THE CONCEPT OF POSSESSION IN THE COMMON

LAW: FOUNDATIONS FOR A NEW APPROACH

By A. E. S. TAY*

I
'A complete theory of possession,' Sir John Salmond wrote opti-
mistically/ 'falls into two parts: first an analysis of the conception
itself, and secondly an exposition of the manner in which it is
recognized and applied in the actual legal system.' There is no
doubt that Salmond thought such a complete theory possible. Yet
the judges and legal writers to whom we look for an analysis of the
concept of possession in the common law-including Salmond
himself-prove surprisingly disappointing. Conscious that they will
later have to wed the concepts they formulate to the complexities
of the common law, trained to make distinctions rather than to see
connexions, they smother analysis and prevent clarification by a
welter of conceptual terms. In the law made or interpreted by judges
we meet one possessory term after another: 'physical possession',
'actual possession', 'de facto possession' and 'possession'; 'right to
possession', 'right of possession', 'constructive possession' and 'pos-
session'; 'possession in law', 'legal possession' and 'rightful posses-
sion'; 'property', 'special property' and 'limited property'. To make
confusion worse confounded, the distinctions are not rigidly ob-
served, in the course of their judgments, by the very men who have
drawn attention to them. 2
Legal writers, far from cutting through this tangle of special
terms and distinctions to a primary concept or logical 'cluster' of
concepts, add or superimpose their own subdivisions. Bentham dis-
tinguishes physical possession from legal possession, exclusive pos-
session from possession in common, possession of things moveable
• Of Lincoln's Inn, Barrister-at-law; advocate and solicitor, Supreme Court of
the State of Singapore; research scholar in the Department of Law, Institute of
Advanced Studies, Australian National University.
1 Salmond, Jurisprudence (10th ed., 1947) 287.
2 Earl Jowitt has drawn attention to a typical example: 'Under English law
where there is a simple contract of bailment the possession of the goods bailed
passes to the bailee. The bailor has in such a case the right to immediate pos-
session and by reason of this right can exercise those possessory remedies which
are available to the possessor. The person having the right to immediate pos-
session is, however, frequently referred to in English law as being the possessor.'
His Lordship goes on to add, quite correctly: 'In truth English law has never
worked out a completely logical and exhaustive definition of possession.'-United
States of America & Republic Of France v. DollfuS Mieg et Cie. S.A. & Bank of
England [1952] A.C. 582, 605. We shall see below how judges who have distin-
guished 'custody' from 'possession' and 'possession' from 'property' will go on
in their judgments to use one term when they mean the other.
NOVEMBER I1}64] Possession in the Common L= 477

from possession of things immoveable, possession of services and

possession of fictitious entities. 3 Continental Romanists and Civilians
bring out allegedly fundamental contrasts between possessio and
detentio and between possession with animus domini and possession
with animus possidendi, as well as the distinction between possessio
naturalis, possessio civilis and possessio ad interdicta. 4 Traditional
common law writers stress the importance of keeping apart 'pos-
session as a law concept', (de facto possession' and 'possession in law'
or (alternatively) 'actual possession', 'civil possession' and 'construc-
tive possession'; some of them counterpose 'corporeal possession' to
'incorporeal possession', 'immediate possession' to 'mediate pos-
session'.5
The unsatisfactory state of conceptual analysis and juristic for-
mulation in the field of possession is now widely recognized. Some
ascribe it to the difficulties inherent in whatever basic concept of
possession there may be; others to confusions of terminology, a
priori imposition of theory and a misguided endeavour to reduce
decisions that have developed in the context of specific branches of
law and of separate remedies to smooth components of a coherent
system. 'In the whole range of legal theory there is no conception
more difficult than that of possession,' Sir John Salmond writes at
the beginning of his discussion of possession. 6 'Possession', we read
in the cases, 'is a word of ambiguous meaning'7-'there is, perhaps,
no legal conception more open to a variety of meanings than "pos-
session",'8 Winfield, characteristically optimistic, puts the main
weight on confusions of terminology: 'Our law has a fairly good
working scheme of possession although it has not indulged in much
scientific dissection of the idea. Its weakest spot is its slovenly ter-
minology.'9 Dias and Hughes, all too ready to resolve the problem
into matters of convenience and policy, prefer to blame their less
3 Bentham, General View of a Complete Code of Laws, in Collected Works, iii, 188.
4 These distinctions form the crux of Savigny's influential Treatise on Possession.
5 This is the barrage of distinctions with which one emerges after reading the
main older works on possession: Pollock and Wright, Possession in the Common
Law (1888), Lightwood, A Treatise on Possession of Land (1894), Salmond, Juris·
prudence (1902), ch. 13, 14 and Terry, 'Possession' (1918-1919) 13 Illinois Law
Review 312, Wigmore Celebration Legal Essays, 170. Keeton's The Elementary
Principles of Jurisprudence (1st ed., 1930), somewhat surprisingly for a book by
a modern writer, accepts most of Salmond's distinctions.
6 Salmond, Jurisprudence (10th ed., 1947) 285.
7 Per Erie C.]., in Bourne v. Fosbrooke (1865) 18 C.B. (N.S.) 515, 526; 144 E.R.

545, 549·
8 Per Fry L.J., in Lyell v. Kennedy (1887) 18 Q.B.D. 796, 813.
9 Winfield, Law of Tort (4th ed., 1948) 306. Note that Winfield's editor is less
optimistic and has changed the passage to: 'Our law has not worked out a con-
sistent theory of possession in any of its branches and it has not indulged in
much scientific dissection of the idea. Its weakest spot is its slovenly terminology:'
6th ed. (edited by T. Ellis Lewis) 364. For similar criticisms of judicial terminology
see Pollock and Wright, op cit. 2, Kocourek, Jural Relations (2nd ed., 1951) 365
and Paton, A Text-Book of Jurisprudence (2nd ed.) 454.
478 Melbourne University Law Review [VOLUME 4
'empirical' forerunners. 'If a topic has ever suffered from too much
theorising,' they write, 10 ~it is that of possession, and nowhere else
is the danger of an a priori approach to jurisprudence better illus-
trated. The actual working of the law has not only been obscured
by a fog of speculation, but, what is worse, decisions have been
falsified so as to fit them into some preconceived theory.'
The period since the First World War has seen a marked decline
in the confidence that men have in universal intellectual systems
and in the pervasive application of fundamental 'rational' principles.
The Western world, and especially the English-speaking world, has
tnoved into an age of ad hoc adjustment and manipulation, of piece-
meal social engineering, of the utilitarian subordination of prin-
ciples, systems and traditions to the requirements of men living
in specific circumstances at a specific time. The most 'modern' legal
writers put less and less weight on the conception of law as a syste-
matic development of principles striving toward the highest pos-
sible degree of coherence and more and more emphasis upon law
as an instrument in the service of competing masters. For rational
coherence with its alleged rigidity they prefer a substitute ad hoc
flexibility, for 'principles of law' attitudes, presumptions and
policies, for concepts specific rules established for specific situations.
The worst of the 'modems' seek, where at all plausible, to reduce
law to disparate rules resulting solely from judicial policy and social
requirements; the best of them argue that the rationality of the
legal system lies in its formulation of open-ended principles and
defeasible concepts. Thus Dias and Hughes, on the one hand,
assure us that 'the idea of possession is no longer tied to fact, and
it has become a concept of the utmost technicality',H which, to them,
means that each branch of the law has made up its own rules. Mr
D. R. Harris, on the other hand, after referring with approval to
Professor H. L. A. Hart's view that legal concepts cannot be defined,
but only described/ 2 argues that there are a number of 'factors rele-
vant to possession' recognized by the courts which nevertheless can-
not serve to define possession because no single factor is decisive and
because not all factors are always relevant. 13 Their relevance or ir-
relevance, like the defences to Professor Hart's defeasible prin-
ciples,a cannot be subsumed under a general rule.
10 Dias and Hughes, Jurisprudence, 308.
11 Ibid. 317.
12 Hart, 'Definition and Theory in Jurisprudence' (1954) 70 Law Quarterly
Review, 37.
13 Harris, 'The Concept of Possession in English Law' in Oxford Essays in Juris-
prudence, 69-106.
14 Hart, 'The Ascription of Responsibility and Rights' (1948-1949) 49 Proceedings
of the Aristotelian Society, 17I. Professor Hart's view has been effectively criticized
by J. L. Mackie, 'Responsibility and Language' (1955) 33 Australasian Journal of
Philosophy 143.
 Possession in the Common Law 479

A living body of law cannot be tied into the strait-jacket of an

a priori conceptual system: but to insist, as a matter of principle,
that we should not ask for general conceptions underlying what
appear to be specific rules separating one possession from another,
is to live in the intellectual Ice Age in which the first forms of
action were born. The rigid procedural requirements of the early
law, by their very emphasis on certain common aspects of recurring
situations, introduce conceptual distinctions into the law and force
it to develop them. It is the recognition of intellectual connexion that
distinguishes the Leges Henrici Primi from the Laws of Aethelred;
it is the presence of conceptual development and continuity that
distinguishes the history of law from a mere chronicle of Acts. In
law, we have to recognize unity as well as separation, the historical
development of a system struggling to accommodate new require-
ments and competing demands. Litigants, in the first place, look not
to contract, or tort, or agency, or sale, but to the law: it is as
aspects of the law that its distinct branches have developed; even if
we were to put aside, for the moment, the extent to which judges
handing down decisions in one field are influenced by concepts and
decisions that they and other judges have laid down in another, we
can gain full understanding of what remain as the specific require-
ments of specific fields only by contrast and comparison with other
legal requirements and other legal remedies.
The defect of the traditional writings on possession is not their
search for a rational understanding of the law as a whole. Neither
were they necessarily wrong, simply because their search for such
understanding forced them to make a plethora of distinctions. The
point is rather that these distinctions did not arise convincingly out
of the course of their argument, but confront us as ad hoc distinc-
tions, as saving devices, forced upon them in the process of fitting
their scheme to the law. In Salmond, the process of matching his
conceptual scheme and the law leads to vicious falsification of legal
developments and decisions;I5 in Pollock, a similar process leads to
so many modifications and to so many contrasts between conceptual
and 'legal' possession that we virtually come to forget the hesitant
conceptual analysis with which Pollock began. From Pollock, as
from the consciously 'modern' writers, we end by knowing the rules
without understanding possession.
The fault, I should argue, lies not in the aim of reaching a com-

15 Dias and Hughes find their natural butt in Salmond and expose his falsifica-
cations in some detail, op. cit. 316 et seq. See also Williams, 'Language and the
Law-IV' (1945) 61 Law Quarterly Review 384, 390-391 and (for the distortion
of decisions on finding by Salmond as well as other writers) Goodhart, 'Three
Cases on Possession' (1928) 3 Cambridge Law Journal 195, Essays in Jurisprudence
and the Common Law 75-90.
480 Melbourne University Law Review [VOLUME 4

plete theory of possession that Sir John Salmond had set himself.
The fault lies in his separation of the analysis of the concept from
the study of its working in the legal system. Concepts are concepts
in use, 'possession' is a term with a certain role. To understand
possession, we must look, not at the word, but at the way in which
possession entered our legal system, the parts it was called upon to
play in it, the character and problems of its development. If we do
this, we do emerge with a general concept of possession implicitly
recognized and applied in our law. Only in terms of such a general
concept, I shall argue, can we understand the special problems that
have arisen in specific fields.
One word about the doctrine of stare decisis. In any study of the
law, as opposed to a mere catalogue of decisions, some judgments
will be held by the author better than others, some cases will appear
to him wrongly decided. It is one thing to falsify the law; it is
another to refuse to treat it as a sequence of sacred and immutable
fiats. In tracing the development of the common law, over a period
of nearly 900 years, one is concerned with a chain of legal reasoning
extending over time; the importance of cases is as links in that chain.
It is as links that I have striven to understand them. The loose ends
that need to be severed from the chain do not occur (at least in my
field) as often as one might expect.

II
Writers with a logical or sociological bent have frequently sought
to gain a preliminary understanding of the concept of possession by
contrasting it with, or relating it to, the concept of ownership.
Neither concept, it is clear, can be divorced from social or legal
sanctions and/or physical powers that establish or protect owner-
ship and possession and thus give the terms meaning and force.
In a society in which no one had or claimed the control of anything
to the exclusion of others, the terms 'ownership' and 'possession'
would not be part of the language. But in societies where such
control is found, the two terms will tend to establish themselves
and to display certain general features that transcend the specific
arrangements and definitions of anyone particular legal system.
One such general contrast suggested by the drawing of a distinction
between ownership and possession is the contrast between an ulti-
mate, non-contingent right and a limited, temporary and derivative
one. This contrast, often based on an untechnical view of Roman
Law, has occasionally intruded upon the common law and has
acquired a certain importance in such modern arrangements as the
registration of title and hire-purchase agreements; it does not play,
and has not played, any fundamental role in the development of
NOVEMBER r9l>4] Possession in the Common Law 481
the common law. It is also a legally sophisticated contrast, moving
into the forefront in those societies that have economic and legal
arrangements by which ownership and possession frequently part
company. In early Western societies, as law developed, this was not
so. There, ownership and possession were normally fused in a single
person, the owner-possessor who came before the law demanding
protection or restitution of something he had held by virtue of
unchallenged right and of uncontested physical use and control.
The contrast here is the more primitive one between ownership as
the assertion of a right against others and possession as a physical
fact, as a relation to the thing. That the concepts of ownership and
possession, in this second sense, arise in law as two aspects of a single
situation has been emphasized by RudoIf von Ihering in a brilliant
passage:
Possession is the objective realization of ownership. It is in fact what
ownership is in right. Possession is the de facto exercise of a claim;
ownership is the de jure recognition of one. A thing is owned by me
when my claim to it is maintained by the will of the State as ex-
pressed in the law; it is possessed by me, when my claim to it is
maintained by my own self-assertive will. Ownership is the guarantee
of the law; possession is the guarantee of the facts. It is weIl to have
both forms of security if possible; and indeed they normally co-exist.
But where there is no law, or where the law is against a man, he
must content himself with the precarious security of the facts. Even
when the law is in one's favour, it is well to have the facts on one's side
also. Beati possidentes. Possession, therefore, is the de facto counterpart
of ownership. It is the external form in which rightful claims nor-
mally manifest themselves. The separation of these two things is
an exceptional incident, due to accident, wrong, or the special nature
of the claim in question. Possession without ownership is the body
of fact, uniformed by the spirit of right which usually accompanies
it. Ownership without possession is right, unaccompanied by that
environment of fact in which it normally realizes itself. The two
things tend mutually to coincide. Ownership strives to realize itself
in possession, and possession endeavours to justify itself as ownership.
The law of prescription determines the process by which, through the
influence of time, possession without title ripens into ownership, and
ownership without possession withers away and dies. 16
To the common lawyer, Ihering's analysis seems particularly apt,
for it is amply confirmed by the earlier concept of seisin in the
English law. Seisin, says F. Joiion des Longrais in a great work/ 7
'is an enjoyment pervaded by the elements of right, fused with right
in all its forms and by nature indistinguishable from it.' Such

16 Ueber den Grund des Besitzschutzes 179. (I cite the translation by Salmond,
op. cit. 311-312).
17 La conception anglaise de la saisine du XII- au XIV- siecle i, 45: 'C'est une
jouissance toute pem!tree d'elements de droit, elle se fond avec le droit sous toutes
des formes, et n'en distingue pas sa nature.'
482 Melbourne University Law Review [VOLUME 4
fusion of right and enjoyment, expressed in the paradigm case of
the owner-possessor (as Ihering suggests), is natural to early law:
the situation where enjoyment and right to enjoy coincide in a
single situation is least disturbing to social order and crude con-
ceptions of justice. But disturbances do occur: men lose and acquire
possession through violence, accident, fraud and other wrong. If
society is to vindicate any conception of justice above the mere
ratification of right, it must recognize and concede that enjoyment
and the right to en joy can part company.
The problems that arise from the possible disjunction of enjoy-
ment and the right to enjoy, of possession and the right to possess,
permit of marked differences in legal approach. Ancient Rome, with
its tight and authoritarian familial and social structure, placed
primary emphasis on the conception of right. Historically, rights
may have stemmed from possession as right stems from might;
Roman law is concerned with this only to the extent of permitting
the severely restricted procedure of usucapion and even then it
treats the very candidate under such procedure not as a possessor
but as a candidate for title.l8 For the Roman legislator and juris-
consult, the paradigm right to enjoy is absolute title, good against
the whole world, in principle capable of proof and vindication in
court without any reference to possession whatever. 19 There are
rights based on physical possession as well. On the one hand, they
are sharply marked off from rights based on title and are to be
established by recourse to quite separate remedies, the interdicts;
on the other hand, such possessory rights do not follow from mere
physical control but depend on a view of possession that confines
it to those who in principle could become owners and behave as
though they were. 20
18 Thus it was essential for possession intended to ripen into ownership by
usucapion to begin with a iusta causa or iustus titulus and for the possessor to
have done everything in his power to become owner, requirements that sharply
separate such possession from the possession recognized by the interdicts (infra).
Again, the remedy of one who was only in the process of acquiring title in via
usucapiendi was nevertheless not confined to the interdicts but included a special
action, the actio Publiciana, which must be classed as proprietary: see Buckland
and McNair, Roman Law and Common Law (2nd ed.) esp. 63, 74.
19 This, of course, is the paradigm; in practice, the plaintiff in a vindicatio would
normally be justifying a title open to doubt and might seek to rely on usucapion
as well as the reputation of ownership. But usucapion here is a technical mode of
proving title, not a general claim to rights in virtue of possession and it
is significant that Roman jurists consider that success or failure in a possessory
action is no bar to success or failure in a proprietary one and vice versa. See Buck-
land and McNair, op. cit. 75-76.
20 Thus, 'no one who has a thing in virtue of a contract recognizing the owner-
ship of another person can possess, and the same is true even of a person who has
a real right of limited extent, such as usufruct, though such persons have quasi-
possession, protected by a special interdict. There are, it is true, four types of
persons who are regarded as possessors, even though they hold under a contract,
namely the pledge creditor, the tenant for a perpetual or very long term of years,
the tenant at will, and the stake-holder. But these cases can all be explained away
NOVEMBER 1964] Possession in the Common Law 483

In ancient Germanic and early English law, formulated in con-

ditions of a far looser social structure and amid the far greater pre-
valence of self-help, might and right were and remained more closely
linked. Ownership did not become clearly divorced from possession;
in ancient Germanic and English law it was not possible either to
gain recognition of a right of ultimate possession good against the
whole world or to vindicate any right to possess without reference
to possession itself. The primary concept-indeed, the only concept
available-was the concept of seisin, a concept which emphasizes
that all proprietary rights stem from physical possession but also
insists that these rights cannot be destroyed merely by destroying
possession. In an authoritarian society, especially one welded to-
gether by the primacy of potestas) the tendency is to derive right
from authority; in the looser, more makeshift and hence more de-
mocratic society of central and north-western Europe the tendency
is to derive rights from the facts. The man who is seised has both
possession and right; a right not simply conferred by authority, but
deriving from the facts and therefore seen as part of his seisin.
The original nature of seisin was obscured from view for several
centuries by the fifteenth-century developments that made 'seisin'
a technical term of the law of real property, divorced and distin-
guished from possession of chattels. Thus, in 1757, Lord Mansfield
gave this well-known and long-accepted definition of seisin: 21
'Seisin is a technical term to denote the completion of that investi-
ture by which the tenant was admitted into the tenure, and with-
out which no freehold could be constituted or pass.' But this is true,
as the subsequent work of Maitland has taught us to see, only of
seisin in and after the fifteenth century. Before that century, seisin
was not a term confined to freehold land, but was the only term
used in law for virtually all instances of possession.'2
on practical or historical grounds. In the classical or later laws they must be
treated as exceptions. Conversely, the bailor or lessor usually retains possession of
the thing he has bailed or let:' Buckland and McNair, op. cit. 73.
21 Taylor demo Atkyns v. Horde ('757) I Burr. 60, 107; 97 E.R. 190, 216; 2
Smith's Leading Cases (9th ed.) 633, 700.
22 Joshua Williams, Law of Real Property (23rd ed.) (1920) 154: 'Seisin ... origin-
ally meant any kind of possession, but was not afterwards used to denote any but
freehold possession.' Pollock and Maitland, History of English Law, ii, 31-32: 'In
the first place, it would seem that for at least three centuries after the Norman
Conquest our lawyers had no other word whereby to describe possession. In their
theoretical discussions, they, or such of them as looked to the Roman books as
models of jurisprudence, could use the words possessio and possidere; but these
words are rarely employed in the formal records of litigation, save in one parti-
cular context. The parson of a church is 'in possession' of the church: -but then
this is no matter for our English law or our temporal courts; it is matter for
the canon law and the courts Christian; and it is all the more expedient to find
some other term than 'seised' for the parson, since it may be necessary to con-
trast the rights of the parson who is possessed of the church with those of the
patron who is seised of the advowson.' Maitland notes that 'for a somewhat similar
reason it is not uncommon to speak of a guardian as having possession of the
wardship, while the ward is seised of the land.'
484 Melhourne University Law Review [VOLUME 4
Throughout the thirteenth century and in the most technical docu-
ments men were seised of chattels and in seisin of them, of a fleece of
wool, of a gammon of bacon, of a penny. People were possessed of
these things; law had to recognize and protect their possession; it had
no other word than 'seisin' and therefore used it freely.23
In the Leges Henrici Primi, in Glanvill, Bracton, Fleta and Brit-
ton, in statutes and rolls from the times of Richard I, Henry Ill, the
three Edwards, Richard II and Henry IV, even to the opening years
of the reign of Henry VI, we find consistent reference to the seisin
of chattels. 24 From about 1443, when uncore detient comes to sup-
plant the phrase uncore seisi in respect of chattels, we enter a tran-
sitional period. In Littleton's Tenures, written between 1474 and
1481, the distinction between 'seisin' and 'possession' familiar to
later English lawyers was decisively proclaimed:
Also, when a man [in pleading] will show a feoffment made to him,
or a gift in tail, or a lease for life, of any lands or tenements, then
he shall say, by force of which feoffment, gift, or lease, he was seised,
etc., but where one will plead a lease or grant made to him of a
chattel, real or personal, then he shall say, by force of which he was
possessed, ete. 25
From the latter half of the fifteenth century seisin becomes an in-
creasingly technical term in the law of real property. It is sharply
distinguished, as we have seen, from 'possession' in the law of per-
sonal property; as its uses become more technical it is also distin-
guished, not only from the possession of a leasehold but also from
the 'mere' possession of land. The later conception of seisin thus
moves away from the early concept of seisin toward a notion of
title or right; the connexion between the early concept of seisin
and the later technical use of the word is therefore far less intimate
than the connexion between the early concept of seisin and the
English law on possession which came to replace it.
III
The social and juridical function of the concept of selsm in its
original form is to serve as a basis for resolving the problems that
arise when enjoyment and the right to enjoyment appear to have
parted company. The concept of seisin attempts such resolution
without conceding in principle that rights to enjoy can be established
independently of enjoyment or can long last without it. 26 The
Roman lawyer, forced to recognize the possible bifurcation of en-
joyment and the right to enjoy, was satisfied to look to the question
23 Pollock and Maitland, op. cit. ii, 32.
24 Maitland has presented the detailed evidence in his 'The Seisin of Chattels',
(1885) 1 Law Quarterly Review 324; Collected Papers, i, 330.
25 S. 324; Coke on Litt/eton, 200. h.
26 Thus the right to enjoy (expressed in the form of a right of entry) of a
person who has been ejected from his land lasts only as long as he makes 'con·
NOVEMBER I 964] PO!,session in the Common Law 485

of right, to derive title from previous title and to recognize rights

flowing from physical possession or enjoyment only in so far as
such possession or enjoyment could be treated as a (deficient) form
of ownership, of the assertion of a right. The early English lawyer
looks primarily to the possession or enjoyment itself and sees rights
as flowing from it. The assize of novel disseisin, the assize of mort
d'ancestor and the other special assizes that followed all recognize
only one ultimate way of establishing a right to enjoy-that is to
show that you or your ancestor had enjoyed before the defendant
or his ancestor and that your enjoyment was taken away unlawfully
and without your consent, or before you had time to take seisin
as heir. The extent to which rights were independent of title (in
the sense of rightful claim to the enjoyment) and seen as flowing
from the fact of enjoyment alone may be judged from the rights,
powers and benefits that automatically accrue to the disseisor. 27 He
may make a feoffment and convey an estate in fee simple even
though he has disseised a tenant for life; the rights appendant to
the estate go to him and he forms a stock of descent; his heir will
inherit the property and his widow be entitled to dower, while if
the disseisor be a woman, her husband is entitled to curtesy. The
ordinary incidents of tenure affect the disseisor as though he had
been a lawful tenant, and if the disseisor die without heir, the land
will not escheat as long as the disseisor or his stock are seised. Bereft
of their foundation in enjoyment, rights wither and die away.28 The
disseisee can neither make feoffment nor alienate the estate; he
cannot assign and his heirs cannot inherit his right of action or
his right of entry under continual claim; his widow cannot claim
dower, and, if the disseisee be a woman, her husband is not entitled
to curtesy.29 So strong was the refusal to treat a right as something

tinual claim', i.e. agitates on or about the land. The period for which he could
do so was always limited, and as soon as it or his agitation had expired, he lost
any direct right to enjoyment and was left only a right to an action.
27 (1313) Y.B. 6, 7 Edw. 2 (S.S.) 189, per Scrape J.: 'the disseisor claimeth fee
and right and freehold till his tort be proven.'
28 Partridge v. Strange (1553) Plowden 78, 88; 75 E.R. 123, 140. Mountague C.}.
said: 'At common law, he who was out of possession might not bargain, grant, or
let his right or title, and if he had done it, it should have been void.'
29 Thus, in 1218, a plaintiff who was seised under a tortious feoffment succeeded
in novel disseisin against a true owner who had disseised the plaintiff; the owner,
the Court held, may pursue in another way if he wants to: Eyre Rolls (S.S., vo!.
53), no. 38. For detailed discussion of the nature of disseisin and its effects on the
rights of the parties see Maitland, 'The Mystery of Seisin' (1886) 2 Law Quarterly
Review 481, Call. Pap. I, 358 et seq. and 'The Beatitude of Seisin' (1888) 4 Law
Quarterly Review 24, 286, Call. Pap. I, 407 et seq; Ames, 'The Disseisin of Chattels'
(1890) 3 Harvard Law Review 23, 313, 337, Select Essays in Anglo·American Legal
History, Ill, 541, Lectures on Legal History, 172; Bordwell, 'Property in Chattels'
(1915'1916) 29 Harvard Law Review 374, 501, 731; and Lightwood, A Treatise on
Possession of Land, 42, 56. The assize of mort d'ancestor, in order to fill the lacuna
by which any man might with impunity enter on land still vacant after a death,
486 Melbourne University Law Review [VOLUME 4

abstract, incorporeal and detachable from physical fact, that even

rights other than the right to enjoy could be established at law only
by showing that they had been exercised and enjoyed. As Maitland
puts it: 30
A man is in seisin of land when he is enjoying it or in a position to
enjoy it; he is seised of an advowson (for of 'incorporeal things' there
may be seisin) when he presents a parson who is admitted to the
church; he is seised of freedom from toll when he successfully resists
demand for payment.
The conception that a man may be 'in possession' of a right is not
part of our modern law; it was part of medieval law precisely because
medieval law saw such rights as forms of enjoyment and therefore
in physical terms.
The intimate fusion of right and enjoyment, with enjoyment
forming the primary ground, comes out in the prescribed form
of the transfer of land in ancient Germanic law. The transferor of
land made a public but oral declaration of intention to transfer,
known as the sala. To become effective in any way, the sala had to
be followed by the gewerida: a ritualistic and actual transfer of the
land, including the solemn handing over of material representing
the land and of material symbolizing dominium over the land,
followed by a formal abjuration on the part of the transferor and
an actual entry on to the land and carrying out of acts indicative
of ownership on the part of the transferee. 31 The English counter-
part of these ceremonies, the feoffment and livery of seisin, brings
out the same point; one cannot transfer a right to enjoy without
transferring with it the actual physical enjoyment on which it is
based. The transfer of the enjoyment may, when the law permits, be
symbolic, but what is symbolized is the transfer of actual enjoyment
and not the transfer of a right to enjoy. For it is not only that early

had to recognize the passing of a right to seiSIn to the heir and to make this
independent of the actual passing of seisin to him. But the heir's right was
still grounded in his ancestor's seisin and was itself only the right to an action.
The same is true of the later writs of entry: the demandant's right was grounded
in his own previous seisin or that of his ancestor (or person through whom he
claims).
30 Pollock and Maitland, op. cit. ii, 34.
31 See Thorne, 'Livery of Seisin' (1936) 52 Law Quarterly Review 345, 348, 352, for
a fuller account of the early sala and gewerida and for reference to Continental
and English research; also Pollock and Maitland, op. cit. ii, 84-86. Finch in his
article 'Seisin' (1919) 4 Corn ell Law Quarterly 1, in accounting for the concreteness
of the ritual emphasizes the need in early law for 'some visible and suggestive
ceremony which the transaction witnesses can see, which they can accurately and
readily remember and which supplies the want of record and of writings' (at p. 2),
and traces the effects of this requirement of visibility on all branches of law,
dominated as they were, by the doctrine of seisin. But, as I suggest below, the
valuable point that rights had to be visibly expressed must be supplemented by
the recognition that rights stemmed from enjoyment.
NOVEMBER r964] Possession in the Common Law 487

Germanic and medieval English man cannot conceive of disem-

bodied right; it is also that he insists that right cannot come into
being or be transferred without the enjoyment on which it rests.
Seisin, then, for all its incorporation of a concept of right, is
enjoyment. In asking whether a man has seisin, we ask whether he
has actual physical enjoyment; in transferring seisin we transfer
actual physical enjoyment. Logically, seisin incorporates right
only because in English law enjoyment immediately and directly
gives rise to right, at the least to the right of peaceful and quiet
enjoyment against the trespasser and the thief. It is precisely because
the concept of seisin is not founded on right that the thief himself
can have seisin; it is because all proprietary and possessory rights
ultimately stem from enjoyment that-seisin lies at the very root
of the development of the English law of property.
The logical implication of even fundamental concepts used in the
law can be obscured or distorted by the social climate in which the
law operates. In the twelfth century this might almost have occurred.
The writ of right was in form a 'droitural' writ in which the de-
mandant's pleading alleged a claim in domico suo ut de foedo et
jure; as Breve de Recto it was brought in the Court Baron of the
lord under whom the land was held, where the right could presum-
ably be traced to its source in feudal authority and not simply in
actual possession. The assize of novel disseisin, on the other hand,
specifically focused attention on the fact of possession, and was
tried in the Royal Courts. For a space of years this distinction may
well have carried within it the seeds from which a distinction be-
tween proprietary and possessory actions in the Roman sense might
arise. But the comparative lack and unreliability of records, the
effect of the statutes of limitation in confining the tracing of rights
within a certain period, and the fact that the tenant to a writ of
right could not plead a jus tertii, all combined, even within this
writ, to bring the claim back to evidence that the demandant or
his ancestors had enjoyed seisin, collected rents and profits and had
been wrongfully deprived by the tenant or his ancestors. At the
same time, further assizes and writs of entry were being created-
assizes and writs that invaded more and more of the ground covered
by the writ of right while insisting even more firmly that the only
basis of right was seisin, recent or remote. The writ of right, being
the most technical and cumbersome of the real actions, full of
essoins and liable to interminable delays through vouchers to war-
ranty-fell into disuse and the doctrine that seisin was the basis
of right was thus even more firmly established. 32
32 There must naturally have been some tension between what one might call
the politically democratic concept of seisin and the importance in feudal society
488 Melbourne University Law Review [VOLUME 4

The fact that seisin was fundamentally enjoyment and that right
could arise only from enjoyment is fully confirmed by all the best-
known rules of the law of real property,33 as it is by the nature of
the protection given to chattels in early English law. The ancient
remedy by self-help, the raising of hue and cry, could only be used
by him who had been in possession; trespass lay only for him who
was in possession, and even the bailor-at-will could only sue under
the fiction that he had possession. When the nineteenth century
reformed the law of real property to allow rights to exist and to
pass independently of possession 34 it thereby revolutionized the fun-
damental part of that law.
The early history of seisin is based upon maintaining an intimate
connexion between the fact of possession and the rights that stem
from it, but it treats the former as primary and this invites us to
begin the study of possession in the common law by concentrating
our initial attention firmly on the legally recognized fact of pos-
session. Logically, one must be very careful indeed to keep this fact
of possession clearly distinct from the various rights of or to pos-
session. Rights may hold good against one person and not against
another; they may stem from the present fact of possession, from
a past fact of possession, or from title or previous right. Prima facie,
this is not true of the fact of possession: initially, it enters the law
as a fact, independent of any circumstances outside itself, true or

of acqumng grants or enfeoffments from feudal superiors. There is little doubt

that the charter of a great baron and the charter, word or deed of the King
successfully put an end to rival claims But the interesting point is that such
feudal acts do not seem to become part of the law as a system; they rather suspend
legal process, putting an end to litigation by the authority of power. Even later,
when the law of real property had come to recognize rights existing independently
of enjoyment, the inferiority of 'seisin in deed' to 'seisin in fact' shows the con-
tinuing weakness of right sans enjoyment within the law. Thus Bracton, who
recognized a very high degree of royal power, but strove to bring as much of
this power as possible within the law, wrote that royal charters may be questioned
by no justice or private person, and that questionable interpretation and alleged
falsifications and t:rasures must be brought coram ipso rege: see Bracton, 34, 11,
109-110, and for a general discussion. Miller, 'The Position of the King in Bracton
and Beaumanoir' (1956) 31 Speculum 263.
33 This is the very theme of Joshua Williams' Seisin of the Freehold (1878) and of
Lightwood's Treatise on Possession of Land, where ample confirmation may be
found.
34 The fifteenth century device of conveying land to X to the use of Y (in order
to avoid feudal disabilities, dues and other incidents of tenure) created a situation
in which equity but not common law came to recognize a right to seisin inde-
pendently of having been put in seisin. The sixteenth century Statute of Uses,
permitting greater freedom of open conveyance at the price of registration and
payment of dues, sought to reunite seisin and right by giving the seisin to Y, but
in doing so was in fact conceding a certain primacy of right. The involved history
of its effect on the law, and its failure in the long run to maintain the union of
seisin and right is traced in Plucknett, A Concise History of the Common Law
(5th ed.), ch. 7, 575 and Holdsworth, Historical Introduction to the Land Law
(1927) 151-166.
 Possession in the Common Law 4S9
false against the whole world. It is something the law takes from
life and does not simply impose on life; it is what Kocourek aptly
calls an 'infra-jural relation'.35 Though particular judges have failed
to keep the logical distinction between the fact of possession and
the rights of possession clearly before their minds, at least for the
whole length of the case before them, early and modern cases never-
theless leave no doubt that English law does recognize such a fact
of possession as logically distinct from the right or rights to pos-
sess,36 and that it treats this fact of possession has having, by itself,
a fundamental importance in the common law. 31
The fact of possession on which we are placing such emphasis is
in essence that state of affairs which Pollock calls (de facto posses-
sion' and which judges frequently refer to as 'mere' or 'actual pos-
session'. Pollock defines this de facto possession as 'effective occupa-
tion or control',38 a definition accepted but given some necessary
elaboration by Holmes in his discussion of possession. 'To gain
possession,' he writes,39 'a man must stand in a certain physical re-
lation to the object and to the rest of the world, and must have
a certain intent.' There must be a degree of power over the object,40
35 Kocourek, Jural Relations, esp. 410-413, 419-420. Cf. Holmes, writing on pos-
session in The Common Law at p. 214: 'Every right is a consequence attached
by the law to one or more facts which the law defines . . . . When a group of facts
thus singled out by the law exists in the case of a given person, he is said to
be entitled to the corresponding rights. . . . The word 'possession' denotes such
a group of facts. Hence, when we say a man has possession, we affirm directly
that all the facts of a certain group are true of him, and we convey indirectly or
by implication that the law will give him the advantage of the situation.'
36 See, among numerous cases, Coverdale v. Charlton (1878) 4 Q.B. D. 104, 127
(C.A.) per Cotton L.J.: 'The plaintiff in this case [an action for trepass] cannot rely
upon any question of right independent of enjoyment by mere possession.' In
The Jupiter (No. 3) [1927] 122, 135-136, the distinction is recognized and insisted
upon as one of logical and not merely technical importance (per Hill J.) In fact, one
has only to read some of the decisions in cases arising under the British Bills of
Sale Acts, Rent and Mortgage Interest Restriction Act 1923, and other Acts where
legislative provisions require the court to focus its attention on the question
whether there is 'apparent' or 'actual possession', to see the immediate gain in
clarity and logical consistency that results from the separation of possession from
the right to possess. See, for example, Ancona v. Rogers (1876) L.R. I Ex. Div. 285,
Ex parte Fletcher (1877) 5 Ch. D. 8og, Hall v. Rogers (1925) 133 L.T. 44 and
Holt v. Dawson [1939] 3 All E.R. 635 [1940] I K.B. 46.
31 The term, 'right to possess', like the identification of seisin with enjoyment,
suggests that possession normally appears in the law as an advantage and not as
a burden. Holmes, in the citation above, also tends to treat it as such. But pos-
session is equally important in the law relating to duties. In the relevant tortious
actions, the plaintiff finds it necessary to prove a nexus between the defendant
and the object causing harm: normally, this is possession. Thus, in cases of
liability for damage done by animals where the scienter rule applies, only those
who have possession-control of the animal can be liable: North v. Wood [1914]
I K.B. 629; Knott v. London County Council [1934] I K.B. 126; Salmond, Law of
Torts (12th ed.), 591 and Williams, Liability for Animals (1938) 324-326.
38 Pollock and Wright, Possession in the Common Law 12.
39 Holmes, op. cit. 216.
40 Ibid. 216, 220. The power is the power to enjoy or use the thing and a man
cannot have possession of things over which he has no such power, e.g. of wild
animals roaming his land. But the degree of power is to be considered in the
490 Melbourne University Law Review [VOLUME 4

an intent to exclude others from it 41 and a 'relation of manifested

power to exclude others co-extensive with the intent' to do SO.42
The crucial thing here is the emphasis on power and controL Pos-
session is not mere physical detention-such detention in pristine
form rarely confronts the law (people do not keep their belongings
chained to their wrists); such detention readily shades off into forms
of control (,detaining' in one's house or one's office, for instance)
and is practically useless as a fundamental concept on which to
build a structure of rights and duties. Possession, one might say,
is the present physical power to use, enjoy or deal with a thing, on
one's own behalf and to the exclusion of all others. This definition,
subject to the interpretative qualifications stressed in the preceding
footnotes, comes close to accuracy, but misses one vital element: the
requirement that a man should not only have the power, but should
also will and intend it. The addition of the element of will and
intent converts power into control, interpreted as a conscious, deli-
berate relation-a relation that requires us to know what we are
doing in the same way as truly counting or speaking a language re-
quires us to know what we are doing; a parrot can do neither. Our
definition of possession, as a fundamental and general concept in
the law, thus becomes: Possession is the present control of a thing,
on one's own behalf and to the exclusion of all others.
Attempts to define 'possession' in general terms, and in terms of
its factual content, tend not to command the respect of contem-
porary common lawyers. 'The idea of possession,' we read in Dias
and Hughes,43 'is no longer tied to fact, and it has become a concept
of the utmost technicality'. Mr D. R. Harris makes it the general
thesis of his essay on 'The Concept of Possession in English Law'44
'that the English decisions preclude us from laying down any con-
ditions, such as physical control or a certain kind of intention, as
absolutely essential for a judicial ruling that a man posseses some-
thing.' The former statement, I submit, is false; the latter is true,
but taken, as it is intended to be taken, as a platform and programme

light of the control that can be exercised over the thing in question. Thus in
The Tubantia [1924] P. 78, the limited control possible over a wreck at the bottom
of the sea was recognized as sufficient for possession. As Lord Fitzgerald put it
in another case-Lord Advocate v. Young (1887) 12 App. Cas. 544, 556: 'By pos-
session is meant possession of that character of which the thing is capable.'
41 Ibid. 220.
42 Ibid. 216, 234, 235. The words 'manifested power' are meant to emphasize
that secret power, future power or present intention to exercise future power are
not sufficient to give possession. The law, however, does not use the term 'power'
to mean 'capacity to deal with any possible or foreseeable eventuality.' A man~s
power may be very precarious indeed; as long as it has not been effectively chal-
lenged the law ignores its precariousness and recognizes his control as possession.
This is -the point of Holmes' example of the child gaining possession of a pocket-
book under the nose of a powerful ruffian.
43 Dias and Hughes, op. cit. 317. 44 Oxford Essays in Jurisprudence, 69.
NOVEMBER I g64] Possession in the Common Law 491

for legal inquiry, it is dangerously limiting and misleading. For the

concept of possession that I have attempted to define above is, I
shall argue, a concept of fundamental importance to the common
law-a concept in terms of which, or against the background of
which, all uses of the term 'possession' in the common law are
to be explained and understood. Our emphasis on 'control', no
doubt, simply transfers many of the problems arising in the defini-
tion of 'possession' to a different context: but it is precisely in this
context of control, I should argue, that the problems can best be
understood and the decisions best be made.
The general-factual-concept of possession, I have argued, is
that in terms of which or against the background of which all uses
of the term 'possession' in the common law are best understood. This
is not to say that all uses of the term can simply be reduced to the
general concept of possession with its emphasis on the fact of control.
Firstly, there will be 'constructive' or simply confused uses of the
term 'possession' where judges wish to award certain rights nor-
mally stemming from possession even though such possession does
not exist. 45 Secondly, there will be areas in which technical rules
have been established that modify, distort or make unusually rigid
the criteria of possession used in the general concept-for example,
the rules applied to the bailee breaking bulk or to the servant re-
ceiving from a stranger and in the modern development of larceny
to cover cases where the initial receipt was not a 'taking'.46 Thirdly,
the term 'possession' has often been used, unhappily, in the growing
number of commercial cases where indicia of title and power to
assume control over the disposition of the goods are far more im-
portant than the actual possession of the goods. 41 Fourthly, in a
number of (normally penal) statutes, the term 'possession' will quite
45 The confusion is especially evident in Hotham B., delivering the opinion of the
Twelve Judges that a coach driver has possession of goods in the coach vis-a-vis
thieves but not vis-a-vis the coachmasters: R. v. Deakin & Smith (1800) :1 Leach
862; 168 E.R. 530. Needless complication and confusion could be avoided by grant-
ing the right to prosecute for larceny to one who, through his charge, has a
responsibility to keep the goods. In effect, this is what the Court did in R. v.
Harding (1929) 46 T.L.R. 106, 21 Cr. App. Rep. 166.
46 The modern developments in larceny, however, are best understood as part of
a slow but definite movement away from treating offences against possession as
the paradigm of reprehensible violence toward treating offences against title as
the paradigm of reprehensible dishonesty. The distortions of the term 'possession'
in larceny thus become temporary ways of masking the fact that larceny is be-
coming essentially an offence against title where it was once essentially an
offence against possession.
41 The use of 'possession' here masks the fact that the courts are concerned
not with 'possession' but with 'rights to possess or dispose': the passing of indicia
of title to goods in warehouses, ships, etc. makes the recipient not a bailee, but
a bailor. It should be noted, despite common misinterpretations, that the im-
portant judgments in Hiort v. Bott (1874) L.R. 9 Ex. 86 specifically avoid imputing
'possession' of the goods in question to the recipient of the indicia of title, though
they make him responsible for wrongful disposition as amounting, in the cir-
cumstances, to a type of conversion (which does not seem to require possession).
492 Melbourne University Law Review [VOLUME 4
properly be construed by the courts in the context of the statute
and its purpose, so that only certain types of control will come under
the term. But we will illuminate or even understand none of these
special situations if we treat them as no more than examples of
arbitrary and disparate technical rules: illumination comes, I should
argue, only when we match the decisions or rules against the general
concept of possession and seek to explain why or how a departure
has come about.
Keeping in mind our emphasis that 'posesssion' is an infra-jural
relation, that it centres on the factual concept of control and that
it has bred certain special concepts for limited circumstances, we
are now in a position to account for the implausibility of earlier
writers and the nihilism of modern ones, and to suggest the lines
along which a systematic account of possession in the common law,
which overcomes their difficulties and their criticisms and resolves
the needless complexities of their terminology, can still be given.
Possession is an infra-jural relation. The common tendency to
distinguish a 'lay' sense of possession from the 'legal' sense of
possession, or de facto possession from de jure possession-in so far
as it is not merely an attempt to distinguish the general concept of
possession in law from the special, limited concepts-is therefore
mistaken. One can no more distinguish de facto possession from the
legally recognized fact of possession that makes up our general
concept than one can distinguish the animal in nature from the
animal recognized by zoology. The problem, rather, is that of the
relation of a technical term to a fact-a specific case of the general
problem of the relation between words and the facts they describe.
Language is a classificatory system; it picks out and isolates those
common features of things that men regard as important for some
purpose or other. The things themselves having these features are
infinitely complex, they fall into an infinite variety of other classes
as well, they shade off into border-line areas which no system of
classification can convert into sharp demarcations. The lawyer, like
the zoologist, will want to draw the line somewhere: to that extent,
he will need to make a decision and the decision will be influenced
by the purposes for which his classification is made. But the line
he draws will still be a line between objectively different facts or
situations: the lawyer no more invents the characteristics of pos-
session or control than the zoologist invents the characteristics of
animals. In seeking to classify, both are no doubt forced to recognize
complexities which few laymen have recognized-to that extent
their use of terms like 'possession' or 'animal' displays a degree of
sophistication not found in the layman's use and may diverge from
it in a significant way. But there is not a lay animal and a zoolo-
NOVEMBER 1964] Possession in the Common Law 493
gical animal, just as there is not a lay possession and a legal pos-
session in the common law. 48
The fact that possession is an infra-jural relation accounts for
some of the difficulties that have accompanied the attempt to define
it. The problems of describing and ultimately confining its nature
are like those that arise in determining what is or constitutes con-
sent and what amounts to a tenement, they are not like telling a
man how to become a trustee or how to contract a marriage. Pre-
cisely for this reason we need a definition that is open-textured, that
uses another infra-jural term like 'control' which gives a court
guidance on the criteria to emphasize without tying it in a strait-
jacket of formal definitions and concepts. This open-texturedness
is to be found in other basic concepts in the law, such as that of
'reasonable care' in the law of negligence. A great deal of harm
is done to the deliberate flexibility of the law when judges and
lawyers begin to seek rigid rules of law as substitutes for the
commonsense discretion deliberately left them, when they begin
to erect 'the law'-relating to liability in the handling of lime and
mortar, or to the liability for children playing on a turntable. It
must be said, that in the civil law judges have resisted this temp-
tation in relation to the general concept of possession; one of the
positive harms that could result from jurisprudential treatment
of possession as a set of highly technical terms would be the
gradual undermining of this resistance.
Possession, we have said, is a fact and not a right; it is an infra-jural
and not a jural relation. When Earl Jowitt says that 49 'English law
has never worked out a completely logical and exhaustive definition
of possession' he is rightly drawing attention to the indeterminacy or
'ambiguity' of the term; but this indeterminacy or 'ambiguity' is
the normal indeterminacy and ambiguity of life, not the vicious
indeterminacy and systematic ambiguity of an incoherent system
or an inadequate principle.
To say that possession is a fact, not a right, is not to say that
the existence of other rights, of the legal system and of the sanctions
it imposes, cannot be part of the factual situation the law is con-
sidering. When I keep things in my house or on my land, even
though it be unlocked or unguarded, I am taking advantage of the
sanctions against trespass and the general respect for them; the
existence of legal rights and duties and the moral respect for them
thus become part, or at times even all, of the factual basis of my
48 I have benefited, in this paragraph and the one that follows, from Dr. Eugene
Kamenka's series of lectures on 'Logic, Ethics and the Law', delivered in the
Faculty of Law, University of Singapore, during 1958-1959.
49 United States of America & Republic of France v. Dollfus Mieg et Cie S.A.
& Bank of England [1952] A.C. 582, 605.
494 Melbourne University Law Review [VOLUME 4
control. On the other hand, when I place my goods on another's
land, unbeknown to the occupier, or when I place them on a public
highway, I place my control in jeopardy by neutralising, or turning
against me, the sanctions applied against trespass. I do not thus
automatically lose possession, but I do raise serious doubts about
whether I have it.
Where the facts are uncertain or finely balanced in favour of
either party, decisions still need to be made. Here, one must then
look to normally extraneous considerations, to rights and even to
policy, i.e. to the normative functions of law. Thus, where neither
party has exclusive possession, possession goes to him who has
better title. 50
Possession, we have said, is control. Since there is no logical or
legal reason why control cannot be exercised from a distance, there
is no warrant for drawing a distinction, as though it were funda-
mental or logically necessary, between 'actual' and 'constructive'
possession. Since there is no logical or legal reason why control can-
not be exercised through another as an instrument, there is no
warrant for distinguishing, as two separate types of possession with
separate principles, 'mediate' and 'immediate' possession. Since
'control' is per se self-conscious and willed, there is no warrant for
treating as odd or exceptional those cases in which intent is not
singled out as a separate element, or in which the intent to control
a thing is presumed from its attachment to or incorporation in
something else that the possessor clearly controls and intends to
control.
On our statement of control, it is clear that a servant acting qua
servant has no possession. 51 This, indeed, is the general line taken
by the courts; the tendency, in earlier law, to ascribe possession to
a servant carrying goods for his master abroad, is to be understood
as the courts' recognition of the weakening of the master's control,
of the fact that the servant will now have to meet situations in-
dependently and thus become more like a bailee. 52 The common law

50 Littleton, s. 701, and Maule J.'s well-known dictum in Jones v. Chapman

(1847) 2 Ex. 803, 821 In Wuta-Ofei v. Danquah [1961] 3 All E.R. 596, a Privy
Council appeal from Ghana, their Lordships accepted tenuous evidence for pos-
session on the basis that there was very full evidence of title, though here title
was taken as creating a factual presumption in favour of intention to control.
51 As Coke puts it the servant has no possession but only a 'charge': Third
Institutes, 108; the courts have accepted this principle, but have frequently ob-
scured its force, by saying, in the course of finding possession for the master, that
'the servant's possession is the master's possession': Cordon v. Harper (1796) 7 T.R.
10, 12, 101 E.R. 828, 829, per Grose J.; Ward v. Macauley (1791) 4 T.R. 489, 490,
100 E.R. "35, per Buller J.; Ward v. Turner (1752) 2 Ves. Sen. 431, 28 E.R. 275;
Anonymous 1 Salk. 289, 91 E.R. 256, per Holt C.].
52 The fluctuations in the law on this point directly reflect the relevance of
communications and commercial practice (e.g., suing on uninsured goods) to the
central issue of control. In earlier times, the comparative factual independence of
NOVEMBER 1964] Possession in the Common Law 495

principle that a servant receiving on behalf of his master from a

third person has 'original, exclusive and lawful possession' till he
brings the thing to his master or puts it in 'the place of ultimate
deposit', is less easy, to understand. The 'place of ultimate de-
posit', it is true, suggests our own emphasis on control: its im-
portance is not that the place is in the control of the master (it may
be a wheelbarrow in the hands of the servant) but that the servant's
act in placing it there is visible confirmation that he is acting qua
servant and submitting to the master's control. Would it not be
more consistent to say-admittedly in the face of precedent-that
in receiving the goods qua servant to deal with 'in the course of his
employment' the servant has already made such submission and
brought the goods within his master's possession. 53
The distinction drawn by the courts between receipt from the
master and receipt from a third person no doubt partly also rests on
the initial difficulty they felt in ascribing to the master control over
goods of the existence of which he may not be aware. This difficulty
has been felt in other areas of possession as well. Its intricacies can
hardly be dealt with in this paper, but I should like to submit that
the general approach to the question, least disturbing to the general
concept of possession, can best be based on the proposition that

the servant abroad was reinforced by the desirability, in the master's own interest,
of allowing the servant to sue the trespasser and prosecute the thief without the
delaying resort to the master's instructions; in modern times, even the master
of the ship in radio communication with the owners becomes a servant and not
a bailee. In The Jupiter [1927] P. 122-not following Pitt v. Gaince (1700) 1 Salk.
10, 91 E.R. 10 and Moore v. Robinson (1831) 2 B. and Ad. 817, 109 E.R. 1346---
Hill J. specifically refers to the greater control exercisable in modern conditions as
reason for rejecting the authority of previous cases and finding that the master
of the ship has only the custody of the ship and cargo. The distinction between
the servant and the bailee, and a crucial test of the existence of bailment, is
the bailee's power to exclude the whole world, including the bailor, until the
bailment be determined, while the servant has no power to exclude the master
at any time. See Tay 'Bailment and the Deposit for Safe-Keeping' (1964) 6
Malaya Law Review (December).
53 The view that a servant finding 'in the course of his employment' brings
the thing found directly into the possession of his master has indeed entered the
law. In McDowell v. Ulster Bank (1899) 33 Ir. L. Times 223, where the Court
awarded the Bank possession of a parcel of banknotes found by the porter-
plaintiff when sweeping out the Bank after business hours, Palles C.B., said: 'I
decide [this case] on the ground of the relation of master and servant, and that
it was by reason of the existence of that relationship and in the performance of
the duties of that service, that the plaintiff acquired possession.' (Judges, alas,
will be loose in describing the decision which they have reached with great care:
what Palles C.B., had decided, as he himself makes clear, is that the plaintiff did
not acquire possession.) The Irish decision is quoted with respect and approval
by Dixon J. (as he then was) in the Australian case Willey v. Synan (1937) 57
C.L.R. 200, esp. 216-217, and both cases are aproved in an obiter dictum of McNair
J. in Corporation of London v. Appleyard [1963] 2 All E.R. 834, 838, [1963] 1
W.L.R. 982. See Tay 'Possession, Larceny and Servants: Towards Tidying Up an
Historical Muddle', to be published in the University of Toronto Law Journal,
January 1965.
496 Melbourne University Law Review [VOLUME 4
control of an object not specifically known is possible, through its
entering a wider sphere that is known and controlled. 54
The departure from the factual conception of control can be
seen in the 'special' concept of possession created in the developed
law of larceny. Here, the criminal requirement of mens rea-for-
mulated strongly as 'intent permanently to deprive'-is conjoined
with the requirement of taking from the victims possession, and
since the law holds (with some exceptions) that there is no larceny
when the guilty intention is formed after the taking, judges have
distorted the general concept of possession in order to be able to
postpone the moment of 'taking' to the moment when the guilty
intent is formed and thus convict. They have been able to do so
by going beyond the intent implied in control to the new require-
ment of intention permanently to hold (or, in the case of carriers,
by drawing a wholly unreal distinction between possessing a package
and possessing its contents).
Here, then, we have a true special concept arising from the dis-
torting influences of criminal policy. Another important special
concept arises more directly out of the nature of social life. Pre-
cisely because the law ascribes rights on the basis of the fact of
possession, no matter how acquired, it is anxious not to recognize
non-consensual acquisition of possession (i.e. the force of might) too
quickly. Hence the concept of 'adverse' possession and the special,
stricter, criteria applied in favour of the original possessor-his bene-
fits of continual claim, right to re-entry, etc.-all amounting to the
proposition that 'possession' or 'control' in the case of non-censensual
possession requires not only domination but also submission. 55
54 This proposition is established in Elwes v. Brigg Gas Co. (1886) 33 Ch. 562
and South Staffordshire Water Co. v. Sharman [1896] 2 Q.B. 4+ Precisely because
the finding cases are concerned with difficult decisions relating to possession and
control-cases where there is absence of specific knowledge and doubt relating
to intention-it is in these cases, since Bridges v. Hawkesworth, that the concept
of possession has been most coherently developed. This process of conceptual
development comes out very clearly in a recent Canadian case, Grafstein v. Holme
and Freeman (1958) 12 D.L.R. (2d) 727, where the Court reviews the earlier deve-
lopment as the development of criteria of possession and insists that there is no
special law of finding. Bridges v. Hawkesworth (1851) 21 L.J.Q.B. 75, '5 Jur. 1079,
I have argued elsewhere, fails to concentrate on the issue of possession because it
fails to break loose from formal requirements arising from the early action of
detinue; the learned judge was concerned not with the question of the occupier's
control, but with the absence or presence of a delivery, by which the possession
could pass. As far as the modern English law of finding is concerned, a careful
study of the relevant cases makes it clear that the decisions do form part of the
working-out of a concept of possession, and that those writers who think that these
decisions are to be explained in terms of a 'judicial policy' for or against the
finder on grounds that are not primarily legal simply fail to grasp the character
of the development and the nature of the problem that the judges were in fact
tackling. See Tay, '''Bridges v. Hawkesworth" and the Early Law of Finding' in
(1964) 8 American Journal of Legal History and Tay, 'Possession and the Modem
Law of Finding', (1964) 5 Sydney Law Review.
55 This requirement is discussed in detail by Thayer, 'Possession and Ownership'
(1907) 23 Law Quarterly Review 175, 314.
NOVEMBER I~4] Possession in the Common Law 497
Other special concepts of possession arise from the special re-
quirements of special areas of social life, particularly from pro-
tections required for commercial arrangements. But in each case,
I submit, the resultant criteria cannot be understood unless we take
into account both the distorting requirements and policies and that
general concept of possession as the fact of control which the law
treats as fundamental and primary.