rei the conductor had only detention of the thing. He acquired, in other words, nothing but his right in personam against the locator,² on whom he was therefore entirely dependent. He had no remedy directly against a third party who interfered with his occupation, but could only sue the locator for breach of his duty to accord him undisturbed enjoy­ment. This was so even if the third party were a purchaser from the locator. The conductor had no protection against the purchaser, who could enter as of right, leaving the conductor to his remedy against the locator. The locator would usually insert a term in the contract of sale requiring the purchaser to respect the tenancy of the conductor, but the purchaser’s breach of this term would give a remedy only to the locator. This was expressed succinctly, though inaccurately,^[82] in the medieval maxim ‘sale breaks hire’.

but joint owners are not necessarily partners: they may well, as in the case of joint heirs, have no common purpose.

Each party must make some contribution, either of capital, skill, or labour. Otherwise the transaction will be one not of partnership but of gift. For the same reason the societas leonina (so-called from the fable of the lion and the ass), in which one party was excluded from sharing in the profits, was, after some dispute, held to be void. It was otherwise, however, with a societas in which one party was excluded from bearing any loss: such a transaction was not a gift, since his participation might be worth securing at that price and it was not for the law to assess the value or adequacy of the parties’ contributions. It was in accord with this principle that, unless it was otherwise agreed, the parties shared equally in both profit and loss, what­ever their contributions.

The purpose must, as in all contracts, be lawful. One of a band of robbers could not, we are told, assert a partnership in order to obtain his share of the proceeds.

Incidents of the contract. The relationship between partners was treated as an especially personal one, the relationship between brothers being invoked as an analogy. The law therefore left the incidents of the contract to a considerable extent to be regulated by the broad principle of good faith. Like sale and hire it was perfectly bilateral, but whereas in sale and hire, as the doubling of their names in Latin indicates, the interests of the parties were different, it was of the essence of societas that their interests should be the same. In both sale and hire, therefore, there were two different sets of duties and two different actions, but in societas the duties of all partners were in principle the same and there was accordingly only a single action, the actio pro socio. Moreover litigation between partners was held to be incompatible with the ‘brotherly’ character of the contract, and therefore the bringing of the action by any partner terminated the contract. The action was in consequence a general winding-up action rather than a remedy for the breach of a particular duty. A partnership could never be brought to court until it had been terminated.

The personal character of the contract appeared also in the other ways in which it could be terminated. Any partner could at any time bring the partnership to an end by a unilateral renunciation. He could, it is true, be called to account in the actio pro socio if his purpose was to avoid bringing into the part­nership some expected gain, or to avoid sharing in some expected loss, but the renunciation was nevertheless effective. The con­tract was likewise terminated by the death, capitis deminut io, or bankruptcy of any partner.

To the English lawyer the most surprising feature of societas is that it created relationships only between the partners them­selves. It gave a third party who contracted with one partner no rights against the others, even though they might have ex­pressly authorized the contract. They were bound to make good to the contracting partner their share in any loss he might suffer, but he alone was liable to the creditor. The English creditor, by contrast, can claim the whole debt from any or all of the partners. But the Roman rule was only one aspect of the general principle, discussed below,¹ that obligations were strictly personal.

(4)   Mandate (mandatum)

Mandatum existed when one person (mandatarius) agreed to perform a service gratuitously for another (mandator) at his request. The service might be of any kind but consisted com­monly in entering into a contract or some other legal transac­tion with a third party. Mandate was imperfectly bilateral,² the mandatarius being bound to perform the service, and the mandator being contingently bound to indemnify him for his expenses. The mandatarius was also bound to account to the man­dator for any incidental benefits he derived from the perform­ance of the service. If the service was the making of a contract, he satisfied this duty by assigning³ to the mandator his action on the contract.

To the English lawyer, accustomed to the idea of contract as necessarily a bargain, the surprising feature of mandate is its gratuitous character. The mandatarius could indeed with im­punity withdraw from the contract before he had begun to carry it out, but thereafter he was bound, even though he had no interest. The origins of the contract are hidden, but they lie

² See above, p. 169, n. 1.

¹PP· the distinction between the two methods of proceeding had long since disappeared, one would expect mandate to have ceased to be gratuitous even in this sense, but the classical statement of the law is preserved and we do not know what happened in practice.

Although the mandatarius could claim no reward under the contract he might nevertheless have an interest in its execution. If, for example, he were asked to lend money at interest to a third party, he would derive an advantage to the extent of the interest on the loan. It was in cases such as this that there arose a further problem: how was mandate to be distinguished from mere advice? The answer was that a mandate in the interest of the mandatarius alone was not binding. If, therefore, A told B that he should employ his money by lending it at interest, or by investing it in some other way, and B incurred a loss, A was not liable. But what if A told B to lend his money to a particular person (C) ? If A himself had an interest (e.g. because C in­tended to use the money on some project in which A was con­cerned) there was no difficulty: the mandate was not in the interest of the mandatarius alone. There was greater hesitation where A had no such interest, but it was eventually held to be sufficient that C had an interest. This made possible a useful means of suretyship. For if A gave a mandate to B to lend to C, and C failed to repay the loan, B could claim to have his loss made good by A. A was in effect acting as surety for C.¹

Mandate, like societas, was ended by the death of either party, but the rule was stretched to allow the mandatarius to claim compensation for expenses incurred up to the moment when he learned of the death of the mandator. The consequences would otherwise be harsh, as English law, which makes no such concession, shows.

‘Mandator’ and ‘mandatarius’ may be loosely rendered ‘prin­cipal’ and ‘agent’, but there is this fundamental difference that the English agent when he contracts on behalf of his principal creates a relationship directly between the principal and the third party—a relationship to which he is ordinarily not even in law a party. The strictly personal character of the Roman obligation, however, made agency in this sense impossible, and the mandatarius alone was both liable and entitled on any contract which he made. The extent to which this principle was circumvented is discussed below.²

(c) Innominate Contracts

The four real and four consensual contracts, together with the contracts verbis and literis,³ exhaust the Institutional list of contracts, but the list leaves gaps and uncertainties. It leaves gaps because it excludes several common types of agreement, such as exchange or any agreement which calls for the payment of a reasonable price (e.g. an agreement for the making of repairs the extent of which cannot be exactly foreseen). It leaves uncertainties because, while it may be clear that a given agree­ment is a contract, there may be a doubt as to the particular heading under which it should be placed. We have encountered examples of this on the borderline between sale and hire, and there were others. Of these the transaction called aestimatum is an example. We should call it an agreement for sale or return: one party provides the other with a thing on the terms that if the other sells it he will pay an agreed sum, and that if he does

¹                                                                                                                                         The modern name of this type of suretyship is mandatum qualificatum. Another reason given for its validity is that B would not have lent to C if it had not been for A’s mandate, but this could equally well apply to the case where A simply told B that he should employ his money by lending it at interest (to no one in particular). It leaves the distinction between mandate and advice to the elusive test of the intention of the parties.       ² pp. 201 ff.                            ³ See below, pp. 193 ff.

not he will return it. This has affinities, if not with mandate and societas, at least with sale and with hire (both of a thing and of services), but does not clearly fall within the limits of any.

The stipulation provided, as we have seen, a way of escape from these difficulties, but if this way were not followed the ius civile offered only limited assistance. Where one side of a bargain consisted in the conveyance of a res, and that side had been performed but the other had not, a condictio could be brought for the return of the res on the ground that it had been given for a purpose (the securing of the performance of the other side of the bargain), which purpose had not been achieved, and that therefore the recipient was not justified in retaining it. This is, in Justinian’s classification, a claim quasi ex contractu.^[LXXXIII] It is a claim for restitution and not, as is a contractual action, for dam­ages for non-performance. For example, if the transaction is an ex­change of a horse for a cow, the horse having been conveyed but the cow not, the owner of the horse will secure by the condictio the restitution of the value of the horse, whereas a contractual remedy would have given him what he had lost by not receiving the cow. This will normally, of course, be the value of the cow, but if he has suffered any consequential loss it will include that also. The practical difference between the two types of remedy will therefore depend on the market value of horses and cows at the time when the action is brought, and on whether the plaintiff has suffered any consequential loss.

The condictio thus provided only a partial remedy. It was confined to agreements which had been performed on one side and in which the performance consisted in a conveyance, and it was directed to restitution and not to performance. Where there had been no performance, or the performance consisted in an act rather than a convevance, the aggrieved party could bring a Praetorian actio de dolo,² but he would then have to show that the breach was due to the defendant’s bad faith.

Such was probably the law at the end of the Republic. By the time of Justinian, however, there had grown up a general action which went by various names but which is most often

191 called actio praescriptis verbis. This action lay whenever one party to an agreement had performed his side of the bargain but the other had failed to perform his (provided, of course, that the agreement did not fall within any of the recognized types of contract). A general principle had thus evolved that an agree­ment which is executed on one side is a contract. The principle emerges clearly from texts in the Digest and is important as con­stituting a break from the system of typical contracts, but since Justinian’s compilers did not revise the traditional classification, it remained without express recognition, and the contracts with­out a name. The term ‘innominate (i.e. nameless) contracts’ by which they are now known occurs first in Theophilus.¹

The more important types in fact have names, such as per­mutatio (exchange) and aestimatum, but this must not be allowed to obscure the significance of the innominate contracts, which lies precisely in the generality of the principle which they represent. For this principle goes a long way towards filling the gaps and re­solving the uncertainties left by the system of typical contracts.

The steps by which this general principle was developed are obscure, but the ultimate generalization is the work of Jus­tinian’s compilers. It is for this reason that in the Corpus Juris both the condictio and the actio praescriptis verbis exist side by side as remedies for the same situation. Nowhere else is there this concurrence of quasi-contractual and contractual remedies. In a contract of sale, for example, the unpaid seller cannot elect to bring a condictio for the return of the thing (a course which would be advantageous if the value of the thing increased since the sale); he is confined to his action on the contract. The same rule should logically have applied to innominate contracts, but the work of expunging the condictio from the texts was beyond the capacity of the compilers,² and they simply added the actio praescriptis verbis.

(tf) Pacts

Bare pacts. The original meaning of pactum and related words is that of a compromise—an agreement not to sue. It was a rule

¹  See above, p. 45.

²It would have been extremely difficult in any case, since the condictio ob cau­sam datorum was only one application of the condictio, and the scope even of the con­dictio ob causam datorum was wider than that of the innominate contracts. See below, p. 230.

of the Civil law that such a pactum completely extinguished any obligation arising ex delicto, and the Praetor extended the prin­ciple by allowing an exceptio pacti to be pleaded in bar of any action. Meanwhile the meaning of pactum had widened to in­clude any agreement, and there was coined the maxim, which is fundamental to the Roman system of contracts, that a bare pact (i.e. an agreement which does not fall within the limits of any recognized contract) begets a defence but no action.¹ In the informal contracts this principle was relaxed so far as to allow agreements made at the time of the contract [pacta con­tinua or in continenti facta) to be treated as part of the contract. But all agreements made subsequently [pacta ex intervallo), and in stipulations pacta continua as well, were effective only by way of defence.

Clothed pacts. In a few cases, however, the Praetor, and later the Emperors, gave actions on agreements which did not fall within the list of contracts. In modern terminology these are called pacta vestita (clothed pacts) and are distinguished as pacta prae­toria and pacta legitima respectively. Only one need be mentioned here.

Constitutum debiti was an agreement to pay an existing debt at a fixed time. If, for example, a debt owing under a mutuum were due on i January, but at the debtor’s request the creditor allowed six months’ grace, this agreement would be a con­stitutum. By suing on the constitutum the creditor could obtain payment of the debt and of a penalty of one-half in addition, whereas in the condictio which lay to enforce the original mutuum the penalty was only of one-third.² The difference in effect compensated the plaintiff for his inability to claim interest. The agreement could also, however, relate to a debt owed by a third party to the promisee [constitutum debiti alieni), the promisor becoming in this way a surety for the debt of the third party.³

¹  ‘Nuda pactio obligationem non parit sed parit exceptionem.’

² These penalties, obtained by procedural wagers on the outcome of the action, are a peculiar feature of Roman law, but were confined to a few actions. The wagers took the form of stipulations. Cf. above, p. 161, n. 2.

³   Cf. below, p. 205.