COMMENTS
COMMENTAIRES
The Undisclosed Principle of Undisclosed Principals
At common law, when A[gent] contracts with a T[hird Party]
on behalf of a P[rincipal] whose existence is not disclosed, both
A and P can sue T and both can be sued by T. This doctrine is
now well established: surprisingly so in view of its history of
reluctant public acceptance,’ characterization by the House of Lords
as an anomaly,2 and denunciation by Holmes and Ames as repug-
nant to common sense and legal principle.3 But in this area com-
mercial utility has triumphed. Since P controls A and receives the
benefits of the transaction, there seems to be no harm in making
P directly answerable to T for acts within the scope of A’s dele-
gated authority. And once an obligation flowing from P to T is
established, there seems to be no injustice in allowing P a recipro-
cal right against T, at least in those situations where the personality
of the obligor is not material.
The legal theory behind all this is harder to ascertain. The
context of triangular consensual dealing seems to require a con-
tract, and contract will indeed account for the mutual rights and
duties of T and A. Sometimes the courts go farther and speak of
the contractual relationship between P and T,4 but this seems to
involve the mysterious transubstantiation of the T- A contract into
a T- P contract, with the T- A contract continuing to exist. In the
absence of an objective manifestation of T’s intention to be bound
to P, it is better to eschew the label of contract as analytically
misleading.5 Perhaps the best that can be done is to recognize
I Scrimshire v. Alderton (1743) 2 Str. 1182, 93 E.R. 1114.
2 Keighley, Maxsted & Co. v. Durant [1901] A.C. 240 (H.L.).
30. W. Holmes, “The History of Agency”, Select Essays in Anglo-American
Legal History (1909), 368, 404; James Barr Ames, Undisclosed Principal – His
Rights and Liabilities (1909) 18 Yale L.J. 443.
4 Said v. Butt [1920] 3 K.B. 497; Cooke v. Eshelby (1887) 12 App. Cas. 271
(H.L.).
5 A. L. Goodhart and C. J. Hamson, Undisclosed Principals in Contract (1932)
4 Camb. L.J. 320, 326. Contrast Glanville L. Williams, Mistake as to Party in
the Law of Contract (1945) 23 Can. Bar Rev. 380, 408, who protests that
since P has rights and duties under the T-A contract, he must be a party
to that contract in any meaningful sense of the word “party”. But the dispute
should not revolve around whether both A and P can be called parties but
1975] UNDISCLOSED PRINCIPLE OF UNDISCLOSED PRINCIPALS
299
that we are dealing with an autonomous doctrine in the law of
agency, historically derivable from assumpsit, which is similar to,
but not identical with, the devices of assignment 7 and trusts and
which operates so as to allow P to intervene upon a bargain struck
by T and A.
This formulation may appear to be simple and obvious,9 but
the jurisprudence of this area is strewn with the torsos of cases
mangled by judicial mistreatment. With regard to the problems
of the parol evidence rule’ 0 and set-off,”x the courts now seem
to be aware that the first step is to elucidate the bargain upon
which P is intervening, but no coherent theory, however simple,
seems to underlie the treatment of situations where A has been
paid by either T or P 12 or where T, having sued either A or P,
now wants to sue the other.”
rather whether they can be called parties in the same sense, and whether
calling both A and P “parties to the contract” contributes to the understanding
of the undisclosed principal doctrine and to its intelligent application to
shifting patterns of facts.
6Wm. Draper Lewis, The Liability of an Undisclosed Principal (1909) 9
Colum. L. Rev. 116; Warren Seavey, The Rationale of Agency (1920) 29 Yale
LJ. 65.
7 Goodhart and Hamson, supra, f.n.5.
8 P.F.P. Higgins, The Equity of the Undisclosed Principal (1965) 28 M.
L.R. 167, 170.
9 It is, however, significantly different from two of the currently popular
theories. Goodhart and Hamson, supra, f.n.5, have P intervene upon the T-A
contract (not the T-A bargain), and Diplock L.J. in Freeman and Lockyer v.
Buckhurst Park Properties (Mangaf), Ltd. [1964] 1 All E.R. 630, 644, regards
the P’s power to sue T and be sued by him as a procedural device that avoids
circuity of actions. Neither of these theories can cope with the situation dealt
with in this paper where the T-P relationship may be enforceable although
the T-A one is not.
10 Murphy v. Rae [1967] N.Z. L.R. 103 (S.C.).
11Campbellville Gravel Supply v. Cook Paving (1968) 70 D.L.R. (2d) 355
(Ont. CA.). Contrast Greer v. Downs Supply Co. [1927] 2 K.B. 28 (C.A.) where
T did not have to pay anything, even subject to his set-off against A, because
there was no P-T contract (as if there ever is).
12Butwick v. Grant [1924] All ER. Rep. 274 (K.B.); Drakeford v. Piercy
(1866) 7 B. & S. 515 (Q.B.); Armstrong v. Stokes (1872) 26 L.T. 872 (Q.B.);
Irvine v. Watson (1880) 42 L.T. 800 (C.A.), distinguished from the previous case
although one member of the Court, BramwelL.J., was baffled as to what
was significant about the distinguishing feature.
13The doctrine of “election” is technically justified either on the supposed
inability to assert inconsistent rights or on a theory of merger: e.g., Brennan
v. Thompson (1915) 33 O.L.R. 465 (S.C. App. Div.). But there is nothing inconsis-
tent in T’s suing A on the contract and then P on the bargain embodied by the
T-A contract, and merger wrongly presupposes that there is a single obligation
in contract inherent in both the T-P and the T-A relationship. The arguments
McGILL LAW JOURNAL
[Vol. 21
The purpose of this discussion, however, is not to poke again
through these well-known cans of worms but to vindicate the utility
of the formulation that P intervenes on the T – A bargain through
discussion of several seemingly unrelated English and Canadian
cases. These cases are somewhat trivial, isolated, and atypical, and
their treatment by the courts has been understandably piecemeal,
if not cursory. They have never been treated as a group or with
the recognition that they might be relevant to the elucidation of
a coherent approach to the problem of the undisclosed principal.
Yet, each of these cases illustrates that the incidents of the P- T
relationship differ radically from those of the P-A relationship,
and this characteristic makes this long-ignored group of cases con-
ceptually very interesting.
To begin, let us consider the case of Rhinelander v. Mont.14
Here P, wishing to avoid publicity due to a pending divorce action,
had his house insured with insurer T through A, who did not dis-
close the agency. When the house burned down, T duly paid the
proceeds to A, but the latter refused to pay the money over to P.
The Nova Scotia Supreme Court dismissed P’s suit against A on
the grounds that
… the defendant insured property which was not his and in which he
had no interest. On this policy he could legally recover nothing, and
cannot be said to hold what he did recover in trust for the plaintiff. The
money really belongs to the insurer who paid it.15
This short passage raises several difficulties. First, the action at
hand is between P and A, and as between these parties P has the
superior right by virtue of the general rule of agency that A must
pay over to P money received for P’s use. P’s suit is not based
on a proprietary right to the specific proceeds, to which the de-
fence of jus tertii might (or might not) be available, but is rather
the claim of a creditor against a debtor.’ Secondly, an insurance
policy not founded on an insurable interest is considered to be
illegal as well as unenforceable, with the result that the insured
is not entitled to restitution of the premiums paid to the insurer.17
of Maurice H. Merrill seem unanswerable whether expressed in prose as in
Election between Agent and Undisclosed Principal (1933) 12 Neb. L. Bull. 100,
or in poetry as in Election (Undisclosed Principal) Revisited (1955) 34 Neb.
L. Rev. 613. For an interesting example of a court reaching the right result
while groaning beneath the weight of precedent, see Swanton Seed Service
Ltd. v. Kulba (1968) 64 W.W.R. 161 (Man. Q.B.).
14 [1933] 2 D.L.R. 508 (N.S.S.C.).
15 Ibid., 511 per Mellish I.
16Hydro-Electric Power Commission of Ontario v. Brown (1959) 21 D.L.R.
(2d) 551 (Ont. C.A.).
17Harse v. Pearl Life Assurance Co. [1904] 1 K.B. 558 (C.A.).
1975] UNDISCLOSED PRINCIPLE OF UNDISCLOSED PRINCIPALS
301
By parity of reasoning, the insurer should not be able to recover
proceeds which have passed into the hands of the insured.
The real significance for present purposes, however, does not
lie in these narrow and technical points, but in the court’s un-
mistakable implication that because the insurer T could have with-
held the proceeds from A, it could have withheld them from P.
The reasoning is presumably that P’s right of action is derived from
A’s and therefore can be no more efficacious than A’s, and that
P’s intervention upon the unenforceable T- A contract must itself
result in unenforceability. This is too mechanical however. The
requirement of an insurable interest is a reflection of the indem-
nity nature of property insurance. It is intended to safeguard
against the moral hazard of deliberate destruction and to discour-
age gambling. 8 However pertinent these considerations are when
the claimant is suing on his own behalf for a loss suffered by
someone else, they would not be forwarded by striking down an
insurance policy made at P’s ultimate initiative to cover P’s own
property. After all, there is an insurable interest underlying the
T – P relationship. Once the T – A bargain has been struck, P should
be able to intervene regardless of the enforceability of that bar-
gain by A.
Indeed, if one starts with the proposition that A
must hold the proceeds for P, even A should be allowed to sue.
Since there will be no personal profit by A, there is no reason
to invoke the insurable interest requirement against him.’ 9
A more sensitive treatment of a similar problem appears in
Danziger v. Thompson.20 Here the landlord T was claiming arrears
of rent both from” P and from A, who had signed the lease but
who was in fact P’s infant daughter. The case differs from Rhine-
lander in that here the issue was enforceability by T rather than
enforceability as against T. But if Rhinelander had indicated a
viable general approach to the undisclosed principal conundrum,
that approach could have been applicable here: the T-P obliga-
tion being consequential upon the T-A one, the unenforceability
of the latter because of A’s incapacity as an infant would be fatal
Is Edwin W. Patterson, Essentials of Insurance Law 2d ed. (1957), 109.
19These remarks are directed only toward the court’s handling of the lack
of insurable interest. On the facts one suspects that an insurer might also
have been able to argue concealment, which would raise problems similar
to those adumbrated below regarding misrepresentation, or the personal
nature of the insurance contract (Rayner v. Preston (1881) 18 Ch. D. 1 (CA.);
Springfield Fire & Marine Inc. Co. v. Maxim [1946] S.C.R. 604) which would
require consideration of Said v. Butt, supra, f.n.4 and Dyster v. Randall [1926]
1 Ch. 932.
20 [1944] K.B. 654, [1944] 2 All E.R. 151.
McGILL LAW JOURNAL
[Vol. 21
to, a claim by T against P. The court’s solution, however, was to
grant the action against P but not against A. 21 The personal im-
munity of A from suit on the T – A contract did not unravel, the
T – P obligation, which was collateral
to the T – A bargain. The
policy of the law in protecting infants even at the expense of
contractual expectations did not require, and would not have been
forwarded by, its serving as a shield for the non-infant P. The
result is equitable, as the infancy of A would not be a bar to the
enforcement of the contract against T at the instance of A 22 and
therefore (on the Rhinelander premises) at the instance of P.
Mutuality requires at least the subsistence of P’s obligation to T.2V 3
This brings us to the nice problem posed by the recent case
of Commonwealth Trust Co. v. Dewitt,2′ which involved the rela-
tionship of the doctrines of undisclosed principal and ultra vires.
A was a trust company which had contracted on behalf of an
undisclosed P to purchase T’s shares in real estate and travel
agency enterprises. The objects of corporations such as A were
set out in the Trust Companies Act,24a and while this statute allowed
A to act as an agent, it did not include investments of the sort
made by A in this case. When A purported to repudiate its con-
tract with T, T sued A for damages. Berger J. had no doubt that
the T – A contract standing alone was ultra vires and thus void.
But when he came to consider the significance of the fact that
A was an agent for an undisclosed principal, he quoted from
Bowstead on Agency to the effect that under the undisclosed prin-
cipal doctrine both P and A are liable to T, and concluded:
I think I should apply that here, I do not think it is circumventing the
Trust Companies Act to do so. The Act says a trust company can act
as an agent. It does not say that it can act as an agent, except for
an undisclosed principal. So if it does act as an agent for an undisclosed
principal, a trust company is liable on the contract itself, even if it
21The facts and disposition of the case are actually very confused because
it was reported only for its treatment of the parol evidence rule. It is not
certain that P was actually undisclosed but, from the cases cited, the court
seems to have treated him as if he was. Also the court seems to have
given
to have admitted
her liability. In any event this article proposes to adumbrate a general
approach to a particular problem rather than to recapitulate existing law,
and the cases are accordingly offered as paradigms, not authorities.
22C.G. Cheshire, C.H.S. Fifoot, and M.P. Furmston, The Law of Contract
in favour of A although A seems
judgment
8th ed. (1972), 403.
23The case is in accord with the Restatement of Agency (2d), s.203.
24(1973) 40 D.L.R. (3d) 113 (B.C.S.C.).
24a R.S.B.C. 1960, c.389, Schedules A and B; as amended by S.B.C. 1962, c.65,
ss.4 and 5.
1975J UNDISCLOSED PRINCIPLE OF UNDISCLOSED PRINCIPALS
303
is one which the trust company could not, as principal, enter into. Once
the Act clothes a trust company with power to act as an agent, the law
of agency applies to any transaction that a trust company enters into.A
Thus, A was held liable for breaching its contract with T.
This remarkable denouement deserves re-examination at several
points. First, is Berger J. justified in discounting the possible
circumvention to which he alludes? The Court had itself identified
the purpose of the ultra vires rule as the protection of the expecta-
tions of the investing public: they are entitled to assume that the
funds they deposit will not be used for the purchase of travel
agencies and real estate businesses. 6 It would seem to be cold
comfort to a depositing member of the public to be told that the
transaction is void if the trust company is the sole obligor but is
valid as against the trust company if a second entity is also an
obligor.
Secondly, it does not follow that if the statute allows A to act for
an undisclosed P, A must be liable itself. The thrust of the law of
agency is to indicate that there is a mechanism whereby P can sue
and be sued. To ascribe A’s liability to the law of agency is insuffi-
cient since on this point the law of agency incorporates and reflects
the concepts of contract. The T- A relationship is contractual, and
this contract, like other ultra vires contracts, should be considered
void.
Finally, if the otherwise void T – A contract is enforceable in the
presence of an undisclosed principal, the structure of the reasoning
must be: (i) the invalidity of the T – A contract does not per se
invalidate the T – P obligation; and (ii) the valid T – P obligation
per se validates the T – A obligation. The first branch of this argument
is sound for the same underlying reason as in the Danziger case. The
disqualifying factor is one that is personal to A and should not
enure to the benefit of P. The policy of protecting A’s investors
would not be promoted by excusing P from the contract. P should
thus be able to intervene on the T – A bargain even in the absence of
an enforceable T – A contract. But the second branch seems less
acceptable: the T – P obligation is always an offspring of the bargain
between T and A, and accordingly it cannot infuse vitality into its
creator. If the original T – A relationship does not stand on its own
feet, it cannot be propped up by the collateral T – P one.
The court in Commonwealth Trust seems to have assumed that
under the doctrine of undisclosed principal, T will have the lia-
25 Supra, f.n.24, 125.
26 Ibid., 122.
McGILL LAW JOURNAL
[Vol. 21
bility either of both A and P, or of neither. This is indeed usual,
but it is not inevitable. There are two separate obligations involved:
one founded on contract and the other on the sanctioning by tlie law
of agency of P’s intervention on A’s bargain. Either obligation can
have disqualifying features peculiar to it which leave t1te other
obligation unaffected.2 7 This does not mean that the two obligations
must travel in water-tight compartments. Both are based on the
bargain between P and A, and this bargain can be influenced by
factors external to each individual relationship. It is, for instance,
conceivable that the T- A contract was formed as a result of a
misrepresentation by P operative on T’s mind or a misrepresentation
by T operative on P’s mind.
Thus, in Garnac Grain Co. Inc. v. H.M.F. Faure & Fairclough Ltd.
and Bunge Corp.28 a supplier of vegetable oil was able to force
merchants with whom it dealt to advance loans to it because of its
position of near monopoly. Usually the supplier set up a chain of
contracts ending with Faure, who sold to the ultimate consumer;
but whenever the transaction with the ultimate consumer fell
through the supplier would repurchase the oil from Faure, thus
preserving goodwill and in effect allowing Faure interest on the
loan. On the occasion in question the supplier, in order to extract
advances of capital, set up the circular chain of supplier – first
purchaser – Garnac – Faure – supplier, but secured Garnac’s particip-
ation in the chain by fraudulently misinforming Garnac of the
circular nature of the series of contracts. The supplier was not able
to provide the oil, but Faure insisted upon Garnac’s performance of
their individual contract. Garnac’s defense was that Faure was really
contracting on behalf of the supplier, who was an undisclosed
principal, and that the supplier’s fraud entitled Garnac to rescission.
The case was eventually decided by the Court of Appeal and the
House of Lords in favour of Faure on the grounds that the alleged
agency relationship between Faure and the supplier did not exist.
But Megaw J. at trial,29 and more tentatively, Diplock L.J. on appeal, 0
were prepared to consider the novel legal question raised by Garnac’s
defence, namely whether T could rescind the T – A contract if he was
induced to enter that contract by a fraudulent misrepresentation
27 In all three cases discussed the peculiarity attached to A, but it may
equally attach to P so as to leave the T-A contract intact while striking down
the T-P one. For instance, a transaction might yield a valid T-A contract but
be ultra vires with regard to P. Compare Job v. Lamb (1856) 11 Exch. 539.
28 [1967] 2 All E.R. 353 (H.L.), aff’g [1965] 3 All E.R. 273 (C.A.).
29 [1965] 1 All E.R. 47 (Q.B.).
30 [1965] 3 All E.R. 273, 286.
1975) UNDISCLOSED PRINCIPLE OF UNDISCLOSED PRINCIPALS
305
made by the undisclosed P. They inclined to the view that he could,
because otherwise A would be answerable to P for the fruits of the
T – A contract and P would thus benefit from his fraud. The result is
sound, but at stake is a principle of wider application than to fraud
alone. If the bargain which T has concluded with A and which lies at
the heart of both the T – A and T – P relationships has been tainted,
both those obligations should be unenforceable against T. In the
context of enforceability as against T there is, in view of P’s control
over A and the benefit that P derives from A, no need to treat A and
P as autonomous entities in matters that affect the bargain. Regard-
less of whether the misrepresentation was fraudulent or innocent,3′
and regardless of whether the misrepresentation was made by P or A,
T should not be held to the unfair bargain thus achieved.
If misrepresentation by either P or A should allow T to rescind,
one would expect that conversely T’s misrepresentation would sup-
port rescission by both P and A. Such slender authority as exists,
however, indicates that this is not so. In Collins v. Associated Grey-
hound Racecourses Ltd.32 T issued a prospectus that contained an
innocent misrepresentation. Relying upon this misrepresentation, P
instructed A to purchase shares without disclosing the agency. In
denying P’s claim for rescission, Russell L.J. pointed to the existence
of the contract between T and A and concluded that “rescission of
that contract between [T] and [A] could only be obtained at the
suit of [A] and upon proof of an allegation that [A] had been misled
by the prospectus”. . Here, both elements were absent. But what if it
had been A who was misled and who obtained rescission? Presumably
the effect of rescission of the T – A contract would be to unravel the
T – P obligation as well, for here we are not concerned with a factor
personal to A, such as infancy or the effects of an ultra vires contract,
but with a factor that vitiates the entire bargain. T, prevented from
enforcing a bargain induced by his misrepresentation against A,
should not be allowed to do so against P.
Two consequences seem to follow from this. First of all, if T can
secure the destruction of his obligation through A’s suit, there is no
reason, aside from a misplaced conceptual formalism, to prevent T
from securing rescission directly. Secondly, the bargain on which
slSupra, f.n.29; MegawJ. cautiously restricted his holding to a misrepre-
sentation that is fraudulent.
32 [1930] 1 Ch. 1. See the lucid treatment of this case in the classic article
by Goodhart and Hamson, supra, f.n.5, 352. Those authors accept the decision
as “undoubtedly correct” (at 355). Cf. also Hyslop v. Morel (1891) 7 T.L.R.
263 (Ch. D).
33 Ibid., 37.
McGILL LAW JOURNAL
[Vol. 21
both the T – A and T – P relationship is based seems equally tainted
whether it is A or P who is misled. Russell L.J. perceived here the
presence of a T – A contract and confined the remedy to one available
against a contracting party. But the existence of an obligation colla-
teral to the contract should have pointed to the possibility of an
additional and collateral remedy. The device of rescission for inno-
cent misrepresentation should be looked upon as illustrating, not ex-
hausting, the law’s attitude toward the proprieties of the bargaining
process.
The purpose of this note has been to stress that the legal position
of the undisclosed principal is derived from the T – A bargain even
if that bargain is not clothed with the jural status of contract. Thus,
the unenforceability of the obligations between T and A should have
no effect on the T – P relationship unless the policy behind T – A un-
enforceability would be forwarded if applied as between P and T.
Indeed, on occasion an alertness to the purpose served by an ap-
parently invalidating condition, such as the insurable interest re-
quirement, may indicate that both the T – A and the T – P obligation
should nonetheless be sustained. Moreover, an excuse
that is
personal to A, such as infancy or corporate incapacity, should neither
have any effect on the T – P relationship nor be itself stultified by
the mere existence of P, as in the Commonwealth Trust case. Finally,
since both the T – A contract and the T – P relationship are founded
on the bargain struck by T and A, factors such as misrepresentation
that taint the bargaining process should have the effect of invalidat-
ing both the T – A and the T – P obligations.
To a large extent we have been concerned here with the pathology
of the undisclosed principal doctrine. The cases chosen for illustra-
tion have raised the possibility of T – P and T – A relationships that
are not mirror images of each other. Such cases are rare, isolated,
and to a certain extent freakish. But when one is dealing with a
creature as enigmatic as the undisclosed principal, it may be pre-
cisely such cases that reveal the most about the anatomy of the
underlying institutions.
Ernest J. Weinrib*
* Associate Professor, Faculty of Law, University of Toronto.
