WHAT WAS THE ACTIO ONERIS AVERSI?

Abstract

D.19.2.31 contains a reply to a question of law attributed to the late-Republican jurist P. Alfenus Varus. Several people had delivered grain to a carrier which was shot into a common pile in the hold of his ship. Subsequently the carrier returned a share of the grain to one of them before the ship went down. The question is asked if the others can proceed against the carrier in respect of their share by raising an action for onus aversum. This article provides a new insight into the scope and application of this otherwise obscure Roman action, by reference to the role of the tort of conversion in analogous cases at common law.

I. Introduction

D.19.2.31 contains a reply to a question of law which has been excerpted by the compilers of Justinian’s Digest from Paul’s epitome of a collection of responsa attributed to the late-Republican jurist P. Alfenus Varus. Several people had delivered grain to a carrier called Saufeius which was shot into a common pile in the hold of his ship. Subsequently Saufeius returned their share of the grain to one of them before the ship went down. The question is asked if the others can proceed against the carrier in respect of their share by raising an action for onus aversum (“aversion” of a cargo). This article inquires into the scope and application of the otherwise obscure actio oneris aversi. The problem lies at the intersection of the laws of property, contract and delict, providing at the same time insights into the development of the Roman system of actions. More specifically, it focuses upon the issues arising from the transfer of fungible goods under contract: both for the legal consequences of the transfer in terms of ownership and the liability of the recipient for goods in their charge. Since we have no other evidence for the action other than the Republican responsum, the approach is to make a comparison with analogous cases at common law and particularly the role and function of the English tort of conversion. The discussion is divided into three parts: (1) representing D.19.2.31; (2) the logical and rhetorical structure of the responsum; and (3) inquiry into the scope and application of the actio oneris aversi by comparison with the role of the tort of conversion in analogous cases at common law.

II. Representing D.19.2.31

Text:

[casus] In navem Saufeii cum complures frumentum confuderant, Saufeius uni ex his frumentum reddiderat de communi et navis perierat:

[quaestio] quaesitum est, an ceteri pro sua parte frumenti cum nauta agere possunt oneris aversi actione.

[responsum] respondit rerum locatarum duo genera esse, ut aut idem redderetur (sicuti cum vestimenta fulloni curanda locarentur) aut eiusdem generis redderetur (veluti cum argentum pusulatum fabro daretur, ut vasa fierent, aut aurum, ut anuli): ex superiore causa rem domini manere, ex posteriore in creditum iri. idem iuris esse in deposito: nam si quis pecuniam numeratam ita deposuisset, ut neque clusam neque obsignatam traderet, sed adnumeraret, nihil alius eum debere apud quem deposita esset, nisi tantundem pecuniae solveret. secundum quae videri triticum factum Saufeii et recte datum. quod si separatim tabulis aut heronibus aut in alia cupa clusum uniuscuiusque triticum fuisset, ita ut internosci posset quid cuiusque esset, non potuisse nos permutationem facere, sed tum posse eum cuius fuisset triticum quod nauta solvisset vindicare. et ideo se improbare actiones oneris aversi: quia sive eius generis essent merces, quae nautae traderentur, ut continuo eius fierent et mercator in creditum iret, non videretur onus esse aversum, quippe quod nautae fuisset: sive eadem res, quae tradita esset, reddi deberet, furti esse actionem locatori et ideo supervacuum esse iudicium oneris aversi.

[A later addition?] sed si ita datum esset, ut in simili re solvi possit, conductorem culpam dumtaxat debere (nam in re, quae utriusque causa contraheretur, culpam deberi) neque omnimodo culpam esse, quod uni reddidisset ex frumento, quoniam alicui primum reddere eum necesse fuisset, tametsi meliorem eius condicionem faceret quam ceterorum.

Translation:

[casus] Several people shot their grain together into Saufeius’s ship, after which the latter returned his share of the grain to one of them out of the common pile and the vessel was lost.

[quaestio] The question was asked whether the others could proceed against the nauta with respect to their share of the grain by raising an action for onus aversum?

[responsum] He responded that there were two kinds of things placed out [in virtue of a contract of letting and hiring], either on terms that the very same thing is given back (such as when clothes are placed out to a fuller for cleaning) or property of the same kind (as when refined silver is given to a smith to make vases or gold to make rings): in the former case the thing remains the property of the owner, whereas in the latter he becomes in creditum. The same principle exists in relation to depositum: for if someone made a deposit of a certain amount of money and neither enclosed it nor handed it over under seal, but rather by counting it out, the person with whom the deposit was made was bound to do nothing more than to deliver back an equivalent sum. Accordingly, it would appear that the grain was made Saufeius’s and had been handed over in an appropriate way. Now if each person’s grain had been separately enclosed by means of partitions or wicker baskets or some other kind of container, so that the consignment of each could be told apart, we are not able to make a substitution, but rather the person to whom the grain belongs can bring a vindicatio to recover what the nauta had delivered. And so he rejected actions for onus aversum, because if, on the one hand, the goods were of such a kind that, on being handed over to the nauta, they immediately became his and the merchant in creditum, it did not appear to be a case of onus aversum, inasmuch as they belonged to the nauta; but if, on the other hand, he was obliged to give back the same thing that was handed over, the actio furti would lie for the locator, so that an iudicium for onus aversum was unnecessary.

[A later addition?] If then the goods were handed over in such a way that they could be delivered back in kind, the conductor is liable only to the extent of his fault (this much being owed in matters contracted for the benefit of both parties); and it is hardly blameworthy that he [i.e. the nauta] restored the grain to one of them out of the common pile, seeing that it was necessary for him to give to one or other person first, even though he made the position of the one better than that of the others by doing so.

D.19.2.31 consists of two parts: first, a juristic responsum given in reply to a hypothetical casus and quaestio; and second, an additional comment (beginning sed si). Although the reply has been drawn from an anthology of responsa attributed to the late-Republican jurist P. Alfenus Varus, it is possible that the author was originally his teacher, Ser. Sulpicius Rufus. Notwithstanding that the text has frequently been suspected of interpolation, the authenticity of the reply as a product of late-Republican jurisprudence is indicated by (among other things) its rhetorical style. ¹ As for the additional comment, it has been suggested that the use of sed si to signal a change of direction is typical of Alfenus and therefore that it should be treated as continuous with the preceding text. ² Considering however the chiastic structure of the main body of the responsum (for which see Section III below), it would appear that this section was added by a later hand. The compromise of these two considerations might suggest that the reply was originally given by Servius and that Alfenus added the latter section. ³ However it cannot be excluded that the main body of the responsum was composed by Servius/Alfenus and the final comment added by Paul, or even the postclassical epitomators or compilers. ⁴ Finally, it is possible that the phrase “nam in re … culpam deberi” in the section beginning sed si is a gloss.

III. The Logical and Rhetorical Structure of the Responsum

The reply has been composed in the intermediate rhetorical style. According to Cicero, the intermediate style was situated (as the name suggests) between the plain style, aimed at instruction and the grand style, suited to speeches on elevated topics. ⁵ The intermediate style combined elements of both, intending not only to instruct (docere) but also to delight (delectare). ⁶ From a legal perspective, the most important feature of the reply’s composition is the elaboration of the jurist’s reasoning within a chiastic structure. Chiasmus is a rhetorical device which consists of “a form of inverted parallelism … that presents subjects in the order A, B, C and then discusses them C, B, A … The seed of chiasmus is to be found wherever framing devices, cyclic form, or symmetry are used”. ⁷ In D.19.2.31, the chiastic device is used to frame a series of syllogisms, which are made to turn on a single axial statement with the conclusions following in reverse order (i.e. A–B–C–B–A).

While the reply’s rhetorical features have been treated elsewhere, more remains to be said about the respondent’s argumentative technique. Formally, the response employs Aristotelian and Platonic methods of classification as a basis for constructing mixed hypothetical syllogisms (a feature of Stoic logic) that are combined within a chiastic structure. The reply may therefore be represented as follows:

He responded

(A) that there were two kinds of things placed out [in virtue of a contract of letting and hiring],

(A1) either on terms that the very same thing is given back (such as when clothes are placed out to a fuller for cleaning)

(A2) or property of the same kind (as when refined silver is given to a smith to make vases or gold to make rings):

in the former case the thing remains the property of the owner, whereas in the latter he becomes in creditum.

(B) The same principle exists in relation to depositum:

(B1, B2) for if someone made a deposit of a certain amount of money and neither enclosed it nor handed it over under seal, but rather by counting it out, the person with whom the deposit was made was bound to do nothing more than to deliver back an equivalent sum.

(C) Accordingly, it would appear that the grain was made Saufeius’s and had been handed over in an appropriate way.

(A2’:B2’) Now if each person’s grain had been separately enclosed by means of partitions or wicker baskets or some other kind of container, so that the consignment of each could be told apart, we are not able to make a substitution, but rather the person to whom the grain belongs can bring a vindicatio to recover what the nauta had delivered.

(A2’:B1’) And so he rejected actions for onus aversum, because if, on the one hand, the goods were of such a kind that, on being handed over to the nauta, they immediately became his and the merchant in creditum, it did not appear to be a case of onus aversum, inasmuch as they belonged to the nauta;

(A1’) but if, on the other hand, he was obliged to give back the same thing that was handed over, the actio furti would lie for the locator, so that an iudicium for onus aversum was unnecessary.

A. Classification

The author’s starting point is to find a categorical definition (horos, definitio) of the grain. ⁸ According to Plato (and Aristotle), to find the definition of a thing is to discover its essence. Only a species (eidos) can have an essence; and these are defined by giving the kind to which the species belongs (genus) and the differentiating feature (differentia) that characterises it within its genus. The method for locating a species is by division (divisio), which requires that the inquirer first locate the most general category into which the grain falls; then to divide that category into two parts and decide which of those the grain falls into. This method is then repeated until the grain has been properly defined.

The responsum is composed against the background of contracts of affreightment concluded between Saufeius and each of the merchants, which were interpreted by the Roman jurists as a form of letting and hiring (locatio conductio). Accordingly, the respondent begins by distinguishing between two categories of res locatae: first, things handed over on terms that the identical object should be returned (e.g. a garment delivered to a fuller for cleaning); and second, goods handed over on terms that the recipient was only bound to restore property of the same kind (e.g. fungible goods, res quae pondere numero mensura constant, namely things which are weighed, counted or measured). In (B) a further distinction is made between deposited things according to the nature of the datio, namely between coins that have been counted out or handed over enclosed or under seal. The implication is that, by analogy with deposited things, the nature of the datio is similarly relevant in the case of res locatae handed over on terms that the recipient was obliged only to give back property of the same kind (A2). In other words, when fungible goods are handed over under a contract of letting and hiring, the obligation to give back either the very same property or the equivalent in quantity of like quality is determined by whether it has been handed over as a discrete unit or counted out, weighed or measured. The result is to produce an exhaustive classification of res locatae falling into three categories: (A1) an object to be returned in specie, in which the transferor retains ownership; (A2:B2) fungible property handed over as a discrete unit, in which the transferor similarly retains ownership; and (A2:B1) fungible property either counted out, weighed or measured, in which the transferee acquires ownership so that the transferor becomes in creditum. ⁹

B. Mixed Hypothetical Syllogisms in a Chiastic Structure

The argument of the responsum is constructed using Stoic forms of argumentation, centred on the use of propositional logic. This is significant because at the time of the reply’s composition, the formal validity of hypothetical syllogisms was hotly disputed by competing philosophical schools. ¹⁰

In Stoic logic, arguments (logoi) are a compound or system of premisses (lēmmata) and a conclusion. ¹¹ Stoic arguments were labelled “hypothetical” because their component parts contained at least one hypothetical proposition. One valid form of argument according to the Stoic system was the so-called modus ponens, in which the leading premiss consisted of a conditional (sunemmenon), defined as an assertable formed by means of the connective “if” (ei). ¹² In its standard form, the conditional can be expressed as “If p, then q”, where p is the antecedent and q the consequent. The complete form, including the co-assumption and conclusion, is normally written as:

If p, then q;

now p,

therefore q.

For the Stoics, the conditional announced a relation of consequence, namely that the consequent followed from the antecedent. Fundamentally in a modus ponens, the argument works by the co-assumption (or minor premiss) affirming the antecedent in the conditional leading premiss, from which the conclusion follows. It is syllogistic because the reasoning used to reach the conclusion is deductive.

In D.19.2.31, the syllogisms which form the argument are each examples of mixed hypothetical structures in the form of a modus ponens. The leading premisses each consist of a conditional, containing antecedent terms from which it follows that ownership in the grain either will or will not transfer with the related consequence for the person making the datio that they either will or will not become in creditum (A, B). The co-assumption (C, the minor premiss) serves to register agreement with the antecedents of the conditional. The conclusions then follow. For res locatae (A) the conditional is therefore that: if (p) fungible property is handed over on terms that the recipient was only bound to restore the same quantity of material of like quality; and (q) the property was counted out, weighed or measured; then (s) the conductor acquires ownership and the locator becomes in creditum. Correspondingly if either of the antecedents (p) or (q) is denied, the locator retains ownership in the goods. In the context of the responsum, which has been composed against the background of the standard-form contracts of affreightment that were used in commercial practice, the analogy drawn with deposit distinguishes between three categories of res locatae depending both upon the kind of property at issue and the terms on which it was handed over, those terms to be inferred from the nature of the datio.

The co-assumption or minor premiss (C) affirms for each leading premiss the antecedents that: (p) the property in question is grain, which is fungible; and (q) the datio was performed in an appropriate way (i.e. by metering out the grain). The chiastic structure of the reply enables this single axial statement to affirm the antecedents of the conditionals.

The conclusions concerning the availability of an action for onus aversum then follow in reverse order as required by the chiasmus. First, at A2’:B2’, the focus is on grain that has been stowed in separate compartments, where the antecedent (p) is affirmed but (q) is denied. It follows that the merchants retain ownership in the grain and can bring rei vindicatio. Second, in A2’:B1’, the grain has been metered out, so that Saufeius acquires ownership in the goods and the merchants becoming in creditum. Here Saufeius’s delivery of a portion of the grain to only one of the merchants is not aversio. Neither is there a remedy for breach of duties arising ex fide bona, because Saufeius is not at fault (culpa) for his disbursement of only a portion of the grain before the ship went down. Third, the author concludes by dealing with the case A1’, in respect of res locatae in which the antecedent (p) is denied. Here, it follows that there has been no transfer of ownership. However, unlike in A2’;B2’, where an action for onus aversum was disapproved because the transferor already had rei vindicatio, in this case we are told that the locator also has actio furti. The final conclusion is that, since the merchants have a remedy in every event (i.e. in which each conditional is either satisfied or fails, taking account of the terms of the contract, the type of property involved and the nature of the datio), for the praetor to grant a trial for onus aversum is unnecessary.

IV. Actio Oneris Aversi

The objective of this section is to establish, so far as possible, the scope and application of the actio oneris aversi and the reason for Servius/Alfenus’s disapproval. There are a wide range of theories concerning the nature and content of the action. On one view the actio either never existed ¹³ or was another action by a different name. ¹⁴ Otto Lenel assumed that, if the action ever existed, it was probably penal, but that in any case the absence of any juristic commentary in the surviving materials indicates that it had been removed or fallen into desuetude by the time of the consolidation of the Edict under Hadrian. ¹⁵ By contrast, others have argued that the action was reipersecutory, such as: that it lay against the carrier for breach of a promise, secured by a penal stipulation, to keep the cargo safe (Adam Wiliński); ¹⁶ that it was given on the analogy of the actio locati (Paul Huvelin); ¹⁷ or that it was (J.A.C. Thomas):

a reipersecutory action available to an owner for misappropriation (in the broadest sense) by magister navis where cargo was taken into a common hold, so making the consignors owners in common: in short that it lay where several merchants were owners in common of a cargo of which the nauta made a wrong apportionment in the discharge of the cargo. ¹⁸

The argument of this section is that the actio did exist as a discrete remedy, ostensibly for cargo theft, with features similar to those of the common law tort of conversion. Ultimately though it was made obsolete by the application and expansion of existing remedies, most especially the actio furti. The approach is: first, to examine the semantic field of onus aversum, in both its legal and trading context; second, to consider the problem from the perspective of the common law; and third, to reconsider the Roman law by comparison with the treatment of analogous cases in the common law.

A. The Meaning of “Onus Aversum”

“Onus” (also sometimes “honus”) especially denotes a cargo or freight, particularly but not exclusively in a shipping context. ¹⁹ Besides D.19.2.31, it appears together with the verb avertere in two other texts. ²⁰ First, in a tablet preserved in the archive of the Sulpicii (first century A.D.), the honus of a vessel with the name Notus, of 18,000 modii burden, appears alongside the verb avertisset. ²¹ Later in the tablet, there are indications that the ship together with its cargo had been made the subject of an auction sale (sub praecone), perhaps in conflict with the status of the local authorities as a privileged creditor for the collection of customs duties (protopraxia). ²² Second, in a passage in the fifth book of Tertullian’s Adversus Marcionem, composed about A.D. 208, Tertullian challenges Marcion, “that shipmaster out of Pontus”, to advance his reasons for accepting Paul as an apostle of Christ, with the same fidelity as he presumably exercised his responsibilities as nauclerus, “supposing that he had never received onto his ship smuggled or illicit goods, nor misappropriated nor adulterated any cargoes” (onus avertisti vel adulterasti). ²³ We know from Galen that the adulteration of goods was a common fraud perpetrated in the olive oil trade, in which unscrupulous merchants cut their stock with animal lard, degrading the quality to increase the saleable bulk (the same problem exists today). As for onus aversum, though its juxtaposition to cargo adulteration implies that it too was a kind of fraud, no further insight into either the common or legal meaning of the expression is forthcoming. ²⁴

The first appearance of the verb avertere in a specific legal context is in the lex repetendarum preserved in the Urbino fragments, possibly dating to the tribunates of C. Gracchus (123/ 122 B.C.). ²⁵ The lex introduced penalties for provincial officials “pecuniam auferre capere cogere conciliare avertere” (usually reduced to “pecunia capta”) and provided for double damages estimated at the value of everything “seized, extorted, taken, averted or procured” following the law’s enactment. ²⁶ Despite attempts to establish more precise meanings for the verbs used in the lex, principally by examining Cicero’s interpretations in the Verrine orations, ²⁷ most scholars continue to adopt a sceptical position, regarding them as generally targeted at the extortion of property by government officials and their illegal enrichment at provincials’ expense.

The term also appears several times in the juristic texts. In D.14.4.7.3, the actio tributoria lies against a Master who with dolus malus lessens the amount available for distribution by turning goods (merces) over to another person. Similarly in D.15.1.21 pr., it is dolus for a Master who, anticipating proceedings against him de peculio, hands property belonging to the fund over to a third party. Here Ulpian distinguishes the payment of a debt from the fund, the satisfaction of an existing creditor not amounting to dolus. In D.47.4.1.13, the same jurist states that the praetorian action against a slave who, having been directed in his Master’s Will to become free, commits theft or damage to property in respect of items belonging to the estate, after the death of his Master but before the inheritance had been accepted, lies in respect of everything in which the heir had an interest in its not being removed (non esse aversa, e.g. a thing held by the deceased in pledge). Finally, in D.47.20.3.1, it is stellionatus (a crime) if a person substitutes goods which he has sold or averted (averterit) elsewhere. What all these usages have in common is that in each case it is actionable (albeit in different contexts) for a person to dispose of property dolo malo, to the detriment of a person who has a legitimate interest in the thing. The only tentative conclusion that may be drawn from this brief survey is that the actio oneris aversi likely lay against carriers for the disposal of a cargo (e.g. to a third party), possibly dolo malo, in which another person or people had an interest.

B. The Problem at Common Law

Nothing in the survey given above indicates anything about the nature of the actio, the grounds for liability, whether it was penal or reipersecutory, the extent and measurement of damages and so forth. We have no means of direct access to this information through contemporary sources. Nonetheless, an inspection of the treatment of roughly analogous cases in the common law provides an insight into the possible function of the actio oneris aversi and the reasons for its disapproval. I propose to examine the problem through the lens of the so-called wheat cases, which address the problem of determining the ownership of grain delivered under contract and shot by multiple contributors into a common pile. At common law, the implication of the conclusion on this point is that if the person taking receipt of the grain was held to have acquired beneficial ownership, no bailment would have arisen imposing duties on the recipient as bailee. In the second edition of Palmer on Bailment, the wheat cases are introduced as follows:

The essence of bailment is that the bailed property should be returned to the bailor or applied in accordance with his instructions when the bailment terminates. The goods need not be in their precise original form when this event occurs in order for the transaction to qualify as a bailment; if this rule were imposed, it would remove main bailments (such as those for repair or alteration) from the sphere of that relation altogether. What is necessary is that the goods themselves, whether in altered or original form, should be returnable and not merely some other goods of equivalent character or value. There must be a clear physical heredity between what has been delivered to the bailee and what must be returned.

This rule has given rise to difficulty in the area of milling or storage contracts where consumable goods are delivered by their owners for a process of treatment which necessarily involves the intermingling of those goods with similar merchandise belonging to other parties. ²⁸

1. The wheat cases

Rather than give a schematic overview of the current law, I propose to focus on two of the leading cases which together expound the basic common law principles. This approach also has the benefit of enabling scrutiny both of the particular facts and the judicial reasoning, which in the first of our cases is indirectly informed by the response contained in D.19.2.31.

South Australian Insurance Co. v Randell. ²⁹ The Randell brothers were two millers. To insure against the risk of fire they took out a policy with the South Australia Insurance Company covering their stock-in-trade (e.g. grain, flour, sacks). It was a term of the policy that “goods held in trust or on commission” were required to be insured as such. Six months after the commencement of the policy, a fire destroyed the Randells’s mill together with the stock contained within it. Subsequently the brothers attempted to enforce a claim under the policy in respect of an amount of wheat and flour which had been destroyed in the fire. The insurer’s defence was that since the millers had received the grain from farmers who had contracted to store it in the mill, they held the wheat as mere bailees such that it fell outside the scope of the policy.

The case was heard on appeal from the Supreme Court of the province of South Australia by the Judicial Committee of the Privy Council. From the evidence of the millers and their foreman, it was inferred that the Randells held the grain under contract on the following terms. According to the established trade practice, the farmers were accustomed to shoot their wheat together into large hutches in the mill. This was done under the supervision of one of the millers, who issued the farmer with a receipt in terms that the grain had been “Received, &c., to store”. Upon discharge into the hutch, the wheat became part of the millers’ current stock, available to them to use at their sole discretion, principally either to sell on the market or to grind into flour. From time to time, different farmers availed themselves of the facility offered by the millers, such that the stock of grain was constantly fluctuating. After a certain period of time had elapsed, the millers levied a storage charge of one farthing per bushel per month. The farmers meanwhile were entitled to demand of the millers at any time the payment of a sum of money equivalent to the market value of a like quantity and quality of grain to that which they had originally discharged, fixed at the prevailing market price on the day of the demand. The millers then had the option either of paying the market price or delivering up to the farmer an equivalent quantity of grain of like quality from their store. ³⁰

It was accepted by the Privy Council that the substantive question was whether the millers were beneficial owners of the wheat insured or merely possessed it as bailees. For the basic law the Privy Council cited Sir William Jones’s An Essay on the Law of Bailments, for the proposition that (in the words of the court): “Whe’ever there is a delivery of property on a contract for an equivalent in money or some other valuable commodity, and not for the return of his identical subject matter in its original or an altered form, this is a transfer of property for value – it is a sale and not a bailment.” ³¹ The court further observed that the same principle was approved in James Kent’s Commentaries on American Law, in which it was described as the “true and settled doctrine”. ³²

Applying this principle to the problem of the intermixture of fungible goods, it is possible to distinguish two discrete cases. The first is that of storage simpliciter, in which the grain, though it had been shot into a common heap, was kept intact, so that the farmer was entitled to demand the return of either the identical wheat or an aliquot share in the specific bulk in which his grain was mixed with his consent. If the grain was commingled with that of others so that each person’s contribution was indistinguishable from the next, the pile would be owned in common by the farmers, each according to the share of their contribution to the commixtion. The result is that the farmers would retain property in the grain or in common in the heap in respect of their share, the transfer of possession amounting to a bailment with the millers possessing the grain as bailees.

The second case concerns wheat shot by farmers into a common pile but with a right only to demand payment of the value at the prevailing market price or the return of an equivalent quantity of grain of like quality. The court distinguished this from storage simpliciter on two grounds: first, that the stock of grain was constantly fluctuating, so that no one person’s grain or any specific bulk of grain, could be identified in which it was possible to say that the farmer retained a proprietary interest; and second, that the millers’ obligation was not to provide storage and therefore to restore the identical grain or a share in a specific bulk, but rather to elect on the farmer’s demand to pay either the market price or to restore an equivalent quantity of like quality. This break in the physical heredity between what was delivered and what was required to be restored made this type of case analogous to that of a deposit banker, who receives money on the understanding that it will go towards his current capital to be used for his own purposes. ³³ In the Privy Council’s judgment, since there was no sound basis upon which to distinguish the millers’ use of the grain as trading capital (i.e. as a “wheatbank”) from that of money received by a deposit banker, the proper characterisation of the relationship was as between lender and borrower in a loan for consumption or mutuum. ³⁴

Consequently, in answer to the question of who would bear the loss of the grain occasioned by fire (putting the insurance to one side), the answer was that this would be shouldered by the millers as beneficial owners on the principle res suo perit domino; and that notwithstanding the destruction of the stock, they were still obliged to pay the market price to the farmers on demand. The Privy Council’s conclusion was that the transactions between the millers and the farmers ought therefore to be interpreted consistently with Jones as a sale and not a bailment. ³⁵ Each was therefore a sale by the farmers to the millers of the grain with the right to the return of the equivalent in money’s worth of a quantity of like quality on demand; and since the millers were beneficial owners of the grain, its value fell to be recovered under the policy of insurance.

Mercer v Craven Grain Storage Ltd. ³⁶ In Mercer, the House of Lords considered in more detail the relationship of the parties in respect of grain handed over for storage simpliciter. Several growers had discharged grain into a common store, the quantity of which fluctuated with the addition and withdrawal of amounts of grain belonging to other growers from time to time. The grain was delivered under a storage contract to Craven Grain Storage Ltd., whose primary function was to provide storage and drying facilities and then to release the grain to a second company (Craven Grain Ltd.) for sale. It was a term of the contract that the growers retained title to the grain and they further instructed that their consignment was not to be sold by the selling agent for less than £160 per tonne. In the event, Craven Grain Ltd. withdrew and sold all but a small quantity of grain for less than the agreed amount and then became insolvent. Subsequently the growers demanded the return of their grain, of which the storage company was only able to restore a fraction of the original quantity. They therefore raised an action for damages representing the value of the amount that had been lost on grounds that the defendant’s sale of the grain at less than the agreed price and failure to restore was conversion.

Reversing the decision of the Court of Appeal, ³⁷ the House of Lords accepted the argument advanced by counsel for the claimants, that the growers were owners in common of the stored grain in proportion to the quantity that they had delivered, which varied with the extent of the bulk. ³⁸ Consequently, the claimants were entitled to the return of their share on demand: the grain was received for storage simpliciter, the pile becoming a consensual mixture owned in common by the various contributors in proportion to their share held by the storage provider as bailee. Consequently, the disposal of the grain together with their inability to meet the demand made the defendants liable in conversion for which they were required to pay damages.

From the discussion of these cases it is possible to reconstruct the common law doctrine on the following lines. When property is delivered under contract, this is either on terms that the recipient returns the very same thing in its original or altered form; or the equivalent in money’s worth or quantity of the same goods of like quality. In the first case, the delivery gives rise to a bailment, the recipient holding the property as bailee; in the second, the recipient acquires property in the goods, the transaction being a sale or loan (mutuum) subject to the right of the person who made the delivery to demand the equivalent quantity of like quality. In the case of fungible goods, the delivery will give rise to a bailment if the goods are kept separately or, though they are intermixed with others’ goods such that no one person’s can be told apart, the whole quantity comprises a specific bulk in which each contributor retains a proportionate share as owner in common. ³⁹ If, however, the goods are discharged into a common store, on terms that the person making delivery has only the right to demand the equivalent in money’s worth or an equal quantity of like quality, the transaction will either be sale or mutuum and property in the goods will transfer. In this latter case, which covers by way of example deposit banking as well as the operation of a wheatbank, the transferor stands in relation as a creditor, the recipient acquiring title to use the property as he sees fit, but also assuming the whole risk on the principle res suo perit domino. Finally, in the event that the delivery gives rise to a bailment, disposal of the property by the bailee without the bailor’s consent will be conversion, for which the bailee will be liable personally in damages for the value of the goods converted. ⁴⁰

2. The role of conversion in the wheat cases

The role of conversion in the wheat cases provides a useful insight into the likely scope and application of the Roman action for onus aversum. Conversion is a so-called property tort, concerned with the protection of property rights in chattels. ⁴¹ The gist of the action is denial of title, the claimant needing only to have a relatively superior claim to possess the property than the defendant. ⁴² Although the scope of the tort has been reformed by the Torts (Interference with Goods) Act 1977, it still retains its essential characteristics. A good definition of the tort before the intervention of the 1977 Act was given by Atkin J. in Lancashire and Yorkshire Railway Co. v MacNicoll:

It appears to me plain that dealing with goods in a manner inconsistent with the right of the true owner amounts to a conversion, provided [that] it is … established that there is an intention on the part of the defendant in so doing to deny the owner’s right or to assert a right which is inconsistent with the owner’s right. ⁴³

At the heart of the tort of conversion is the performance by one person of an act asserting control over a chattel inconsistent with the rights of a person with superior possessory title. Liability is strict. ⁴⁴ There is no requirement on the claimant to show that the defendant’s conduct was culpable; only that they intended to perform the offending act and that it was done without the claimant’s consent. Examples of acts of conversion include the delivery of a chattel under a contract of sale; ⁴⁵ and the delivery by a carrier or other bailee to the wrong person, even though the misdelivery was unintentional. ⁴⁶ Goods can also be converted in a maritime context. ⁴⁷ In Motis Exports Ltd. v Dampskibsselskabet AF 1912, A/S and another, ⁴⁸ a carrier made delivery to fraudsters on presentation of forged bills of lading. It is a fundamental obligation in connection with a bill of lading that the carrier must deliver the goods only to a party presenting an original bill. Notwithstanding that the defendant did not know that the bills of lading were forged, they were liable for conversion, it being sufficient that the delivery was intentional and that it had interfered with the claimant’s rights. As for damages, it was a rigid rule (before the 1977 Act) that these were assessed with reference to the value of the chattel at the time of the conversion. Since the claim is in the law of obligations the defendant in any given case need not themselves be in possession. ⁴⁹

These features make the tort liable to generate particular hardship for defendants who have innocently delivered a thing to a third party and are therefore liable to pay damages notwithstanding that they are no longer in possession of the property. ⁵⁰ According to Cleasby B., the imposition of strict, personal liability was “a salutary rule for the protection of property, namely, that persons deal with the property in chattels or exercise acts of ownership over them at their peril”. ⁵¹ The justification for these features becomes especially apparent when one considers the function of the tort in the common law system, which unlike Roman law lacks a remedy equivalent to the rei vindicatio, especially for the vindication of a proprietary right in a chattel. ⁵² In this respect, conversion has been regarded as performing a similar function in English law to the civilian vindicatio, at least in as far as both remedies enable a person asserting (superior) title to recover damages equivalent to the value of a thing of which they are not in possession.

C. Roman Law

The common law approach to resolving the so-called wheat cases and the liability of bailees in conversion can be used as a basis upon which to reconsider the likely characteristics of the actio oneris aversi and its place in the Roman system of actions. In D.19.2.31, the hypothetical case is that several people had delivered grain under contract to Saufeius which they shot into a common pile in the ship’s hold. The question was, if after having returned the grain to one of them the ship went down, could Saufeius be held liable for onus aversum by those who had not received their share?

The grain has been delivered to Saufeius against the background of a contract of affreightment. For res locatae, the first distinction is the same as that elaborated by Jones: either the res is delivered on terms that the very same thing is returned (A1) or property of the same kind (A2). For money deposits, there is a further distinction in the nature of the datio, between money counted out (B1) or handed over in a sealed packet (B2). In the former, ownership transfers and the depositor becomes in creditum; in the latter, the depositor retains ownership in the money and the banker is obliged to return the very same coins. By analogy with money deposits, the terms on which property is handed over in a contract of letting and hiring can similarly be inferred from the nature of the datio: if it has been counted out, weighed or measured then the recipient acquires ownership and the transferor becomes in creditum. In all other cases the transferor retains ownership. In Randell, the consequence of transferring ownership and becoming in creditum was that the recipient acquired title to the property and could use it at his sole discretion, though at his own risk, while the transferor acquired a personal right to the return of the equivalent or money’s worth on demand. The farmers (just as a person making a money deposit with a banker) therefore stood in relation to the recipient as an unsecured creditor. This is especially important when either the property is destroyed by vis maior or the recipient becomes insolvent: in both cases the transferor has only an unsecured personal claim, which in the case of insolvency may only be worth a fraction of the goods’ actual value.

We are now in a position to assess the respondent’s conclusions as regards the Roman law. A2’:B2’ presents facts broadly analogous to those in Mercer (though without the element of substitution): since each person’s grain has been handed over to Saufeius as a separate parcel the case is equivalent to storage simpliciter. The analogy is drawn with the example of coins handed over to a banker in a sealed packet, ownership remaining with the depositor. In Mercer, the result of the growers’ retention of ownership in the grain was that the bailee’s disposal of it to a third party interfered with his rights and was therefore conversion. By contrast, the response in D.19.2.31 is that the merchant has rei vindicatio (an actio in rem) to sue for its value. ⁵³ The implication is that the misdelivery of grain that has been stowed in a separate compartment amounts to onus aversum, but that the vindicatio affords sufficient protection. To give both actions would therefore provide the merchant with the opportunity for double recovery: first in rem using rei vindicatio, either achieving the return of the grain or its value; and second personally against the carrier for onus aversum. The jurist therefore disapproves (improbare) an action for onus aversum.

A2’:B1’ presents facts analogous to those in Randell. Each person’s grain has been shot under contract into a common pile. We are told that it is not aversio for the carrier to dispose of a portion of the common pile to one creditor ahead of others, the implication being that a potential claimant having only a personal right cannot claim that property out of which his debt could have been satisfied has been averted when it was used to satisfy another creditor’s legitimate demand. ⁵⁴ The observation is also made that the carrier does not show fault by delivering to one of the merchants ahead of the others, excluding liability under the actio locati. Obviously, they have no vindicatio since they do not retain a right in rem. The question however arises whether the merchants, being in creditum, can recover personally against Saufeius for the money’s worth of their contribution notwithstanding the destruction of the remainder through vis maior?

At first blush it would appear that carriers receiving grain for transport were treated analogously to deposit bankers or wheatbank operators at common law: in both of those cases, the recipient acquires ownership of the grain subject to a generic debt to return the equivalent in money’s worth or quantity of the same goods of like quality. This is the standard Pandectist interpretation, which designs the transaction as locatio conductio irregularis, to be distinguished from regular instances of letting and hiring by the conductor’s acquisition of ownership. ⁵⁵ Although some have sought to distinguish Saufeius’s case by arguing that he acquired peregrine rather than civil law ownership, the more widely held view is that he became owner of the grain at civil law. ⁵⁶ This being accepted, for a long time it was held that ordinary civil law principles would apply, namely that the carrier assumed the whole risk for the goods on the principle res suo perit domino subject to a generic debt owed to each of the merchants. ⁵⁷ Indeed this was the decision in Randell, in which the millers assumed the full risk for the grain in the common store as owners and were personally liable to the farmers notwithstanding its destruction by fire. Ultimately however the millers were able to claim against their insurers, because as owners of the grain it was not excluded from the policy as property held on trust. The problem is therefore to determine to what extent the analogy between deposit banking and wheatbanks can be extended to grain transport, particularly in respect of the distribution of risk.

As Èva Jakab has observed, there are two related issues: (1) whether it was usual for carriers to assume the whole risk for the goods they transported; and (2) whether the debt owed by a carrier to a merchant was generic or restricted to the return of either the same property or grain from a specific bulk. ⁵⁸ To advance our understanding in this area Jakab made a study of the only available evidence for contracts of affreightment from the period, which are preserved in the Graeco-Roman papyri. ⁵⁹ Assuming that the standard clauses in those contracts are broadly representative of Mediterranean practice, she identified three that were common to private grain transport: (1) a clause in which the carrier undertook to keep the cargo “dry, pure, undamaged, unadulterated, and complete”; (2) a clause excluding the carrier’s liability in certain cases of force majeure (e.g. storm, piracy); and (3) a “navigation clause”, prohibiting the carrier from, for example, sailing at night, in bad weather or stopping at certain ports. The papyri also provide insights into commercial practice. In private transactions the grain was typically measured out to the carrier, who issued the merchant with a receipt. On completion of the voyage, the grain was then measured back to the merchant, presumably on presentation of the receipt. In addition to details about the quantity of grain etc., receipts could be accompanied by a penalty clause which imposed liability on the carrier for any shortfall, to be compensated at either the market price or for a predetermined sum per unit. It seems therefore that it was standard practice in private transport contracts for the parties to vary the default incidence of risk by excluding carriers’ liability for the loss of goods due to vis maior. In addition, while Roman carriers did acquire ownership of the grain at civil law, they were not free to dispose of it subject to a generic debt owed to the merchant, but rather strictly liable to return the quantity acknowledged in the receipt from the specific bulk with an obligation to pay for any shortfall. ⁶⁰

Jakab’s findings also solve an economic problem created by the straightforward application of the civil law principles. In Randell the millers were able to shift the risk of loss due to vis maior to insurers. For Roman carriers, however, there is no evidence of any mechanism that would perform the same function as modern marine insurance. By contrast, it was a standard clause of maritime loan contracts (nautikai syngraphai) that creditors who lent to merchants to finance the acquisition of cargoes assumed the risk of their destruction by force majeure, in return for which they demanded an increased yield. ⁶¹ In the hypothetical case in which the merchants had acquired their grain with money borrowed on maritime terms, the clause excluding liability for loss due to vis maior in the contract of affreightment would shift the risk from the carrier to the merchant; and the maritime loan contract would shift the risk again from the merchant to the financier. Ultimately it was the financiers who assumed the marine risk (periculum maris), typically by lending to merchants on maritime terms.

Returning to the final conclusion A1’, we are told that an object (e.g. an item of clothing given to a fuller) has been delivered under a contract of letting and hiring, on terms that the very same thing should be returned. Ownership is retained by the locator, the conductor becoming obliged to furnish custodia (in a way analogous to a bailee). ⁶² In the event of misdelivery, the original locator must have rei vindicatio. If the recipient’s conduct was wilful, he also has actio furti, which it is implied is sufficiently wide to meet the case. ⁶³ This action was penal and at least in duplum, measured by the value of the thing.

The ultimate conclusion is that to grant a trial for onus aversum is unnecessary. To what extent can the common law action for conversion be treated as a model for understanding the scope and application of the actio oneris aversi? The common law action lies for interference with the rights of a person with superior title and is strict, personal and for simple damages measured at the value of the thing at the time the conversion took place. A Roman law action that shared these features would fit neatly into the conclusions recorded in D.19.2.31. The act of onus aversum would then align with the original act of conversion in English law, namely unlawful disposal of another’s property. It is further instructive that in A2’:B2’ there is onus aversum, but it is stated that when a cargo has been stored in a separate compartment the action ought to be disapproved because the vindicatio is sufficient. The implication is that the intention behind the actio oneris aversi was to protect the cargo-owner’s proprietary interest against wrongful interference. A further consideration is that in typical cases the defendants in each proceeding would be different individuals: the recipient of the goods in vindication (so far as they wished to defend the action) and the carrier for onus aversum. This alone makes a convincing case for the disapproval of the action in cases analogous to storage simpliciter because the availability of both actions, each involving different defendants, provided the opportunity for double recovery. The justification for the disapproval of the actio oneris aversi in Roman law would then mirror that given for the survival of conversion in English law: in Roman law, the availability of the proprietary action makes conversion unnecessary; in English law, the absence of a proprietary action justifies its retention.

However, the overlap between the actions should not be pressed too far. First, unlike conversion, the Roman action was not generally available in respect of movables but lay only in respect of cargoes (though it is not obvious that this was confined to shipments stowed in a vessel and may have extended to cargoes stored in, e.g., warehouses). This restriction also curtails the possibility, by contrast with the common law, that multiple conversions could take place in respect of the same goods by different defendants. Second, the availability of the action for conversion to a person with superior possessory title does not straightforwardly carry over to the Roman context. Although the actio oneris aversi did lie to the owner (which would explain, for example, its disapproval next to the actio furti), it may also have been available to non-owners. This would be the case if, as in A2’:B1’, a merchant with a personal interest in the distribution of the common pile might claim aversio if the carrier disposed of the grain in fraud of creditors (in contrast to the loss of the goods through force majeure). Finally, whereas the English action is conceived as a strict liability property tort, it is not necessarily the case that onus aversum either imposed strict liability or was penal. However, the disapplication of the action next to the more expansive actio furti suggests that it too was penal and may well have required proof of dolus. If this was the case, a merchant who had not received the grain he was due could bring a reipersecutory action for the shortfall and a penal action for aversio if the reason for the shortfall was the carrier’s fraud. ⁶⁴

The scope and application of the actio oneris aversi may therefore have been as follows. The gist of the action seems to have been for conversion of a cargo. It was probably penal, requiring proof of dolus and lay either for simple damages or at most in duplum. Damages were probably measured at the value of the goods at the time of the wrongful act. As for its origins, it would not be surprising if it was an actio honoraria that was framed in factum. Although it cannot be excluded that the action was a creation of the civil law (e.g. with a source in statute or jurisprudence), a praetorian action in factum would be consistent with the other edictal actions directed at maritime trade from broadly the same period (particularly the actio de recepto and so-called actio furti adversus nautas, both of which were in factum conceptum). If this is right then the tenor of the jurist’s response is that the honorary action ought to be disapproved because the case is adequately dealt with by the civil law (all the other actions discussed in the text arising ex iure civile). In particular, the application of the actio furti to cases of theft of use, in which a person receiving property under contract handled it in a manner contrary to the will of the owner, will have expanded its scope to capture the situations covered by the more context-specific actio oneris aversi. The reply may therefore be an example of juristic rationalisation intended to show the redundancy of an action introduced through the ius honorarium and justifying its removal from the Edict by reference to the comprehensiveness of the civil law.

V. Concluding Remarks

There is an intellectual thread that runs from Servius/Alfenus, through Paul, the compilers of Justinian’s Digest, the civilian jurists, to Sir William Jones (1746–94) and the wheat cases adjudicated at common law. The origins of the responsum contained in D.19.2.31 are in the later Roman Republic, with either Ser. Sulpicius Rufus or his pupil P. Alfenus Varus. The responsum was later epitomised by the late-classical jurist Paul, the compilers choosing to include the epitomised version of the text in the Digest title on locatio conductio (D.19.2). Notwithstanding that the actio oneris aversi at the heart of the quaestio was by this time a legal “fossil”, the compilers probably recognised the responsum as a source of sound principles in respect of property handed over under contract (particularly locatio conductio but also depositum). ⁶⁵ In addition, there is the attractive possibility that the compilers held the response in high esteem for its rhetorical qualities.

Following the rediscovery of the Digest in the West in the eleventh century, the text quickly became a favourite of the Mediaeval lawyers and has remained a mainstay of the civilian tradition ever since. In the eighteenth century the text was taken over to the common law by Jones in his Essay, who refers to it as “the Celebrated law of Alfenus”. ⁶⁶ Jones cites the text alongside the observations of the Roman-Dutch jurist Cornelius van Bynkershoek (1673–1743) in support of the proposition that the distinction made in respect of loans (viz. the Roman commodatum and mutuum), between the obligation to restore the specific thing or only its equivalent, holds equally when property is transferred under a contract of letting and hiring or deposit. ⁶⁷ This supports the proposition that, at common law, the delivery of property on a contract for an equivalent in money or some other valuable commodity produces a sale and not a bailment, which was subsequently relied upon in the adjudication of cases such as Randell.

The responsum contained in D.19.2.31 has been treated both at civil and common law as the fons et origo of the principles governing the relationship between the parties in certain cases involving the delivery of goods under contract. This has led to a consistency in doctrine, at least so far as these principles help to distinguish cases in which the recipient’s obligation is either to restore the identical thing or only its equivalent. In the first case, the transferor retains ownership or title, while the recipient owes certain duties to care for the property, in Roman law as an aspect of good faith under the contract and at common law as bailee. In the second case, the recipient acquires ownership of the property, while the transferor stands in relation as a creditor. In both cases, the principle res suo perit domino applies, with the attendant consequence for which of the parties bears the risk in the case of accidental destruction of the property or the insolvency of the recipient. At Rome, however, at least in the context of contracts of affreightment, it was typical for the parties to agree by the inclusion of standard clauses that the merchant should assume the risk for the loss of the goods due to force majeure.

It is in the alternative case in which the transferor retains a proprietary interest that the difference between the Roman and common law is most pronounced. In classical Roman law, the transferor is given both vindicatio for the protection of their right in rem and actio furti for a penalty, in the event that (among other things) the recipient either deliberately uses the thing contrary to the agreement or wilfully deprives the owner of his usus. The actio oneris aversi was disapproved, on the one hand, because the owner of a cargo had vindicatio to protect his proprietary interest and, on the other, because its field had been taken over by the expansion of the actio furti. At common law, by contrast, the transfer gives rise to a bailment. In the specific case of misdelivery by the bailee, the bailor has no equivalent action to the Roman vindicatio to enforce a right in rem. Instead, the bailor’s protection is achieved by holding the bailee liable for conversion. Roman law likely recognised a broadly similar action in respect of cargoes, in the form of the actio oneris aversi. However, since its scope was circumscribed by the vindicatio and actio furti, it was disapproved and seems ultimately to have been removed from the Praetor’s Edict altogether.