Possession and the Rights of Finders

at Common Law

In Chapter 2 we located the origin of our modern theory of possession in the writings of Holmes and Pollock late in the nineteenth century. We noted that each owed a heavy debt to Roman law, at least insofar as they conceived of possession as depending on proof of two discrete factual elements, ie some degree of factual control over the object in question, and some degree of intention to exclude others from interfering with it. It cannot be emphasised strongly enough that each of these writers regarded these twin elements as facts. Their legal significance was that when demonstrably present in a given person they pointed to the existence of a ‘definite legal relation’ between that person and the thing possessed.[4] In addition, Pollock’s account disclosed a further distinction between possession acquired with the consent of some previous possessor and possession acquired without such consent.[5] Pollock was careful to suggest that consensual changes of possession were ‘not fully determined’ by the facts of control and intention, inasmuch as the consent of the transferor was ‘a real element’ in the transferee’s ‘power of enjoyment and control’.[6] The exact import of this observation is a matter of some doubt,[7] but for now it is enough to note that a finder who acquired possession was certainly in the latter category. Her acquisition might be a justified trespass if she took with intention of saving the thing for its owner;[8] it would be wrongful, and subject to the law on larceny, if she took for herself;[9] but in neither case did it occur with the express consent of the loser.[10] On Pollock’s thesis, then, a finder can establish a ‘definite legal relation’ to the object of her find insofar as (and because) she can establish the twin possessory facts of control and intention, and this of course is borne out by the cases decided in modern times, as we saw in Chapter 2.

In arguing for the existence of this ‘definite legal relation’, Pollock’s major premise is that possession makes available legal protection. His treatment of ‘title by possession’ begins with this proposition:

Existing possession, however acquired, is protected against any interference by a mere wrongdoer; and the wrongdoer cannot defend himself by showing a better title than the plaintiff’s in some third person through or under whom he does not himself claim or justify.[11]

Pollock thought that this proposition was true equally of goods and land, in the actions of trover and ejectment respectively,[12] and treated trover and ejectment coterminously is his consideration of the right resulting from possession.[13] Since the accuracy of his whole argument depends on the truth of this assimilation, we consider first the extent to which it was correct to regard possession as the basis of trover and ejectment respectively, before moving to discuss Pollock’s substantive account of the resultant right.

Possession as the Basis of Trover

To the modern lawyer, there is nothing startling in the proposition that actual possession gives title to sue in conversion.[14] Accordingly, it seems a fairly orthodox rationalisation of trover to say that it lay to protect a right to possess goods, and that it was available to a possessor on proof of possession alone. However, it is likely that our understanding of the action in these terms owes a heavier debt to Pollock and his contemporaries than we might realise. A good deal of intellectual rationalisation of the cases occurred at the end of the nineteenth century. For the proposition that one who is able to establish possession of goods can (on that basis) sue for conversion, our modern books typically rely on Armory v Delamirie, yet curiously Armory was not cited at all by Pollock.[15] Instead, Pollock relied on Jeffries v Great Western Railway Co,[16] or more accurately, on the following proposition from Serjeant Williams’ notes to Wilbraham v Snow,[17] which was approved by the judges of the Exchequer Chamber in Jeffries:

Possession with an assertion of title, or even possession alone, gives the possessor such a property as will enable him to maintain this action [trover] against a wrongdoer.[18]

The latter part of this passage so closely resonates with Armory that it seems strange Pratt CJ’s judgment was not cited directly by Pollock. Stranger still is that in Jeffries the question whether proof of possession was sufficient to allow trover against a wrongdoer was treated as a novelty. The plaintiff proved possession of certain trucks which had been seized by the defendants, and which the plaintiff claimed to have held under assignment from one Owen (a bankrupt). The defendants claimed under a subsequent assignment from Owen, and produced evidence tending to show that the assignment to plaintiff was fraudulent and void,[19] prompting at least one of the judges to treat the plaintiff as a wrongdoer in possession.[20] Counsel for the plaintiff argued that ‘mere possession is sufficient title against a wrongdoer’, but the authorities given were roundly denied;[21] and when, in reliance on Williams’ notes, Wightman J put the same proposition to the defendants, counsel retorted: ‘There is no decision to that effect’.[22] Yet ultimately the court accepted the proposition, and framed its decision in language which looks familiar in light of our understanding of Armory v Delamirie. So Lord Campbell CJ confirmed that ‘against a wrongdoer possession is title’;[23] Wightman J spoke of ‘the prima facie right arising from possession’;[24] all of their Lordships approved the statement from the notes to Wilbraham v Snow;[25] and all denied to the defendants the ability to plead a jus tertii defence.[26]

It seems possible that, as it was originally understood, Armory was not about possession, or at least that the plaintiff’s right did not arise as a result of possession in the factual sense employed by Pollock. The short judgment in Armory makes no express mention of possession as a source of the right,[27] and neither do most reports of Wilbraham v Snow.[28] Instead, Wilbraham depended on a concept of special property, according to which a non-owner in possession of goods with the consent of their owner was allowed to bring trover for the goods because he was accountable to the owner for their value. In Wilbraham itself, a sheriff was allowed to recover goods taken out of his possession following their lawful seizure. The court reached its decision by analogy with the legal position of carriers, allowing trover to the sheriff because by law he was accountable for the value of goods seized.[29] In such cases the plaintiff’s right appears to be premised on his legal responsibility for the goods, and that responsibility in turn premised on the lawful nature of his possession.[30] It seems possible that, to the lawyers of its day, Armory would have been understood in the same way. Significantly, in Williams’ notes to Wilbraham v Snow, the chimney sweep’s boy was treated as a ‘lawful’ possessor,[31] the only conceivable reason (recalling the law of charity discussed in Chapter 1) being a supposition that the boy had taken the goods to save them for their owner such that he would have had a defence to proceedings for trespass. In other words, perhaps the chimney sweep’s boy could maintain trover because he had made himself responsible for the safety of the jewel. There are traces of this understanding in Bridges v Hawkesworth, where, in disputing the claim of the plaintiff finder, counsel urged the court to consider whether he did ‘take to himself the charge of these notes, or make himself liable for the advertisements’.[32] In Jeffries, though, where the evidence tended to suggest that the plaintiff was a wrongdoer, this special property construction could not technically have been possible. In this sense, it could be that it was the fact that the plaintiff was a wrongdoer in possession which was the novelty in Jeffries,[33] and which explains why Armory v Delamirie was not cited to the court. However, once the ability of a wrongdoer to maintain trover has been recognised, it immediately becomes meaningless to insist on consent or accountability as the basis of recovery. If the wrongdoer in possession without consent can have trover, a fortiori the literal bailee or similar possessor with a positive duty to account to some other for the value of goods. In other words, after Jeffries, the right of a person actually in possession to proceed in trover for some conversion of goods need only depend on proof of possession in the bare factual sense. Correlatively, there is no need to lay emphasis on the lawful nature of special property. This latter at least was the view of Holmes, who argued specifically that special property ‘did not mean anything more’ than possession.[34]

Consistently with these last suggestions we find that authorities in the late nineteenth and early twentieth centuries treat Armory and the earlier cases on trover as depending on proof of possession alone. In South Staffs Water Co v Sharman, the county court judge found for the plaintiffs on the basis of Armory,[35] and (as we have seen) that verdict was upheld by the Divisional Court. In The Winkfield, Collins MR referred to Armory and Jeffries in the same paragraph to support the view that ‘a long series of authorities’ established that in trover ‘possession is good against a wrongdoer’.[36] In a like manner, Holdsworth treated Armory as a possession case,[37] as did Cyprian Williams in the later editions of Williams on Personal Property.[38] No doubt this process of rationalisation was not completely linear: there are instances much earlier than Jeffries where it appears that Armory was treated as a matter of mere possession.[39] But we can be sure that legal writing in the nineteenth and twentieth centuries had a tremendously clarifying effect on this branch of law, and after Pollock no one seems to doubt that proof of possession alone is sufficient to bring proceedings in trover.[40]

Possession as the Basis of Ejectment

It is instructive to note that in the land law of the nineteenth century we find a parallel development, as possession came to be treated as the basis of entitlement to sue for ejectment. By the middle of the seventeenth century, the action of ejectment had become in practice the principal means for protecting freehold interests in land, and this position was consolidated by the abolition of the old real actions in the nineteenth century.[41] The common law orthodoxy was that the generation of freehold estates in land (and therefore the ability to try a freehold estate in ejectment), depended on seisin.[42] It has been suggested that in its original sense seisin bore a close factual resemblance to possession,[43] but equally certain is that it came to bear a technical legal meaning.[44] Relevant proof of this, were it needed, is found in the origins of the action for ejectment itself. Very late in the fifteenth century, a species of the writ of trespass called ejectio firmae was allowed to a termor in order that he might achieve specific recovery of his term.[45] Previously, specific recovery had been unavailable to him because he could not have the assize of novel disseisin for the very reason that he lacked seisin.[46] According to Maitland, the word ‘possession’ eventually appeared in the English texts to explain the interest of the termor, and distinguish it from that of the person seised.[47] The termor had possession, and therefore ejectment as originally understood; the person seised had an estate, and had entry or his action if he was disseised.[48]

By means of an elaborate fiction, the action of ejectment came to be available to freeholders.[49] It was trite that the action was relative,[50] but the plaintiff must prove his case. He recovered on the strength of his own title, and could not rely on any defect or weakness in the title of the defendant.[51] Possession became an important index of proof. The facts of possession were taken as evidence of seisin in fee,[52] and early in the nineteenth century they were accepted as conclusive in the absence of evidence to the contrary. In Doe v Dyeball,[53] the plaintiff proved a lease to him and a year’s possession, and was allowed to maintain ejectment against the defendant who forcibly turned him out of a room. Against the objection that the plaintiff proved no title, Lord Tenterden CJ replied ‘there is ample proof; the plaintiff is in possession and you come and turn him out, you must show your title’.[54] In Doe v Cooke,[55] the plaintiff proved 23 years’ possession under a lease prior to the last 10 years, when it was not disputed that the defendant had been in possession. The court took it for granted that the plaintiff’s earlier possession established a prima facie case, and held that the defendant could only succeed on proof of a title ‘of a higher description’ than that of the plaintiff.[56] In the absence of this evidence the court favoured the earlier possession, and found for the plaintiff.[57]

In the middle of the nineteenth century, some judges were prepared to depart from the orthodox notion that possession was merely a mechanism for proving seisin. The language of the judgments began to suggest alternatively that possession itself was a source of freehold estates in land. In Asher v Whitlock,[58] one Thomas enclosed land from the waste and built a cottage. By his will he devised the land and cottage to his wife for life or until she remarried, remainder to his only daughter ‘in fee’. At the date of Thomas’ death, the limitation period had not expired. The plaintiff claimed as heir at law of the daughter. The defendant was the second husband of the wife, who continued to reside in the cottage after the wife’s death. On the question whether Thomas could have maintained ejectment against the defendant,[59] Cockburn CJ thought it ‘too clear to admit of doubt’ that he could have done so,[60] but framed the proposition solely in terms of Thomas’ possession:

‘On the simple ground that possession is good title against all but the true owner, I think the plaintiffs entitled to succeed, and that the rule should be discharged’.[61]

That this was an unorthodox approach becomes clear through the contrasting opinion of Mellor J, who was more faithful to the dogma. His Lordship agreed that Thomas acquired an estate interest in the land, but this was because ‘the fact of possession is primâ facie evidence of seisin in fee’.[62] Similarly, in the slightly earlier case of Doe v Barnard,[63] Patteson J held that possession raised a presumption of seisin in fee, and that this fee would descend to an heir.[64] In argument, counsel for the plaintiff attempted to rest the necessary right on possession alone, but the proposition was roundly denied:

the ground [of saying that an actual possessor can maintain ejectment] would not be that possession alone is sufficient in ejectment (as it is in trespass) to maintain the action; but that such possession is prima facie evidence of title, and, no other interest appearing in proof, evidence of seisin in fee.[65]

As is apparent from such diverging opinions, it seems that in the nineteenth century there was real debate about whether seisin or possession was the proper basis of a freehold estate in land, and therefore of title to sue for ejectment. Evidence of equivocation on the issue is provided by the various judgments of the Court of Appeal in Rosenberg v Cook.[66] The defendant had purchased certain land from a railway company via a conveyance that was ultra vires the company and void. The plaintiff contracted to purchase the land from the defendant, but later he learned of the defect in the defendant’s title and brought an action to recover the deposit paid under the contract. The plaintiff argued that the contract should be set aside because the defendant had contracted to sell a freehold interest in the land whereas he held only a revocable licence to go on to the land.[67] The Court of Appeal rejected that claim, holding that the defendant held an original freehold estate in the land, but the language of the judgments oscillates between seisin and possession as the source of the defendant’s right. Jessel MR noted first that a ‘disseisor’ acquired ‘a title good against all the world, except against those who might be proved to have a better one’,[68] but framed his conclusion in terms of possession:

The simple fact is that the [defendant] had a possession in this case, so that a fair sale of that possession is perfectly good.[69]

Brett LJ seems to have estimated the defendant’s right in terms of possession, and reasoned clearly that ‘possession would give a title’ before rather ambiguously adding that the title in question was ‘similar to that of a disseisor or a trespasser’.[70] Cotton LJ, on the other hand, firmly attributed possession as the cause of the defendant’s title, and did not at all employ the language of seisin. The defendant had ‘actual possession’, and a ‘sufficient possession to ripen into a title under the Statute of Limitations’.[71]

The commentators too were divided. Hargreaves argued that developments in the action of ejectment had ‘not altered the principles derived from the real actions’.[72] He was very clear that, as a matter of fact, a plaintiff in ejectment must prove seisin in a technical sense,[73] and whilst he accepted that proof of possession was good evidence of seisin in fee in the absence of an alternative explanation for the plaintiff’s possession,[74] abridgements of this proposition (such as those found in Asher) that connected possession with the generation of an estate were to be rejected as hinging on ‘loose terminology’ attributable to ‘the sentimental liberalism of the [eighteen] fifties’.[75] Holdsworth took an opposing position, considering possession the foundation of a general right, and connecting this with an emergent recognition in English law of an absolute conception of ownership like the Roman dominium.[76] Others explored the middle ground. The first Williams on Real Property to cite Asher was the seventeenth edition, prepared by TC Williams.[77] It acknowledged the orthodoxy that possession was evidence of seisin in fee, but equally spoke of ‘title’ being ‘founded on possession’, and ‘wrongful’ possession as generating a fee simple estate.[78] Lightwood thought it impossible to deny that possession ‘is itself the source of a right of possession’,[79] but his description seemed to indicate that this right was qualitatively different from that of the person seised in fee,[80] and he continued to insist on the truth of the established doctrine.[81] As we move through the twentieth century it becomes more common to treat possession alone as the basis of title: the second edition of Cheshire is to this effect;[82] and Megarry and Wade took the view from the first.[83] But the point seems never authoritatively to have been resolved,[84] and even in the 1980s an eminent commentator regarded it as unsettled.[85] Pollock for his part seemed heavily influenced by Maitland’s opinion that possession and seisin were originally synonymous,[86] and so whilst in places his essay holds closely to the dogma,[87] ultimately he conflated possession and seisin, arguing that ‘seisin is the legal result of de facto possession’.[88] It seems plain that at one stage the law was much more technical than this, but Pollock thought it ‘unnecessary to speak of the details of the old law’,[89] and in considering Asher and the related authorities he spoke only in the language of possession and dispossession.[90]

From Remedy to Right

Having thus consigned the difficulties of seisin, Pollock could easily speak of
trover and ejectment as equivalent procedures. Accordingly we read that:

Under the old procedure an actual possessor might sue either in trespass for the wrong to his possession, or in a form of action founded on right to possess (ejectment for land, trover for goods). In the latter alternative, his right, being derived from his own actual possession, was still not allowed to be disputed by a wrongdoer, and he had the same advantages as if he had sued in trespass. In other words, possession is equivalent to title as against a mere wrongdoer, and this is a substantive rule of law not affected by forms of action’.[91]

As we have seen, already this conclusion is the result of considerable (and potentially novel) rationalisation,[92] but having in these terms identified a general ‘right to possess’, Pollock switched his focus from the remedial procedures of trover and ejectment to an elucidation of the legal rights of potential claimants. His starting point was the generality of the possessor’s protection. Since the cases made clear that the possessor’s right was exigible against any defendant, and not vulnerable to a defence of jus tertii,[93] it followed that:

possession confers more than a personal right to be protected against wrongdoers; it confers a qualified right to possess, a right in the nature of property which is valid against every one who cannot show a prior and better right.[94]

Once the law had gone so far as to recognise this right accruing to the possessor, Pollock thought that necessarily it must go further, offering protection not just to the possessor, but also to her successors and assigns.[95] The possessor would deal with her goods. She would sell them, give them, or bequeath them, and her transferees should be afforded the same degree of legal protection, because ‘the general reasons of policy are at least as strong in their favour … their case at least as meritorious’.[96] Accordingly, Pollock thought that the right arising from possession must also be capable of transfer,[97] and his statement of the right took on a very substantive flavour:

As against strangers, the right founded on possession has the incidents of ownership and is transmissible according to the nature of the subject matter: we may say compendiously that Possession is a root of title.[98]

This statement is a good deal broader than the label ‘right to possess’ suggests. Pollock is here arguing that possession generates an entitlement, that it creates a property right in the fullest sense, attracting the standard advantages of ownership, and differing from ownership only insofar as it will yield to one who can prove a better (because earlier) right to possess.[99] In other words, the right founded on possession differs from ownership only by degree and not by quality: it has the content of ownership, but it will be vulnerable to a commensurately better right. Holmes ventured a similar, though conversely framed proposition,[100] and certainly the position finds support in the ejectment cases we have been considering. In Asher v Whitlock, having decided that Thomas could have maintained ejectment against the defendant, the court held further that Thomas’ interest was devisable and that ejectment could be maintained by his daughter and her heirs.[101] Doe v Jauncey and Doe v Barnard also endorse the general principle that the right acquired in these circumstances is devisable even before the expiry of the relevant limitation period,[102] but it is the general structure of the law that lends the greatest support to Pollock’s argument. It was well settled that in its fictional form ejectment put at issue the freehold title of the true plaintiff. To the extent that possession generated a right to proceed for ejectment, it must also generate a freehold estate if a successful suit was to have any meaning. In other words, the substantive effect of possession, ie that it created a freehold estate in land, was already clear from the procedural learning on ejectment.[103] The achievement of Pollock was to rationalise the matter from the other end, considering the nature of the right in isolation from the procedure of the remedy.[104] In this sense no novel proposition of law is involved, but our focus is laid squarely on the right, and the rhetoric of freehold estates confirms rather than denies that it ‘has the incidents of ownership and is transmissible according to the nature of the subject matter’.[105]

We might question whether the learning on goods lent itself as readily to this conclusion. Instructively, in his eight-page treatment of ‘title by possession’, Pollock referred only to one other case on goods (which was Buckley v Gross)[106] for the proposition (discussed below) that a right generated by possession could not avail to the possessor once it had been lawfully divested.[107] It is clear that the majority of his argument hinged on the land law. It is possible that the notion of special property, drawn from the earlier trover cases, did not lend itself to correlativity with a freehold estate in land, though Holmes probably would have disputed this.[108] Nevertheless, as far as goods are concerned, the relative lack of cited decisions gives Pollock’s account the appearance of novelty, and it would not be unreasonable to conclude that the ejectment cases were the greater influence on his substantive account of the right arising by possession.[109]

Seeming to operate generally against Pollock’s conclusions on the rights of possessors is the tendency of the courts, both in ejectment and trover, to regard their conclusions in favour of successful plaintiffs as presumptions. In Jeffries, Lord Campbell CJ spoke of ‘a presumption of law … that the person who has possession has the property’,[110] and similar dicta are found in the ejectment cases.[111] Understood literally, a presumption that a possessor had title need not involve any commitment to the nature of the rights actually held by the possessor. In theory we could affirm that ‘as against’ a particular defendant a particular possessor could be regarded as having property (for the sake of dispute resolution), but at the same time assert that in reality he had no rights.[112] The better view, however, is that these cases do not depend on presumptions.[113] As has been suggested recently, presumptions ‘properly-so-called are methods of proof of facts, and no fact is here in issue’.[114] In the ejectment cases considered above, and very clearly in Jeffries, the plaintiff had actually proven possession before the court.[115] Accordingly, the question in the cases concerned the legal inference to be drawn from facts already proven by evidence. In these circumstances, it is not technically possible for a claimant to benefit from any presumption,[116] and it seems analytically more accurate to regard possession as the source of a definite legal relation.

Whilst the idea that possession generates a relatively good property right had a heavy influence on later commentators,[117] it is worth noting that the finding cases themselves contain little by way of express comment on the issue; indeed the language of the judgments is entirely equivocal. Although the most recent are framed in terms of the rights of the parties, they disclose no analysis of the nature of those rights. In truth, the question has been avoided in reliance on Armory v Delamirie or, as in Lord Russell CJ’s pivotal judgment in Sharman, on some general rehearsal of Pollock’s view. Thus in Parker, despite a conviction that there was no decision binding on the Court of Appeal, and the accordingly portentous rehearsal of the court’s ‘duty to extend and adapt the common law in the light of established principles and the current needs of the community’,[118] Donaldson LJ did little more than reproduce the 250-year-old judgment in Armory:

a finder of a chattel, whilst not acquiring any absolute property or ownership in the chattel, acquires a right to keep it against all but the true owner or those in a position to claim through the true owner or one who can assert a prior right to keep the chattel which was subsisting at the time when the finder took the chattel into his care and control.[119]

Of course, this dictum was a qualification of Armory. It made plain that there were right-holders other than the ‘true owner’ to whom the finder’s rights would yield. But (a) this was not a new qualification;[120] and (b) the dictum wholly failed to discuss or answer the interesting question about the nature of this ‘right to keep’. We are told no more about the right in Parker. We might have expected a Court of Appeal freed from the constraints of the forms of action, and exercising its duty to adapt common law authorities to modern life, to give more thought to the substance of the right it was recognising. As it is, insofar as they employ the language of rights, the judgments in Parker and the other cases remain heavily dependent on the work of Pollock, though since Sharman they are at least consistent with Pollock’s general view, ie that a finder of goods, because of her possession, acquires a general property right to the object of her find, which right attracts the standard advantages of ownership and will avail against all those incapable of proving a better, because earlier, right.

Some questions must remain as to the authoritative basis of Pollock’s thesis. It is entirely clear that the core of the argument was an assimilation of trover and ejectment, a deep conviction that since each depended solely on proof of possession each raised a right to possess protected to the full extent of the law. This proposition could equally well be regarded as heresy or genius rationalisation. Certainly, insofar as it related to goods it depended to some extent on a resolution of the old cases on special property; on ejectment, it depended on an outright conflation of possession and seisin. Accordingly it would seem at least to be arguable (a) that trover and ejectment were not technically as synonymous as Pollock’s major premise would suggest; and (b) that the role of possession in each was not as clear as the premise would suggest. We have tried to show that the general trend of the cases does lend some support to Pollock’s view, but much further research is needed in this area. For now, and as we seek later to show that English law as currently understood produces a satisfactory law of finders, it will be enough to have established that our finding cases depend on Pollock’s understanding of possession, and Pollock thought that possessors acquired relatively good property rights in the things they possessed. To this extent, that most pervasive idea in our law of finds, ie that a finder does acquire some property in the object of her find, continues to depend for its validity on the old learning of the forms of action, notwithstanding their long abolition from the common law.

The Extent of the Finder’s Right

Accepting that by possession a finder acquires a property right in her find, a second question arises as to the extent of that right. It is plain that for so long as she remains in actual possession of a chattel, a finder will be able to resist any wrongful interference with it. Much more difficult is the question whether a finder’s rights outlive the facts of possession. Would these rights survive a subsequent loss of the chattel? Can a finder sue for a subsequent conversion when she is not also the possessor at the time of that conversion? Does possession ‘confer a right to possession which outlives possession itself?’[121]

At the extreme end of a spectrum of negative answers to this question is the opinion of Winfield, who seems to hold that a finder cannot sue for conversion unless she was actually in possession of the goods at the time of the alleged conversion.[122] The immediate difficulty for this opinion is that the decided cases are against it. Indeed, in almost all the cases where a finder successfully has asserted a right in respect of some conversion of goods, the finder has been out of the possession at the time of the conversion. In Armory v Delamirie, the finder had passed the ring to the goldsmith’s apprentice, who passed it in turn to his master, after which the conversion occurred. In Bridges v Hawkesworth the notes were likewise surrendered to the shop owner, who kept them safely and advertised their loss. The finder had been out of actual possession for three years before he asserted his right against the defendant. Parker follows more or less the same course, and yet a clearer example is Hannah v Peel