England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
ADMIRALTY COURT

		Royal Courts of Justice
		15th July 1997

Phillips L.J.

On the 26th February 1989 a disastrous shipping casualty occurred. The "River Gurara" was on a laden voyage from Africa to Europe, when she suffered an engine breakdown. She stranded on the coast of Portugal and subsequently broke up with a loss of life and a total loss of cargo. Consignees of the cargo have sued on the bills of lading. Those bills were subject to the Hague Rules and the issue raised on this appeal is the manner of calculating the limit of the shipowners' liability under those Rules. That issue arises in relation to cargo that was containerised. The bills of lading describe the cargo that was said by the shippers to be within the containers as constituting a specified number of "bales" or "parcels" or "bags" or "bundles" or "crates" or "cartons" or "pallets". Article IV rule 5 of the Hague Rules limits shipowners' liability to "£100 per package or unit". The principal issue to be resolved is whether, in the circumstances of this case, the packages on which the limit is to be calculated are the containers, or the individual items within them. Such an issue has been the subject of judicial decision in many other jurisdictions, but this is the first time that it has arisen for determination in this country.

The Bills of Lading

The bills of lading were on the form of the UK West Africa Line. Under that form the carriage of the goods was, by a clause paramount, made subject to the Hague Rules if they formed part of the law of the place of shipment. The law of the places of shipment of the "River Gurara"'s cargo incorporated the Hague Rules, in their unamended form, and thus the contract of carriage was rendered subject to those Rules.

Clause 9(B) of the bills of lading provides:

"Shipper packed containers

If a Container has not been packed or filled by or on behalf of the Carrier

(B) notwithstanding any provision of law to the contrary the Container shall be considered a package or unit even though it has been used to consolidate the Goods the number of packages or units constituting which have been enumerated on the face hereof as having been packed therein by or on behalf of the Merchant and the liability of the Carrier (if any) shall be calculated accordingly.

It is the shipowners' case that the containers in any event constitute the package or unit for the purpose of calculating the Hague Rules limit but that, should there be any doubt about this, Clause 9(B) resolves it in their favour. It is cargo owners' case that, on the true interpretation of Article IV rule 5 it is the items within the containers that constitute the relevant packages or units and that Clause 9(B) is rendered ineffective by Article III rule 8 of the Hague Rules, which provides:

Any clause, covenant or agreement in a contract of carriage relieving the carrier or ship from liability for loss or damage in connection with goods arising from negligence, fault or failure in the duties and obligations provided by this Article, or lessening such liability shall be null and void and of no effect.

On the application of the cargo owners, an Order was made for the trial of the following preliminary issues:

(a) Whether clause 9 of the UKWAL form of the bill of lading is contrary to Article III r.8 of the Hague Rules and hence void;

(b) If the answer to (a) is no, whether the burden of proving that any particular container is "shipper packed" rests upon the Plaintiffs or the Defendants;

(c) In circumstances where a container or pallet has been used to consolidate goods and the bill of lading states not only the number of containers and/or pallets, but also quantifies the number of goods loaded therein or thereon, whether the Defendants are entitled to limit their liability pursuant to Article IV, r.5 of the Hague Rules by referenced to (a) the number of containers or (b) the number of pallets or (c) the number of goods described by the bill of lading as having been loaded therein or thereon.

To answer issue (a) it is necessary first to address issue (c), in order to see whether there is any conflict between Clause 9 of the bills of lading and Article III rule 8.

That was the approach adopted by the trial Judge, Colman J. He held that it was the number of items described by the bill of lading as being within the containers, rather than the number of the containers themselves, that was the basis for calculation of the limit. He further held that, insofar as Clause 9 provided to the contrary, it was contrary to Article III rule 8, and therefore void. In those circumstances issue (b) did not arise. The shipowners now appeal against the Judge's decision.

Question (c) is phrased on the premise that the Hague Rules limit falls to be calculated on the cargo, as described in the bill of lading. Colman J's decision proceeded on that basis. In the course of argument it became apparent that there was an important issue between the parties as to the effect of the description of the goods in the bill of lading. For the shipowners, Mr Kay Q.C., argued that, for the purpose of Hague Rules limitation, the description of the goods in the bill of lading was the definitive basis for calculating the limit. For cargo owners, Mr Russell, Q.C., argued that the limit fell to be calculated on the cargo as actually loaded. The bill of lading would normally be of evidential value, whether simply as prima facie evidence or as a result of estoppel, as to what was loaded, but it was not necessarily conclusive.

This issue is not merely academic. It goes to the root of the approach to the calculation of the limit in this case for which Mr Kay contends. In these circumstances the parties agreed that issue (c) required to be reformulated into a number of separate questions and suggested how this should be done. Varying slightly their suggestion, I propose to consider the following questions:

(1) Where packages are shipped in containers does the Hague Rules limit, on its true construction, fall to be calculated on the number of packages or the number of containers?

(2) What is the effect of the description in the bill of lading on the basis for calculating the Hague Rules limit? In particular,

(3) What is the effect of qualifying the description of the contents of a container by the words "said to contain".

The approach to the construction of the Hague Rules

The Hague Rules were the product of international convention. They were incorporated into the domestic legislation of a large number of seagoing nations and became widely used as the terms which governed the international carriage of goods by sea. Two considerations follow from this. First, it is legitimate when construing the Rules to have regard to their objects, as disclosed by the travaux preparatoires of the Convention. Second, particular respect should be paid to decisions of other jurisdictions in respect of the meaning of the Rules, for the stated object of the Convention was the unification of the domestic laws of the contracting states relating to bills of lading - see Stag Line v Foscolo Mango & Co [1932] AC 328 at 342 per Lord Atkin and at 350 per Lord Macmillan; The Hollandia [1938] AC 565 at 572 per Lord Diplock.

The objects of the limitation provisions of the Hague Rules are considered in a number of the United States authorities to which we have been referred and in a number of learned articles. For present purposes it is helpful to note that:

"one of the main purposes of limitation was to benefit cargo owners..... The intention of the Hague Rules was to give cargo a liberal limit of liability so as to preclude shipowners from inserting clauses in their bills of lading purporting to limit liability to ridiculously low figures"

'The Hague-Visby Rules' by Anthony Diamond Q.C. 1978 Lloyds M.C.L.Q. 225.

Mr Kay did not seek to gainsay this purpose of limitation, but he submitted that there was another purpose, to which he sought to give paramount effect. He contended that one of the objects underlying the Rules was to ensure that the shipowner was able to verify the extent of his liability. Where the nature and value of the goods inside a package were not specifically declared, the limit of liability would attach to the package itself. The number of packages would be apparent to the shipowner so that he could verify the limit of his liability. It followed that if a number of smaller packages were encased in a larger package, the appropriate package for limitation purposes was the larger one, for that was the only one that the shipowner could verify. Applying this principle, where packages were put inside a container, the container was the appropriate package for limitation purposes.

Colman J. was not attracted by this argument, nor am I. The verification principle is not apparent from consideration of the travaux preparatoires of the Convention. Furthermore, as Colman J. observed, rules 3 to 5 of Article III, to which I shall refer in due course, envisage circumstances in which the shipowner will not be able to verify the number of packages shipped.

Mr Russell submitted that, when the Convention was concluded in 1924, a figure of £100 represented a fair figure for the average value of a package shipped. To apply the same figure to a huge container stuffed with many packages would defeat the object of preventing shipowners from limiting their liability to sums that were absurdly low by reference to the average values of cargoes shipped. I consider that there is force in this submission. If Mr Kay is correct, the change in the method of stowing and carrying cargo that occurred when containerisation was introduced effected a radical change in the limitation regime. I would not readily reach such a conclusion.

Mr Russell further submitted that to describe a container as a package was to strain the natural meaning of that word. With this also I agree. In Bekol BV v Terracina Shipping Corporation (unreported - 13th July 1988), which seems to be the only recorded case in which the English court has considered the meaning of "package" in the Hague Rules, Leggatt J. referred to the definition of that word in the Oxford English Dictionary: "a bundle of things packed up, whether in a box or receptacle, or merely compactly tied up". A huge metal container stuffed with goods which will normally themselves be made up in individual packages is not naturally described as a package.

These two considerations alone would lead me, in the absence of authority, to conclude that where the Hague Rules limit falls to be computed in relation to parcels of cargo which are loaded in containers, it is the parcels, and not the containers, which constitute the relevant packages.

When I turn to consider decisions on the point in other jurisdictions, they reinforce my conclusion. They attach significance to the manner in which the bill of lading describes the cargo - and that is a matter which I shall come to in due course - but where the bill of lading describes the cargo by reference to a specified number of packages loaded inside a specified number of containers, the Courts have calculated the Hague Rules limit on the basis of the packages and not the containers. The reasons for so doing have been the same as those which have led me to favour this approach. Thus, in The "Mormaclynx" [1971] 2 Lloyds Rep. 476, in a passage of his judgment which was repeatedly cited in later cases, Chief Judge Friendly, sitting in the United States Court of Appeals, Second Circuit, said at p. 486:

"Still we cannot escape the belief that the purpose of sect. 4(5) of COGSA was to set a reasonable figure below which the carrier should not be permitted to limit his liability and that "package" is thus more sensibly related to the unit in which the shipper packed the goods and described them than to a large metal object, functionally a part of the ship, in which the carrier caused them to be "contained"".

The preference for the packages, rather than the containers in which they are stuffed, at least where the bill of lading states the number of each, has been shown, not merely by the Courts of the United States, but by those of Canada, Australia, France, Holland, Italy and Sweden.

The weight of this international authority, coupled with my provisional conclusion formed independently of it, leaves me in no doubt that in the present case the shipowners' limit of liability should be calculated on the basis of the number of packages carried in the containers rather than the number of containers, unless the manner in which the cargo has been described in the bills of lading requires a contrary approach. That, however, is precisely the position for which Mr Kay contends, and it is helpful at this stage to summarise his submissions as to the effect of the bills of lading.

The effect of the description of the cargo in the bill of lading

Shipowners' Submissions

Mr Kay's submission is that whether the containers or the packages within them provide the basis for calculating the limit of liability depends upon the agreement of the parties, as embodied in the bills of lading. If the bills of lading state the number of containers, but not the number of packages within them, they choose the containers as the basis for limitation. In the present case most of the bills of lading itemise the contents of the containers, but subject to the qualification of 'stc' ('said to contain'). Mr Kay submits that the effect of this qualification is to rob the statement of the contents of the containers of all evidential significance, so that all that the bills of lading do is to enumerate the containers. It follows that it is the number of containers which forms the basis for computing the limit of liability. In so submitting, Mr Kay has relied heavily on decisions of the United States Courts, and those of other jurisdictions which have followed them, and I now turn to review those authorities.

The United States Authorities

The United States enacted the relevant Hague Rules limit of liability in S.4 (5) of the U.S. Carriage of Goods by Sea Act 1936 in the following form:

"Neither the carrier nor the ship shall in any event be or become liable for any loss or damage to or in connection with the transportation of goods in an amount exceeding $500 per package lawful money of the United States, or in the case of goods not shipped in packages, the customary freight unit...."

In 1968 a Protocol was agreed by a number of the States which had enacted the Hague Rules in an attempt to resolve the very problem with which this appeal is concerned. The Rules, as amended by this Protocol, are known as the Hague-Visby Rules. The material provision is Article IV rule 5(c), which provides:

"where a container, pallet or similar article of transport is used to consolidate goods, the number of packages or units enumerated in the bill of lading as packed in such article of transport shall be deemed the number of packages or units for the purpose of this paragraph as far as these packages or units are concerned. Except as aforesaid such article of transport shall be considered the package or unit."

The United States did not enact this Protocol but, as will become apparent, it had some influence on some of the relevant decisions nonetheless.

In Standard Electrica S.A. v Hamburg [1967] 2 Lloyds Rep., one finds in the decision of the Court of Appeals (Second Circuit) a theme that recurs in most of the subsequent decisions - the importance of the description of the goods on the face of the bill of lading. In that case seven pallets were lost, each of which contained six cartons, and the issue was whether the pallets or the cartons constituted "packages" for the purpose of limitation. Chief Judge Lumbard commented at p.195:

"The dock receipt, the bill of lading, and libellant's claim letter all indicated that the parties regarded each pallet as a package....

Inasmuch as we are not faced with a case where the parties have attempted to define the word package by their agreement in a manner that might be repugnant to the Act.... we think that such characterisations are entitled to considerable weight in that the parties each had the same understanding as to what constitutes a "package" and reflect the meaning given that term by the custom and usage of the trade"

The Court held that the pallets constituted packages for limitation purposes.

I have already referred to The Mormaclynx, another decision of the Second Circuit. In that case the bill of lading described the goods shipped as "1 container s.t.c. 99 bales of leather". The bill of lading had a clause which provided:

"shipper hereby agrees that carrier's liability is limited to $500 with respect to the entire contents of each container..."

The Court held that each of the 99 bales constituted a separate package and that the clause was an invalid limitation of liability as it was in conflict with the Act. That decision was followed by the same Circuit in The "Pioneer Moon" [1975] 2 Lloyds Rep. 199 where at p.201 Judge Friendly commented on an argument that a clause in the bill of lading deemed portable tanks to be packages:

"If, as we hold, the tanks furnished by the carrier were not packages, the quoted provision from the bill of lading could not make them so."

Meanwhile, in The "Kulmerland" [1973] 2 Lloyds Rep. 428, the Second Circuit had introduced a new concept. The bill of lading described the goods shipped as "1 container said to contain machinery". The evidence showed that the container contained 350 adding machines, each in a cardboard carton. The issue was whether the limit of liability fell to be computed on the basis of 1 or 350 packages. Judge Oakes advanced the following proposition at p.431:

"The statutory purpose here leads us to suggest what for want of a better term we will call the functional economics test. In this regard, the first question in any container case is whether the contents of the container could have feasibly been shipped overseas in the individual packages or cartons in which they were packed by the shipper. Here it is plain that they could not...

When, as here, the shipper's own individual units are not functional or usable for overseas shipment the burden shifts to the shipper to show why the container should not be treated as the "package"...

Absent shipment in a functional packing unit, the burden is on the shipper to show by other evidence that his units are themselves "packages". Only then does custom and usage in the trade, the parties' own characterisation or treatment of the items being shipped in supporting documentation, or otherwise, and any other factor bearing on the parties' intent become relevant".

For a while the functional economics test was adopted, at least on the Second Circuit, coupled with the approach of attempting to deduce the intention of the parties as to what was to constitute a "package" from the description of the goods used in the bill of lading - see Nichimen Company v. M.V.Farland 462 F.2d 319 and The "Container Forwarder" [1974] 1Lloyds Rep. 119. But in 1976 the scene shifted to the Pacific seaboard where, in The "Aegis Spirit" [1977] 1 Lloyds Rep. 93, District Judge Beeks, an experienced Admiralty practitioner, sitting on the Ninth Circuit in Seattle, subjected it to a powerful attack. In that case, the bill of lading stated under the heading "No. of Containers or Pkgs" "2 containers". It went on to state under the heading "Kind of packages; description of goods":

"SHIPPER'S LOAD COUNT AND SEAL. Said to contain ...601 cartons."

Judge Beeks held that the cartons and not the containers constituted the packages for limitation purposes. He began his analysis by stating at p.99:

"To be satisfactory, a test for determining whether a container is a package must reflect the realities of the maritime industry of today while remaining faithful to the express language and the legislative policy embodied in the pertinent COGSA provisions."

He then went on to observe that it was impossible to derive the functional economics test from the legislation. He added at p.100:

Second, the "functional economics" test suffers from a more serious flaw; it makes the intent of the parties, as revealed by the available evidence, the touchstone in applying the COGSA liability to a containerized shipment. This is true because evidence of intent will serve to rebut the initial presumption and identify conclusively, it would appear, the COGSA package. The better and more traditional approach, which I adopt, is to conscientiously construe the legislation in the factual context seeking to effectuate the legislative, not the parties' intent and purpose. The undoubted objective of 46 U.S.C. s.1304(5) was to establish a minimum floor below which carriers subject to the act could not reduce their liability for cargo damage. If carriers alone, or even carriers and shippers together, are allowed to christen something a "package" which distorts or belies the plain meaning of this word as used in the statute, then the liability floor becomes illusory and negotiable. The package limitation provision serves no purpose whatsoever if the Courts' function in applying it is to merely identify and uphold the parties' private definition of COGSA package. Of course, the parties, characterization may often be wholly reasonable and consistent with the language and purpose of the statute, but the point to be made is that it is not the parties' characterization of the shipment, but the Court's interpretation of the statute, that controls.

A further undesirable side effect of a rule based upon the parties' intentions is its obvious potential for impairing the value and negotiability of ocean bills of lading, due to uncertainty in the allocation of risks with respect to the cargo. The holder of the bill can never be sure what the shipper and carrier "intended" to treat as a package, except to the extent that said intent can be deduced from the four corners of the bill itself. Bills of lading, though, are hardly appropriate vehicles for such expressions of mutual intent, because their contractual terms are commonly the product of unilateral draftsmanship by the carrier incorporating largely self-serving provisions.

Judge Beeks' disapproval of the functional economics test swiftly received the endorsement of other Circuits. Thus in Yeramex International v. S.S.Tendo [1977] A.M.C. 1807 District Judge Kellam, sitting in Virginia, followed Judge Beeks in rejecting the functional economics test and holding that "packages" must be given their normal meaning. Then, in Mitsui & Co. Ltd. v. American Export Lines Inc. 636 F. 2d 807, a decision of the Court of Appeals on the Second Circuit, Judge Friendly commended Judge Beeks' judgment in the Aegis Spirit and rejected the functional economics test. At p.821, after referring to the Protocol, he said:

"Even if the language and purposes of COGSA left us in doubt as to whether carrier-furnished containers whose contents are disclosed should be treated as packages, the interest in securing international uniformity would thus suggest that they should not be so treated."

This decision was followed by a number of cases in which the Courts treated the unamended provisions of COGSA as having the same effect if they had been amended in accordance with the Protocol.

In Binladen BSB Landscaping v. "Nedloyd Rotterdam" (1985) F.2d 1006, a decision of the Second Circuit, Judge Mansfield set out a number of "basic principles", the first of which was:

"that the touchstone of our analysis should be the contractual agreement between the parties as set forth in the bill of lading".

After reviewing the authorities, he propounded the following rule which "not only accords with the Brussels Protocol but has the virtue of certainty":

"when the bill of lading does not clearly indicate an alternative number of packages, the container must be considered as a COGSA package if it is listed as a package on the bill of lading ..."

He stated, however, that because this rule might produce unexpected results it would not be applied retroactively, but only to bills of lading prepared after the date of the decision.

Binladen was followed by the 11th Circuit in Hayes-Leger Associates Inc. v "Oriental Knight". In that case, Judge Kravitch summarised the law in the following two rules:

(1) when a bill of lading discloses the number of COGSA packages in a container, the liability limitation applies to those packages; but (2) when a bill of lading lists the number of containers as the number of packages, and fails to disclose the number of COGSA packages within each container, the liability limitation of S.4(5) applies to the containers themselves.

Once again the Court said that the second rule would only be applied prospectively.

Other foreign decisions

A similar approach to that of the recent American decisions has been adopted by the Canadian Courts - see The Tindefjell [1973] 2 Lloyds Rep. 253 and Hauerkate v. Toronto Harbour Commissioners (1986) 30 DLR 4th 125 and by the New South Wales Supreme Court in P.S. Chellaram & Co. Ltd v. China Ocean Shipping Co [1989] 1 Lloyds Rep. 413. The Courts of Holland and France and Sweden appear to have adopted the same approach. Italy seems to strike a discordant note. In Comesmar v. Carniti, a decision of the Court of Cassation of the 27th April 1984, No.2643, the Court held:

"The container, if used by the carrier, is, in fact, only a limitation of space used to organise the on-board goods - packaged or loose - and is equivalent to loading in a hold.

If, as in this case, the shipper uses a container there are two possibilities: either the packages contained in it will be indicated, in which case the packages will be the unit of measurement; or there will be no indications, and in this case according to the ratio of the convention regulations, the unit of measurement will not be the container, understood as a "package", but the freight unit of the goods contained in it".

Colman J. held that, adopting a purposive approach to construction, it was appropriate to follow the approach of the American, Canadian, Australian, French and Dutch courts. His conclusions were as follows:

I therefore hold that where:

(i) separately packed items have been loaded into a container by the Shipper or his agents and the carrier has had no opportunity to tally or verify the contents of the container; and

(ii) the carrier or his agent signs a bill of lading which, as here, describes under the heading "container No.s" the identification numbers of the various containers received and stages under the heading "Number and Kind of Packages; Description of Goods" words such as "1 x 20" container stc: 8 cases" of goods.

there are for the purposes of Article IV rule 5, 8 packages and not one.

Moreover, if the contents of the container are described by words which leave it unclear whether they are separately packed for transportation, the container will be the package and not the individual items.

If the contents of the container are described in the bill of lading as said to contain so many separately packed items which in turn are said to contain a specified number of separately packed items, the number of packages will be the smallest category of separately packed items so described. For example, in the present case, typical of several bill os lading is one (interest 32) which bears the words "1 x 20' container STC: 8 pallets STC: 1855 bundles Ghana Makore and Sapele Veneer". The correct approach is clearly to treat the bundles and not the pallets as the packages under Article IV rule 5. Once the verification principle has been rejected and it is accepted, following the authorities to which I have referred, that it is the intention of the parties as expressed in the bill of lading which is the main determinant of what is to be treated as a package, there is no logical justification for confining consideration as packages to the larger separately-packed items identified in the bill of lading. The insertion in the bill of lading of the lesser separately-packed items, although they are bound together on pallets or in similar consolidated groups, is a clear indication that those lesser items are to be treated as the unit of measurement for limitation purposes.

Mr Kay's principal criticism of the Judge's conclusions is that he treated the bills of lading as enumerating the packages stowed within the containers, notwithstanding that the bills were qualified by the letters "stc". Mr Kay submits that the Judge should have held that this qualification had the effect that there was no effective enumeration of the contents of the containers, so that the containers fell to be treated as the packages for limitation purposes.

Conclusions

While I appreciate the desirability of international uniformity, I am unable to accept that the basis of limitation under the unamended Hague Rules depends upon the agreement of the parties as to what constitute the relevant "packages", as represented by the description of the cargo on the face of the bill of lading. In according to the unamended Hague Rules the same effect as the Protocol the American courts were, in effect, legislating, as was recognised by the remarkable decision of the Second and Eleventh Circuits to apply their interpretation prospectively only. I do not believe that it is an interpretation that can properly be given to the unamended Hague Rules. My reasons echo those of Judge Beeks in the Aegis Spirit:

1. ) The Hague Rules limitation provisions were designed to prevent shipowners imposing on shippers unrealistically low limits of liability. If the parties are permitted to agree their own definition of "packages", shipowners will, by applying that definition to containers, succeed in evading the minimum limit of liability that the Hague Rules aimed to secure. The American courts repeatedly held that so-called 'boilerplate' clauses in bills of lading, stating that containers were to be deemed to be "packages", were ineffective as being in conflict with the COGSA limit. I find it illogical that they held that shipowners could achieve the same result by ensuring that the number of containers, but not of the packages within them, should appear on the face of the bill of lading.
2. ) Statements in a bill of lading describing the cargo shipped do not constitute an agreement between the parties as to the identity of that cargo.

The effect of the bill of lading

The Hague Rules deal specifically with the effect of the description of the goods in the bill of lading. Article III provides:

3. After receiving the goods into his charge, the carrier, or the master or agent of the carrier, shall, on demand of the shipper, issue to the shipper a bill of lading showing among other things -

(a) The leading marks necessary for identification of the goods as the same are furnished in writing by the shipper before the loading of such goods starts, provided such marks are stamped or otherwise shown clearly upon the goods if uncovered, or on the cases or coverings in which such goods are contained, in such a manner as should ordinarily remain legible until the end of the voyage;

(b) Either the number of packages or pieces, or the quantity, or weight, as the case may be, as furnished in writing by the shipper;

(c) The apparent order and condition of the goods:

Provided that no carrier, master or agent of the carrier, shall be bound to state or show in the bill of lading any marks, number, quantity, or weight which he has reasonable ground for suspecting not accurately to represent the goods actually received, or which he has had no reasonable means of checking.

4. Such a bill of lading shall be prima facie evidence of the receipt by the carrier of the goods as therein described in accordance with paragraph 3(a), (b) and (c).

5. The shipper shall be deemed to have guaranteed to the carrier the accuracy at the time of shipment of the marks, number, quantity, and weight, as furnished by him, and the shipper shall indemnify the carrier against all loss, damages, and expenses arising or resulting from inaccuracies in such particulars. The right of the carrier to such indemnity shall in no way limit his responsibility and liability under the contract of carriage to any person other than the shipper.

Thus, under the Hague Rules, an unqualified description of the goods in the bill of lading does not constitute a binding agreement between the shipper and the carrier that the goods have been shipped as stated, but merely prima facie evidence of that fact. Furthermore, even before the initiation of container carriage, the Rules catered for the possibility that the carrier would have no reasonable means of checking the "number of packages or pieces, or the quantity or the weight", in which case he was under no obligation to record these matters on the bill of lading.

Nor, prior to the Carriage of Goods by Sea Act 1992 did a statement in a bill of lading give rise to an estoppel against the shipowner - Grant v. Norway (1851) 10 CB 665.

It is important to bear in mind that, before any question of limitation of liability can arise, the onus is on the cargo owner to prove his loss. Where he does so by reliance on the bill of lading as prima facie evidence of what is shipped, the description of the goods in the bill of lading will also form the basis of calculation of the Hague Rules limit of liability. Where, however, the shipowner discharges the heavy onus of displacing the evidential effect of the bill of lading, or the cargo owner establishes his claim to damages by reference to evidence extrinsic to the bill of lading, then I am of opinion that the Hague Rules limit of liability falls to be calculated by reference to the particulars of the cargo and its packaging as it is proved to have been on loading, not by reference to the description in the bill of lading. A consignee who proves that he has lost a container containing 3 packages will be subject to a limit calculated by reference to those packages, even though they were not enumerated in the bill of lading. If the shipowner finds himself exposed to a greater limit than that which would have resulted from the goods as described in the bill of lading, he may have a claim for breach of warranty against the shipper under Article III Rule 5.

The position that I have depicted is inconvenient and can lead to uncertainty, which the United States decisions avoid, but that cannot justify an interpretation which the Rules cannot bear. Furthermore, in most cases the problem is theoretical rather than practical, for the parties usually accept that the description in the bill of lading is accurate and base the limit upon it. The Protocol has largely removed the problem in those cases to which the Hague-Visby Rules apply, by substituting a very different limitation regime.

The effect of qualifying the description in the bill of lading

Under the proviso to Article III Rule 3 of the Hague Rules the carrier is not required to state on the bill of lading the number of packages received when he has had no reasonable means of checking this. The proviso plainly applies in relation to packages stuffed in containers by the shipper or his agent. In such circumstances it is commonplace for the bill of lading to state the number of packages as furnished by the shipper, but to qualify the statement with the words "weight, number and quantity unknown". Where the bill of lading is so qualified it does not even constitute prima facie evidence that the goods detailed by the shipper have been shipped:- New Chinese Antimony Co. Ltd. v. Ocean Steamship Co. Ltd., [1917] 2 K.B.664; Attorney-General of Ceylon v. Scindia Steam Navigation Co. Ltd., [1962] AC 60: The "Atlas" [1996] 1 Lloyds Rep. 642. In such circumstances the onus is on a claimant to prove by extrinsic evidence the shipment of any goods which he claims have been lost or damaged. The shipowner similarly has to rely on extrinsic evidence - probably the same evidence - to demonstrate the number of packages upon the basis of which his limitation of liability falls to be computed.

In the present case, the description of the goods in the bills of lading was qualified simply be the letters "stc". Mr Kay submitted that such a qualification has the same effect as qualifying the description of the goods "weight, quantity, number unknown". Assuming that this is correct, I do not see how Mr Kay could properly contend that the parties had agreed that containers were to be the relevant packages for limitation purposes. Where the shipper gives details of the packages shipped and the carrier clauses the bills of lading to indicate that he does not accept those details, the bills of lading manifestly do not indicate any agreement at all as to the description of what has been shipped. Thus even if Mr Kay were correct in his contention that the American authorities should be followed, this would not lead to the conclusion that limitation fell to be calculated on the basis of the number of containers rather than the number of packages proved to have been lost.

The effect of 'stc'

Mr Russell conceded in argument that the effect of qualifying a description of the goods within a container by the words "said to contain" was the same as using the qualification "contents unknown" or "weight, number, quantity unknown". If this concession is rightly made, it will follow, as Mr Russell accepted, that his clients will have the onus of showing, by extrinsic evidence, what cargo was loaded aboard the "Gurara" and lost with the ship, and that the shipowners' limit of liability will similarly fall to be calculated on the basis of extrinsic evidence.

The effect of the qualification "said to contain" does not seem to have been canvassed in argument in the court below. Colman J. proceeded on the basis that that qualification had no effect on the evidential status of the bills of lading. The American decisions equally do not seem to have attached any significance to that qualification. While I appreciate the reasoning that led Mr Russell to make his concession, it seems to me at least arguable that the words "said to contain" do no more than make plain that the carrier is, as required by Article III Rule 3, stating on the bill the "number of packages... as furnished in writing by the shipper" without dissenting from the description, so that the description can be relied upon as providing prima facie evidence as to what was within the containers.

For the purposes of resolving the preliminary issues brought before the Court, the governing principle is that the shipowners limit of liability under the Hague Rules falls to be calculated on the number of packages that are proved to have been loaded within the containers and not upon the number of containers.

Before Colman J. a subsidiary issue also arose as to whether what he described as "obscure wording" on bills of lading which referred to containers loaded with pallets of bales constituted agreement between the parties that the pallets or the bales should constitute "packages" for the purpose of limitation. On a narrow balance he came to the conclusion that the bales were to be treated as packages. I have dissented from the approach of attempting to deduce an agreement of the parties as to limitation from the face of the bill of lading. In the light of Mr Russell's concession, there is also now doubt as to whether the description in the bills of lading is any evidence of the contents of the containers. If one assumes, however, that the bills of lading do constitute such evidence, and that the Court has no other evidence, I would agree with the conclusion of the Judge that the bales rather than the pallets constitute "packages" for limitation purposes.

Clause 9(B) of the Bill of Lading

Mr Kay did not seek to challenge the Judge's decision that, on the premise that packages rather than containers provided the basis for the calculation of the Hague Rules limit, Clause 9(B) of the bill of lading was rendered ineffective by virtue of Article III Rule 8 of the Hague Rules. The Order of Coleman J must be varied so as to accord with this judgment.

The effect of this judgment must be to dismiss the appeal, albeit that neither the issues dealt with nor the answers given to them precisely accord with the decision of the Judge.

Mummery L.J.

I agree with the Judgment of Phillips L.J.

Hirst L.J.

I agree that this appeal should be dismissed, though I for my part would do so on precisely the same grounds as those adopted by Colman J, i.e. applying the approach now adopted by the American, Canadian, Australian, French and Dutch courts, and in effect treating the unamended Hague Rules as having the same effect as the Protocol.

I fully recognise the shortcomings of this approach as demonstrated by Phillips L.J., but in my judgment these are outweighed by the need for international uniformity.

Order: Appeal dismissed; costs in the Court of Appeal to be the respondents' costs; application for leave to appeal to the House of Lords refused; order as per agreed minute of order as provided by counsel.